netdevice.h 139.1 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9
/*
 * INET		An implementation of the TCP/IP protocol suite for the LINUX
 *		operating system.  INET is implemented using the  BSD Socket
 *		interface as the means of communication with the user level.
 *
 *		Definitions for the Interfaces handler.
 *
 * Version:	@(#)dev.h	1.0.10	08/12/93
 *
10
 * Authors:	Ross Biro
L
Linus Torvalds 已提交
11 12 13
 *		Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG>
 *		Corey Minyard <wf-rch!minyard@relay.EU.net>
 *		Donald J. Becker, <becker@cesdis.gsfc.nasa.gov>
14
 *		Alan Cox, <alan@lxorguk.ukuu.org.uk>
L
Linus Torvalds 已提交
15 16 17 18 19 20 21 22 23 24 25 26 27
 *		Bjorn Ekwall. <bj0rn@blox.se>
 *              Pekka Riikonen <priikone@poseidon.pspt.fi>
 *
 *		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.
 *
 *		Moved to /usr/include/linux for NET3
 */
#ifndef _LINUX_NETDEVICE_H
#define _LINUX_NETDEVICE_H

A
Al Viro 已提交
28
#include <linux/timer.h>
29
#include <linux/bug.h>
30
#include <linux/delay.h>
A
Arun Sharma 已提交
31
#include <linux/atomic.h>
32
#include <linux/prefetch.h>
L
Linus Torvalds 已提交
33 34 35 36
#include <asm/cache.h>
#include <asm/byteorder.h>

#include <linux/percpu.h>
37
#include <linux/rculist.h>
38
#include <linux/workqueue.h>
T
Tom Herbert 已提交
39
#include <linux/dynamic_queue_limits.h>
L
Linus Torvalds 已提交
40

41
#include <linux/ethtool.h>
42
#include <net/net_namespace.h>
J
Jeff Kirsher 已提交
43
#ifdef CONFIG_DCB
44 45
#include <net/dcbnl.h>
#endif
46
#include <net/netprio_cgroup.h>
47
#include <net/xdp.h>
48

49
#include <linux/netdev_features.h>
50
#include <linux/neighbour.h>
51
#include <uapi/linux/netdevice.h>
52
#include <uapi/linux/if_bonding.h>
53
#include <uapi/linux/pkt_cls.h>
54
#include <linux/hashtable.h>
55

56
struct netpoll_info;
57
struct device;
58
struct phy_device;
59
struct dsa_port;
60

61 62
/* 802.11 specific */
struct wireless_dev;
63 64
/* 802.15.4 specific */
struct wpan_dev;
R
Robert Shearman 已提交
65
struct mpls_dev;
66 67
/* UDP Tunnel offloads */
struct udp_tunnel_info;
68
struct bpf_prog;
69
struct xdp_buff;
L
Linus Torvalds 已提交
70

71 72
void netdev_set_default_ethtool_ops(struct net_device *dev,
				    const struct ethtool_ops *ops);
73

74 75 76 77
/* Backlog congestion levels */
#define NET_RX_SUCCESS		0	/* keep 'em coming, baby */
#define NET_RX_DROP		1	/* packet dropped */

78 79 80 81 82 83 84 85 86 87 88 89
/*
 * Transmit return codes: transmit return codes originate from three different
 * namespaces:
 *
 * - qdisc return codes
 * - driver transmit return codes
 * - errno values
 *
 * Drivers are allowed to return any one of those in their hard_start_xmit()
 * function. Real network devices commonly used with qdiscs should only return
 * the driver transmit return codes though - when qdiscs are used, the actual
 * transmission happens asynchronously, so the value is not propagated to
B
Bjorn Helgaas 已提交
90 91
 * higher layers. Virtual network devices transmit synchronously; in this case
 * the driver transmit return codes are consumed by dev_queue_xmit(), and all
92 93 94 95 96
 * others are propagated to higher layers.
 */

/* qdisc ->enqueue() return codes. */
#define NET_XMIT_SUCCESS	0x00
97 98 99
#define NET_XMIT_DROP		0x01	/* skb dropped			*/
#define NET_XMIT_CN		0x02	/* congestion notification	*/
#define NET_XMIT_MASK		0x0f	/* qdisc flags in net/sch_generic.h */
L
Linus Torvalds 已提交
100

101 102 103
/* NET_XMIT_CN is special. It does not guarantee that this packet is lost. It
 * indicates that the device will soon be dropping packets, or already drops
 * some packets of the same priority; prompting us to send less aggressively. */
104
#define net_xmit_eval(e)	((e) == NET_XMIT_CN ? 0 : (e))
L
Linus Torvalds 已提交
105 106
#define net_xmit_errno(e)	((e) != NET_XMIT_CN ? -ENOBUFS : 0)

107
/* Driver transmit return codes */
108
#define NETDEV_TX_MASK		0xf0
109

110
enum netdev_tx {
111
	__NETDEV_TX_MIN	 = INT_MIN,	/* make sure enum is signed */
112 113
	NETDEV_TX_OK	 = 0x00,	/* driver took care of packet */
	NETDEV_TX_BUSY	 = 0x10,	/* driver tx path was busy*/
114 115 116
};
typedef enum netdev_tx netdev_tx_t;

117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134
/*
 * Current order: NETDEV_TX_MASK > NET_XMIT_MASK >= 0 is significant;
 * hard_start_xmit() return < NET_XMIT_MASK means skb was consumed.
 */
static inline bool dev_xmit_complete(int rc)
{
	/*
	 * Positive cases with an skb consumed by a driver:
	 * - successful transmission (rc == NETDEV_TX_OK)
	 * - error while transmitting (rc < 0)
	 * - error while queueing to a different device (rc & NET_XMIT_MASK)
	 */
	if (likely(rc < NET_XMIT_MASK))
		return true;

	return false;
}

L
Linus Torvalds 已提交
135
/*
B
Bjorn Helgaas 已提交
136
 *	Compute the worst-case header length according to the protocols
L
Linus Torvalds 已提交
137 138
 *	used.
 */
G
Graf Yang 已提交
139

140 141 142
#if defined(CONFIG_HYPERV_NET)
# define LL_MAX_HEADER 128
#elif defined(CONFIG_WLAN) || IS_ENABLED(CONFIG_AX25)
143 144 145 146 147
# if defined(CONFIG_MAC80211_MESH)
#  define LL_MAX_HEADER 128
# else
#  define LL_MAX_HEADER 96
# endif
L
Linus Torvalds 已提交
148
#else
149
# define LL_MAX_HEADER 32
L
Linus Torvalds 已提交
150 151
#endif

152 153
#if !IS_ENABLED(CONFIG_NET_IPIP) && !IS_ENABLED(CONFIG_NET_IPGRE) && \
    !IS_ENABLED(CONFIG_IPV6_SIT) && !IS_ENABLED(CONFIG_IPV6_TUNNEL)
L
Linus Torvalds 已提交
154 155 156 157 158 159
#define MAX_HEADER LL_MAX_HEADER
#else
#define MAX_HEADER (LL_MAX_HEADER + 48)
#endif

/*
160 161
 *	Old network device statistics. Fields are native words
 *	(unsigned long) so they can be read and written atomically.
L
Linus Torvalds 已提交
162
 */
G
Graf Yang 已提交
163

E
Eric Dumazet 已提交
164
struct net_device_stats {
165 166 167 168 169 170 171 172 173
	unsigned long	rx_packets;
	unsigned long	tx_packets;
	unsigned long	rx_bytes;
	unsigned long	tx_bytes;
	unsigned long	rx_errors;
	unsigned long	tx_errors;
	unsigned long	rx_dropped;
	unsigned long	tx_dropped;
	unsigned long	multicast;
L
Linus Torvalds 已提交
174 175
	unsigned long	collisions;
	unsigned long	rx_length_errors;
176 177 178 179 180
	unsigned long	rx_over_errors;
	unsigned long	rx_crc_errors;
	unsigned long	rx_frame_errors;
	unsigned long	rx_fifo_errors;
	unsigned long	rx_missed_errors;
L
Linus Torvalds 已提交
181 182 183 184 185 186 187 188 189 190 191 192 193
	unsigned long	tx_aborted_errors;
	unsigned long	tx_carrier_errors;
	unsigned long	tx_fifo_errors;
	unsigned long	tx_heartbeat_errors;
	unsigned long	tx_window_errors;
	unsigned long	rx_compressed;
	unsigned long	tx_compressed;
};


#include <linux/cache.h>
#include <linux/skbuff.h>

194
#ifdef CONFIG_RPS
195 196
#include <linux/static_key.h>
extern struct static_key rps_needed;
E
Eric Dumazet 已提交
197
extern struct static_key rfs_needed;
198 199
#endif

L
Linus Torvalds 已提交
200 201 202 203
struct neighbour;
struct neigh_parms;
struct sk_buff;

204 205 206 207
struct netdev_hw_addr {
	struct list_head	list;
	unsigned char		addr[MAX_ADDR_LEN];
	unsigned char		type;
J
Jiri Pirko 已提交
208 209 210 211
#define NETDEV_HW_ADDR_T_LAN		1
#define NETDEV_HW_ADDR_T_SAN		2
#define NETDEV_HW_ADDR_T_SLAVE		3
#define NETDEV_HW_ADDR_T_UNICAST	4
212 213
#define NETDEV_HW_ADDR_T_MULTICAST	5
	bool			global_use;
214
	int			sync_cnt;
215
	int			refcount;
216
	int			synced;
217 218 219
	struct rcu_head		rcu_head;
};

220 221 222 223 224
struct netdev_hw_addr_list {
	struct list_head	list;
	int			count;
};

225 226 227 228
#define netdev_hw_addr_list_count(l) ((l)->count)
#define netdev_hw_addr_list_empty(l) (netdev_hw_addr_list_count(l) == 0)
#define netdev_hw_addr_list_for_each(ha, l) \
	list_for_each_entry(ha, &(l)->list, list)
229

230 231 232 233
#define netdev_uc_count(dev) netdev_hw_addr_list_count(&(dev)->uc)
#define netdev_uc_empty(dev) netdev_hw_addr_list_empty(&(dev)->uc)
#define netdev_for_each_uc_addr(ha, dev) \
	netdev_hw_addr_list_for_each(ha, &(dev)->uc)
234

235 236
#define netdev_mc_count(dev) netdev_hw_addr_list_count(&(dev)->mc)
#define netdev_mc_empty(dev) netdev_hw_addr_list_empty(&(dev)->mc)
237
#define netdev_for_each_mc_addr(ha, dev) \
238
	netdev_hw_addr_list_for_each(ha, &(dev)->mc)
239

E
Eric Dumazet 已提交
240
struct hh_cache {
241
	unsigned int	hh_len;
242
	seqlock_t	hh_lock;
L
Linus Torvalds 已提交
243 244 245 246

	/* cached hardware header; allow for machine alignment needs.        */
#define HH_DATA_MOD	16
#define HH_DATA_OFF(__len) \
J
Jiri Benc 已提交
247
	(HH_DATA_MOD - (((__len - 1) & (HH_DATA_MOD - 1)) + 1))
L
Linus Torvalds 已提交
248 249 250 251 252
#define HH_DATA_ALIGN(__len) \
	(((__len)+(HH_DATA_MOD-1))&~(HH_DATA_MOD - 1))
	unsigned long	hh_data[HH_DATA_ALIGN(LL_MAX_HEADER) / sizeof(long)];
};

B
Bjorn Helgaas 已提交
253
/* Reserve HH_DATA_MOD byte-aligned hard_header_len, but at least that much.
L
Linus Torvalds 已提交
254 255 256 257 258 259 260 261
 * Alternative is:
 *   dev->hard_header_len ? (dev->hard_header_len +
 *                           (HH_DATA_MOD - 1)) & ~(HH_DATA_MOD - 1) : 0
 *
 * We could use other alignment values, but we must maintain the
 * relationship HH alignment <= LL alignment.
 */
#define LL_RESERVED_SPACE(dev) \
262
	((((dev)->hard_header_len+(dev)->needed_headroom)&~(HH_DATA_MOD - 1)) + HH_DATA_MOD)
L
Linus Torvalds 已提交
263
#define LL_RESERVED_SPACE_EXTRA(dev,extra) \
264
	((((dev)->hard_header_len+(dev)->needed_headroom+(extra))&~(HH_DATA_MOD - 1)) + HH_DATA_MOD)
L
Linus Torvalds 已提交
265

266 267 268
struct header_ops {
	int	(*create) (struct sk_buff *skb, struct net_device *dev,
			   unsigned short type, const void *daddr,
269
			   const void *saddr, unsigned int len);
270
	int	(*parse)(const struct sk_buff *skb, unsigned char *haddr);
271
	int	(*cache)(const struct neighbour *neigh, struct hh_cache *hh, __be16 type);
272 273 274
	void	(*cache_update)(struct hh_cache *hh,
				const struct net_device *dev,
				const unsigned char *haddr);
275
	bool	(*validate)(const char *ll_header, unsigned int len);
276 277
};

L
Linus Torvalds 已提交
278
/* These flag bits are private to the generic network queueing
B
Bjorn Helgaas 已提交
279
 * layer; they may not be explicitly referenced by any other
L
Linus Torvalds 已提交
280 281 282
 * code.
 */

E
Eric Dumazet 已提交
283
enum netdev_state_t {
L
Linus Torvalds 已提交
284 285 286
	__LINK_STATE_START,
	__LINK_STATE_PRESENT,
	__LINK_STATE_NOCARRIER,
S
Stefan Rompf 已提交
287 288
	__LINK_STATE_LINKWATCH_PENDING,
	__LINK_STATE_DORMANT,
L
Linus Torvalds 已提交
289 290 291 292
};


/*
B
Bjorn Helgaas 已提交
293
 * This structure holds boot-time configured netdevice settings. They
G
Graf Yang 已提交
294
 * are then used in the device probing.
L
Linus Torvalds 已提交
295 296 297 298 299 300 301
 */
struct netdev_boot_setup {
	char name[IFNAMSIZ];
	struct ifmap map;
};
#define NETDEV_BOOT_SETUP_MAX 8

302
int __init netdev_boot_setup(char *str);
L
Linus Torvalds 已提交
303

304 305 306 307 308 309 310
/*
 * Structure for NAPI scheduling similar to tasklet but with weighting
 */
struct napi_struct {
	/* The poll_list must only be managed by the entity which
	 * changes the state of the NAPI_STATE_SCHED bit.  This means
	 * whoever atomically sets that bit can add this napi_struct
B
Bjorn Helgaas 已提交
311
	 * to the per-CPU poll_list, and whoever clears that bit
312 313 314 315 316 317
	 * can remove from the list right before clearing the bit.
	 */
	struct list_head	poll_list;

	unsigned long		state;
	int			weight;
318
	unsigned int		gro_count;
319 320 321 322
	int			(*poll)(struct napi_struct *, int);
#ifdef CONFIG_NETPOLL
	int			poll_owner;
#endif
H
Herbert Xu 已提交
323
	struct net_device	*dev;
324
	struct sk_buff		*gro_list;
H
Herbert Xu 已提交
325
	struct sk_buff		*skb;
326
	struct hrtimer		timer;
327
	struct list_head	dev_list;
E
Eliezer Tamir 已提交
328 329
	struct hlist_node	napi_hash_node;
	unsigned int		napi_id;
330 331
};

E
Eric Dumazet 已提交
332
enum {
333
	NAPI_STATE_SCHED,	/* Poll is scheduled */
E
Eric Dumazet 已提交
334
	NAPI_STATE_MISSED,	/* reschedule a napi */
D
David S. Miller 已提交
335
	NAPI_STATE_DISABLE,	/* Disable pending */
336
	NAPI_STATE_NPSVC,	/* Netpoll - don't dequeue from poll_list */
E
Eric Dumazet 已提交
337 338
	NAPI_STATE_HASHED,	/* In NAPI hash (busy polling possible) */
	NAPI_STATE_NO_BUSY_POLL,/* Do not add in napi_hash, no busy polling */
339 340 341 342
	NAPI_STATE_IN_BUSY_POLL,/* sk_busy_loop() owns this NAPI */
};

enum {
E
Eric Dumazet 已提交
343 344 345 346 347 348 349
	NAPIF_STATE_SCHED	 = BIT(NAPI_STATE_SCHED),
	NAPIF_STATE_MISSED	 = BIT(NAPI_STATE_MISSED),
	NAPIF_STATE_DISABLE	 = BIT(NAPI_STATE_DISABLE),
	NAPIF_STATE_NPSVC	 = BIT(NAPI_STATE_NPSVC),
	NAPIF_STATE_HASHED	 = BIT(NAPI_STATE_HASHED),
	NAPIF_STATE_NO_BUSY_POLL = BIT(NAPI_STATE_NO_BUSY_POLL),
	NAPIF_STATE_IN_BUSY_POLL = BIT(NAPI_STATE_IN_BUSY_POLL),
350 351
};

352
enum gro_result {
353 354 355 356 357
	GRO_MERGED,
	GRO_MERGED_FREE,
	GRO_HELD,
	GRO_NORMAL,
	GRO_DROP,
358
	GRO_CONSUMED,
359
};
360
typedef enum gro_result gro_result_t;
361

362 363 364 365 366 367 368
/*
 * enum rx_handler_result - Possible return values for rx_handlers.
 * @RX_HANDLER_CONSUMED: skb was consumed by rx_handler, do not process it
 * further.
 * @RX_HANDLER_ANOTHER: Do another round in receive path. This is indicated in
 * case skb->dev was changed by rx_handler.
 * @RX_HANDLER_EXACT: Force exact delivery, no wildcard.
B
Bjorn Helgaas 已提交
369
 * @RX_HANDLER_PASS: Do nothing, pass the skb as if no rx_handler was called.
370 371 372 373 374 375 376 377 378 379 380 381 382 383
 *
 * rx_handlers are functions called from inside __netif_receive_skb(), to do
 * special processing of the skb, prior to delivery to protocol handlers.
 *
 * Currently, a net_device can only have a single rx_handler registered. Trying
 * to register a second rx_handler will return -EBUSY.
 *
 * To register a rx_handler on a net_device, use netdev_rx_handler_register().
 * To unregister a rx_handler on a net_device, use
 * netdev_rx_handler_unregister().
 *
 * Upon return, rx_handler is expected to tell __netif_receive_skb() what to
 * do with the skb.
 *
B
Bjorn Helgaas 已提交
384
 * If the rx_handler consumed the skb in some way, it should return
385
 * RX_HANDLER_CONSUMED. This is appropriate when the rx_handler arranged for
B
Bjorn Helgaas 已提交
386
 * the skb to be delivered in some other way.
387 388 389 390 391
 *
 * If the rx_handler changed skb->dev, to divert the skb to another
 * net_device, it should return RX_HANDLER_ANOTHER. The rx_handler for the
 * new device will be called if it exists.
 *
B
Bjorn Helgaas 已提交
392
 * If the rx_handler decides the skb should be ignored, it should return
393
 * RX_HANDLER_EXACT. The skb will only be delivered to protocol handlers that
394
 * are registered on exact device (ptype->dev == skb->dev).
395
 *
B
Bjorn Helgaas 已提交
396
 * If the rx_handler didn't change skb->dev, but wants the skb to be normally
397 398 399 400 401 402 403 404 405 406 407 408 409 410
 * delivered, it should return RX_HANDLER_PASS.
 *
 * A device without a registered rx_handler will behave as if rx_handler
 * returned RX_HANDLER_PASS.
 */

enum rx_handler_result {
	RX_HANDLER_CONSUMED,
	RX_HANDLER_ANOTHER,
	RX_HANDLER_EXACT,
	RX_HANDLER_PASS,
};
typedef enum rx_handler_result rx_handler_result_t;
typedef rx_handler_result_t rx_handler_func_t(struct sk_buff **pskb);
411

412
void __napi_schedule(struct napi_struct *n);
413
void __napi_schedule_irqoff(struct napi_struct *n);
414

415
static inline bool napi_disable_pending(struct napi_struct *n)
D
David S. Miller 已提交
416 417 418 419
{
	return test_bit(NAPI_STATE_DISABLE, &n->state);
}

E
Eric Dumazet 已提交
420
bool napi_schedule_prep(struct napi_struct *n);
421 422 423

/**
 *	napi_schedule - schedule NAPI poll
B
Bjorn Helgaas 已提交
424
 *	@n: NAPI context
425 426 427 428 429 430 431 432 433 434
 *
 * Schedule NAPI poll routine to be called if it is not already
 * running.
 */
static inline void napi_schedule(struct napi_struct *n)
{
	if (napi_schedule_prep(n))
		__napi_schedule(n);
}

435 436
/**
 *	napi_schedule_irqoff - schedule NAPI poll
B
Bjorn Helgaas 已提交
437
 *	@n: NAPI context
438 439 440 441 442 443 444 445 446
 *
 * Variant of napi_schedule(), assuming hard irqs are masked.
 */
static inline void napi_schedule_irqoff(struct napi_struct *n)
{
	if (napi_schedule_prep(n))
		__napi_schedule_irqoff(n);
}

447
/* Try to reschedule poll. Called by dev->poll() after napi_complete().  */
448
static inline bool napi_reschedule(struct napi_struct *napi)
449 450 451
{
	if (napi_schedule_prep(napi)) {
		__napi_schedule(napi);
452
		return true;
453
	}
454
	return false;
455 456
}

457
bool napi_complete_done(struct napi_struct *n, int work_done);
458 459
/**
 *	napi_complete - NAPI processing complete
B
Bjorn Helgaas 已提交
460
 *	@n: NAPI context
461 462
 *
 * Mark NAPI processing as complete.
463
 * Consider using napi_complete_done() instead.
464
 * Return false if device should avoid rearming interrupts.
465
 */
466
static inline bool napi_complete(struct napi_struct *n)
467 468 469
{
	return napi_complete_done(n, 0);
}
470

E
Eliezer Tamir 已提交
471 472
/**
 *	napi_hash_del - remove a NAPI from global table
B
Bjorn Helgaas 已提交
473
 *	@napi: NAPI context
E
Eliezer Tamir 已提交
474
 *
B
Bjorn Helgaas 已提交
475
 * Warning: caller must observe RCU grace period
476 477
 * before freeing memory containing @napi, if
 * this function returns true.
478
 * Note: core networking stack automatically calls it
B
Bjorn Helgaas 已提交
479
 * from netif_napi_del().
480
 * Drivers might want to call this helper to combine all
B
Bjorn Helgaas 已提交
481
 * the needed RCU grace periods into a single one.
E
Eliezer Tamir 已提交
482
 */
483
bool napi_hash_del(struct napi_struct *napi);
E
Eliezer Tamir 已提交
484

485 486
/**
 *	napi_disable - prevent NAPI from scheduling
B
Bjorn Helgaas 已提交
487
 *	@n: NAPI context
488 489 490 491
 *
 * Stop NAPI from being scheduled on this context.
 * Waits till any outstanding processing completes.
 */
492
void napi_disable(struct napi_struct *n);
493 494 495

/**
 *	napi_enable - enable NAPI scheduling
B
Bjorn Helgaas 已提交
496
 *	@n: NAPI context
497 498 499 500 501 502 503
 *
 * Resume NAPI from being scheduled on this context.
 * Must be paired with napi_disable.
 */
static inline void napi_enable(struct napi_struct *n)
{
	BUG_ON(!test_bit(NAPI_STATE_SCHED, &n->state));
504
	smp_mb__before_atomic();
505
	clear_bit(NAPI_STATE_SCHED, &n->state);
506
	clear_bit(NAPI_STATE_NPSVC, &n->state);
507 508
}

509 510
/**
 *	napi_synchronize - wait until NAPI is not running
B
Bjorn Helgaas 已提交
511
 *	@n: NAPI context
512 513 514 515 516 517 518
 *
 * Wait until NAPI is done being scheduled on this context.
 * Waits till any outstanding processing completes but
 * does not disable future activations.
 */
static inline void napi_synchronize(const struct napi_struct *n)
{
519 520 521 522 523
	if (IS_ENABLED(CONFIG_SMP))
		while (test_bit(NAPI_STATE_SCHED, &n->state))
			msleep(1);
	else
		barrier();
524 525
}

E
Eric Dumazet 已提交
526
enum netdev_queue_state_t {
527 528
	__QUEUE_STATE_DRV_XOFF,
	__QUEUE_STATE_STACK_XOFF,
529
	__QUEUE_STATE_FROZEN,
530
};
531 532 533 534 535 536 537 538 539 540 541

#define QUEUE_STATE_DRV_XOFF	(1 << __QUEUE_STATE_DRV_XOFF)
#define QUEUE_STATE_STACK_XOFF	(1 << __QUEUE_STATE_STACK_XOFF)
#define QUEUE_STATE_FROZEN	(1 << __QUEUE_STATE_FROZEN)

#define QUEUE_STATE_ANY_XOFF	(QUEUE_STATE_DRV_XOFF | QUEUE_STATE_STACK_XOFF)
#define QUEUE_STATE_ANY_XOFF_OR_FROZEN (QUEUE_STATE_ANY_XOFF | \
					QUEUE_STATE_FROZEN)
#define QUEUE_STATE_DRV_XOFF_OR_FROZEN (QUEUE_STATE_DRV_XOFF | \
					QUEUE_STATE_FROZEN)

542 543 544 545 546 547 548 549 550
/*
 * __QUEUE_STATE_DRV_XOFF is used by drivers to stop the transmit queue.  The
 * netif_tx_* functions below are used to manipulate this flag.  The
 * __QUEUE_STATE_STACK_XOFF flag is used by the stack to stop the transmit
 * queue independently.  The netif_xmit_*stopped functions below are called
 * to check if the queue has been stopped by the driver or stack (either
 * of the XOFF bits are set in the state).  Drivers should not need to call
 * netif_xmit*stopped functions, they should only be using netif_tx_*.
 */
551

552
struct netdev_queue {
553
/*
B
Bjorn Helgaas 已提交
554
 * read-mostly part
555
 */
556
	struct net_device	*dev;
557
	struct Qdisc __rcu	*qdisc;
558
	struct Qdisc		*qdisc_sleeping;
559
#ifdef CONFIG_SYSFS
T
Tom Herbert 已提交
560 561
	struct kobject		kobj;
#endif
562 563 564
#if defined(CONFIG_XPS) && defined(CONFIG_NUMA)
	int			numa_node;
#endif
565 566 567 568 569 570
	unsigned long		tx_maxrate;
	/*
	 * Number of TX timeouts for this queue
	 * (/sys/class/net/DEV/Q/trans_timeout)
	 */
	unsigned long		trans_timeout;
571
/*
B
Bjorn Helgaas 已提交
572
 * write-mostly part
573 574 575
 */
	spinlock_t		_xmit_lock ____cacheline_aligned_in_smp;
	int			xmit_lock_owner;
576
	/*
F
Florian Westphal 已提交
577
	 * Time (in jiffies) of last Tx
578 579
	 */
	unsigned long		trans_start;
580

T
Tom Herbert 已提交
581 582 583 584 585
	unsigned long		state;

#ifdef CONFIG_BQL
	struct dql		dql;
#endif
586
} ____cacheline_aligned_in_smp;
587

588 589 590 591 592
static inline int netdev_queue_numa_node_read(const struct netdev_queue *q)
{
#if defined(CONFIG_XPS) && defined(CONFIG_NUMA)
	return q->numa_node;
#else
593
	return NUMA_NO_NODE;
594 595 596 597 598 599 600 601 602 603
#endif
}

static inline void netdev_queue_numa_node_write(struct netdev_queue *q, int node)
{
#if defined(CONFIG_XPS) && defined(CONFIG_NUMA)
	q->numa_node = node;
#endif
}

E
Eric Dumazet 已提交
604
#ifdef CONFIG_RPS
T
Tom Herbert 已提交
605 606 607 608 609 610 611 612 613
/*
 * This structure holds an RPS map which can be of variable length.  The
 * map is an array of CPUs.
 */
struct rps_map {
	unsigned int len;
	struct rcu_head rcu;
	u16 cpus[0];
};
614
#define RPS_MAP_SIZE(_num) (sizeof(struct rps_map) + ((_num) * sizeof(u16)))
T
Tom Herbert 已提交
615

T
Tom Herbert 已提交
616
/*
617 618 619
 * The rps_dev_flow structure contains the mapping of a flow to a CPU, the
 * tail pointer for that CPU's input queue at the time of last enqueue, and
 * a hardware filter index.
T
Tom Herbert 已提交
620 621 622
 */
struct rps_dev_flow {
	u16 cpu;
623
	u16 filter;
T
Tom Herbert 已提交
624 625
	unsigned int last_qtail;
};
626
#define RPS_NO_FILTER 0xffff
T
Tom Herbert 已提交
627 628 629 630 631 632 633 634 635 636

/*
 * The rps_dev_flow_table structure contains a table of flow mappings.
 */
struct rps_dev_flow_table {
	unsigned int mask;
	struct rcu_head rcu;
	struct rps_dev_flow flows[0];
};
#define RPS_DEV_FLOW_TABLE_SIZE(_num) (sizeof(struct rps_dev_flow_table) + \
637
    ((_num) * sizeof(struct rps_dev_flow)))
T
Tom Herbert 已提交
638 639 640 641

/*
 * The rps_sock_flow_table contains mappings of flows to the last CPU
 * on which they were processed by the application (set in recvmsg).
B
Bjorn Helgaas 已提交
642 643
 * Each entry is a 32bit value. Upper part is the high-order bits
 * of flow hash, lower part is CPU number.
644
 * rps_cpu_mask is used to partition the space, depending on number of
B
Bjorn Helgaas 已提交
645 646
 * possible CPUs : rps_cpu_mask = roundup_pow_of_two(nr_cpu_ids) - 1
 * For example, if 64 CPUs are possible, rps_cpu_mask = 0x3f,
647
 * meaning we use 32-6=26 bits for the hash.
T
Tom Herbert 已提交
648 649
 */
struct rps_sock_flow_table {
650
	u32	mask;
651 652

	u32	ents[0] ____cacheline_aligned_in_smp;
T
Tom Herbert 已提交
653
};
654
#define	RPS_SOCK_FLOW_TABLE_SIZE(_num) (offsetof(struct rps_sock_flow_table, ents[_num]))
T
Tom Herbert 已提交
655 656 657

#define RPS_NO_CPU 0xffff

658 659 660
extern u32 rps_cpu_mask;
extern struct rps_sock_flow_table __rcu *rps_sock_flow_table;

T
Tom Herbert 已提交
661 662 663 664
static inline void rps_record_sock_flow(struct rps_sock_flow_table *table,
					u32 hash)
{
	if (table && hash) {
665 666
		unsigned int index = hash & table->mask;
		u32 val = hash & ~rps_cpu_mask;
T
Tom Herbert 已提交
667

B
Bjorn Helgaas 已提交
668
		/* We only give a hint, preemption can change CPU under us */
669
		val |= raw_smp_processor_id();
T
Tom Herbert 已提交
670

671 672
		if (table->ents[index] != val)
			table->ents[index] = val;
T
Tom Herbert 已提交
673 674 675
	}
}

676
#ifdef CONFIG_RFS_ACCEL
677 678
bool rps_may_expire_flow(struct net_device *dev, u16 rxq_index, u32 flow_id,
			 u16 filter_id);
679
#endif
680
#endif /* CONFIG_RPS */
681

T
Tom Herbert 已提交
682 683
/* This structure contains an instance of an RX queue. */
struct netdev_rx_queue {
684
#ifdef CONFIG_RPS
E
Eric Dumazet 已提交
685 686
	struct rps_map __rcu		*rps_map;
	struct rps_dev_flow_table __rcu	*rps_flow_table;
687
#endif
E
Eric Dumazet 已提交
688
	struct kobject			kobj;
T
Tom Herbert 已提交
689
	struct net_device		*dev;
690
	struct xdp_rxq_info		xdp_rxq;
T
Tom Herbert 已提交
691
} ____cacheline_aligned_in_smp;
692 693 694 695 696 697

/*
 * RX queue sysfs structures and functions.
 */
struct rx_queue_attribute {
	struct attribute attr;
698
	ssize_t (*show)(struct netdev_rx_queue *queue, char *buf);
699
	ssize_t (*store)(struct netdev_rx_queue *queue,
700
			 const char *buf, size_t len);
701
};
702

T
Tom Herbert 已提交
703 704 705 706 707 708 709 710 711 712 713
#ifdef CONFIG_XPS
/*
 * This structure holds an XPS map which can be of variable length.  The
 * map is an array of queues.
 */
struct xps_map {
	unsigned int len;
	unsigned int alloc_len;
	struct rcu_head rcu;
	u16 queues[0];
};
714
#define XPS_MAP_SIZE(_num) (sizeof(struct xps_map) + ((_num) * sizeof(u16)))
715 716
#define XPS_MIN_MAP_ALLOC ((L1_CACHE_ALIGN(offsetof(struct xps_map, queues[1])) \
       - sizeof(struct xps_map)) / sizeof(u16))
T
Tom Herbert 已提交
717 718 719 720 721 722

/*
 * This structure holds all XPS maps for device.  Maps are indexed by CPU.
 */
struct xps_dev_maps {
	struct rcu_head rcu;
E
Eric Dumazet 已提交
723
	struct xps_map __rcu *cpu_map[0];
T
Tom Herbert 已提交
724
};
725 726
#define XPS_DEV_MAPS_SIZE(_tcs) (sizeof(struct xps_dev_maps) +		\
	(nr_cpu_ids * (_tcs) * sizeof(struct xps_map *)))
T
Tom Herbert 已提交
727 728
#endif /* CONFIG_XPS */

729 730 731 732 733 734 735 736
#define TC_MAX_QUEUE	16
#define TC_BITMASK	15
/* HW offloaded queuing disciplines txq count and offset maps */
struct netdev_tc_txq {
	u16 count;
	u16 offset;
};

737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753
#if defined(CONFIG_FCOE) || defined(CONFIG_FCOE_MODULE)
/*
 * This structure is to hold information about the device
 * configured to run FCoE protocol stack.
 */
struct netdev_fcoe_hbainfo {
	char	manufacturer[64];
	char	serial_number[64];
	char	hardware_version[64];
	char	driver_version[64];
	char	optionrom_version[64];
	char	firmware_version[64];
	char	model[256];
	char	model_description[256];
};
#endif

754
#define MAX_PHYS_ITEM_ID_LEN 32
755

756 757
/* This structure holds a unique identifier to identify some
 * physical item (port for example) used by a netdevice.
758
 */
759 760
struct netdev_phys_item_id {
	unsigned char id[MAX_PHYS_ITEM_ID_LEN];
761 762 763
	unsigned char id_len;
};

764 765 766 767 768 769 770
static inline bool netdev_phys_item_id_same(struct netdev_phys_item_id *a,
					    struct netdev_phys_item_id *b)
{
	return a->id_len == b->id_len &&
	       memcmp(a->id, b->id, a->id_len) == 0;
}

771 772 773
typedef u16 (*select_queue_fallback_t)(struct net_device *dev,
				       struct sk_buff *skb);

774
enum tc_setup_type {
775
	TC_SETUP_QDISC_MQPRIO,
776
	TC_SETUP_CLSU32,
777
	TC_SETUP_CLSFLOWER,
778
	TC_SETUP_CLSMATCHALL,
779
	TC_SETUP_CLSBPF,
780
	TC_SETUP_BLOCK,
781
	TC_SETUP_QDISC_CBS,
782
	TC_SETUP_QDISC_RED,
783
	TC_SETUP_QDISC_PRIO,
784 785
};

786 787
/* These structures hold the attributes of bpf state that are being passed
 * to the netdevice through the bpf op.
788
 */
789
enum bpf_netdev_command {
790 791 792 793 794 795 796 797
	/* Set or clear a bpf program used in the earliest stages of packet
	 * rx. The prog will have been loaded as BPF_PROG_TYPE_XDP. The callee
	 * is responsible for calling bpf_prog_put on any old progs that are
	 * stored. In case of error, the callee need not release the new prog
	 * reference, but on success it takes ownership and must bpf_prog_put
	 * when it is no longer used.
	 */
	XDP_SETUP_PROG,
798
	XDP_SETUP_PROG_HW,
799
	/* Check if a bpf program is set on the device.  The callee should
800 801
	 * set @prog_attached to one of XDP_ATTACHED_* values, note that "true"
	 * is equivalent to XDP_ATTACHED_DRV.
802 803
	 */
	XDP_QUERY_PROG,
804 805 806 807
	/* BPF program for offload callbacks, invoked at program load time. */
	BPF_OFFLOAD_VERIFIER_PREP,
	BPF_OFFLOAD_TRANSLATE,
	BPF_OFFLOAD_DESTROY,
808 809
	BPF_OFFLOAD_MAP_ALLOC,
	BPF_OFFLOAD_MAP_FREE,
810 811
};

812
struct bpf_prog_offload_ops;
813 814
struct netlink_ext_ack;

815 816
struct netdev_bpf {
	enum bpf_netdev_command command;
817 818
	union {
		/* XDP_SETUP_PROG */
819
		struct {
820
			u32 flags;
821 822 823
			struct bpf_prog *prog;
			struct netlink_ext_ack *extack;
		};
824
		/* XDP_QUERY_PROG */
M
Martin KaFai Lau 已提交
825
		struct {
826
			u8 prog_attached;
M
Martin KaFai Lau 已提交
827
			u32 prog_id;
828 829
			/* flags with which program was installed */
			u32 prog_flags;
M
Martin KaFai Lau 已提交
830
		};
831 832 833
		/* BPF_OFFLOAD_VERIFIER_PREP */
		struct {
			struct bpf_prog *prog;
834
			const struct bpf_prog_offload_ops *ops; /* callee set */
835 836 837 838 839
		} verifier;
		/* BPF_OFFLOAD_TRANSLATE, BPF_OFFLOAD_DESTROY */
		struct {
			struct bpf_prog *prog;
		} offload;
840 841 842 843
		/* BPF_OFFLOAD_MAP_ALLOC, BPF_OFFLOAD_MAP_FREE */
		struct {
			struct bpf_offloaded_map *offmap;
		};
844 845
	};
};
846

847 848 849 850 851 852 853
#ifdef CONFIG_XFRM_OFFLOAD
struct xfrmdev_ops {
	int	(*xdo_dev_state_add) (struct xfrm_state *x);
	void	(*xdo_dev_state_delete) (struct xfrm_state *x);
	void	(*xdo_dev_state_free) (struct xfrm_state *x);
	bool	(*xdo_dev_offload_ok) (struct sk_buff *skb,
				       struct xfrm_state *x);
854
	void	(*xdo_dev_state_advance_esn) (struct xfrm_state *x);
855 856 857
};
#endif

858 859 860 861 862
struct dev_ifalias {
	struct rcu_head rcuhead;
	char ifalias[];
};

863 864
/*
 * This structure defines the management hooks for network devices.
865 866
 * The following hooks can be defined; unless noted otherwise, they are
 * optional and can be filled with a null pointer.
867 868
 *
 * int (*ndo_init)(struct net_device *dev);
B
Bjorn Helgaas 已提交
869 870 871 872
 *     This function is called once when a network device is registered.
 *     The network device can use this for any late stage initialization
 *     or semantic validation. It can fail with an error code which will
 *     be propagated back to register_netdev.
873 874 875 876 877 878
 *
 * void (*ndo_uninit)(struct net_device *dev);
 *     This function is called when device is unregistered or when registration
 *     fails. It is not called if init fails.
 *
 * int (*ndo_open)(struct net_device *dev);
B
Bjorn Helgaas 已提交
879
 *     This function is called when a network device transitions to the up
880 881 882
 *     state.
 *
 * int (*ndo_stop)(struct net_device *dev);
B
Bjorn Helgaas 已提交
883
 *     This function is called when a network device transitions to the down
884 885
 *     state.
 *
886 887
 * netdev_tx_t (*ndo_start_xmit)(struct sk_buff *skb,
 *                               struct net_device *dev);
888
 *	Called when a packet needs to be transmitted.
889 890 891 892
 *	Returns NETDEV_TX_OK.  Can return NETDEV_TX_BUSY, but you should stop
 *	the queue before that can happen; it's for obsolete devices and weird
 *	corner cases, but the stack really does a non-trivial amount
 *	of useless work if you return NETDEV_TX_BUSY.
B
Bjorn Helgaas 已提交
893
 *	Required; cannot be NULL.
894
 *
895 896 897 898 899 900 901 902 903
 * netdev_features_t (*ndo_features_check)(struct sk_buff *skb,
 *					   struct net_device *dev
 *					   netdev_features_t features);
 *	Called by core transmit path to determine if device is capable of
 *	performing offload operations on a given packet. This is to give
 *	the device an opportunity to implement any restrictions that cannot
 *	be otherwise expressed by feature flags. The check is called with
 *	the set of features that the stack has calculated and it returns
 *	those the driver believes to be appropriate.
904
 *
905
 * u16 (*ndo_select_queue)(struct net_device *dev, struct sk_buff *skb,
906
 *                         void *accel_priv, select_queue_fallback_t fallback);
B
Bjorn Helgaas 已提交
907
 *	Called to decide which queue to use when device supports multiple
908 909
 *	transmit queues.
 *
910 911
 * void (*ndo_change_rx_flags)(struct net_device *dev, int flags);
 *	This function is called to allow device receiver to make
B
Bjorn Helgaas 已提交
912
 *	changes to configuration when multicast or promiscuous is enabled.
913 914 915
 *
 * void (*ndo_set_rx_mode)(struct net_device *dev);
 *	This function is called device changes address list filtering.
916
 *	If driver handles unicast address filtering, it should set
B
Bjorn Helgaas 已提交
917
 *	IFF_UNICAST_FLT in its priv_flags.
918 919 920
 *
 * int (*ndo_set_mac_address)(struct net_device *dev, void *addr);
 *	This function  is called when the Media Access Control address
921
 *	needs to be changed. If this interface is not defined, the
B
Bjorn Helgaas 已提交
922
 *	MAC address can not be changed.
923 924 925 926 927
 *
 * int (*ndo_validate_addr)(struct net_device *dev);
 *	Test if Media Access Control address is valid for the device.
 *
 * int (*ndo_do_ioctl)(struct net_device *dev, struct ifreq *ifr, int cmd);
B
Bjorn Helgaas 已提交
928 929
 *	Called when a user requests an ioctl which can't be handled by
 *	the generic interface code. If not defined ioctls return
930 931 932 933
 *	not supported error code.
 *
 * int (*ndo_set_config)(struct net_device *dev, struct ifmap *map);
 *	Used to set network devices bus interface parameters. This interface
B
Bjorn Helgaas 已提交
934
 *	is retained for legacy reasons; new devices should use the bus
935 936 937 938
 *	interface (PCI) for low level management.
 *
 * int (*ndo_change_mtu)(struct net_device *dev, int new_mtu);
 *	Called when a user wants to change the Maximum Transfer Unit
939
 *	of a device.
940
 *
941
 * void (*ndo_tx_timeout)(struct net_device *dev);
B
Bjorn Helgaas 已提交
942
 *	Callback used when the transmitter has not made any progress
943 944
 *	for dev->watchdog ticks.
 *
945 946
 * void (*ndo_get_stats64)(struct net_device *dev,
 *                         struct rtnl_link_stats64 *storage);
947
 * struct net_device_stats* (*ndo_get_stats)(struct net_device *dev);
948
 *	Called when a user wants to get the network device usage
949
 *	statistics. Drivers must do one of the following:
950 951
 *	1. Define @ndo_get_stats64 to fill in a zero-initialised
 *	   rtnl_link_stats64 structure passed by the caller.
952
 *	2. Define @ndo_get_stats to update a net_device_stats structure
953 954 955 956 957
 *	   (which should normally be dev->stats) and return a pointer to
 *	   it. The structure may be changed asynchronously only if each
 *	   field is written atomically.
 *	3. Update dev->stats asynchronously and atomically, and define
 *	   neither operation.
958
 *
959
 * bool (*ndo_has_offload_stats)(const struct net_device *dev, int attr_id)
960 961 962 963 964 965 966
 *	Return true if this device supports offload stats of this attr_id.
 *
 * int (*ndo_get_offload_stats)(int attr_id, const struct net_device *dev,
 *	void *attr_data)
 *	Get statistics for offload operations by attr_id. Write it into the
 *	attr_data pointer.
 *
967
 * int (*ndo_vlan_rx_add_vid)(struct net_device *dev, __be16 proto, u16 vid);
B
Bjorn Helgaas 已提交
968
 *	If device supports VLAN filtering this function is called when a
969
 *	VLAN id is registered.
970
 *
971
 * int (*ndo_vlan_rx_kill_vid)(struct net_device *dev, __be16 proto, u16 vid);
B
Bjorn Helgaas 已提交
972
 *	If device supports VLAN filtering this function is called when a
973
 *	VLAN id is unregistered.
974 975
 *
 * void (*ndo_poll_controller)(struct net_device *dev);
976 977 978
 *
 *	SR-IOV management functions.
 * int (*ndo_set_vf_mac)(struct net_device *dev, int vf, u8* mac);
979 980
 * int (*ndo_set_vf_vlan)(struct net_device *dev, int vf, u16 vlan,
 *			  u8 qos, __be16 proto);
981 982
 * int (*ndo_set_vf_rate)(struct net_device *dev, int vf, int min_tx_rate,
 *			  int max_tx_rate);
983
 * int (*ndo_set_vf_spoofchk)(struct net_device *dev, int vf, bool setting);
H
Hiroshi Shimamoto 已提交
984
 * int (*ndo_set_vf_trust)(struct net_device *dev, int vf, bool setting);
985 986
 * int (*ndo_get_vf_config)(struct net_device *dev,
 *			    int vf, struct ifla_vf_info *ivf);
987
 * int (*ndo_set_vf_link_state)(struct net_device *dev, int vf, int link_state);
988 989
 * int (*ndo_set_vf_port)(struct net_device *dev, int vf,
 *			  struct nlattr *port[]);
990 991 992
 *
 *      Enable or disable the VF ability to query its RSS Redirection Table and
 *      Hash Key. This is needed since on some devices VF share this information
B
Bjorn Helgaas 已提交
993
 *      with PF and querying it may introduce a theoretical security risk.
994
 * int (*ndo_set_vf_rss_query_en)(struct net_device *dev, int vf, bool setting);
995
 * int (*ndo_get_vf_port)(struct net_device *dev, int vf, struct sk_buff *skb);
996
 * int (*ndo_setup_tc)(struct net_device *dev, enum tc_setup_type type,
997
 *		       void *type_data);
F
Florian Fainelli 已提交
998 999 1000 1001
 *	Called to setup any 'tc' scheduler, classifier or action on @dev.
 *	This is always called from the stack with the rtnl lock held and netif
 *	tx queues stopped. This allows the netdevice to perform queue
 *	management safely.
1002
 *
1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032
 *	Fiber Channel over Ethernet (FCoE) offload functions.
 * int (*ndo_fcoe_enable)(struct net_device *dev);
 *	Called when the FCoE protocol stack wants to start using LLD for FCoE
 *	so the underlying device can perform whatever needed configuration or
 *	initialization to support acceleration of FCoE traffic.
 *
 * int (*ndo_fcoe_disable)(struct net_device *dev);
 *	Called when the FCoE protocol stack wants to stop using LLD for FCoE
 *	so the underlying device can perform whatever needed clean-ups to
 *	stop supporting acceleration of FCoE traffic.
 *
 * int (*ndo_fcoe_ddp_setup)(struct net_device *dev, u16 xid,
 *			     struct scatterlist *sgl, unsigned int sgc);
 *	Called when the FCoE Initiator wants to initialize an I/O that
 *	is a possible candidate for Direct Data Placement (DDP). The LLD can
 *	perform necessary setup and returns 1 to indicate the device is set up
 *	successfully to perform DDP on this I/O, otherwise this returns 0.
 *
 * int (*ndo_fcoe_ddp_done)(struct net_device *dev,  u16 xid);
 *	Called when the FCoE Initiator/Target is done with the DDPed I/O as
 *	indicated by the FC exchange id 'xid', so the underlying device can
 *	clean up and reuse resources for later DDP requests.
 *
 * int (*ndo_fcoe_ddp_target)(struct net_device *dev, u16 xid,
 *			      struct scatterlist *sgl, unsigned int sgc);
 *	Called when the FCoE Target wants to initialize an I/O that
 *	is a possible candidate for Direct Data Placement (DDP). The LLD can
 *	perform necessary setup and returns 1 to indicate the device is set up
 *	successfully to perform DDP on this I/O, otherwise this returns 0.
 *
1033 1034 1035 1036 1037 1038 1039
 * int (*ndo_fcoe_get_hbainfo)(struct net_device *dev,
 *			       struct netdev_fcoe_hbainfo *hbainfo);
 *	Called when the FCoE Protocol stack wants information on the underlying
 *	device. This information is utilized by the FCoE protocol stack to
 *	register attributes with Fiber Channel management service as per the
 *	FC-GS Fabric Device Management Information(FDMI) specification.
 *
1040 1041 1042 1043 1044 1045
 * int (*ndo_fcoe_get_wwn)(struct net_device *dev, u64 *wwn, int type);
 *	Called when the underlying device wants to override default World Wide
 *	Name (WWN) generation mechanism in FCoE protocol stack to pass its own
 *	World Wide Port Name (WWPN) or World Wide Node Name (WWNN) to the FCoE
 *	protocol stack to use.
 *
1046 1047 1048 1049 1050 1051
 *	RFS acceleration.
 * int (*ndo_rx_flow_steer)(struct net_device *dev, const struct sk_buff *skb,
 *			    u16 rxq_index, u32 flow_id);
 *	Set hardware filter for RFS.  rxq_index is the target queue index;
 *	flow_id is a flow ID to be passed to rps_may_expire_flow() later.
 *	Return the filter ID on success, or a negative error code.
1052
 *
1053
 *	Slave management functions (for bridge, bonding, etc).
1054 1055 1056 1057 1058
 * int (*ndo_add_slave)(struct net_device *dev, struct net_device *slave_dev);
 *	Called to make another netdev an underling.
 *
 * int (*ndo_del_slave)(struct net_device *dev, struct net_device *slave_dev);
 *	Called to release previously enslaved netdev.
1059 1060
 *
 *      Feature/offload setting functions.
1061 1062 1063 1064 1065 1066
 * netdev_features_t (*ndo_fix_features)(struct net_device *dev,
 *		netdev_features_t features);
 *	Adjusts the requested feature flags according to device-specific
 *	constraints, and returns the resulting flags. Must not modify
 *	the device state.
 *
1067
 * int (*ndo_set_features)(struct net_device *dev, netdev_features_t features);
1068 1069 1070 1071
 *	Called to update device configuration to new features. Passed
 *	feature set might be less than what was returned by ndo_fix_features()).
 *	Must return >0 or -errno if it changed dev->features itself.
 *
1072 1073
 * int (*ndo_fdb_add)(struct ndmsg *ndm, struct nlattr *tb[],
 *		      struct net_device *dev,
1074
 *		      const unsigned char *addr, u16 vid, u16 flags)
1075
 *	Adds an FDB entry to dev for addr.
1076 1077
 * int (*ndo_fdb_del)(struct ndmsg *ndm, struct nlattr *tb[],
 *		      struct net_device *dev,
1078
 *		      const unsigned char *addr, u16 vid)
1079 1080
 *	Deletes the FDB entry from dev coresponding to addr.
 * int (*ndo_fdb_dump)(struct sk_buff *skb, struct netlink_callback *cb,
1081
 *		       struct net_device *dev, struct net_device *filter_dev,
1082
 *		       int *idx)
1083 1084
 *	Used to add FDB entries to dump requests. Implementers should add
 *	entries to skb and update idx with the number of entries.
J
John Fastabend 已提交
1085
 *
1086 1087
 * int (*ndo_bridge_setlink)(struct net_device *dev, struct nlmsghdr *nlh,
 *			     u16 flags)
J
John Fastabend 已提交
1088
 * int (*ndo_bridge_getlink)(struct sk_buff *skb, u32 pid, u32 seq,
1089 1090
 *			     struct net_device *dev, u32 filter_mask,
 *			     int nlflags)
1091 1092
 * int (*ndo_bridge_dellink)(struct net_device *dev, struct nlmsghdr *nlh,
 *			     u16 flags);
J
Jiri Pirko 已提交
1093 1094 1095 1096 1097 1098 1099 1100
 *
 * int (*ndo_change_carrier)(struct net_device *dev, bool new_carrier);
 *	Called to change device carrier. Soft-devices (like dummy, team, etc)
 *	which do not represent real hardware may define this to allow their
 *	userspace components to manage their virtual carrier state. Devices
 *	that determine carrier state from physical hardware properties (eg
 *	network cables) or protocol-dependent mechanisms (eg
 *	USB_CDC_NOTIFY_NETWORK_CONNECTION) should NOT implement this function.
1101 1102
 *
 * int (*ndo_get_phys_port_id)(struct net_device *dev,
1103
 *			       struct netdev_phys_item_id *ppid);
1104 1105 1106
 *	Called to get ID of physical port of this device. If driver does
 *	not implement this, it is assumed that the hw is not able to have
 *	multiple net devices on single physical port.
1107
 *
1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120
 * void (*ndo_udp_tunnel_add)(struct net_device *dev,
 *			      struct udp_tunnel_info *ti);
 *	Called by UDP tunnel to notify a driver about the UDP port and socket
 *	address family that a UDP tunnel is listnening to. It is called only
 *	when a new port starts listening. The operation is protected by the
 *	RTNL.
 *
 * void (*ndo_udp_tunnel_del)(struct net_device *dev,
 *			      struct udp_tunnel_info *ti);
 *	Called by UDP tunnel to notify the driver about a UDP port and socket
 *	address family that the UDP tunnel is not listening to anymore. The
 *	operation is protected by the RTNL.
 *
1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132
 * void* (*ndo_dfwd_add_station)(struct net_device *pdev,
 *				 struct net_device *dev)
 *	Called by upper layer devices to accelerate switching or other
 *	station functionality into hardware. 'pdev is the lowerdev
 *	to use for the offload and 'dev' is the net device that will
 *	back the offload. Returns a pointer to the private structure
 *	the upper layer will maintain.
 * void (*ndo_dfwd_del_station)(struct net_device *pdev, void *priv)
 *	Called by upper layer device to delete the station created
 *	by 'ndo_dfwd_add_station'. 'pdev' is the net device backing
 *	the station and priv is the structure returned by the add
 *	operation.
1133 1134 1135 1136
 * int (*ndo_set_tx_maxrate)(struct net_device *dev,
 *			     int queue_index, u32 maxrate);
 *	Called when a user wants to set a max-rate limitation of specific
 *	TX queue.
1137 1138
 * int (*ndo_get_iflink)(const struct net_device *dev);
 *	Called to get the iflink value of this device.
1139
 * void (*ndo_change_proto_down)(struct net_device *dev,
B
Bjorn Helgaas 已提交
1140
 *				 bool proto_down);
1141 1142 1143
 *	This function is used to pass protocol port error state information
 *	to the switch driver. The switch driver can react to the proto_down
 *      by doing a phys down on the associated switch port.
1144 1145 1146 1147
 * int (*ndo_fill_metadata_dst)(struct net_device *dev, struct sk_buff *skb);
 *	This function is used to get egress tunnel information for given skb.
 *	This is useful for retrieving outer tunnel header parameters while
 *	sampling packet.
1148 1149 1150 1151
 * void (*ndo_set_rx_headroom)(struct net_device *dev, int needed_headroom);
 *	This function is used to specify the headroom that the skb must
 *	consider when allocation skb during packet reception. Setting
 *	appropriate rx headroom value allows avoiding skb head copy on
B
Bjorn Helgaas 已提交
1152
 *	forward. Setting a negative value resets the rx headroom to the
1153
 *	default value.
1154
 * int (*ndo_bpf)(struct net_device *dev, struct netdev_bpf *bpf);
1155
 *	This function is used to set or query state related to XDP on the
1156 1157
 *	netdevice and manage BPF offload. See definition of
 *	enum bpf_netdev_command for details.
1158 1159 1160
 * int (*ndo_xdp_xmit)(struct net_device *dev, struct xdp_buff *xdp);
 *	This function is used to submit a XDP packet for transmit on a
 *	netdevice.
1161
 * void (*ndo_xdp_flush)(struct net_device *dev);
1162 1163
 *	This function is used to inform the driver to flush a particular
 *	xdp tx queue. Must be called on same CPU as xdp_xmit.
1164 1165 1166 1167 1168 1169
 */
struct net_device_ops {
	int			(*ndo_init)(struct net_device *dev);
	void			(*ndo_uninit)(struct net_device *dev);
	int			(*ndo_open)(struct net_device *dev);
	int			(*ndo_stop)(struct net_device *dev);
1170 1171 1172 1173 1174
	netdev_tx_t		(*ndo_start_xmit)(struct sk_buff *skb,
						  struct net_device *dev);
	netdev_features_t	(*ndo_features_check)(struct sk_buff *skb,
						      struct net_device *dev,
						      netdev_features_t features);
1175
	u16			(*ndo_select_queue)(struct net_device *dev,
1176
						    struct sk_buff *skb,
1177 1178
						    void *accel_priv,
						    select_queue_fallback_t fallback);
1179 1180 1181 1182 1183 1184 1185 1186 1187 1188
	void			(*ndo_change_rx_flags)(struct net_device *dev,
						       int flags);
	void			(*ndo_set_rx_mode)(struct net_device *dev);
	int			(*ndo_set_mac_address)(struct net_device *dev,
						       void *addr);
	int			(*ndo_validate_addr)(struct net_device *dev);
	int			(*ndo_do_ioctl)(struct net_device *dev,
					        struct ifreq *ifr, int cmd);
	int			(*ndo_set_config)(struct net_device *dev,
					          struct ifmap *map);
1189 1190 1191 1192
	int			(*ndo_change_mtu)(struct net_device *dev,
						  int new_mtu);
	int			(*ndo_neigh_setup)(struct net_device *dev,
						   struct neigh_parms *);
1193 1194
	void			(*ndo_tx_timeout) (struct net_device *dev);

1195 1196
	void			(*ndo_get_stats64)(struct net_device *dev,
						   struct rtnl_link_stats64 *storage);
1197
	bool			(*ndo_has_offload_stats)(const struct net_device *dev, int attr_id);
1198 1199 1200
	int			(*ndo_get_offload_stats)(int attr_id,
							 const struct net_device *dev,
							 void *attr_data);
1201 1202
	struct net_device_stats* (*ndo_get_stats)(struct net_device *dev);

1203
	int			(*ndo_vlan_rx_add_vid)(struct net_device *dev,
1204
						       __be16 proto, u16 vid);
1205
	int			(*ndo_vlan_rx_kill_vid)(struct net_device *dev,
1206
						        __be16 proto, u16 vid);
1207 1208
#ifdef CONFIG_NET_POLL_CONTROLLER
	void                    (*ndo_poll_controller)(struct net_device *dev);
H
Herbert Xu 已提交
1209
	int			(*ndo_netpoll_setup)(struct net_device *dev,
1210
						     struct netpoll_info *info);
1211
	void			(*ndo_netpoll_cleanup)(struct net_device *dev);
1212
#endif
1213 1214 1215
	int			(*ndo_set_vf_mac)(struct net_device *dev,
						  int queue, u8 *mac);
	int			(*ndo_set_vf_vlan)(struct net_device *dev,
1216 1217
						   int queue, u16 vlan,
						   u8 qos, __be16 proto);
1218 1219 1220
	int			(*ndo_set_vf_rate)(struct net_device *dev,
						   int vf, int min_tx_rate,
						   int max_tx_rate);
1221 1222
	int			(*ndo_set_vf_spoofchk)(struct net_device *dev,
						       int vf, bool setting);
H
Hiroshi Shimamoto 已提交
1223 1224
	int			(*ndo_set_vf_trust)(struct net_device *dev,
						    int vf, bool setting);
1225 1226 1227
	int			(*ndo_get_vf_config)(struct net_device *dev,
						     int vf,
						     struct ifla_vf_info *ivf);
1228 1229
	int			(*ndo_set_vf_link_state)(struct net_device *dev,
							 int vf, int link_state);
1230 1231 1232 1233
	int			(*ndo_get_vf_stats)(struct net_device *dev,
						    int vf,
						    struct ifla_vf_stats
						    *vf_stats);
1234 1235 1236 1237 1238
	int			(*ndo_set_vf_port)(struct net_device *dev,
						   int vf,
						   struct nlattr *port[]);
	int			(*ndo_get_vf_port)(struct net_device *dev,
						   int vf, struct sk_buff *skb);
1239 1240 1241
	int			(*ndo_set_vf_guid)(struct net_device *dev,
						   int vf, u64 guid,
						   int guid_type);
1242 1243 1244
	int			(*ndo_set_vf_rss_query_en)(
						   struct net_device *dev,
						   int vf, bool setting);
1245
	int			(*ndo_setup_tc)(struct net_device *dev,
1246
						enum tc_setup_type type,
1247
						void *type_data);
1248
#if IS_ENABLED(CONFIG_FCOE)
1249 1250
	int			(*ndo_fcoe_enable)(struct net_device *dev);
	int			(*ndo_fcoe_disable)(struct net_device *dev);
1251 1252 1253 1254 1255 1256
	int			(*ndo_fcoe_ddp_setup)(struct net_device *dev,
						      u16 xid,
						      struct scatterlist *sgl,
						      unsigned int sgc);
	int			(*ndo_fcoe_ddp_done)(struct net_device *dev,
						     u16 xid);
1257 1258 1259 1260
	int			(*ndo_fcoe_ddp_target)(struct net_device *dev,
						       u16 xid,
						       struct scatterlist *sgl,
						       unsigned int sgc);
1261 1262
	int			(*ndo_fcoe_get_hbainfo)(struct net_device *dev,
							struct netdev_fcoe_hbainfo *hbainfo);
1263 1264
#endif

1265
#if IS_ENABLED(CONFIG_LIBFCOE)
1266 1267 1268 1269
#define NETDEV_FCOE_WWNN 0
#define NETDEV_FCOE_WWPN 1
	int			(*ndo_fcoe_get_wwn)(struct net_device *dev,
						    u64 *wwn, int type);
1270
#endif
1271

1272 1273 1274 1275 1276 1277
#ifdef CONFIG_RFS_ACCEL
	int			(*ndo_rx_flow_steer)(struct net_device *dev,
						     const struct sk_buff *skb,
						     u16 rxq_index,
						     u32 flow_id);
#endif
1278
	int			(*ndo_add_slave)(struct net_device *dev,
D
David Ahern 已提交
1279 1280
						 struct net_device *slave_dev,
						 struct netlink_ext_ack *extack);
1281 1282
	int			(*ndo_del_slave)(struct net_device *dev,
						 struct net_device *slave_dev);
1283 1284
	netdev_features_t	(*ndo_fix_features)(struct net_device *dev,
						    netdev_features_t features);
1285
	int			(*ndo_set_features)(struct net_device *dev,
1286
						    netdev_features_t features);
1287 1288 1289 1290
	int			(*ndo_neigh_construct)(struct net_device *dev,
						       struct neighbour *n);
	void			(*ndo_neigh_destroy)(struct net_device *dev,
						     struct neighbour *n);
1291 1292

	int			(*ndo_fdb_add)(struct ndmsg *ndm,
1293
					       struct nlattr *tb[],
1294
					       struct net_device *dev,
1295
					       const unsigned char *addr,
1296
					       u16 vid,
1297 1298
					       u16 flags);
	int			(*ndo_fdb_del)(struct ndmsg *ndm,
1299
					       struct nlattr *tb[],
1300
					       struct net_device *dev,
1301 1302
					       const unsigned char *addr,
					       u16 vid);
1303 1304 1305
	int			(*ndo_fdb_dump)(struct sk_buff *skb,
						struct netlink_callback *cb,
						struct net_device *dev,
1306
						struct net_device *filter_dev,
1307
						int *idx);
J
John Fastabend 已提交
1308 1309

	int			(*ndo_bridge_setlink)(struct net_device *dev,
1310 1311
						      struct nlmsghdr *nlh,
						      u16 flags);
J
John Fastabend 已提交
1312 1313
	int			(*ndo_bridge_getlink)(struct sk_buff *skb,
						      u32 pid, u32 seq,
1314
						      struct net_device *dev,
1315 1316
						      u32 filter_mask,
						      int nlflags);
1317
	int			(*ndo_bridge_dellink)(struct net_device *dev,
1318 1319
						      struct nlmsghdr *nlh,
						      u16 flags);
J
Jiri Pirko 已提交
1320 1321
	int			(*ndo_change_carrier)(struct net_device *dev,
						      bool new_carrier);
1322
	int			(*ndo_get_phys_port_id)(struct net_device *dev,
1323
							struct netdev_phys_item_id *ppid);
1324 1325
	int			(*ndo_get_phys_port_name)(struct net_device *dev,
							  char *name, size_t len);
1326 1327 1328 1329
	void			(*ndo_udp_tunnel_add)(struct net_device *dev,
						      struct udp_tunnel_info *ti);
	void			(*ndo_udp_tunnel_del)(struct net_device *dev,
						      struct udp_tunnel_info *ti);
1330 1331 1332 1333 1334
	void*			(*ndo_dfwd_add_station)(struct net_device *pdev,
							struct net_device *dev);
	void			(*ndo_dfwd_del_station)(struct net_device *pdev,
							void *priv);

1335
	int			(*ndo_get_lock_subclass)(struct net_device *dev);
1336 1337 1338
	int			(*ndo_set_tx_maxrate)(struct net_device *dev,
						      int queue_index,
						      u32 maxrate);
1339
	int			(*ndo_get_iflink)(const struct net_device *dev);
1340 1341
	int			(*ndo_change_proto_down)(struct net_device *dev,
							 bool proto_down);
1342 1343
	int			(*ndo_fill_metadata_dst)(struct net_device *dev,
						       struct sk_buff *skb);
1344 1345
	void			(*ndo_set_rx_headroom)(struct net_device *dev,
						       int needed_headroom);
1346 1347
	int			(*ndo_bpf)(struct net_device *dev,
					   struct netdev_bpf *bpf);
1348 1349
	int			(*ndo_xdp_xmit)(struct net_device *dev,
						struct xdp_buff *xdp);
1350
	void			(*ndo_xdp_flush)(struct net_device *dev);
1351 1352
};

1353 1354 1355 1356 1357
/**
 * enum net_device_priv_flags - &struct net_device priv_flags
 *
 * These are the &struct net_device, they are only set internally
 * by drivers and used in the kernel. These flags are invisible to
B
Bjorn Helgaas 已提交
1358
 * userspace; this means that the order of these flags can change
1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381
 * during any kernel release.
 *
 * You should have a pretty good reason to be extending these flags.
 *
 * @IFF_802_1Q_VLAN: 802.1Q VLAN device
 * @IFF_EBRIDGE: Ethernet bridging device
 * @IFF_BONDING: bonding master or slave
 * @IFF_ISATAP: ISATAP interface (RFC4214)
 * @IFF_WAN_HDLC: WAN HDLC device
 * @IFF_XMIT_DST_RELEASE: dev_hard_start_xmit() is allowed to
 *	release skb->dst
 * @IFF_DONT_BRIDGE: disallow bridging this ether dev
 * @IFF_DISABLE_NETPOLL: disable netpoll at run-time
 * @IFF_MACVLAN_PORT: device used as macvlan port
 * @IFF_BRIDGE_PORT: device used as bridge port
 * @IFF_OVS_DATAPATH: device used as Open vSwitch datapath port
 * @IFF_TX_SKB_SHARING: The interface supports sharing skbs on transmit
 * @IFF_UNICAST_FLT: Supports unicast filtering
 * @IFF_TEAM_PORT: device used as team port
 * @IFF_SUPP_NOFCS: device supports sending custom FCS
 * @IFF_LIVE_ADDR_CHANGE: device supports hardware address
 *	change when it's running
 * @IFF_MACVLAN: Macvlan device
1382 1383 1384 1385
 * @IFF_XMIT_DST_RELEASE_PERM: IFF_XMIT_DST_RELEASE not taking into account
 *	underlying stacked devices
 * @IFF_IPVLAN_MASTER: IPvlan master device
 * @IFF_IPVLAN_SLAVE: IPvlan slave device
1386
 * @IFF_L3MDEV_MASTER: device is an L3 master device
1387
 * @IFF_NO_QUEUE: device can run without qdisc attached
1388
 * @IFF_OPENVSWITCH: device is a Open vSwitch master
D
David Ahern 已提交
1389
 * @IFF_L3MDEV_SLAVE: device is enslaved to an L3 master device
J
Jiri Pirko 已提交
1390
 * @IFF_TEAM: device is a team device
1391
 * @IFF_RXFH_CONFIGURED: device has had Rx Flow indirection table configured
1392 1393
 * @IFF_PHONY_HEADROOM: the headroom value is controlled by an external
 *	entity (i.e. the master device for bridged veth)
1394
 * @IFF_MACSEC: device is a MACsec device
1395 1396 1397 1398
 */
enum netdev_priv_flags {
	IFF_802_1Q_VLAN			= 1<<0,
	IFF_EBRIDGE			= 1<<1,
1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416
	IFF_BONDING			= 1<<2,
	IFF_ISATAP			= 1<<3,
	IFF_WAN_HDLC			= 1<<4,
	IFF_XMIT_DST_RELEASE		= 1<<5,
	IFF_DONT_BRIDGE			= 1<<6,
	IFF_DISABLE_NETPOLL		= 1<<7,
	IFF_MACVLAN_PORT		= 1<<8,
	IFF_BRIDGE_PORT			= 1<<9,
	IFF_OVS_DATAPATH		= 1<<10,
	IFF_TX_SKB_SHARING		= 1<<11,
	IFF_UNICAST_FLT			= 1<<12,
	IFF_TEAM_PORT			= 1<<13,
	IFF_SUPP_NOFCS			= 1<<14,
	IFF_LIVE_ADDR_CHANGE		= 1<<15,
	IFF_MACVLAN			= 1<<16,
	IFF_XMIT_DST_RELEASE_PERM	= 1<<17,
	IFF_IPVLAN_MASTER		= 1<<18,
	IFF_IPVLAN_SLAVE		= 1<<19,
1417
	IFF_L3MDEV_MASTER		= 1<<20,
1418 1419
	IFF_NO_QUEUE			= 1<<21,
	IFF_OPENVSWITCH			= 1<<22,
D
David Ahern 已提交
1420
	IFF_L3MDEV_SLAVE		= 1<<23,
J
Jiri Pirko 已提交
1421
	IFF_TEAM			= 1<<24,
1422
	IFF_RXFH_CONFIGURED		= 1<<25,
1423
	IFF_PHONY_HEADROOM		= 1<<26,
1424
	IFF_MACSEC			= 1<<27,
1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443
};

#define IFF_802_1Q_VLAN			IFF_802_1Q_VLAN
#define IFF_EBRIDGE			IFF_EBRIDGE
#define IFF_BONDING			IFF_BONDING
#define IFF_ISATAP			IFF_ISATAP
#define IFF_WAN_HDLC			IFF_WAN_HDLC
#define IFF_XMIT_DST_RELEASE		IFF_XMIT_DST_RELEASE
#define IFF_DONT_BRIDGE			IFF_DONT_BRIDGE
#define IFF_DISABLE_NETPOLL		IFF_DISABLE_NETPOLL
#define IFF_MACVLAN_PORT		IFF_MACVLAN_PORT
#define IFF_BRIDGE_PORT			IFF_BRIDGE_PORT
#define IFF_OVS_DATAPATH		IFF_OVS_DATAPATH
#define IFF_TX_SKB_SHARING		IFF_TX_SKB_SHARING
#define IFF_UNICAST_FLT			IFF_UNICAST_FLT
#define IFF_TEAM_PORT			IFF_TEAM_PORT
#define IFF_SUPP_NOFCS			IFF_SUPP_NOFCS
#define IFF_LIVE_ADDR_CHANGE		IFF_LIVE_ADDR_CHANGE
#define IFF_MACVLAN			IFF_MACVLAN
1444
#define IFF_XMIT_DST_RELEASE_PERM	IFF_XMIT_DST_RELEASE_PERM
1445 1446
#define IFF_IPVLAN_MASTER		IFF_IPVLAN_MASTER
#define IFF_IPVLAN_SLAVE		IFF_IPVLAN_SLAVE
1447
#define IFF_L3MDEV_MASTER		IFF_L3MDEV_MASTER
1448
#define IFF_NO_QUEUE			IFF_NO_QUEUE
1449
#define IFF_OPENVSWITCH			IFF_OPENVSWITCH
1450
#define IFF_L3MDEV_SLAVE		IFF_L3MDEV_SLAVE
J
Jiri Pirko 已提交
1451
#define IFF_TEAM			IFF_TEAM
1452
#define IFF_RXFH_CONFIGURED		IFF_RXFH_CONFIGURED
1453
#define IFF_MACSEC			IFF_MACSEC
1454

1455 1456
/**
 *	struct net_device - The DEVICE structure.
1457 1458 1459 1460
 *
 *	Actually, this whole structure is a big mistake.  It mixes I/O
 *	data with strictly "high-level" data, and it has to know about
 *	almost every data structure used in the INET module.
1461 1462 1463
 *
 *	@name:	This is the first field of the "visible" part of this structure
 *		(i.e. as seen by users in the "Space.c" file).  It is the name
1464
 *		of the interface.
1465 1466 1467 1468 1469 1470 1471 1472 1473 1474
 *
 *	@name_hlist: 	Device name hash chain, please keep it close to name[]
 *	@ifalias:	SNMP alias
 *	@mem_end:	Shared memory end
 *	@mem_start:	Shared memory start
 *	@base_addr:	Device I/O address
 *	@irq:		Device IRQ number
 *
 *	@state:		Generic network queuing layer state, see netdev_state_t
 *	@dev_list:	The global list of network devices
B
Bjorn Helgaas 已提交
1475 1476 1477 1478
 *	@napi_list:	List entry used for polling NAPI devices
 *	@unreg_list:	List entry  when we are unregistering the
 *			device; see the function unregister_netdev
 *	@close_list:	List entry used when we are closing the device
1479 1480
 *	@ptype_all:     Device-specific packet handlers for all protocols
 *	@ptype_specific: Device-specific, protocol-specific packet handlers
1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496
 *
 *	@adj_list:	Directly linked devices, like slaves for bonding
 *	@features:	Currently active device features
 *	@hw_features:	User-changeable features
 *
 *	@wanted_features:	User-requested features
 *	@vlan_features:		Mask of features inheritable by VLAN devices
 *
 *	@hw_enc_features:	Mask of features inherited by encapsulating devices
 *				This field indicates what encapsulation
 *				offloads the hardware is capable of doing,
 *				and drivers will need to set them appropriately.
 *
 *	@mpls_features:	Mask of features inheritable by MPLS
 *
 *	@ifindex:	interface index
B
Bjorn Helgaas 已提交
1497
 *	@group:		The group the device belongs to
1498 1499 1500 1501 1502 1503 1504 1505
 *
 *	@stats:		Statistics struct, which was left as a legacy, use
 *			rtnl_link_stats64 instead
 *
 *	@rx_dropped:	Dropped packets by core network,
 *			do not use this in drivers
 *	@tx_dropped:	Dropped packets by core network,
 *			do not use this in drivers
1506 1507
 *	@rx_nohandler:	nohandler dropped packets by core network on
 *			inactive devices, do not use this in drivers
1508 1509
 *	@carrier_up_count:	Number of times the carrier has been up
 *	@carrier_down_count:	Number of times the carrier has been down
1510 1511 1512 1513 1514 1515 1516 1517 1518
 *
 *	@wireless_handlers:	List of functions to handle Wireless Extensions,
 *				instead of ioctl,
 *				see <net/iw_handler.h> for details.
 *	@wireless_data:	Instance data managed by the core of wireless extensions
 *
 *	@netdev_ops:	Includes several pointers to callbacks,
 *			if one wants to override the ndo_*() functions
 *	@ethtool_ops:	Management operations
1519 1520
 *	@ndisc_ops:	Includes callbacks for different IPv6 neighbour
 *			discovery handling. Necessary for e.g. 6LoWPAN.
E
Eric W. Biederman 已提交
1521
 *	@header_ops:	Includes callbacks for creating,parsing,caching,etc
1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533
 *			of Layer 2 headers.
 *
 *	@flags:		Interface flags (a la BSD)
 *	@priv_flags:	Like 'flags' but invisible to userspace,
 *			see if.h for the definitions
 *	@gflags:	Global flags ( kept as legacy )
 *	@padded:	How much padding added by alloc_netdev()
 *	@operstate:	RFC2863 operstate
 *	@link_mode:	Mapping policy to operstate
 *	@if_port:	Selectable AUI, TP, ...
 *	@dma:		DMA channel
 *	@mtu:		Interface MTU value
1534 1535
 *	@min_mtu:	Interface Minimum MTU value
 *	@max_mtu:	Interface Maximum MTU value
1536
 *	@type:		Interface hardware type
1537
 *	@hard_header_len: Maximum hardware header length.
1538
 *	@min_header_len:  Minimum hardware header length
1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550
 *
 *	@needed_headroom: Extra headroom the hardware may need, but not in all
 *			  cases can this be guaranteed
 *	@needed_tailroom: Extra tailroom the hardware may need, but not in all
 *			  cases can this be guaranteed. Some cases also use
 *			  LL_MAX_HEADER instead to allocate the skb
 *
 *	interface address info:
 *
 * 	@perm_addr:		Permanent hw address
 * 	@addr_assign_type:	Hw address assignment type
 * 	@addr_len:		Hardware address length
1551
 *	@neigh_priv_len:	Used in neigh_alloc()
1552 1553 1554 1555 1556
 * 	@dev_id:		Used to differentiate devices that share
 * 				the same link layer address
 * 	@dev_port:		Used to differentiate devices that share
 * 				the same function
 *	@addr_list_lock:	XXX: need comments on this one
B
Bjorn Helgaas 已提交
1557
 *	@uc_promisc:		Counter that indicates promiscuous mode
1558 1559 1560
 *				has been enabled due to the need to listen to
 *				additional unicast addresses in a device that
 *				does not implement ndo_set_rx_mode()
1561 1562 1563 1564
 *	@uc:			unicast mac addresses
 *	@mc:			multicast mac addresses
 *	@dev_addrs:		list of device hw addresses
 *	@queues_kset:		Group of all Kobjects in the Tx and RX queues
B
Bjorn Helgaas 已提交
1565 1566 1567
 *	@promiscuity:		Number of times the NIC is told to work in
 *				promiscuous mode; if it becomes 0 the NIC will
 *				exit promiscuous mode
1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589
 *	@allmulti:		Counter, enables or disables allmulticast mode
 *
 *	@vlan_info:	VLAN info
 *	@dsa_ptr:	dsa specific data
 *	@tipc_ptr:	TIPC specific data
 *	@atalk_ptr:	AppleTalk link
 *	@ip_ptr:	IPv4 specific data
 *	@dn_ptr:	DECnet specific data
 *	@ip6_ptr:	IPv6 specific data
 *	@ax25_ptr:	AX.25 specific data
 *	@ieee80211_ptr:	IEEE 802.11 specific data, assign before registering
 *
 *	@dev_addr:	Hw address (before bcast,
 *			because most packets are unicast)
 *
 *	@_rx:			Array of RX queues
 *	@num_rx_queues:		Number of RX queues
 *				allocated at register_netdev() time
 *	@real_num_rx_queues: 	Number of RX queues currently active in device
 *
 *	@rx_handler:		handler for received packets
 *	@rx_handler_data: 	XXX: need comments on this one
1590 1591
 *	@miniq_ingress:		ingress/clsact qdisc specific data for
 *				ingress processing
1592 1593 1594
 *	@ingress_queue:		XXX: need comments on this one
 *	@broadcast:		hw bcast address
 *
1595 1596 1597 1598 1599 1600
 *	@rx_cpu_rmap:	CPU reverse-mapping for RX completion interrupts,
 *			indexed by RX queue number. Assigned by driver.
 *			This must only be set if the ndo_rx_flow_steer
 *			operation is defined
 *	@index_hlist:		Device index hash chain
 *
1601 1602 1603 1604 1605 1606 1607 1608
 *	@_tx:			Array of TX queues
 *	@num_tx_queues:		Number of TX queues allocated at alloc_netdev_mq() time
 *	@real_num_tx_queues: 	Number of TX queues currently active in device
 *	@qdisc:			Root qdisc from userspace point of view
 *	@tx_queue_len:		Max frames per queue allowed
 *	@tx_global_lock: 	XXX: need comments on this one
 *
 *	@xps_maps:	XXX: need comments on this one
1609 1610
 *	@miniq_egress:		clsact qdisc specific data for
 *				egress processing
1611
 *	@watchdog_timeo:	Represents the timeout that is used by
B
Bjorn Helgaas 已提交
1612
 *				the watchdog (see dev_watchdog())
1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623
 *	@watchdog_timer:	List of timers
 *
 *	@pcpu_refcnt:		Number of references to this device
 *	@todo_list:		Delayed register/unregister
 *	@link_watch_list:	XXX: need comments on this one
 *
 *	@reg_state:		Register/unregister state machine
 *	@dismantle:		Device is going to be freed
 *	@rtnl_link_state:	This enum represents the phases of creating
 *				a new link
 *
1624 1625
 *	@needs_free_netdev:	Should unregister perform free_netdev?
 *	@priv_destructor:	Called from unregister
1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651
 *	@npinfo:		XXX: need comments on this one
 * 	@nd_net:		Network namespace this network device is inside
 *
 * 	@ml_priv:	Mid-layer private
 * 	@lstats:	Loopback statistics
 * 	@tstats:	Tunnel statistics
 * 	@dstats:	Dummy statistics
 * 	@vstats:	Virtual ethernet statistics
 *
 *	@garp_port:	GARP
 *	@mrp_port:	MRP
 *
 *	@dev:		Class/net/name entry
 *	@sysfs_groups:	Space for optional device, statistics and wireless
 *			sysfs groups
 *
 *	@sysfs_rx_queue_group:	Space for optional per-rx queue attributes
 *	@rtnl_link_ops:	Rtnl_link_ops
 *
 *	@gso_max_size:	Maximum size of generic segmentation offload
 *	@gso_max_segs:	Maximum number of segments that can be passed to the
 *			NIC for GSO
 *
 *	@dcbnl_ops:	Data Center Bridging netlink ops
 *	@num_tc:	Number of traffic classes in the net device
 *	@tc_to_txq:	XXX: need comments on this one
R
Randy Dunlap 已提交
1652
 *	@prio_tc_map:	XXX: need comments on this one
1653 1654 1655 1656 1657 1658 1659
 *
 *	@fcoe_ddp_xid:	Max exchange id for FCoE LRO by ddp
 *
 *	@priomap:	XXX: need comments on this one
 *	@phydev:	Physical device may attach itself
 *			for hardware timestamping
 *
1660 1661
 *	@qdisc_tx_busylock: lockdep class annotating Qdisc->busylock spinlock
 *	@qdisc_running_key: lockdep class annotating Qdisc->running seqcount
1662
 *
1663 1664 1665 1666
 *	@proto_down:	protocol port state information can be sent to the
 *			switch driver and used to set the phys state of the
 *			switch port.
 *
L
Linus Torvalds 已提交
1667 1668 1669 1670
 *	FIXME: cleanup struct net_device such that network protocol info
 *	moves out.
 */

E
Eric Dumazet 已提交
1671
struct net_device {
L
Linus Torvalds 已提交
1672
	char			name[IFNAMSIZ];
1673
	struct hlist_node	name_hlist;
1674
	struct dev_ifalias	__rcu *ifalias;
L
Linus Torvalds 已提交
1675 1676 1677 1678
	/*
	 *	I/O specific fields
	 *	FIXME: Merge these and struct ifmap into one
	 */
1679 1680 1681 1682
	unsigned long		mem_end;
	unsigned long		mem_start;
	unsigned long		base_addr;
	int			irq;
L
Linus Torvalds 已提交
1683 1684

	/*
1685 1686
	 *	Some hardware also needs these fields (state,dev_list,
	 *	napi_list,unreg_list,close_list) but they are not
L
Linus Torvalds 已提交
1687 1688 1689 1690 1691
	 *	part of the usual set specified in Space.c.
	 */

	unsigned long		state;

1692
	struct list_head	dev_list;
1693
	struct list_head	napi_list;
1694
	struct list_head	unreg_list;
1695
	struct list_head	close_list;
1696 1697
	struct list_head	ptype_all;
	struct list_head	ptype_specific;
1698 1699 1700 1701 1702 1703

	struct {
		struct list_head upper;
		struct list_head lower;
	} adj_list;

1704 1705 1706 1707
	netdev_features_t	features;
	netdev_features_t	hw_features;
	netdev_features_t	wanted_features;
	netdev_features_t	vlan_features;
1708
	netdev_features_t	hw_enc_features;
S
Simon Horman 已提交
1709
	netdev_features_t	mpls_features;
1710
	netdev_features_t	gso_partial_features;
1711

L
Linus Torvalds 已提交
1712
	int			ifindex;
1713
	int			group;
L
Linus Torvalds 已提交
1714

R
Rusty Russell 已提交
1715
	struct net_device_stats	stats;
1716 1717 1718

	atomic_long_t		rx_dropped;
	atomic_long_t		tx_dropped;
1719
	atomic_long_t		rx_nohandler;
L
Linus Torvalds 已提交
1720

1721 1722 1723 1724
	/* Stats to monitor link on/off, flapping */
	atomic_t		carrier_up_count;
	atomic_t		carrier_down_count;

1725
#ifdef CONFIG_WIRELESS_EXT
B
Bjorn Helgaas 已提交
1726 1727
	const struct iw_handler_def *wireless_handlers;
	struct iw_public_data	*wireless_data;
1728
#endif
1729
	const struct net_device_ops *netdev_ops;
1730
	const struct ethtool_ops *ethtool_ops;
S
Scott Feldman 已提交
1731
#ifdef CONFIG_NET_SWITCHDEV
J
Jiri Pirko 已提交
1732
	const struct switchdev_ops *switchdev_ops;
S
Scott Feldman 已提交
1733
#endif
1734 1735 1736
#ifdef CONFIG_NET_L3_MASTER_DEV
	const struct l3mdev_ops	*l3mdev_ops;
#endif
1737 1738 1739
#if IS_ENABLED(CONFIG_IPV6)
	const struct ndisc_ops *ndisc_ops;
#endif
L
Linus Torvalds 已提交
1740

1741
#ifdef CONFIG_XFRM_OFFLOAD
1742 1743 1744
	const struct xfrmdev_ops *xfrmdev_ops;
#endif

1745 1746
	const struct header_ops *header_ops;

1747 1748 1749
	unsigned int		flags;
	unsigned int		priv_flags;

L
Linus Torvalds 已提交
1750
	unsigned short		gflags;
1751
	unsigned short		padded;
L
Linus Torvalds 已提交
1752

1753 1754
	unsigned char		operstate;
	unsigned char		link_mode;
S
Stefan Rompf 已提交
1755

1756 1757
	unsigned char		if_port;
	unsigned char		dma;
1758

1759
	unsigned int		mtu;
1760 1761
	unsigned int		min_mtu;
	unsigned int		max_mtu;
1762 1763
	unsigned short		type;
	unsigned short		hard_header_len;
1764
	unsigned char		min_header_len;
L
Linus Torvalds 已提交
1765

1766 1767 1768
	unsigned short		needed_headroom;
	unsigned short		needed_tailroom;

L
Linus Torvalds 已提交
1769
	/* Interface address info. */
1770 1771 1772
	unsigned char		perm_addr[MAX_ADDR_LEN];
	unsigned char		addr_assign_type;
	unsigned char		addr_len;
1773
	unsigned short		neigh_priv_len;
1774 1775
	unsigned short          dev_id;
	unsigned short          dev_port;
J
Jiri Pirko 已提交
1776
	spinlock_t		addr_list_lock;
1777 1778
	unsigned char		name_assign_type;
	bool			uc_promisc;
1779 1780 1781 1782
	struct netdev_hw_addr_list	uc;
	struct netdev_hw_addr_list	mc;
	struct netdev_hw_addr_list	dev_addrs;

1783 1784 1785
#ifdef CONFIG_SYSFS
	struct kset		*queues_kset;
#endif
1786 1787
	unsigned int		promiscuity;
	unsigned int		allmulti;
L
Linus Torvalds 已提交
1788 1789


B
Bjorn Helgaas 已提交
1790
	/* Protocol-specific pointers */
1791

1792
#if IS_ENABLED(CONFIG_VLAN_8021Q)
1793
	struct vlan_info __rcu	*vlan_info;
1794
#endif
1795
#if IS_ENABLED(CONFIG_NET_DSA)
1796
	struct dsa_port		*dsa_ptr;
1797 1798
#endif
#if IS_ENABLED(CONFIG_TIPC)
1799
	struct tipc_bearer __rcu *tipc_ptr;
1800
#endif
1801 1802
	void 			*atalk_ptr;
	struct in_device __rcu	*ip_ptr;
1803
#if IS_ENABLED(CONFIG_DECNET)
1804
	struct dn_dev __rcu     *dn_ptr;
1805
#endif
1806
	struct inet6_dev __rcu	*ip6_ptr;
1807
#if IS_ENABLED(CONFIG_AX25)
1808
	void			*ax25_ptr;
1809
#endif
1810
	struct wireless_dev	*ieee80211_ptr;
1811
	struct wpan_dev		*ieee802154_ptr;
R
Robert Shearman 已提交
1812 1813 1814
#if IS_ENABLED(CONFIG_MPLS_ROUTING)
	struct mpls_dev __rcu	*mpls_ptr;
#endif
L
Linus Torvalds 已提交
1815

1816
/*
E
Eric Dumazet 已提交
1817
 * Cache lines mostly used on receive path (including eth_type_trans())
1818 1819
 */
	/* Interface address info used in eth_type_trans() */
1820
	unsigned char		*dev_addr;
1821

T
Tom Herbert 已提交
1822 1823
	struct netdev_rx_queue	*_rx;
	unsigned int		num_rx_queues;
1824
	unsigned int		real_num_rx_queues;
T
Tom Herbert 已提交
1825

1826
	struct bpf_prog __rcu	*xdp_prog;
1827
	unsigned long		gro_flush_timeout;
1828 1829
	rx_handler_func_t __rcu	*rx_handler;
	void __rcu		*rx_handler_data;
1830

1831
#ifdef CONFIG_NET_CLS_ACT
1832
	struct mini_Qdisc __rcu	*miniq_ingress;
1833
#endif
1834
	struct netdev_queue __rcu *ingress_queue;
1835
#ifdef CONFIG_NETFILTER_INGRESS
1836
	struct nf_hook_entries __rcu *nf_hooks_ingress;
1837
#endif
1838

1839
	unsigned char		broadcast[MAX_ADDR_LEN];
1840 1841 1842 1843
#ifdef CONFIG_RFS_ACCEL
	struct cpu_rmap		*rx_cpu_rmap;
#endif
	struct hlist_node	index_hlist;
E
Eric Dumazet 已提交
1844 1845 1846 1847

/*
 * Cache lines mostly used on transmit path
 */
1848 1849
	struct netdev_queue	*_tx ____cacheline_aligned_in_smp;
	unsigned int		num_tx_queues;
1850
	unsigned int		real_num_tx_queues;
1851
	struct Qdisc		*qdisc;
1852 1853 1854
#ifdef CONFIG_NET_SCHED
	DECLARE_HASHTABLE	(qdisc_hash, 4);
#endif
1855
	unsigned int		tx_queue_len;
1856
	spinlock_t		tx_global_lock;
1857
	int			watchdog_timeo;
E
Eric Dumazet 已提交
1858

T
Tom Herbert 已提交
1859
#ifdef CONFIG_XPS
E
Eric Dumazet 已提交
1860
	struct xps_dev_maps __rcu *xps_maps;
T
Tom Herbert 已提交
1861
#endif
D
Daniel Borkmann 已提交
1862
#ifdef CONFIG_NET_CLS_ACT
1863
	struct mini_Qdisc __rcu	*miniq_egress;
D
Daniel Borkmann 已提交
1864
#endif
1865

1866 1867 1868
	/* These may be needed for future network-power-down code. */
	struct timer_list	watchdog_timer;

E
Eric Dumazet 已提交
1869
	int __percpu		*pcpu_refcnt;
L
Linus Torvalds 已提交
1870 1871
	struct list_head	todo_list;

1872
	struct list_head	link_watch_list;
1873

L
Linus Torvalds 已提交
1874
	enum { NETREG_UNINITIALIZED=0,
1875
	       NETREG_REGISTERED,	/* completed register_netdevice */
L
Linus Torvalds 已提交
1876 1877 1878
	       NETREG_UNREGISTERING,	/* called unregister_netdevice */
	       NETREG_UNREGISTERED,	/* completed unregister todo */
	       NETREG_RELEASED,		/* called free_netdev */
1879
	       NETREG_DUMMY,		/* dummy device for NAPI poll */
1880 1881
	} reg_state:8;

1882
	bool dismantle;
1883 1884 1885 1886 1887

	enum {
		RTNL_LINK_INITIALIZED,
		RTNL_LINK_INITIALIZING,
	} rtnl_link_state:16;
L
Linus Torvalds 已提交
1888

1889 1890
	bool needs_free_netdev;
	void (*priv_destructor)(struct net_device *dev);
L
Linus Torvalds 已提交
1891 1892

#ifdef CONFIG_NETPOLL
1893
	struct netpoll_info __rcu	*npinfo;
L
Linus Torvalds 已提交
1894
#endif
1895

1896
	possible_net_t			nd_net;
1897

D
David S. Miller 已提交
1898
	/* mid-layer private */
E
Eric Dumazet 已提交
1899
	union {
1900 1901
		void					*ml_priv;
		struct pcpu_lstats __percpu		*lstats;
1902
		struct pcpu_sw_netstats __percpu	*tstats;
1903 1904
		struct pcpu_dstats __percpu		*dstats;
		struct pcpu_vstats __percpu		*vstats;
E
Eric Dumazet 已提交
1905
	};
1906

1907
#if IS_ENABLED(CONFIG_GARP)
E
Eric Dumazet 已提交
1908
	struct garp_port __rcu	*garp_port;
1909 1910
#endif
#if IS_ENABLED(CONFIG_MRP)
1911
	struct mrp_port __rcu	*mrp_port;
1912
#endif
L
Linus Torvalds 已提交
1913

B
Bjorn Helgaas 已提交
1914
	struct device		dev;
1915
	const struct attribute_group *sysfs_groups[4];
1916
	const struct attribute_group *sysfs_rx_queue_group;
P
Patrick McHardy 已提交
1917 1918

	const struct rtnl_link_ops *rtnl_link_ops;
1919

1920 1921 1922
	/* for setting kernel sock attribute on TCP connection setup */
#define GSO_MAX_SIZE		65536
	unsigned int		gso_max_size;
1923 1924
#define GSO_MAX_SEGS		65535
	u16			gso_max_segs;
1925

J
Jeff Kirsher 已提交
1926
#ifdef CONFIG_DCB
1927
	const struct dcbnl_rtnl_ops *dcbnl_ops;
1928
#endif
B
Bjorn Helgaas 已提交
1929 1930 1931
	u8			num_tc;
	struct netdev_tc_txq	tc_to_txq[TC_MAX_QUEUE];
	u8			prio_tc_map[TC_BITMASK + 1];
1932

1933
#if IS_ENABLED(CONFIG_FCOE)
1934
	unsigned int		fcoe_ddp_xid;
1935
#endif
1936
#if IS_ENABLED(CONFIG_CGROUP_NET_PRIO)
1937
	struct netprio_map __rcu *priomap;
1938
#endif
B
Bjorn Helgaas 已提交
1939 1940
	struct phy_device	*phydev;
	struct lock_class_key	*qdisc_tx_busylock;
1941
	struct lock_class_key	*qdisc_running_key;
B
Bjorn Helgaas 已提交
1942
	bool			proto_down;
L
Linus Torvalds 已提交
1943
};
1944
#define to_net_dev(d) container_of(d, struct net_device, dev)
L
Linus Torvalds 已提交
1945

D
David S. Miller 已提交
1946 1947 1948 1949 1950 1951 1952
static inline bool netif_elide_gro(const struct net_device *dev)
{
	if (!(dev->features & NETIF_F_GRO) || dev->xdp_prog)
		return true;
	return false;
}

L
Linus Torvalds 已提交
1953 1954
#define	NETDEV_ALIGN		32

1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970
static inline
int netdev_get_prio_tc_map(const struct net_device *dev, u32 prio)
{
	return dev->prio_tc_map[prio & TC_BITMASK];
}

static inline
int netdev_set_prio_tc_map(struct net_device *dev, u8 prio, u8 tc)
{
	if (tc >= dev->num_tc)
		return -EINVAL;

	dev->prio_tc_map[prio & TC_BITMASK] = tc & TC_BITMASK;
	return 0;
}

1971
int netdev_txq_to_tc(struct net_device *dev, unsigned int txq);
1972 1973 1974
void netdev_reset_tc(struct net_device *dev);
int netdev_set_tc_queue(struct net_device *dev, u8 tc, u16 count, u16 offset);
int netdev_set_num_tc(struct net_device *dev, u8 num_tc);
1975 1976 1977 1978 1979 1980 1981

static inline
int netdev_get_num_tc(struct net_device *dev)
{
	return dev->num_tc;
}

1982 1983 1984 1985 1986 1987 1988
static inline
struct netdev_queue *netdev_get_tx_queue(const struct net_device *dev,
					 unsigned int index)
{
	return &dev->_tx[index];
}

1989 1990 1991 1992 1993 1994
static inline struct netdev_queue *skb_get_tx_queue(const struct net_device *dev,
						    const struct sk_buff *skb)
{
	return netdev_get_tx_queue(dev, skb_get_queue_mapping(skb));
}

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
static inline void netdev_for_each_tx_queue(struct net_device *dev,
					    void (*f)(struct net_device *,
						      struct netdev_queue *,
						      void *),
					    void *arg)
{
	unsigned int i;

	for (i = 0; i < dev->num_tx_queues; i++)
		f(dev, &dev->_tx[i], arg);
}

2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023
#define netdev_lockdep_set_classes(dev)				\
{								\
	static struct lock_class_key qdisc_tx_busylock_key;	\
	static struct lock_class_key qdisc_running_key;		\
	static struct lock_class_key qdisc_xmit_lock_key;	\
	static struct lock_class_key dev_addr_list_lock_key;	\
	unsigned int i;						\
								\
	(dev)->qdisc_tx_busylock = &qdisc_tx_busylock_key;	\
	(dev)->qdisc_running_key = &qdisc_running_key;		\
	lockdep_set_class(&(dev)->addr_list_lock,		\
			  &dev_addr_list_lock_key); 		\
	for (i = 0; i < (dev)->num_tx_queues; i++)		\
		lockdep_set_class(&(dev)->_tx[i]._xmit_lock,	\
				  &qdisc_xmit_lock_key);	\
}

2024
struct netdev_queue *netdev_pick_tx(struct net_device *dev,
2025 2026
				    struct sk_buff *skb,
				    void *accel_priv);
2027

2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047
/* returns the headroom that the master device needs to take in account
 * when forwarding to this dev
 */
static inline unsigned netdev_get_fwd_headroom(struct net_device *dev)
{
	return dev->priv_flags & IFF_PHONY_HEADROOM ? 0 : dev->needed_headroom;
}

static inline void netdev_set_rx_headroom(struct net_device *dev, int new_hr)
{
	if (dev->netdev_ops->ndo_set_rx_headroom)
		dev->netdev_ops->ndo_set_rx_headroom(dev, new_hr);
}

/* set the device rx headroom to the dev's default */
static inline void netdev_reset_rx_headroom(struct net_device *dev)
{
	netdev_set_rx_headroom(dev, -1);
}

2048 2049 2050 2051 2052 2053
/*
 * Net namespace inlines
 */
static inline
struct net *dev_net(const struct net_device *dev)
{
E
Eric Dumazet 已提交
2054
	return read_pnet(&dev->nd_net);
2055 2056 2057
}

static inline
2058
void dev_net_set(struct net_device *dev, struct net *net)
2059
{
2060
	write_pnet(&dev->nd_net, net);
2061 2062
}

2063 2064 2065 2066 2067 2068
/**
 *	netdev_priv - access network device private data
 *	@dev: network device
 *
 * Get network device private data
 */
2069
static inline void *netdev_priv(const struct net_device *dev)
L
Linus Torvalds 已提交
2070
{
2071
	return (char *)dev + ALIGN(sizeof(struct net_device), NETDEV_ALIGN);
L
Linus Torvalds 已提交
2072 2073 2074 2075 2076
}

/* Set the sysfs physical device reference for the network logical device
 * if set prior to registration will cause a symlink during initialization.
 */
2077
#define SET_NETDEV_DEV(net, pdev)	((net)->dev.parent = (pdev))
L
Linus Torvalds 已提交
2078

2079
/* Set the sysfs device type for the network logical device to allow
2080
 * fine-grained identification of different network device types. For
B
Bjorn Helgaas 已提交
2081
 * example Ethernet, Wireless LAN, Bluetooth, WiMAX etc.
2082 2083 2084
 */
#define SET_NETDEV_DEVTYPE(net, devtype)	((net)->dev.type = (devtype))

E
Eric Dumazet 已提交
2085 2086 2087 2088 2089
/* Default NAPI poll() weight
 * Device drivers are strongly advised to not use bigger value
 */
#define NAPI_POLL_WEIGHT 64

2090
/**
B
Bjorn Helgaas 已提交
2091
 *	netif_napi_add - initialize a NAPI context
2092
 *	@dev:  network device
B
Bjorn Helgaas 已提交
2093
 *	@napi: NAPI context
2094 2095 2096
 *	@poll: polling function
 *	@weight: default weight
 *
B
Bjorn Helgaas 已提交
2097 2098
 * netif_napi_add() must be used to initialize a NAPI context prior to calling
 * *any* of the other NAPI-related functions.
2099
 */
2100 2101
void netif_napi_add(struct net_device *dev, struct napi_struct *napi,
		    int (*poll)(struct napi_struct *, int), int weight);
2102

E
Eric Dumazet 已提交
2103
/**
B
Bjorn Helgaas 已提交
2104
 *	netif_tx_napi_add - initialize a NAPI context
E
Eric Dumazet 已提交
2105
 *	@dev:  network device
B
Bjorn Helgaas 已提交
2106
 *	@napi: NAPI context
E
Eric Dumazet 已提交
2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122
 *	@poll: polling function
 *	@weight: default weight
 *
 * This variant of netif_napi_add() should be used from drivers using NAPI
 * to exclusively poll a TX queue.
 * This will avoid we add it into napi_hash[], thus polluting this hash table.
 */
static inline void netif_tx_napi_add(struct net_device *dev,
				     struct napi_struct *napi,
				     int (*poll)(struct napi_struct *, int),
				     int weight)
{
	set_bit(NAPI_STATE_NO_BUSY_POLL, &napi->state);
	netif_napi_add(dev, napi, poll, weight);
}

2123
/**
B
Bjorn Helgaas 已提交
2124 2125
 *  netif_napi_del - remove a NAPI context
 *  @napi: NAPI context
2126
 *
B
Bjorn Helgaas 已提交
2127
 *  netif_napi_del() removes a NAPI context from the network device NAPI list
2128
 */
2129 2130 2131
void netif_napi_del(struct napi_struct *napi);

struct napi_gro_cb {
2132
	/* Virtual address of skb_shinfo(skb)->frags[0].page + offset. */
B
Bjorn Helgaas 已提交
2133
	void	*frag0;
2134

2135 2136 2137
	/* Length of frag0. */
	unsigned int frag0_len;

2138
	/* This indicates where we are processing relative to skb->data. */
B
Bjorn Helgaas 已提交
2139
	int	data_offset;
2140

2141
	/* This is non-zero if the packet cannot be merged with the new skb. */
2142 2143 2144 2145
	u16	flush;

	/* Save the IP ID here and check when we get to the transport layer */
	u16	flush_id;
2146 2147

	/* Number of segments aggregated. */
2148 2149
	u16	count;

2150 2151 2152
	/* Start offset for remote checksum offload */
	u16	gro_remcsum_start;

2153 2154
	/* jiffies when first packet was created/queued */
	unsigned long age;
2155

T
Tom Herbert 已提交
2156
	/* Used in ipv6_gro_receive() and foo-over-udp */
2157 2158
	u16	proto;

2159 2160 2161
	/* This is non-zero if the packet may be of the same flow. */
	u8	same_flow:1;

2162 2163
	/* Used in tunnel GRO receive */
	u8	encap_mark:1;
T
Tom Herbert 已提交
2164 2165 2166 2167

	/* GRO checksum is valid */
	u8	csum_valid:1;

2168 2169
	/* Number of checksums via CHECKSUM_UNNECESSARY */
	u8	csum_cnt:3;
2170

2171 2172 2173 2174 2175
	/* Free the skb? */
	u8	free:2;
#define NAPI_GRO_FREE		  1
#define NAPI_GRO_FREE_STOLEN_HEAD 2

2176 2177 2178
	/* Used in foo-over-udp, set in udp[46]_gro_receive */
	u8	is_ipv6:1;

2179 2180 2181
	/* Used in GRE, set in fou/gue_gro_receive */
	u8	is_fou:1;

2182 2183 2184
	/* Used to determine if flush_id can be ignored */
	u8	is_atomic:1;

S
Sabrina Dubroca 已提交
2185 2186 2187 2188
	/* Number of gro_receive callbacks this packet already went through */
	u8 recursion_counter:4;

	/* 1 bit hole */
2189

2190 2191 2192
	/* used to support CHECKSUM_COMPLETE for tunneling protocols */
	__wsum	csum;

2193 2194
	/* used in skb_gro_receive() slow path */
	struct sk_buff *last;
2195 2196 2197
};

#define NAPI_GRO_CB(skb) ((struct napi_gro_cb *)(skb)->cb)
2198

S
Sabrina Dubroca 已提交
2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232
#define GRO_RECURSION_LIMIT 15
static inline int gro_recursion_inc_test(struct sk_buff *skb)
{
	return ++NAPI_GRO_CB(skb)->recursion_counter == GRO_RECURSION_LIMIT;
}

typedef struct sk_buff **(*gro_receive_t)(struct sk_buff **, struct sk_buff *);
static inline struct sk_buff **call_gro_receive(gro_receive_t cb,
						struct sk_buff **head,
						struct sk_buff *skb)
{
	if (unlikely(gro_recursion_inc_test(skb))) {
		NAPI_GRO_CB(skb)->flush |= 1;
		return NULL;
	}

	return cb(head, skb);
}

typedef struct sk_buff **(*gro_receive_sk_t)(struct sock *, struct sk_buff **,
					     struct sk_buff *);
static inline struct sk_buff **call_gro_receive_sk(gro_receive_sk_t cb,
						   struct sock *sk,
						   struct sk_buff **head,
						   struct sk_buff *skb)
{
	if (unlikely(gro_recursion_inc_test(skb))) {
		NAPI_GRO_CB(skb)->flush |= 1;
		return NULL;
	}

	return cb(sk, head, skb);
}

L
Linus Torvalds 已提交
2233
struct packet_type {
D
David S. Miller 已提交
2234 2235 2236 2237 2238 2239
	__be16			type;	/* This is really htons(ether_type). */
	struct net_device	*dev;	/* NULL is wildcarded here	     */
	int			(*func) (struct sk_buff *,
					 struct net_device *,
					 struct packet_type *,
					 struct net_device *);
2240 2241
	bool			(*id_match)(struct packet_type *ptype,
					    struct sock *sk);
L
Linus Torvalds 已提交
2242 2243 2244 2245
	void			*af_packet_priv;
	struct list_head	list;
};

2246
struct offload_callbacks {
2247
	struct sk_buff		*(*gso_segment)(struct sk_buff *skb,
2248
						netdev_features_t features);
2249
	struct sk_buff		**(*gro_receive)(struct sk_buff **head,
2250
						 struct sk_buff *skb);
2251
	int			(*gro_complete)(struct sk_buff *skb, int nhoff);
2252 2253 2254 2255
};

struct packet_offload {
	__be16			 type;	/* This is really htons(ether_type). */
2256
	u16			 priority;
2257 2258
	struct offload_callbacks callbacks;
	struct list_head	 list;
L
Linus Torvalds 已提交
2259 2260
};

B
Bjorn Helgaas 已提交
2261
/* often modified stats are per-CPU, other are shared (netdev->stats) */
2262 2263 2264 2265 2266 2267 2268 2269
struct pcpu_sw_netstats {
	u64     rx_packets;
	u64     rx_bytes;
	u64     tx_packets;
	u64     tx_bytes;
	struct u64_stats_sync   syncp;
};

2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281
#define __netdev_alloc_pcpu_stats(type, gfp)				\
({									\
	typeof(type) __percpu *pcpu_stats = alloc_percpu_gfp(type, gfp);\
	if (pcpu_stats)	{						\
		int __cpu;						\
		for_each_possible_cpu(__cpu) {				\
			typeof(type) *stat;				\
			stat = per_cpu_ptr(pcpu_stats, __cpu);		\
			u64_stats_init(&stat->syncp);			\
		}							\
	}								\
	pcpu_stats;							\
2282 2283
})

2284
#define netdev_alloc_pcpu_stats(type)					\
2285
	__netdev_alloc_pcpu_stats(type, GFP_KERNEL)
2286

2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299
enum netdev_lag_tx_type {
	NETDEV_LAG_TX_TYPE_UNKNOWN,
	NETDEV_LAG_TX_TYPE_RANDOM,
	NETDEV_LAG_TX_TYPE_BROADCAST,
	NETDEV_LAG_TX_TYPE_ROUNDROBIN,
	NETDEV_LAG_TX_TYPE_ACTIVEBACKUP,
	NETDEV_LAG_TX_TYPE_HASH,
};

struct netdev_lag_upper_info {
	enum netdev_lag_tx_type tx_type;
};

2300 2301 2302 2303 2304
struct netdev_lag_lower_state_info {
	u8 link_up : 1,
	   tx_enabled : 1;
};

L
Linus Torvalds 已提交
2305 2306
#include <linux/notifier.h>

2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319
/* netdevice notifier chain. Please remember to update the rtnetlink
 * notification exclusion list in rtnetlink_event() when adding new
 * types.
 */
#define NETDEV_UP	0x0001	/* For now you can't veto a device up/down */
#define NETDEV_DOWN	0x0002
#define NETDEV_REBOOT	0x0003	/* Tell a protocol stack a network interface
				   detected a hardware crash and restarted
				   - we can use this eg to kick tcp sessions
				   once done */
#define NETDEV_CHANGE	0x0004	/* Notify device state change */
#define NETDEV_REGISTER 0x0005
#define NETDEV_UNREGISTER	0x0006
2320
#define NETDEV_CHANGEMTU	0x0007 /* notify after mtu change happened */
2321 2322 2323 2324 2325 2326 2327 2328 2329
#define NETDEV_CHANGEADDR	0x0008
#define NETDEV_GOING_DOWN	0x0009
#define NETDEV_CHANGENAME	0x000A
#define NETDEV_FEAT_CHANGE	0x000B
#define NETDEV_BONDING_FAILOVER 0x000C
#define NETDEV_PRE_UP		0x000D
#define NETDEV_PRE_TYPE_CHANGE	0x000E
#define NETDEV_POST_TYPE_CHANGE	0x000F
#define NETDEV_POST_INIT	0x0010
2330
#define NETDEV_UNREGISTER_FINAL 0x0011
2331 2332 2333
#define NETDEV_RELEASE		0x0012
#define NETDEV_NOTIFY_PEERS	0x0013
#define NETDEV_JOIN		0x0014
2334
#define NETDEV_CHANGEUPPER	0x0015
2335
#define NETDEV_RESEND_IGMP	0x0016
2336
#define NETDEV_PRECHANGEMTU	0x0017 /* notify before mtu change happened */
2337
#define NETDEV_CHANGEINFODATA	0x0018
2338
#define NETDEV_BONDING_INFO	0x0019
2339
#define NETDEV_PRECHANGEUPPER	0x001A
2340
#define NETDEV_CHANGELOWERSTATE	0x001B
2341
#define NETDEV_UDP_TUNNEL_PUSH_INFO	0x001C
2342
#define NETDEV_UDP_TUNNEL_DROP_INFO	0x001D
2343
#define NETDEV_CHANGE_TX_QUEUE_LEN	0x001E
2344

2345 2346
int register_netdevice_notifier(struct notifier_block *nb);
int unregister_netdevice_notifier(struct notifier_block *nb);
2347 2348

struct netdev_notifier_info {
2349 2350
	struct net_device	*dev;
	struct netlink_ext_ack	*extack;
2351 2352
};

2353 2354 2355 2356 2357
struct netdev_notifier_change_info {
	struct netdev_notifier_info info; /* must be first */
	unsigned int flags_changed;
};

2358 2359 2360 2361
struct netdev_notifier_changeupper_info {
	struct netdev_notifier_info info; /* must be first */
	struct net_device *upper_dev; /* new upper dev */
	bool master; /* is upper dev master */
B
Bjorn Helgaas 已提交
2362
	bool linking; /* is the notification for link or unlink */
2363
	void *upper_info; /* upper dev info */
2364 2365
};

2366 2367 2368 2369 2370
struct netdev_notifier_changelowerstate_info {
	struct netdev_notifier_info info; /* must be first */
	void *lower_state_info; /* is lower dev state */
};

2371 2372 2373 2374
static inline void netdev_notifier_info_init(struct netdev_notifier_info *info,
					     struct net_device *dev)
{
	info->dev = dev;
2375
	info->extack = NULL;
2376 2377
}

2378 2379 2380 2381 2382 2383
static inline struct net_device *
netdev_notifier_info_to_dev(const struct netdev_notifier_info *info)
{
	return info->dev;
}

2384 2385 2386 2387 2388 2389
static inline struct netlink_ext_ack *
netdev_notifier_info_to_extack(const struct netdev_notifier_info *info)
{
	return info->extack;
}

2390
int call_netdevice_notifiers(unsigned long val, struct net_device *dev);
2391 2392


L
Linus Torvalds 已提交
2393 2394
extern rwlock_t				dev_base_lock;		/* Device list lock */

2395 2396
#define for_each_netdev(net, d)		\
		list_for_each_entry(d, &(net)->dev_base_head, dev_list)
2397 2398
#define for_each_netdev_reverse(net, d)	\
		list_for_each_entry_reverse(d, &(net)->dev_base_head, dev_list)
2399 2400
#define for_each_netdev_rcu(net, d)		\
		list_for_each_entry_rcu(d, &(net)->dev_base_head, dev_list)
2401 2402 2403 2404
#define for_each_netdev_safe(net, d, n)	\
		list_for_each_entry_safe(d, n, &(net)->dev_base_head, dev_list)
#define for_each_netdev_continue(net, d)		\
		list_for_each_entry_continue(d, &(net)->dev_base_head, dev_list)
2405 2406
#define for_each_netdev_continue_rcu(net, d)		\
	list_for_each_entry_continue_rcu(d, &(net)->dev_base_head, dev_list)
2407 2408
#define for_each_netdev_in_bond_rcu(bond, slave)	\
		for_each_netdev_rcu(&init_net, slave)	\
2409
			if (netdev_master_upper_dev_get_rcu(slave) == (bond))
2410
#define net_device_entry(lh)	list_entry(lh, struct net_device, dev_list)
2411

2412 2413 2414 2415 2416
static inline struct net_device *next_net_device(struct net_device *dev)
{
	struct list_head *lh;
	struct net *net;

2417
	net = dev_net(dev);
2418 2419 2420 2421
	lh = dev->dev_list.next;
	return lh == &net->dev_base_head ? NULL : net_device_entry(lh);
}

2422 2423 2424 2425 2426 2427
static inline struct net_device *next_net_device_rcu(struct net_device *dev)
{
	struct list_head *lh;
	struct net *net;

	net = dev_net(dev);
E
Eric Dumazet 已提交
2428
	lh = rcu_dereference(list_next_rcu(&dev->dev_list));
2429 2430 2431
	return lh == &net->dev_base_head ? NULL : net_device_entry(lh);
}

2432 2433 2434 2435 2436
static inline struct net_device *first_net_device(struct net *net)
{
	return list_empty(&net->dev_base_head) ? NULL :
		net_device_entry(net->dev_base_head.next);
}
2437

E
Eric Dumazet 已提交
2438 2439 2440 2441 2442 2443 2444
static inline struct net_device *first_net_device_rcu(struct net *net)
{
	struct list_head *lh = rcu_dereference(list_next_rcu(&net->dev_base_head));

	return lh == &net->dev_base_head ? NULL : net_device_entry(lh);
}

2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456
int netdev_boot_setup_check(struct net_device *dev);
unsigned long netdev_boot_base(const char *prefix, int unit);
struct net_device *dev_getbyhwaddr_rcu(struct net *net, unsigned short type,
				       const char *hwaddr);
struct net_device *dev_getfirstbyhwtype(struct net *net, unsigned short type);
struct net_device *__dev_getfirstbyhwtype(struct net *net, unsigned short type);
void dev_add_pack(struct packet_type *pt);
void dev_remove_pack(struct packet_type *pt);
void __dev_remove_pack(struct packet_type *pt);
void dev_add_offload(struct packet_offload *po);
void dev_remove_offload(struct packet_offload *po);

2457
int dev_get_iflink(const struct net_device *dev);
2458
int dev_fill_metadata_dst(struct net_device *dev, struct sk_buff *skb);
2459 2460
struct net_device *__dev_get_by_flags(struct net *net, unsigned short flags,
				      unsigned short mask);
2461 2462 2463 2464 2465
struct net_device *dev_get_by_name(struct net *net, const char *name);
struct net_device *dev_get_by_name_rcu(struct net *net, const char *name);
struct net_device *__dev_get_by_name(struct net *net, const char *name);
int dev_alloc_name(struct net_device *dev, const char *name);
int dev_open(struct net_device *dev);
2466 2467
void dev_close(struct net_device *dev);
void dev_close_many(struct list_head *head, bool unlink);
2468
void dev_disable_lro(struct net_device *dev);
2469
int dev_loopback_xmit(struct net *net, struct sock *sk, struct sk_buff *newskb);
2470
int dev_queue_xmit(struct sk_buff *skb);
2471
int dev_queue_xmit_accel(struct sk_buff *skb, void *accel_priv);
2472 2473 2474
int register_netdevice(struct net_device *dev);
void unregister_netdevice_queue(struct net_device *dev, struct list_head *head);
void unregister_netdevice_many(struct list_head *head);
2475 2476 2477 2478 2479
static inline void unregister_netdevice(struct net_device *dev)
{
	unregister_netdevice_queue(dev, NULL);
}

2480 2481
int netdev_refcnt_read(const struct net_device *dev);
void free_netdev(struct net_device *dev);
2482
void netdev_freemem(struct net_device *dev);
2483 2484
void synchronize_net(void);
int init_dummy_netdev(struct net_device *dev);
2485

2486
DECLARE_PER_CPU(int, xmit_recursion);
2487 2488
#define XMIT_RECURSION_LIMIT	10

2489 2490 2491 2492 2493
static inline int dev_recursion_level(void)
{
	return this_cpu_read(xmit_recursion);
}

2494 2495 2496
struct net_device *dev_get_by_index(struct net *net, int ifindex);
struct net_device *__dev_get_by_index(struct net *net, int ifindex);
struct net_device *dev_get_by_index_rcu(struct net *net, int ifindex);
2497
struct net_device *dev_get_by_napi_id(unsigned int napi_id);
2498 2499 2500
int netdev_get_name(struct net *net, char *name, int ifindex);
int dev_restart(struct net_device *dev);
int skb_gro_receive(struct sk_buff **head, struct sk_buff *skb);
2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516

static inline unsigned int skb_gro_offset(const struct sk_buff *skb)
{
	return NAPI_GRO_CB(skb)->data_offset;
}

static inline unsigned int skb_gro_len(const struct sk_buff *skb)
{
	return skb->len - NAPI_GRO_CB(skb)->data_offset;
}

static inline void skb_gro_pull(struct sk_buff *skb, unsigned int len)
{
	NAPI_GRO_CB(skb)->data_offset += len;
}

2517 2518
static inline void *skb_gro_header_fast(struct sk_buff *skb,
					unsigned int offset)
2519
{
2520 2521
	return NAPI_GRO_CB(skb)->frag0 + offset;
}
2522

2523 2524 2525 2526
static inline int skb_gro_header_hard(struct sk_buff *skb, unsigned int hlen)
{
	return NAPI_GRO_CB(skb)->frag0_len < hlen;
}
2527

2528 2529 2530 2531 2532 2533
static inline void skb_gro_frag0_invalidate(struct sk_buff *skb)
{
	NAPI_GRO_CB(skb)->frag0 = NULL;
	NAPI_GRO_CB(skb)->frag0_len = 0;
}

2534 2535 2536
static inline void *skb_gro_header_slow(struct sk_buff *skb, unsigned int hlen,
					unsigned int offset)
{
2537 2538 2539
	if (!pskb_may_pull(skb, hlen))
		return NULL;

2540
	skb_gro_frag0_invalidate(skb);
2541
	return skb->data + offset;
2542
}
L
Linus Torvalds 已提交
2543

H
Herbert Xu 已提交
2544 2545
static inline void *skb_gro_network_header(struct sk_buff *skb)
{
2546 2547
	return (NAPI_GRO_CB(skb)->frag0 ?: skb->data) +
	       skb_network_offset(skb);
H
Herbert Xu 已提交
2548 2549
}

2550 2551 2552
static inline void skb_gro_postpull_rcsum(struct sk_buff *skb,
					const void *start, unsigned int len)
{
T
Tom Herbert 已提交
2553
	if (NAPI_GRO_CB(skb)->csum_valid)
2554 2555 2556 2557
		NAPI_GRO_CB(skb)->csum = csum_sub(NAPI_GRO_CB(skb)->csum,
						  csum_partial(start, len, 0));
}

T
Tom Herbert 已提交
2558 2559 2560 2561 2562 2563 2564
/* GRO checksum functions. These are logical equivalents of the normal
 * checksum functions (in skbuff.h) except that they operate on the GRO
 * offsets and fields in sk_buff.
 */

__sum16 __skb_gro_checksum_complete(struct sk_buff *skb);

2565 2566
static inline bool skb_at_gro_remcsum_start(struct sk_buff *skb)
{
2567
	return (NAPI_GRO_CB(skb)->gro_remcsum_start == skb_gro_offset(skb));
2568 2569
}

T
Tom Herbert 已提交
2570 2571 2572 2573
static inline bool __skb_gro_checksum_validate_needed(struct sk_buff *skb,
						      bool zero_okay,
						      __sum16 check)
{
2574 2575 2576
	return ((skb->ip_summed != CHECKSUM_PARTIAL ||
		skb_checksum_start_offset(skb) <
		 skb_gro_offset(skb)) &&
2577
		!skb_at_gro_remcsum_start(skb) &&
2578
		NAPI_GRO_CB(skb)->csum_cnt == 0 &&
T
Tom Herbert 已提交
2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595
		(!zero_okay || check));
}

static inline __sum16 __skb_gro_checksum_validate_complete(struct sk_buff *skb,
							   __wsum psum)
{
	if (NAPI_GRO_CB(skb)->csum_valid &&
	    !csum_fold(csum_add(psum, NAPI_GRO_CB(skb)->csum)))
		return 0;

	NAPI_GRO_CB(skb)->csum = psum;

	return __skb_gro_checksum_complete(skb);
}

static inline void skb_gro_incr_csum_unnecessary(struct sk_buff *skb)
{
2596 2597 2598 2599 2600 2601 2602 2603 2604
	if (NAPI_GRO_CB(skb)->csum_cnt > 0) {
		/* Consume a checksum from CHECKSUM_UNNECESSARY */
		NAPI_GRO_CB(skb)->csum_cnt--;
	} else {
		/* Update skb for CHECKSUM_UNNECESSARY and csum_level when we
		 * verified a new top level checksum or an encapsulated one
		 * during GRO. This saves work if we fallback to normal path.
		 */
		__skb_incr_checksum_unnecessary(skb);
T
Tom Herbert 已提交
2605 2606 2607 2608 2609 2610 2611 2612 2613 2614
	}
}

#define __skb_gro_checksum_validate(skb, proto, zero_okay, check,	\
				    compute_pseudo)			\
({									\
	__sum16 __ret = 0;						\
	if (__skb_gro_checksum_validate_needed(skb, zero_okay, check))	\
		__ret = __skb_gro_checksum_validate_complete(skb,	\
				compute_pseudo(skb, proto));		\
2615
	if (!__ret)							\
T
Tom Herbert 已提交
2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629
		skb_gro_incr_csum_unnecessary(skb);			\
	__ret;								\
})

#define skb_gro_checksum_validate(skb, proto, compute_pseudo)		\
	__skb_gro_checksum_validate(skb, proto, false, 0, compute_pseudo)

#define skb_gro_checksum_validate_zero_check(skb, proto, check,		\
					     compute_pseudo)		\
	__skb_gro_checksum_validate(skb, proto, true, check, compute_pseudo)

#define skb_gro_checksum_simple_validate(skb)				\
	__skb_gro_checksum_validate(skb, 0, false, 0, null_compute_pseudo)

2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649
static inline bool __skb_gro_checksum_convert_check(struct sk_buff *skb)
{
	return (NAPI_GRO_CB(skb)->csum_cnt == 0 &&
		!NAPI_GRO_CB(skb)->csum_valid);
}

static inline void __skb_gro_checksum_convert(struct sk_buff *skb,
					      __sum16 check, __wsum pseudo)
{
	NAPI_GRO_CB(skb)->csum = ~pseudo;
	NAPI_GRO_CB(skb)->csum_valid = 1;
}

#define skb_gro_checksum_try_convert(skb, proto, check, compute_pseudo)	\
do {									\
	if (__skb_gro_checksum_convert_check(skb))			\
		__skb_gro_checksum_convert(skb, check,			\
					   compute_pseudo(skb, proto));	\
} while (0)

2650 2651 2652 2653 2654 2655 2656
struct gro_remcsum {
	int offset;
	__wsum delta;
};

static inline void skb_gro_remcsum_init(struct gro_remcsum *grc)
{
2657
	grc->offset = 0;
2658 2659 2660
	grc->delta = 0;
}

2661 2662 2663 2664 2665
static inline void *skb_gro_remcsum_process(struct sk_buff *skb, void *ptr,
					    unsigned int off, size_t hdrlen,
					    int start, int offset,
					    struct gro_remcsum *grc,
					    bool nopartial)
2666 2667
{
	__wsum delta;
2668
	size_t plen = hdrlen + max_t(size_t, offset + sizeof(u16), start);
2669 2670 2671

	BUG_ON(!NAPI_GRO_CB(skb)->csum_valid);

2672
	if (!nopartial) {
2673 2674 2675 2676 2677 2678 2679 2680 2681
		NAPI_GRO_CB(skb)->gro_remcsum_start = off + hdrlen + start;
		return ptr;
	}

	ptr = skb_gro_header_fast(skb, off);
	if (skb_gro_header_hard(skb, off + plen)) {
		ptr = skb_gro_header_slow(skb, off + plen, off);
		if (!ptr)
			return NULL;
2682 2683
	}

2684 2685
	delta = remcsum_adjust(ptr + hdrlen, NAPI_GRO_CB(skb)->csum,
			       start, offset);
2686 2687 2688

	/* Adjust skb->csum since we changed the packet */
	NAPI_GRO_CB(skb)->csum = csum_add(NAPI_GRO_CB(skb)->csum, delta);
2689

2690
	grc->offset = off + hdrlen + offset;
2691
	grc->delta = delta;
2692 2693

	return ptr;
2694 2695
}

2696 2697 2698
static inline void skb_gro_remcsum_cleanup(struct sk_buff *skb,
					   struct gro_remcsum *grc)
{
2699 2700 2701
	void *ptr;
	size_t plen = grc->offset + sizeof(u16);

2702 2703 2704
	if (!grc->delta)
		return;

2705 2706 2707 2708 2709 2710 2711 2712
	ptr = skb_gro_header_fast(skb, grc->offset);
	if (skb_gro_header_hard(skb, grc->offset + sizeof(u16))) {
		ptr = skb_gro_header_slow(skb, plen, grc->offset);
		if (!ptr)
			return;
	}

	remcsum_unadjust((__sum16 *)ptr, grc->delta);
2713
}
2714

2715 2716 2717 2718 2719 2720 2721
#ifdef CONFIG_XFRM_OFFLOAD
static inline void skb_gro_flush_final(struct sk_buff *skb, struct sk_buff **pp, int flush)
{
	if (PTR_ERR(pp) != -EINPROGRESS)
		NAPI_GRO_CB(skb)->flush |= flush;
}
#else
2722 2723 2724 2725
static inline void skb_gro_flush_final(struct sk_buff *skb, struct sk_buff **pp, int flush)
{
	NAPI_GRO_CB(skb)->flush |= flush;
}
2726
#endif
2727

2728 2729
static inline int dev_hard_header(struct sk_buff *skb, struct net_device *dev,
				  unsigned short type,
2730
				  const void *daddr, const void *saddr,
2731
				  unsigned int len)
2732
{
2733
	if (!dev->header_ops || !dev->header_ops->create)
2734
		return 0;
2735 2736

	return dev->header_ops->create(skb, dev, type, daddr, saddr, len);
2737 2738
}

S
Stephen Hemminger 已提交
2739 2740 2741 2742 2743
static inline int dev_parse_header(const struct sk_buff *skb,
				   unsigned char *haddr)
{
	const struct net_device *dev = skb->dev;

2744
	if (!dev->header_ops || !dev->header_ops->parse)
S
Stephen Hemminger 已提交
2745
		return 0;
2746
	return dev->header_ops->parse(skb, haddr);
S
Stephen Hemminger 已提交
2747 2748
}

2749 2750 2751 2752 2753 2754
/* ll_header must have at least hard_header_len allocated */
static inline bool dev_validate_header(const struct net_device *dev,
				       char *ll_header, int len)
{
	if (likely(len >= dev->hard_header_len))
		return true;
2755 2756
	if (len < dev->min_header_len)
		return false;
2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768

	if (capable(CAP_SYS_RAWIO)) {
		memset(ll_header + len, 0, dev->hard_header_len - len);
		return true;
	}

	if (dev->header_ops && dev->header_ops->validate)
		return dev->header_ops->validate(ll_header, len);

	return false;
}

2769 2770
typedef int gifconf_func_t(struct net_device * dev, char __user * bufptr,
			   int len, int size);
2771
int register_gifconf(unsigned int family, gifconf_func_t *gifconf);
L
Linus Torvalds 已提交
2772 2773 2774 2775 2776
static inline int unregister_gifconf(unsigned int family)
{
	return register_gifconf(family, NULL);
}

2777
#ifdef CONFIG_NET_FLOW_LIMIT
2778
#define FLOW_LIMIT_HISTORY	(1 << 7)  /* must be ^2 and !overflow buckets */
2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789
struct sd_flow_limit {
	u64			count;
	unsigned int		num_buckets;
	unsigned int		history_head;
	u16			history[FLOW_LIMIT_HISTORY];
	u8			buckets[];
};

extern int netdev_flow_limit_table_len;
#endif /* CONFIG_NET_FLOW_LIMIT */

L
Linus Torvalds 已提交
2790
/*
B
Bjorn Helgaas 已提交
2791
 * Incoming packets are placed on per-CPU queues
L
Linus Torvalds 已提交
2792
 */
E
Eric Dumazet 已提交
2793
struct softnet_data {
L
Linus Torvalds 已提交
2794
	struct list_head	poll_list;
2795
	struct sk_buff_head	process_queue;
L
Linus Torvalds 已提交
2796

C
Changli Gao 已提交
2797
	/* stats */
2798 2799 2800
	unsigned int		processed;
	unsigned int		time_squeeze;
	unsigned int		received_rps;
2801
#ifdef CONFIG_RPS
E
Eric Dumazet 已提交
2802
	struct softnet_data	*rps_ipi_list;
E
Eric Dumazet 已提交
2803 2804 2805 2806 2807 2808 2809
#endif
#ifdef CONFIG_NET_FLOW_LIMIT
	struct sd_flow_limit __rcu *flow_limit;
#endif
	struct Qdisc		*output_queue;
	struct Qdisc		**output_queue_tailp;
	struct sk_buff		*completion_queue;
2810 2811 2812
#ifdef CONFIG_XFRM_OFFLOAD
	struct sk_buff_head	xfrm_backlog;
#endif
E
Eric Dumazet 已提交
2813
#ifdef CONFIG_RPS
2814 2815 2816 2817 2818 2819
	/* input_queue_head should be written by cpu owning this struct,
	 * and only read by other cpus. Worth using a cache line.
	 */
	unsigned int		input_queue_head ____cacheline_aligned_in_smp;

	/* Elements below can be accessed between CPUs for RPS/RFS */
2820
	call_single_data_t	csd ____cacheline_aligned_in_smp;
E
Eric Dumazet 已提交
2821 2822
	struct softnet_data	*rps_ipi_next;
	unsigned int		cpu;
2823
	unsigned int		input_queue_tail;
2824
#endif
2825
	unsigned int		dropped;
T
Tom Herbert 已提交
2826
	struct sk_buff_head	input_pkt_queue;
2827
	struct napi_struct	backlog;
2828

L
Linus Torvalds 已提交
2829 2830
};

2831
static inline void input_queue_head_incr(struct softnet_data *sd)
T
Tom Herbert 已提交
2832 2833
{
#ifdef CONFIG_RPS
2834 2835 2836 2837 2838 2839 2840 2841 2842
	sd->input_queue_head++;
#endif
}

static inline void input_queue_tail_incr_save(struct softnet_data *sd,
					      unsigned int *qtail)
{
#ifdef CONFIG_RPS
	*qtail = ++sd->input_queue_tail;
T
Tom Herbert 已提交
2843 2844 2845
#endif
}

T
Tom Herbert 已提交
2846
DECLARE_PER_CPU_ALIGNED(struct softnet_data, softnet_data);
L
Linus Torvalds 已提交
2847

2848
void __netif_schedule(struct Qdisc *q);
2849
void netif_schedule_queue(struct netdev_queue *txq);
2850

2851 2852 2853 2854 2855 2856 2857 2858
static inline void netif_tx_schedule_all(struct net_device *dev)
{
	unsigned int i;

	for (i = 0; i < dev->num_tx_queues; i++)
		netif_schedule_queue(netdev_get_tx_queue(dev, i));
}

2859
static __always_inline void netif_tx_start_queue(struct netdev_queue *dev_queue)
2860
{
2861
	clear_bit(__QUEUE_STATE_DRV_XOFF, &dev_queue->state);
2862 2863
}

2864 2865 2866 2867 2868 2869
/**
 *	netif_start_queue - allow transmit
 *	@dev: network device
 *
 *	Allow upper layers to call the device hard_start_xmit routine.
 */
L
Linus Torvalds 已提交
2870 2871
static inline void netif_start_queue(struct net_device *dev)
{
2872
	netif_tx_start_queue(netdev_get_tx_queue(dev, 0));
L
Linus Torvalds 已提交
2873 2874
}

2875 2876 2877 2878 2879 2880 2881 2882 2883 2884
static inline void netif_tx_start_all_queues(struct net_device *dev)
{
	unsigned int i;

	for (i = 0; i < dev->num_tx_queues; i++) {
		struct netdev_queue *txq = netdev_get_tx_queue(dev, i);
		netif_tx_start_queue(txq);
	}
}

2885
void netif_tx_wake_queue(struct netdev_queue *dev_queue);
2886

2887 2888 2889 2890 2891 2892 2893
/**
 *	netif_wake_queue - restart transmit
 *	@dev: network device
 *
 *	Allow upper layers to call the device hard_start_xmit routine.
 *	Used for flow control when transmit resources are available.
 */
2894 2895
static inline void netif_wake_queue(struct net_device *dev)
{
2896
	netif_tx_wake_queue(netdev_get_tx_queue(dev, 0));
L
Linus Torvalds 已提交
2897 2898
}

2899 2900 2901 2902 2903 2904 2905 2906 2907 2908
static inline void netif_tx_wake_all_queues(struct net_device *dev)
{
	unsigned int i;

	for (i = 0; i < dev->num_tx_queues; i++) {
		struct netdev_queue *txq = netdev_get_tx_queue(dev, i);
		netif_tx_wake_queue(txq);
	}
}

2909
static __always_inline void netif_tx_stop_queue(struct netdev_queue *dev_queue)
2910
{
2911
	set_bit(__QUEUE_STATE_DRV_XOFF, &dev_queue->state);
2912 2913
}

2914 2915 2916 2917 2918 2919 2920
/**
 *	netif_stop_queue - stop transmitted packets
 *	@dev: network device
 *
 *	Stop upper layers calling the device hard_start_xmit routine.
 *	Used for flow control when transmit resources are unavailable.
 */
L
Linus Torvalds 已提交
2921 2922
static inline void netif_stop_queue(struct net_device *dev)
{
2923
	netif_tx_stop_queue(netdev_get_tx_queue(dev, 0));
L
Linus Torvalds 已提交
2924 2925
}

2926
void netif_tx_stop_all_queues(struct net_device *dev);
2927

2928
static inline bool netif_tx_queue_stopped(const struct netdev_queue *dev_queue)
2929
{
2930
	return test_bit(__QUEUE_STATE_DRV_XOFF, &dev_queue->state);
2931 2932
}

2933 2934 2935 2936 2937 2938
/**
 *	netif_queue_stopped - test if transmit queue is flowblocked
 *	@dev: network device
 *
 *	Test if transmit queue on device is currently unable to send.
 */
2939
static inline bool netif_queue_stopped(const struct net_device *dev)
L
Linus Torvalds 已提交
2940
{
2941
	return netif_tx_queue_stopped(netdev_get_tx_queue(dev, 0));
L
Linus Torvalds 已提交
2942 2943
}

2944
static inline bool netif_xmit_stopped(const struct netdev_queue *dev_queue)
2945
{
2946 2947 2948
	return dev_queue->state & QUEUE_STATE_ANY_XOFF;
}

2949 2950
static inline bool
netif_xmit_frozen_or_stopped(const struct netdev_queue *dev_queue)
2951 2952 2953 2954
{
	return dev_queue->state & QUEUE_STATE_ANY_XOFF_OR_FROZEN;
}

2955 2956 2957 2958 2959 2960
static inline bool
netif_xmit_frozen_or_drv_stopped(const struct netdev_queue *dev_queue)
{
	return dev_queue->state & QUEUE_STATE_DRV_XOFF_OR_FROZEN;
}

2961 2962 2963 2964 2965
/**
 *	netdev_txq_bql_enqueue_prefetchw - prefetch bql data for write
 *	@dev_queue: pointer to transmit queue
 *
 * BQL enabled drivers might use this helper in their ndo_start_xmit(),
B
Bjorn Helgaas 已提交
2966
 * to give appropriate hint to the CPU.
2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979
 */
static inline void netdev_txq_bql_enqueue_prefetchw(struct netdev_queue *dev_queue)
{
#ifdef CONFIG_BQL
	prefetchw(&dev_queue->dql.num_queued);
#endif
}

/**
 *	netdev_txq_bql_complete_prefetchw - prefetch bql data for write
 *	@dev_queue: pointer to transmit queue
 *
 * BQL enabled drivers might use this helper in their TX completion path,
B
Bjorn Helgaas 已提交
2980
 * to give appropriate hint to the CPU.
2981 2982 2983 2984 2985 2986 2987 2988
 */
static inline void netdev_txq_bql_complete_prefetchw(struct netdev_queue *dev_queue)
{
#ifdef CONFIG_BQL
	prefetchw(&dev_queue->dql.limit);
#endif
}

2989 2990 2991
static inline void netdev_tx_sent_queue(struct netdev_queue *dev_queue,
					unsigned int bytes)
{
T
Tom Herbert 已提交
2992 2993
#ifdef CONFIG_BQL
	dql_queued(&dev_queue->dql, bytes);
2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009

	if (likely(dql_avail(&dev_queue->dql) >= 0))
		return;

	set_bit(__QUEUE_STATE_STACK_XOFF, &dev_queue->state);

	/*
	 * The XOFF flag must be set before checking the dql_avail below,
	 * because in netdev_tx_completed_queue we update the dql_completed
	 * before checking the XOFF flag.
	 */
	smp_mb();

	/* check again in case another CPU has just made room avail */
	if (unlikely(dql_avail(&dev_queue->dql) >= 0))
		clear_bit(__QUEUE_STATE_STACK_XOFF, &dev_queue->state);
T
Tom Herbert 已提交
3010
#endif
3011 3012
}

3013 3014 3015 3016 3017 3018 3019 3020 3021
/**
 * 	netdev_sent_queue - report the number of bytes queued to hardware
 * 	@dev: network device
 * 	@bytes: number of bytes queued to the hardware device queue
 *
 * 	Report the number of bytes queued for sending/completion to the network
 * 	device hardware queue. @bytes should be a good approximation and should
 * 	exactly match netdev_completed_queue() @bytes
 */
3022 3023 3024 3025 3026 3027
static inline void netdev_sent_queue(struct net_device *dev, unsigned int bytes)
{
	netdev_tx_sent_queue(netdev_get_tx_queue(dev, 0), bytes);
}

static inline void netdev_tx_completed_queue(struct netdev_queue *dev_queue,
3028
					     unsigned int pkts, unsigned int bytes)
3029
{
T
Tom Herbert 已提交
3030
#ifdef CONFIG_BQL
3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047
	if (unlikely(!bytes))
		return;

	dql_completed(&dev_queue->dql, bytes);

	/*
	 * Without the memory barrier there is a small possiblity that
	 * netdev_tx_sent_queue will miss the update and cause the queue to
	 * be stopped forever
	 */
	smp_mb();

	if (dql_avail(&dev_queue->dql) < 0)
		return;

	if (test_and_clear_bit(__QUEUE_STATE_STACK_XOFF, &dev_queue->state))
		netif_schedule_queue(dev_queue);
T
Tom Herbert 已提交
3048
#endif
3049 3050
}

3051 3052 3053 3054 3055 3056 3057 3058 3059 3060
/**
 * 	netdev_completed_queue - report bytes and packets completed by device
 * 	@dev: network device
 * 	@pkts: actual number of packets sent over the medium
 * 	@bytes: actual number of bytes sent over the medium
 *
 * 	Report the number of bytes and packets transmitted by the network device
 * 	hardware queue over the physical medium, @bytes must exactly match the
 * 	@bytes amount passed to netdev_sent_queue()
 */
3061
static inline void netdev_completed_queue(struct net_device *dev,
3062
					  unsigned int pkts, unsigned int bytes)
3063 3064 3065 3066 3067 3068
{
	netdev_tx_completed_queue(netdev_get_tx_queue(dev, 0), pkts, bytes);
}

static inline void netdev_tx_reset_queue(struct netdev_queue *q)
{
T
Tom Herbert 已提交
3069
#ifdef CONFIG_BQL
3070
	clear_bit(__QUEUE_STATE_STACK_XOFF, &q->state);
T
Tom Herbert 已提交
3071 3072
	dql_reset(&q->dql);
#endif
3073 3074
}

3075 3076 3077 3078 3079 3080 3081
/**
 * 	netdev_reset_queue - reset the packets and bytes count of a network device
 * 	@dev_queue: network device
 *
 * 	Reset the bytes and packet count of a network device and clear the
 * 	software flow control OFF bit for this network device
 */
3082 3083 3084
static inline void netdev_reset_queue(struct net_device *dev_queue)
{
	netdev_tx_reset_queue(netdev_get_tx_queue(dev_queue, 0));
3085 3086
}

3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106
/**
 * 	netdev_cap_txqueue - check if selected tx queue exceeds device queues
 * 	@dev: network device
 * 	@queue_index: given tx queue index
 *
 * 	Returns 0 if given tx queue index >= number of device tx queues,
 * 	otherwise returns the originally passed tx queue index.
 */
static inline u16 netdev_cap_txqueue(struct net_device *dev, u16 queue_index)
{
	if (unlikely(queue_index >= dev->real_num_tx_queues)) {
		net_warn_ratelimited("%s selects TX queue %d, but real number of TX queues is %d\n",
				     dev->name, queue_index,
				     dev->real_num_tx_queues);
		return 0;
	}

	return queue_index;
}

3107 3108 3109 3110 3111 3112
/**
 *	netif_running - test if up
 *	@dev: network device
 *
 *	Test if the device has been brought up.
 */
3113
static inline bool netif_running(const struct net_device *dev)
L
Linus Torvalds 已提交
3114 3115 3116 3117
{
	return test_bit(__LINK_STATE_START, &dev->state);
}

3118
/*
B
Bjorn Helgaas 已提交
3119
 * Routines to manage the subqueues on a device.  We only need start,
3120 3121 3122 3123
 * stop, and a check if it's stopped.  All other device management is
 * done at the overall netdevice level.
 * Also test the device if we're multiqueue.
 */
3124 3125 3126 3127 3128 3129 3130 3131

/**
 *	netif_start_subqueue - allow sending packets on subqueue
 *	@dev: network device
 *	@queue_index: sub queue index
 *
 * Start individual transmit queue of a device with multiple transmit queues.
 */
3132 3133
static inline void netif_start_subqueue(struct net_device *dev, u16 queue_index)
{
3134
	struct netdev_queue *txq = netdev_get_tx_queue(dev, queue_index);
3135 3136

	netif_tx_start_queue(txq);
3137 3138
}

3139 3140 3141 3142 3143 3144 3145
/**
 *	netif_stop_subqueue - stop sending packets on subqueue
 *	@dev: network device
 *	@queue_index: sub queue index
 *
 * Stop individual transmit queue of a device with multiple transmit queues.
 */
3146 3147
static inline void netif_stop_subqueue(struct net_device *dev, u16 queue_index)
{
3148
	struct netdev_queue *txq = netdev_get_tx_queue(dev, queue_index);
3149
	netif_tx_stop_queue(txq);
3150 3151
}

3152 3153 3154 3155 3156 3157 3158
/**
 *	netif_subqueue_stopped - test status of subqueue
 *	@dev: network device
 *	@queue_index: sub queue index
 *
 * Check individual transmit queue of a device with multiple transmit queues.
 */
3159 3160
static inline bool __netif_subqueue_stopped(const struct net_device *dev,
					    u16 queue_index)
3161
{
3162
	struct netdev_queue *txq = netdev_get_tx_queue(dev, queue_index);
3163 3164

	return netif_tx_queue_stopped(txq);
3165 3166
}

3167 3168
static inline bool netif_subqueue_stopped(const struct net_device *dev,
					  struct sk_buff *skb)
3169 3170 3171
{
	return __netif_subqueue_stopped(dev, skb_get_queue_mapping(skb));
}
3172

3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185
/**
 *	netif_wake_subqueue - allow sending packets on subqueue
 *	@dev: network device
 *	@queue_index: sub queue index
 *
 * Resume individual transmit queue of a device with multiple transmit queues.
 */
static inline void netif_wake_subqueue(struct net_device *dev, u16 queue_index)
{
	struct netdev_queue *txq = netdev_get_tx_queue(dev, queue_index);

	netif_tx_wake_queue(txq);
}
3186

3187
#ifdef CONFIG_XPS
3188
int netif_set_xps_queue(struct net_device *dev, const struct cpumask *mask,
3189
			u16 index);
3190 3191
#else
static inline int netif_set_xps_queue(struct net_device *dev,
3192
				      const struct cpumask *mask,
3193 3194 3195 3196 3197 3198
				      u16 index)
{
	return 0;
}
#endif

J
Jiri Pirko 已提交
3199 3200 3201
u16 __skb_tx_hash(const struct net_device *dev, struct sk_buff *skb,
		  unsigned int num_tx_queues);

3202 3203 3204 3205 3206
/*
 * Returns a Tx hash for the given packet when dev->real_num_tx_queues is used
 * as a distribution range limit for the returned value.
 */
static inline u16 skb_tx_hash(const struct net_device *dev,
3207
			      struct sk_buff *skb)
3208 3209 3210 3211
{
	return __skb_tx_hash(dev, skb, dev->real_num_tx_queues);
}

3212 3213 3214 3215 3216 3217
/**
 *	netif_is_multiqueue - test if device has multiple transmit queues
 *	@dev: network device
 *
 * Check if device has multiple transmit queues
 */
3218
static inline bool netif_is_multiqueue(const struct net_device *dev)
3219
{
E
Eric Dumazet 已提交
3220
	return dev->num_tx_queues > 1;
3221
}
L
Linus Torvalds 已提交
3222

3223
int netif_set_real_num_tx_queues(struct net_device *dev, unsigned int txq);
3224

3225
#ifdef CONFIG_SYSFS
3226
int netif_set_real_num_rx_queues(struct net_device *dev, unsigned int rxq);
3227 3228 3229 3230 3231 3232 3233 3234
#else
static inline int netif_set_real_num_rx_queues(struct net_device *dev,
						unsigned int rxq)
{
	return 0;
}
#endif

3235 3236 3237 3238 3239 3240
static inline struct netdev_rx_queue *
__netif_get_rx_queue(struct net_device *dev, unsigned int rxq)
{
	return dev->_rx + rxq;
}

3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252
#ifdef CONFIG_SYSFS
static inline unsigned int get_netdev_rx_queue_index(
		struct netdev_rx_queue *queue)
{
	struct net_device *dev = queue->dev;
	int index = queue - dev->_rx;

	BUG_ON(index >= dev->num_rx_queues);
	return index;
}
#endif

3253
#define DEFAULT_MAX_NUM_RSS_QUEUES	(8)
3254
int netif_get_num_default_rss_queues(void);
3255

3256 3257 3258 3259 3260 3261 3262
enum skb_free_reason {
	SKB_REASON_CONSUMED,
	SKB_REASON_DROPPED,
};

void __dev_kfree_skb_irq(struct sk_buff *skb, enum skb_free_reason reason);
void __dev_kfree_skb_any(struct sk_buff *skb, enum skb_free_reason reason);
L
Linus Torvalds 已提交
3263

3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280 3281
/*
 * It is not allowed to call kfree_skb() or consume_skb() from hardware
 * interrupt context or with hardware interrupts being disabled.
 * (in_irq() || irqs_disabled())
 *
 * We provide four helpers that can be used in following contexts :
 *
 * dev_kfree_skb_irq(skb) when caller drops a packet from irq context,
 *  replacing kfree_skb(skb)
 *
 * dev_consume_skb_irq(skb) when caller consumes a packet from irq context.
 *  Typically used in place of consume_skb(skb) in TX completion path
 *
 * dev_kfree_skb_any(skb) when caller doesn't know its current irq context,
 *  replacing kfree_skb(skb)
 *
 * dev_consume_skb_any(skb) when caller doesn't know its current irq context,
 *  and consumed a packet. Used in place of consume_skb(skb)
L
Linus Torvalds 已提交
3282
 */
3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301
static inline void dev_kfree_skb_irq(struct sk_buff *skb)
{
	__dev_kfree_skb_irq(skb, SKB_REASON_DROPPED);
}

static inline void dev_consume_skb_irq(struct sk_buff *skb)
{
	__dev_kfree_skb_irq(skb, SKB_REASON_CONSUMED);
}

static inline void dev_kfree_skb_any(struct sk_buff *skb)
{
	__dev_kfree_skb_any(skb, SKB_REASON_DROPPED);
}

static inline void dev_consume_skb_any(struct sk_buff *skb)
{
	__dev_kfree_skb_any(skb, SKB_REASON_CONSUMED);
}
L
Linus Torvalds 已提交
3302

3303 3304
void generic_xdp_tx(struct sk_buff *skb, struct bpf_prog *xdp_prog);
int do_xdp_generic(struct bpf_prog *xdp_prog, struct sk_buff *skb);
3305 3306
int netif_rx(struct sk_buff *skb);
int netif_rx_ni(struct sk_buff *skb);
3307
int netif_receive_skb(struct sk_buff *skb);
3308
int netif_receive_skb_core(struct sk_buff *skb);
3309 3310 3311 3312
gro_result_t napi_gro_receive(struct napi_struct *napi, struct sk_buff *skb);
void napi_gro_flush(struct napi_struct *napi, bool flush_old);
struct sk_buff *napi_get_frags(struct napi_struct *napi);
gro_result_t napi_gro_frags(struct napi_struct *napi);
3313 3314
struct packet_offload *gro_find_receive_by_type(__be16 type);
struct packet_offload *gro_find_complete_by_type(__be16 type);
3315 3316 3317 3318 3319 3320 3321

static inline void napi_free_frags(struct napi_struct *napi)
{
	kfree_skb(napi->skb);
	napi->skb = NULL;
}

M
Mahesh Bandewar 已提交
3322
bool netdev_is_rx_handler_busy(struct net_device *dev);
3323 3324 3325 3326 3327 3328
int netdev_rx_handler_register(struct net_device *dev,
			       rx_handler_func_t *rx_handler,
			       void *rx_handler_data);
void netdev_rx_handler_unregister(struct net_device *dev);

bool dev_valid_name(const char *name);
3329 3330
int dev_ioctl(struct net *net, unsigned int cmd, struct ifreq *ifr,
		bool *need_copyout);
3331
int dev_ifconf(struct net *net, struct ifconf *, int);
3332 3333 3334 3335
int dev_ethtool(struct net *net, struct ifreq *);
unsigned int dev_get_flags(const struct net_device *);
int __dev_change_flags(struct net_device *, unsigned int flags);
int dev_change_flags(struct net_device *, unsigned int);
3336 3337
void __dev_notify_flags(struct net_device *, unsigned int old_flags,
			unsigned int gchanges);
3338 3339
int dev_change_name(struct net_device *, const char *);
int dev_set_alias(struct net_device *, const char *, size_t);
3340
int dev_get_alias(const struct net_device *, char *, size_t);
3341
int dev_change_net_namespace(struct net_device *, struct net *, const char *);
3342
int __dev_set_mtu(struct net_device *, int);
3343
int dev_set_mtu(struct net_device *, int);
3344
int dev_change_tx_queue_len(struct net_device *, unsigned long);
3345 3346 3347 3348
void dev_set_group(struct net_device *, int);
int dev_set_mac_address(struct net_device *, struct sockaddr *);
int dev_change_carrier(struct net_device *, bool new_carrier);
int dev_get_phys_port_id(struct net_device *dev,
3349
			 struct netdev_phys_item_id *ppid);
3350 3351
int dev_get_phys_port_name(struct net_device *dev,
			   char *name, size_t len);
3352
int dev_change_proto_down(struct net_device *dev, bool proto_down);
3353
struct sk_buff *validate_xmit_skb_list(struct sk_buff *skb, struct net_device *dev, bool *again);
3354 3355
struct sk_buff *dev_hard_start_xmit(struct sk_buff *skb, struct net_device *dev,
				    struct netdev_queue *txq, int *ret);
3356

3357
typedef int (*bpf_op_t)(struct net_device *dev, struct netdev_bpf *bpf);
3358 3359
int dev_change_xdp_fd(struct net_device *dev, struct netlink_ext_ack *extack,
		      int fd, u32 flags);
3360 3361
void __dev_xdp_query(struct net_device *dev, bpf_op_t xdp_op,
		     struct netdev_bpf *xdp);
3362

H
Herbert Xu 已提交
3363
int __dev_forward_skb(struct net_device *dev, struct sk_buff *skb);
3364
int dev_forward_skb(struct net_device *dev, struct sk_buff *skb);
3365 3366
bool is_skb_forwardable(const struct net_device *dev,
			const struct sk_buff *skb);
L
Linus Torvalds 已提交
3367

3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382
static __always_inline int ____dev_forward_skb(struct net_device *dev,
					       struct sk_buff *skb)
{
	if (skb_orphan_frags(skb, GFP_ATOMIC) ||
	    unlikely(!is_skb_forwardable(dev, skb))) {
		atomic_long_inc(&dev->rx_dropped);
		kfree_skb(skb);
		return NET_RX_DROP;
	}

	skb_scrub_packet(skb, true);
	skb->priority = 0;
	return 0;
}

3383 3384
void dev_queue_xmit_nit(struct sk_buff *skb, struct net_device *dev);

3385
extern int		netdev_budget;
3386
extern unsigned int	netdev_budget_usecs;
L
Linus Torvalds 已提交
3387 3388

/* Called by rtnetlink.c:rtnl_unlock() */
3389
void netdev_run_todo(void);
L
Linus Torvalds 已提交
3390

3391 3392 3393 3394
/**
 *	dev_put - release reference to device
 *	@dev: network device
 *
3395
 * Release reference to device to allow it to be freed.
3396
 */
L
Linus Torvalds 已提交
3397 3398
static inline void dev_put(struct net_device *dev)
{
3399
	this_cpu_dec(*dev->pcpu_refcnt);
L
Linus Torvalds 已提交
3400 3401
}

3402 3403 3404 3405
/**
 *	dev_hold - get reference to device
 *	@dev: network device
 *
3406
 * Hold reference to device to keep it from being freed.
3407
 */
3408 3409
static inline void dev_hold(struct net_device *dev)
{
3410
	this_cpu_inc(*dev->pcpu_refcnt);
3411
}
L
Linus Torvalds 已提交
3412 3413 3414 3415

/* Carrier loss detection, dial on demand. The functions netif_carrier_on
 * and _off may be called from IRQ context, but it is caller
 * who is responsible for serialization of these calls.
S
Stefan Rompf 已提交
3416 3417 3418 3419
 *
 * The name carrier is inappropriate, these functions should really be
 * called netif_lowerlayer_*() because they represent the state of any
 * kind of lower layer not just hardware media.
L
Linus Torvalds 已提交
3420 3421
 */

3422 3423 3424
void linkwatch_init_dev(struct net_device *dev);
void linkwatch_fire_event(struct net_device *dev);
void linkwatch_forget_dev(struct net_device *dev);
L
Linus Torvalds 已提交
3425

3426 3427 3428 3429 3430 3431
/**
 *	netif_carrier_ok - test if carrier present
 *	@dev: network device
 *
 * Check if carrier is present on device
 */
3432
static inline bool netif_carrier_ok(const struct net_device *dev)
L
Linus Torvalds 已提交
3433 3434 3435 3436
{
	return !test_bit(__LINK_STATE_NOCARRIER, &dev->state);
}

3437
unsigned long dev_trans_start(struct net_device *dev);
3438

3439
void __netdev_watchdog_up(struct net_device *dev);
L
Linus Torvalds 已提交
3440

3441
void netif_carrier_on(struct net_device *dev);
L
Linus Torvalds 已提交
3442

3443
void netif_carrier_off(struct net_device *dev);
L
Linus Torvalds 已提交
3444

3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456
/**
 *	netif_dormant_on - mark device as dormant.
 *	@dev: network device
 *
 * Mark device as dormant (as per RFC2863).
 *
 * The dormant state indicates that the relevant interface is not
 * actually in a condition to pass packets (i.e., it is not 'up') but is
 * in a "pending" state, waiting for some external event.  For "on-
 * demand" interfaces, this new state identifies the situation where the
 * interface is waiting for events to place it in the up state.
 */
S
Stefan Rompf 已提交
3457 3458 3459 3460 3461 3462
static inline void netif_dormant_on(struct net_device *dev)
{
	if (!test_and_set_bit(__LINK_STATE_DORMANT, &dev->state))
		linkwatch_fire_event(dev);
}

3463 3464 3465 3466 3467 3468
/**
 *	netif_dormant_off - set device as not dormant.
 *	@dev: network device
 *
 * Device is not in dormant state.
 */
S
Stefan Rompf 已提交
3469 3470 3471 3472 3473 3474
static inline void netif_dormant_off(struct net_device *dev)
{
	if (test_and_clear_bit(__LINK_STATE_DORMANT, &dev->state))
		linkwatch_fire_event(dev);
}

3475
/**
3476
 *	netif_dormant - test if device is dormant
3477 3478
 *	@dev: network device
 *
3479
 * Check if device is dormant.
3480
 */
3481
static inline bool netif_dormant(const struct net_device *dev)
S
Stefan Rompf 已提交
3482 3483 3484 3485 3486
{
	return test_bit(__LINK_STATE_DORMANT, &dev->state);
}


3487 3488 3489 3490 3491 3492
/**
 *	netif_oper_up - test if device is operational
 *	@dev: network device
 *
 * Check if carrier is operational
 */
3493
static inline bool netif_oper_up(const struct net_device *dev)
E
Eric Dumazet 已提交
3494
{
S
Stefan Rompf 已提交
3495 3496 3497 3498
	return (dev->operstate == IF_OPER_UP ||
		dev->operstate == IF_OPER_UNKNOWN /* backward compat */);
}

3499 3500 3501 3502 3503 3504
/**
 *	netif_device_present - is device available or removed
 *	@dev: network device
 *
 * Check if device has not been removed from system.
 */
3505
static inline bool netif_device_present(struct net_device *dev)
L
Linus Torvalds 已提交
3506 3507 3508 3509
{
	return test_bit(__LINK_STATE_PRESENT, &dev->state);
}

3510
void netif_device_detach(struct net_device *dev);
L
Linus Torvalds 已提交
3511

3512
void netif_device_attach(struct net_device *dev);
L
Linus Torvalds 已提交
3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523 3524 3525 3526 3527 3528 3529 3530 3531 3532 3533 3534 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

/*
 * Network interface message level settings
 */

enum {
	NETIF_MSG_DRV		= 0x0001,
	NETIF_MSG_PROBE		= 0x0002,
	NETIF_MSG_LINK		= 0x0004,
	NETIF_MSG_TIMER		= 0x0008,
	NETIF_MSG_IFDOWN	= 0x0010,
	NETIF_MSG_IFUP		= 0x0020,
	NETIF_MSG_RX_ERR	= 0x0040,
	NETIF_MSG_TX_ERR	= 0x0080,
	NETIF_MSG_TX_QUEUED	= 0x0100,
	NETIF_MSG_INTR		= 0x0200,
	NETIF_MSG_TX_DONE	= 0x0400,
	NETIF_MSG_RX_STATUS	= 0x0800,
	NETIF_MSG_PKTDATA	= 0x1000,
	NETIF_MSG_HW		= 0x2000,
	NETIF_MSG_WOL		= 0x4000,
};

#define netif_msg_drv(p)	((p)->msg_enable & NETIF_MSG_DRV)
#define netif_msg_probe(p)	((p)->msg_enable & NETIF_MSG_PROBE)
#define netif_msg_link(p)	((p)->msg_enable & NETIF_MSG_LINK)
#define netif_msg_timer(p)	((p)->msg_enable & NETIF_MSG_TIMER)
#define netif_msg_ifdown(p)	((p)->msg_enable & NETIF_MSG_IFDOWN)
#define netif_msg_ifup(p)	((p)->msg_enable & NETIF_MSG_IFUP)
#define netif_msg_rx_err(p)	((p)->msg_enable & NETIF_MSG_RX_ERR)
#define netif_msg_tx_err(p)	((p)->msg_enable & NETIF_MSG_TX_ERR)
#define netif_msg_tx_queued(p)	((p)->msg_enable & NETIF_MSG_TX_QUEUED)
#define netif_msg_intr(p)	((p)->msg_enable & NETIF_MSG_INTR)
#define netif_msg_tx_done(p)	((p)->msg_enable & NETIF_MSG_TX_DONE)
#define netif_msg_rx_status(p)	((p)->msg_enable & NETIF_MSG_RX_STATUS)
#define netif_msg_pktdata(p)	((p)->msg_enable & NETIF_MSG_PKTDATA)
#define netif_msg_hw(p)		((p)->msg_enable & NETIF_MSG_HW)
#define netif_msg_wol(p)	((p)->msg_enable & NETIF_MSG_WOL)

static inline u32 netif_msg_init(int debug_value, int default_msg_enable_bits)
{
	/* use default */
	if (debug_value < 0 || debug_value >= (sizeof(u32) * 8))
		return default_msg_enable_bits;
	if (debug_value == 0)	/* no output */
		return 0;
	/* set low N bits */
	return (1 << debug_value) - 1;
}

3563
static inline void __netif_tx_lock(struct netdev_queue *txq, int cpu)
H
Herbert Xu 已提交
3564
{
3565 3566
	spin_lock(&txq->_xmit_lock);
	txq->xmit_lock_owner = cpu;
3567 3568
}

3569 3570 3571 3572 3573 3574 3575 3576 3577 3578 3579
static inline bool __netif_tx_acquire(struct netdev_queue *txq)
{
	__acquire(&txq->_xmit_lock);
	return true;
}

static inline void __netif_tx_release(struct netdev_queue *txq)
{
	__release(&txq->_xmit_lock);
}

3580 3581 3582 3583 3584 3585
static inline void __netif_tx_lock_bh(struct netdev_queue *txq)
{
	spin_lock_bh(&txq->_xmit_lock);
	txq->xmit_lock_owner = smp_processor_id();
}

3586
static inline bool __netif_tx_trylock(struct netdev_queue *txq)
3587
{
3588
	bool ok = spin_trylock(&txq->_xmit_lock);
3589 3590 3591 3592 3593 3594 3595 3596 3597 3598 3599 3600 3601 3602 3603 3604 3605
	if (likely(ok))
		txq->xmit_lock_owner = smp_processor_id();
	return ok;
}

static inline void __netif_tx_unlock(struct netdev_queue *txq)
{
	txq->xmit_lock_owner = -1;
	spin_unlock(&txq->_xmit_lock);
}

static inline void __netif_tx_unlock_bh(struct netdev_queue *txq)
{
	txq->xmit_lock_owner = -1;
	spin_unlock_bh(&txq->_xmit_lock);
}

E
Eric Dumazet 已提交
3606 3607 3608 3609 3610 3611
static inline void txq_trans_update(struct netdev_queue *txq)
{
	if (txq->xmit_lock_owner != -1)
		txq->trans_start = jiffies;
}

3612 3613 3614
/* legacy drivers only, netdev_start_xmit() sets txq->trans_start */
static inline void netif_trans_update(struct net_device *dev)
{
F
Florian Westphal 已提交
3615 3616 3617 3618
	struct netdev_queue *txq = netdev_get_tx_queue(dev, 0);

	if (txq->trans_start != jiffies)
		txq->trans_start = jiffies;
3619 3620
}

3621 3622 3623 3624 3625 3626
/**
 *	netif_tx_lock - grab network device transmit lock
 *	@dev: network device
 *
 * Get network device transmit lock
 */
3627 3628
static inline void netif_tx_lock(struct net_device *dev)
{
3629
	unsigned int i;
3630
	int cpu;
3631

3632 3633
	spin_lock(&dev->tx_global_lock);
	cpu = smp_processor_id();
3634 3635
	for (i = 0; i < dev->num_tx_queues; i++) {
		struct netdev_queue *txq = netdev_get_tx_queue(dev, i);
3636 3637 3638 3639 3640 3641 3642

		/* We are the only thread of execution doing a
		 * freeze, but we have to grab the _xmit_lock in
		 * order to synchronize with threads which are in
		 * the ->hard_start_xmit() handler and already
		 * checked the frozen bit.
		 */
3643
		__netif_tx_lock(txq, cpu);
3644 3645
		set_bit(__QUEUE_STATE_FROZEN, &txq->state);
		__netif_tx_unlock(txq);
3646
	}
H
Herbert Xu 已提交
3647 3648 3649 3650
}

static inline void netif_tx_lock_bh(struct net_device *dev)
{
3651 3652
	local_bh_disable();
	netif_tx_lock(dev);
H
Herbert Xu 已提交
3653 3654 3655 3656
}

static inline void netif_tx_unlock(struct net_device *dev)
{
3657 3658 3659 3660
	unsigned int i;

	for (i = 0; i < dev->num_tx_queues; i++) {
		struct netdev_queue *txq = netdev_get_tx_queue(dev, i);
3661

3662 3663 3664 3665 3666
		/* No need to grab the _xmit_lock here.  If the
		 * queue is not stopped for another reason, we
		 * force a schedule.
		 */
		clear_bit(__QUEUE_STATE_FROZEN, &txq->state);
3667
		netif_schedule_queue(txq);
3668 3669
	}
	spin_unlock(&dev->tx_global_lock);
H
Herbert Xu 已提交
3670 3671 3672 3673
}

static inline void netif_tx_unlock_bh(struct net_device *dev)
{
3674 3675
	netif_tx_unlock(dev);
	local_bh_enable();
H
Herbert Xu 已提交
3676 3677
}

3678
#define HARD_TX_LOCK(dev, txq, cpu) {			\
3679
	if ((dev->features & NETIF_F_LLTX) == 0) {	\
3680
		__netif_tx_lock(txq, cpu);		\
3681 3682
	} else {					\
		__netif_tx_acquire(txq);		\
3683 3684 3685
	}						\
}

3686 3687 3688
#define HARD_TX_TRYLOCK(dev, txq)			\
	(((dev->features & NETIF_F_LLTX) == 0) ?	\
		__netif_tx_trylock(txq) :		\
3689
		__netif_tx_acquire(txq))
3690

3691
#define HARD_TX_UNLOCK(dev, txq) {			\
3692
	if ((dev->features & NETIF_F_LLTX) == 0) {	\
3693
		__netif_tx_unlock(txq);			\
3694 3695
	} else {					\
		__netif_tx_release(txq);		\
3696 3697 3698
	}						\
}

L
Linus Torvalds 已提交
3699 3700
static inline void netif_tx_disable(struct net_device *dev)
{
3701
	unsigned int i;
3702
	int cpu;
3703

3704 3705
	local_bh_disable();
	cpu = smp_processor_id();
3706 3707
	for (i = 0; i < dev->num_tx_queues; i++) {
		struct netdev_queue *txq = netdev_get_tx_queue(dev, i);
3708 3709

		__netif_tx_lock(txq, cpu);
3710
		netif_tx_stop_queue(txq);
3711
		__netif_tx_unlock(txq);
3712
	}
3713
	local_bh_enable();
L
Linus Torvalds 已提交
3714 3715
}

3716 3717 3718 3719 3720
static inline void netif_addr_lock(struct net_device *dev)
{
	spin_lock(&dev->addr_list_lock);
}

3721 3722
static inline void netif_addr_lock_nested(struct net_device *dev)
{
3723 3724 3725 3726 3727 3728
	int subclass = SINGLE_DEPTH_NESTING;

	if (dev->netdev_ops->ndo_get_lock_subclass)
		subclass = dev->netdev_ops->ndo_get_lock_subclass(dev);

	spin_lock_nested(&dev->addr_list_lock, subclass);
3729 3730
}

3731 3732 3733 3734 3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745
static inline void netif_addr_lock_bh(struct net_device *dev)
{
	spin_lock_bh(&dev->addr_list_lock);
}

static inline void netif_addr_unlock(struct net_device *dev)
{
	spin_unlock(&dev->addr_list_lock);
}

static inline void netif_addr_unlock_bh(struct net_device *dev)
{
	spin_unlock_bh(&dev->addr_list_lock);
}

3746
/*
3747
 * dev_addrs walker. Should be used only for read access. Call with
3748 3749 3750
 * rcu_read_lock held.
 */
#define for_each_dev_addr(dev, ha) \
3751
		list_for_each_entry_rcu(ha, &dev->dev_addrs.list, list)
3752

L
Linus Torvalds 已提交
3753 3754
/* These functions live elsewhere (drivers/net/net_init.c, but related) */

3755
void ether_setup(struct net_device *dev);
L
Linus Torvalds 已提交
3756 3757

/* Support for loadable net-drivers */
3758
struct net_device *alloc_netdev_mqs(int sizeof_priv, const char *name,
3759
				    unsigned char name_assign_type,
3760 3761
				    void (*setup)(struct net_device *),
				    unsigned int txqs, unsigned int rxqs);
3762 3763 3764
int dev_get_valid_name(struct net *net, struct net_device *dev,
		       const char *name);

3765 3766
#define alloc_netdev(sizeof_priv, name, name_assign_type, setup) \
	alloc_netdev_mqs(sizeof_priv, name, name_assign_type, setup, 1, 1)
T
Tom Herbert 已提交
3767

3768 3769 3770
#define alloc_netdev_mq(sizeof_priv, name, name_assign_type, setup, count) \
	alloc_netdev_mqs(sizeof_priv, name, name_assign_type, setup, count, \
			 count)
T
Tom Herbert 已提交
3771

3772 3773
int register_netdev(struct net_device *dev);
void unregister_netdev(struct net_device *dev);
3774

3775
/* General hardware address lists handling functions */
3776 3777 3778 3779
int __hw_addr_sync(struct netdev_hw_addr_list *to_list,
		   struct netdev_hw_addr_list *from_list, int addr_len);
void __hw_addr_unsync(struct netdev_hw_addr_list *to_list,
		      struct netdev_hw_addr_list *from_list, int addr_len);
3780 3781 3782 3783 3784 3785 3786 3787 3788
int __hw_addr_sync_dev(struct netdev_hw_addr_list *list,
		       struct net_device *dev,
		       int (*sync)(struct net_device *, const unsigned char *),
		       int (*unsync)(struct net_device *,
				     const unsigned char *));
void __hw_addr_unsync_dev(struct netdev_hw_addr_list *list,
			  struct net_device *dev,
			  int (*unsync)(struct net_device *,
					const unsigned char *));
3789
void __hw_addr_init(struct netdev_hw_addr_list *list);
3790

3791
/* Functions used for device addresses handling */
3792 3793 3794 3795 3796 3797
int dev_addr_add(struct net_device *dev, const unsigned char *addr,
		 unsigned char addr_type);
int dev_addr_del(struct net_device *dev, const unsigned char *addr,
		 unsigned char addr_type);
void dev_addr_flush(struct net_device *dev);
int dev_addr_init(struct net_device *dev);
3798 3799

/* Functions used for unicast addresses handling */
3800 3801 3802 3803 3804 3805 3806 3807
int dev_uc_add(struct net_device *dev, const unsigned char *addr);
int dev_uc_add_excl(struct net_device *dev, const unsigned char *addr);
int dev_uc_del(struct net_device *dev, const unsigned char *addr);
int dev_uc_sync(struct net_device *to, struct net_device *from);
int dev_uc_sync_multiple(struct net_device *to, struct net_device *from);
void dev_uc_unsync(struct net_device *to, struct net_device *from);
void dev_uc_flush(struct net_device *dev);
void dev_uc_init(struct net_device *dev);
3808

3809 3810 3811 3812 3813 3814 3815 3816
/**
 *  __dev_uc_sync - Synchonize device's unicast list
 *  @dev:  device to sync
 *  @sync: function to call if address should be added
 *  @unsync: function to call if address should be removed
 *
 *  Add newly added addresses to the interface, and release
 *  addresses that have been deleted.
B
Bjorn Helgaas 已提交
3817
 */
3818 3819 3820 3821 3822 3823 3824 3825 3826 3827
static inline int __dev_uc_sync(struct net_device *dev,
				int (*sync)(struct net_device *,
					    const unsigned char *),
				int (*unsync)(struct net_device *,
					      const unsigned char *))
{
	return __hw_addr_sync_dev(&dev->uc, dev, sync, unsync);
}

/**
3828
 *  __dev_uc_unsync - Remove synchronized addresses from device
3829 3830 3831 3832
 *  @dev:  device to sync
 *  @unsync: function to call if address should be removed
 *
 *  Remove all addresses that were added to the device by dev_uc_sync().
B
Bjorn Helgaas 已提交
3833
 */
3834 3835 3836 3837 3838 3839 3840
static inline void __dev_uc_unsync(struct net_device *dev,
				   int (*unsync)(struct net_device *,
						 const unsigned char *))
{
	__hw_addr_unsync_dev(&dev->uc, dev, unsync);
}

3841
/* Functions used for multicast addresses handling */
3842 3843 3844 3845 3846 3847 3848 3849 3850 3851
int dev_mc_add(struct net_device *dev, const unsigned char *addr);
int dev_mc_add_global(struct net_device *dev, const unsigned char *addr);
int dev_mc_add_excl(struct net_device *dev, const unsigned char *addr);
int dev_mc_del(struct net_device *dev, const unsigned char *addr);
int dev_mc_del_global(struct net_device *dev, const unsigned char *addr);
int dev_mc_sync(struct net_device *to, struct net_device *from);
int dev_mc_sync_multiple(struct net_device *to, struct net_device *from);
void dev_mc_unsync(struct net_device *to, struct net_device *from);
void dev_mc_flush(struct net_device *dev);
void dev_mc_init(struct net_device *dev);
3852

3853 3854 3855 3856 3857 3858 3859 3860
/**
 *  __dev_mc_sync - Synchonize device's multicast list
 *  @dev:  device to sync
 *  @sync: function to call if address should be added
 *  @unsync: function to call if address should be removed
 *
 *  Add newly added addresses to the interface, and release
 *  addresses that have been deleted.
B
Bjorn Helgaas 已提交
3861
 */
3862 3863 3864 3865 3866 3867 3868 3869 3870 3871
static inline int __dev_mc_sync(struct net_device *dev,
				int (*sync)(struct net_device *,
					    const unsigned char *),
				int (*unsync)(struct net_device *,
					      const unsigned char *))
{
	return __hw_addr_sync_dev(&dev->mc, dev, sync, unsync);
}

/**
3872
 *  __dev_mc_unsync - Remove synchronized addresses from device
3873 3874 3875 3876
 *  @dev:  device to sync
 *  @unsync: function to call if address should be removed
 *
 *  Remove all addresses that were added to the device by dev_mc_sync().
B
Bjorn Helgaas 已提交
3877
 */
3878 3879 3880 3881 3882 3883 3884
static inline void __dev_mc_unsync(struct net_device *dev,
				   int (*unsync)(struct net_device *,
						 const unsigned char *))
{
	__hw_addr_unsync_dev(&dev->mc, dev, unsync);
}

3885
/* Functions used for secondary unicast and multicast support */
3886 3887 3888 3889 3890 3891 3892
void dev_set_rx_mode(struct net_device *dev);
void __dev_set_rx_mode(struct net_device *dev);
int dev_set_promiscuity(struct net_device *dev, int inc);
int dev_set_allmulti(struct net_device *dev, int inc);
void netdev_state_change(struct net_device *dev);
void netdev_notify_peers(struct net_device *dev);
void netdev_features_change(struct net_device *dev);
L
Linus Torvalds 已提交
3893
/* Load a device via the kmod */
3894 3895 3896 3897 3898
void dev_load(struct net *net, const char *name);
struct rtnl_link_stats64 *dev_get_stats(struct net_device *dev,
					struct rtnl_link_stats64 *storage);
void netdev_stats_to_stats64(struct rtnl_link_stats64 *stats64,
			     const struct net_device_stats *netdev_stats);
3899

L
Linus Torvalds 已提交
3900
extern int		netdev_max_backlog;
E
Eric Dumazet 已提交
3901
extern int		netdev_tstamp_prequeue;
L
Linus Torvalds 已提交
3902
extern int		weight_p;
3903 3904 3905 3906
extern int		dev_weight_rx_bias;
extern int		dev_weight_tx_bias;
extern int		dev_rx_weight;
extern int		dev_tx_weight;
J
Jiri Pirko 已提交
3907

3908
bool netdev_has_upper_dev(struct net_device *dev, struct net_device *upper_dev);
3909 3910
struct net_device *netdev_upper_get_next_dev_rcu(struct net_device *dev,
						     struct list_head **iter);
3911 3912
struct net_device *netdev_all_upper_get_next_dev_rcu(struct net_device *dev,
						     struct list_head **iter);
3913

3914 3915 3916 3917 3918 3919 3920
/* iterate through upper list, must be called under RCU read lock */
#define netdev_for_each_upper_dev_rcu(dev, updev, iter) \
	for (iter = &(dev)->adj_list.upper, \
	     updev = netdev_upper_get_next_dev_rcu(dev, &(iter)); \
	     updev; \
	     updev = netdev_upper_get_next_dev_rcu(dev, &(iter)))

3921 3922 3923 3924 3925 3926 3927 3928
int netdev_walk_all_upper_dev_rcu(struct net_device *dev,
				  int (*fn)(struct net_device *upper_dev,
					    void *data),
				  void *data);

bool netdev_has_upper_dev_all_rcu(struct net_device *dev,
				  struct net_device *upper_dev);

3929 3930
bool netdev_has_any_upper_dev(struct net_device *dev);

3931 3932 3933 3934
void *netdev_lower_get_next_private(struct net_device *dev,
				    struct list_head **iter);
void *netdev_lower_get_next_private_rcu(struct net_device *dev,
					struct list_head **iter);
3935 3936 3937 3938 3939 3940 3941 3942 3943 3944 3945 3946 3947

#define netdev_for_each_lower_private(dev, priv, iter) \
	for (iter = (dev)->adj_list.lower.next, \
	     priv = netdev_lower_get_next_private(dev, &(iter)); \
	     priv; \
	     priv = netdev_lower_get_next_private(dev, &(iter)))

#define netdev_for_each_lower_private_rcu(dev, priv, iter) \
	for (iter = &(dev)->adj_list.lower, \
	     priv = netdev_lower_get_next_private_rcu(dev, &(iter)); \
	     priv; \
	     priv = netdev_lower_get_next_private_rcu(dev, &(iter)))

3948 3949
void *netdev_lower_get_next(struct net_device *dev,
				struct list_head **iter);
3950

3951
#define netdev_for_each_lower_dev(dev, ldev, iter) \
3952
	for (iter = (dev)->adj_list.lower.next, \
3953 3954 3955 3956
	     ldev = netdev_lower_get_next(dev, &(iter)); \
	     ldev; \
	     ldev = netdev_lower_get_next(dev, &(iter)))

3957 3958 3959 3960 3961
struct net_device *netdev_all_lower_get_next(struct net_device *dev,
					     struct list_head **iter);
struct net_device *netdev_all_lower_get_next_rcu(struct net_device *dev,
						 struct list_head **iter);

3962 3963 3964 3965 3966 3967 3968 3969 3970
int netdev_walk_all_lower_dev(struct net_device *dev,
			      int (*fn)(struct net_device *lower_dev,
					void *data),
			      void *data);
int netdev_walk_all_lower_dev_rcu(struct net_device *dev,
				  int (*fn)(struct net_device *lower_dev,
					    void *data),
				  void *data);

3971
void *netdev_adjacent_get_private(struct list_head *adj_list);
3972
void *netdev_lower_get_first_private_rcu(struct net_device *dev);
3973 3974
struct net_device *netdev_master_upper_dev_get(struct net_device *dev);
struct net_device *netdev_master_upper_dev_get_rcu(struct net_device *dev);
3975 3976
int netdev_upper_dev_link(struct net_device *dev, struct net_device *upper_dev,
			  struct netlink_ext_ack *extack);
3977
int netdev_master_upper_dev_link(struct net_device *dev,
3978
				 struct net_device *upper_dev,
3979 3980
				 void *upper_priv, void *upper_info,
				 struct netlink_ext_ack *extack);
3981 3982
void netdev_upper_dev_unlink(struct net_device *dev,
			     struct net_device *upper_dev);
3983
void netdev_adjacent_rename_links(struct net_device *dev, char *oldname);
3984 3985
void *netdev_lower_dev_get_private(struct net_device *dev,
				   struct net_device *lower_dev);
3986 3987
void netdev_lower_state_changed(struct net_device *lower_dev,
				void *lower_state_info);
3988 3989 3990

/* RSS keys are 40 or 52 bytes long */
#define NETDEV_RSS_KEY_LEN 52
3991
extern u8 netdev_rss_key[NETDEV_RSS_KEY_LEN] __read_mostly;
3992 3993
void netdev_rss_key_fill(void *buffer, size_t len);

3994
int dev_get_nest_level(struct net_device *dev);
3995
int skb_checksum_help(struct sk_buff *skb);
3996
int skb_crc32c_csum_help(struct sk_buff *skb);
3997 3998 3999
int skb_csum_hwoffload_help(struct sk_buff *skb,
			    const netdev_features_t features);

4000 4001 4002 4003
struct sk_buff *__skb_gso_segment(struct sk_buff *skb,
				  netdev_features_t features, bool tx_path);
struct sk_buff *skb_mac_gso_segment(struct sk_buff *skb,
				    netdev_features_t features);
4004

4005 4006 4007 4008 4009 4010 4011 4012 4013 4014 4015 4016 4017
struct netdev_bonding_info {
	ifslave	slave;
	ifbond	master;
};

struct netdev_notifier_bonding_info {
	struct netdev_notifier_info info; /* must be first */
	struct netdev_bonding_info  bonding_info;
};

void netdev_bonding_info_change(struct net_device *dev,
				struct netdev_bonding_info *bonding_info);

4018 4019 4020 4021 4022
static inline
struct sk_buff *skb_gso_segment(struct sk_buff *skb, netdev_features_t features)
{
	return __skb_gso_segment(skb, features, true);
}
4023
__be16 skb_network_protocol(struct sk_buff *skb, int *depth);
4024 4025 4026 4027

static inline bool can_checksum_protocol(netdev_features_t features,
					 __be16 protocol)
{
4028 4029 4030 4031 4032 4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 4043 4044 4045
	if (protocol == htons(ETH_P_FCOE))
		return !!(features & NETIF_F_FCOE_CRC);

	/* Assume this is an IP checksum (not SCTP CRC) */

	if (features & NETIF_F_HW_CSUM) {
		/* Can checksum everything */
		return true;
	}

	switch (protocol) {
	case htons(ETH_P_IP):
		return !!(features & NETIF_F_IP_CSUM);
	case htons(ETH_P_IPV6):
		return !!(features & NETIF_F_IPV6_CSUM);
	default:
		return false;
	}
4046
}
4047

4048
#ifdef CONFIG_BUG
4049
void netdev_rx_csum_fault(struct net_device *dev);
4050 4051 4052 4053 4054
#else
static inline void netdev_rx_csum_fault(struct net_device *dev)
{
}
#endif
L
Linus Torvalds 已提交
4055
/* rx skb timestamps */
4056 4057
void net_enable_timestamp(void);
void net_disable_timestamp(void);
L
Linus Torvalds 已提交
4058

4059
#ifdef CONFIG_PROC_FS
4060
int __init dev_proc_init(void);
4061 4062
#else
#define dev_proc_init() 0
4063 4064
#endif

D
David S. Miller 已提交
4065
static inline netdev_tx_t __netdev_start_xmit(const struct net_device_ops *ops,
4066 4067
					      struct sk_buff *skb, struct net_device *dev,
					      bool more)
D
David S. Miller 已提交
4068
{
4069
	skb->xmit_more = more ? 1 : 0;
4070
	return ops->ndo_start_xmit(skb, dev);
D
David S. Miller 已提交
4071 4072
}

4073
static inline netdev_tx_t netdev_start_xmit(struct sk_buff *skb, struct net_device *dev,
4074
					    struct netdev_queue *txq, bool more)
D
David S. Miller 已提交
4075 4076
{
	const struct net_device_ops *ops = dev->netdev_ops;
4077
	int rc;
D
David S. Miller 已提交
4078

4079
	rc = __netdev_start_xmit(ops, skb, dev, more);
4080 4081 4082 4083
	if (rc == NETDEV_TX_OK)
		txq_trans_update(txq);

	return rc;
D
David S. Miller 已提交
4084 4085
}

4086
int netdev_class_create_file_ns(const struct class_attribute *class_attr,
4087
				const void *ns);
4088
void netdev_class_remove_file_ns(const struct class_attribute *class_attr,
4089
				 const void *ns);
4090

4091
static inline int netdev_class_create_file(const struct class_attribute *class_attr)
4092 4093 4094 4095
{
	return netdev_class_create_file_ns(class_attr, NULL);
}

4096
static inline void netdev_class_remove_file(const struct class_attribute *class_attr)
4097 4098 4099
{
	netdev_class_remove_file_ns(class_attr, NULL);
}
4100

4101
extern const struct kobj_ns_type_operations net_ns_type_operations;
4102

4103
const char *netdev_drivername(const struct net_device *dev);
4104

4105
void linkwatch_run_queue(void);
4106

4107 4108 4109
static inline netdev_features_t netdev_intersect_features(netdev_features_t f1,
							  netdev_features_t f2)
{
4110 4111
	if ((f1 ^ f2) & NETIF_F_HW_CSUM) {
		if (f1 & NETIF_F_HW_CSUM)
4112
			f1 |= (NETIF_F_IP_CSUM|NETIF_F_IPV6_CSUM);
4113
		else
4114
			f2 |= (NETIF_F_IP_CSUM|NETIF_F_IPV6_CSUM);
4115
	}
4116

4117
	return f1 & f2;
4118 4119
}

4120 4121
static inline netdev_features_t netdev_get_wanted_features(
	struct net_device *dev)
4122 4123 4124
{
	return (dev->features & ~dev->hw_features) | dev->wanted_features;
}
4125 4126
netdev_features_t netdev_increment_features(netdev_features_t all,
	netdev_features_t one, netdev_features_t mask);
4127 4128 4129 4130 4131 4132 4133 4134 4135 4136 4137

/* Allow TSO being used on stacked device :
 * Performing the GSO segmentation before last device
 * is a performance improvement.
 */
static inline netdev_features_t netdev_add_tso_features(netdev_features_t features,
							netdev_features_t mask)
{
	return netdev_increment_features(features, NETIF_F_ALL_TSO, mask);
}

4138
int __netdev_update_features(struct net_device *dev);
4139
void netdev_update_features(struct net_device *dev);
4140
void netdev_change_features(struct net_device *dev);
4141

4142 4143 4144
void netif_stacked_transfer_operstate(const struct net_device *rootdev,
					struct net_device *dev);

4145 4146 4147
netdev_features_t passthru_features_check(struct sk_buff *skb,
					  struct net_device *dev,
					  netdev_features_t features);
4148
netdev_features_t netif_skb_features(struct sk_buff *skb);
4149

4150
static inline bool net_gso_ok(netdev_features_t features, int gso_type)
4151
{
4152
	netdev_features_t feature = (netdev_features_t)gso_type << NETIF_F_GSO_SHIFT;
4153 4154 4155 4156 4157

	/* check flags correspondence */
	BUILD_BUG_ON(SKB_GSO_TCPV4   != (NETIF_F_TSO >> NETIF_F_GSO_SHIFT));
	BUILD_BUG_ON(SKB_GSO_DODGY   != (NETIF_F_GSO_ROBUST >> NETIF_F_GSO_SHIFT));
	BUILD_BUG_ON(SKB_GSO_TCP_ECN != (NETIF_F_TSO_ECN >> NETIF_F_GSO_SHIFT));
4158
	BUILD_BUG_ON(SKB_GSO_TCP_FIXEDID != (NETIF_F_TSO_MANGLEID >> NETIF_F_GSO_SHIFT));
4159 4160
	BUILD_BUG_ON(SKB_GSO_TCPV6   != (NETIF_F_TSO6 >> NETIF_F_GSO_SHIFT));
	BUILD_BUG_ON(SKB_GSO_FCOE    != (NETIF_F_FSO >> NETIF_F_GSO_SHIFT));
4161 4162
	BUILD_BUG_ON(SKB_GSO_GRE     != (NETIF_F_GSO_GRE >> NETIF_F_GSO_SHIFT));
	BUILD_BUG_ON(SKB_GSO_GRE_CSUM != (NETIF_F_GSO_GRE_CSUM >> NETIF_F_GSO_SHIFT));
4163 4164
	BUILD_BUG_ON(SKB_GSO_IPXIP4  != (NETIF_F_GSO_IPXIP4 >> NETIF_F_GSO_SHIFT));
	BUILD_BUG_ON(SKB_GSO_IPXIP6  != (NETIF_F_GSO_IPXIP6 >> NETIF_F_GSO_SHIFT));
4165 4166
	BUILD_BUG_ON(SKB_GSO_UDP_TUNNEL != (NETIF_F_GSO_UDP_TUNNEL >> NETIF_F_GSO_SHIFT));
	BUILD_BUG_ON(SKB_GSO_UDP_TUNNEL_CSUM != (NETIF_F_GSO_UDP_TUNNEL_CSUM >> NETIF_F_GSO_SHIFT));
4167
	BUILD_BUG_ON(SKB_GSO_PARTIAL != (NETIF_F_GSO_PARTIAL >> NETIF_F_GSO_SHIFT));
4168
	BUILD_BUG_ON(SKB_GSO_TUNNEL_REMCSUM != (NETIF_F_GSO_TUNNEL_REMCSUM >> NETIF_F_GSO_SHIFT));
M
Marcelo Ricardo Leitner 已提交
4169
	BUILD_BUG_ON(SKB_GSO_SCTP    != (NETIF_F_GSO_SCTP >> NETIF_F_GSO_SHIFT));
S
Steffen Klassert 已提交
4170
	BUILD_BUG_ON(SKB_GSO_ESP != (NETIF_F_GSO_ESP >> NETIF_F_GSO_SHIFT));
4171
	BUILD_BUG_ON(SKB_GSO_UDP != (NETIF_F_GSO_UDP >> NETIF_F_GSO_SHIFT));
4172

4173
	return (features & feature) == feature;
4174 4175
}

4176
static inline bool skb_gso_ok(struct sk_buff *skb, netdev_features_t features)
4177
{
H
Herbert Xu 已提交
4178
	return net_gso_ok(features, skb_shinfo(skb)->gso_type) &&
4179
	       (!skb_has_frag_list(skb) || (features & NETIF_F_FRAGLIST));
4180 4181
}

4182
static inline bool netif_needs_gso(struct sk_buff *skb,
4183
				   netdev_features_t features)
4184
{
4185
	return skb_is_gso(skb) && (!skb_gso_ok(skb, features) ||
4186 4187
		unlikely((skb->ip_summed != CHECKSUM_PARTIAL) &&
			 (skb->ip_summed != CHECKSUM_UNNECESSARY)));
4188 4189
}

4190 4191 4192 4193 4194 4195
static inline void netif_set_gso_max_size(struct net_device *dev,
					  unsigned int size)
{
	dev->gso_max_size = size;
}

4196 4197 4198 4199 4200 4201 4202 4203 4204 4205 4206 4207 4208
static inline void skb_gso_error_unwind(struct sk_buff *skb, __be16 protocol,
					int pulled_hlen, u16 mac_offset,
					int mac_len)
{
	skb->protocol = protocol;
	skb->encapsulation = 1;
	skb_push(skb, pulled_hlen);
	skb_reset_transport_header(skb);
	skb->mac_header = mac_offset;
	skb->network_header = skb->mac_header + mac_len;
	skb->mac_len = mac_len;
}

4209 4210 4211 4212 4213
static inline bool netif_is_macsec(const struct net_device *dev)
{
	return dev->priv_flags & IFF_MACSEC;
}

4214
static inline bool netif_is_macvlan(const struct net_device *dev)
4215 4216 4217 4218
{
	return dev->priv_flags & IFF_MACVLAN;
}

4219
static inline bool netif_is_macvlan_port(const struct net_device *dev)
4220 4221 4222 4223
{
	return dev->priv_flags & IFF_MACVLAN_PORT;
}

4224
static inline bool netif_is_ipvlan(const struct net_device *dev)
4225 4226 4227 4228
{
	return dev->priv_flags & IFF_IPVLAN_SLAVE;
}

4229
static inline bool netif_is_ipvlan_port(const struct net_device *dev)
4230 4231 4232 4233
{
	return dev->priv_flags & IFF_IPVLAN_MASTER;
}

4234
static inline bool netif_is_bond_master(const struct net_device *dev)
4235 4236 4237 4238
{
	return dev->flags & IFF_MASTER && dev->priv_flags & IFF_BONDING;
}

4239
static inline bool netif_is_bond_slave(const struct net_device *dev)
J
Jiri Pirko 已提交
4240 4241 4242 4243
{
	return dev->flags & IFF_SLAVE && dev->priv_flags & IFF_BONDING;
}

4244 4245 4246 4247 4248
static inline bool netif_supports_nofcs(struct net_device *dev)
{
	return dev->priv_flags & IFF_SUPP_NOFCS;
}

4249
static inline bool netif_is_l3_master(const struct net_device *dev)
4250
{
4251
	return dev->priv_flags & IFF_L3MDEV_MASTER;
4252 4253
}

D
David Ahern 已提交
4254 4255 4256 4257 4258
static inline bool netif_is_l3_slave(const struct net_device *dev)
{
	return dev->priv_flags & IFF_L3MDEV_SLAVE;
}

4259 4260 4261 4262 4263
static inline bool netif_is_bridge_master(const struct net_device *dev)
{
	return dev->priv_flags & IFF_EBRIDGE;
}

4264 4265 4266 4267 4268
static inline bool netif_is_bridge_port(const struct net_device *dev)
{
	return dev->priv_flags & IFF_BRIDGE_PORT;
}

4269 4270 4271 4272 4273
static inline bool netif_is_ovs_master(const struct net_device *dev)
{
	return dev->priv_flags & IFF_OPENVSWITCH;
}

J
Jiri Pirko 已提交
4274 4275 4276 4277 4278
static inline bool netif_is_ovs_port(const struct net_device *dev)
{
	return dev->priv_flags & IFF_OVS_DATAPATH;
}

4279
static inline bool netif_is_team_master(const struct net_device *dev)
J
Jiri Pirko 已提交
4280 4281 4282 4283
{
	return dev->priv_flags & IFF_TEAM;
}

4284
static inline bool netif_is_team_port(const struct net_device *dev)
J
Jiri Pirko 已提交
4285 4286 4287 4288
{
	return dev->priv_flags & IFF_TEAM_PORT;
}

4289
static inline bool netif_is_lag_master(const struct net_device *dev)
J
Jiri Pirko 已提交
4290 4291 4292 4293
{
	return netif_is_bond_master(dev) || netif_is_team_master(dev);
}

4294
static inline bool netif_is_lag_port(const struct net_device *dev)
J
Jiri Pirko 已提交
4295 4296 4297 4298
{
	return netif_is_bond_slave(dev) || netif_is_team_port(dev);
}

4299 4300 4301 4302 4303
static inline bool netif_is_rxfh_configured(const struct net_device *dev)
{
	return dev->priv_flags & IFF_RXFH_CONFIGURED;
}

4304 4305 4306 4307 4308 4309
/* This device needs to keep skb dst for qdisc enqueue or ndo_start_xmit() */
static inline void netif_keep_dst(struct net_device *dev)
{
	dev->priv_flags &= ~(IFF_XMIT_DST_RELEASE | IFF_XMIT_DST_RELEASE_PERM);
}

4310 4311 4312 4313 4314 4315 4316
/* return true if dev can't cope with mtu frames that need vlan tag insertion */
static inline bool netif_reduces_vlan_mtu(struct net_device *dev)
{
	/* TODO: reserve and use an additional IFF bit, if we get more users */
	return dev->priv_flags & IFF_MACSEC;
}

4317
extern struct pernet_operations __net_initdata loopback_net_ops;
4318

4319 4320 4321 4322 4323 4324
/* Logging, debugging and troubleshooting/diagnostic helpers. */

/* netdev_printk helpers, similar to dev_printk */

static inline const char *netdev_name(const struct net_device *dev)
{
4325 4326
	if (!dev->name[0] || strchr(dev->name, '%'))
		return "(unnamed net_device)";
4327 4328 4329
	return dev->name;
}

4330 4331 4332 4333 4334
static inline bool netdev_unregistering(const struct net_device *dev)
{
	return dev->reg_state == NETREG_UNREGISTERING;
}

4335 4336 4337 4338 4339 4340 4341 4342 4343 4344 4345 4346 4347 4348 4349
static inline const char *netdev_reg_state(const struct net_device *dev)
{
	switch (dev->reg_state) {
	case NETREG_UNINITIALIZED: return " (uninitialized)";
	case NETREG_REGISTERED: return "";
	case NETREG_UNREGISTERING: return " (unregistering)";
	case NETREG_UNREGISTERED: return " (unregistered)";
	case NETREG_RELEASED: return " (released)";
	case NETREG_DUMMY: return " (dummy)";
	}

	WARN_ONCE(1, "%s: unknown reg_state %d\n", dev->name, dev->reg_state);
	return " (unknown)";
}

4350
__printf(3, 4)
4351 4352
void netdev_printk(const char *level, const struct net_device *dev,
		   const char *format, ...);
4353
__printf(2, 3)
4354
void netdev_emerg(const struct net_device *dev, const char *format, ...);
4355
__printf(2, 3)
4356
void netdev_alert(const struct net_device *dev, const char *format, ...);
4357
__printf(2, 3)
4358
void netdev_crit(const struct net_device *dev, const char *format, ...);
4359
__printf(2, 3)
4360
void netdev_err(const struct net_device *dev, const char *format, ...);
4361
__printf(2, 3)
4362
void netdev_warn(const struct net_device *dev, const char *format, ...);
4363
__printf(2, 3)
4364
void netdev_notice(const struct net_device *dev, const char *format, ...);
4365
__printf(2, 3)
4366
void netdev_info(const struct net_device *dev, const char *format, ...);
4367

4368 4369 4370 4371 4372 4373 4374 4375 4376 4377 4378 4379 4380 4381 4382 4383 4384 4385 4386 4387 4388 4389 4390 4391 4392
#define netdev_level_once(level, dev, fmt, ...)			\
do {								\
	static bool __print_once __read_mostly;			\
								\
	if (!__print_once) {					\
		__print_once = true;				\
		netdev_printk(level, dev, fmt, ##__VA_ARGS__);	\
	}							\
} while (0)

#define netdev_emerg_once(dev, fmt, ...) \
	netdev_level_once(KERN_EMERG, dev, fmt, ##__VA_ARGS__)
#define netdev_alert_once(dev, fmt, ...) \
	netdev_level_once(KERN_ALERT, dev, fmt, ##__VA_ARGS__)
#define netdev_crit_once(dev, fmt, ...) \
	netdev_level_once(KERN_CRIT, dev, fmt, ##__VA_ARGS__)
#define netdev_err_once(dev, fmt, ...) \
	netdev_level_once(KERN_ERR, dev, fmt, ##__VA_ARGS__)
#define netdev_warn_once(dev, fmt, ...) \
	netdev_level_once(KERN_WARNING, dev, fmt, ##__VA_ARGS__)
#define netdev_notice_once(dev, fmt, ...) \
	netdev_level_once(KERN_NOTICE, dev, fmt, ##__VA_ARGS__)
#define netdev_info_once(dev, fmt, ...) \
	netdev_level_once(KERN_INFO, dev, fmt, ##__VA_ARGS__)

4393 4394 4395
#define MODULE_ALIAS_NETDEV(device) \
	MODULE_ALIAS("netdev-" device)

4396
#if defined(CONFIG_DYNAMIC_DEBUG)
4397 4398
#define netdev_dbg(__dev, format, args...)			\
do {								\
4399
	dynamic_netdev_dbg(__dev, format, ##args);		\
4400
} while (0)
4401 4402 4403
#elif defined(DEBUG)
#define netdev_dbg(__dev, format, args...)			\
	netdev_printk(KERN_DEBUG, __dev, format, ##args)
4404 4405 4406 4407 4408 4409 4410 4411 4412 4413 4414 4415 4416 4417 4418 4419 4420 4421 4422 4423 4424 4425 4426 4427 4428 4429
#else
#define netdev_dbg(__dev, format, args...)			\
({								\
	if (0)							\
		netdev_printk(KERN_DEBUG, __dev, format, ##args); \
})
#endif

#if defined(VERBOSE_DEBUG)
#define netdev_vdbg	netdev_dbg
#else

#define netdev_vdbg(dev, format, args...)			\
({								\
	if (0)							\
		netdev_printk(KERN_DEBUG, dev, format, ##args);	\
	0;							\
})
#endif

/*
 * netdev_WARN() acts like dev_printk(), but with the key difference
 * of using a WARN/WARN_ON to get the message out, including the
 * file/line information and a backtrace.
 */
#define netdev_WARN(dev, format, args...)			\
4430
	WARN(1, "netdevice: %s%s: " format, netdev_name(dev),	\
4431
	     netdev_reg_state(dev), ##args)
4432

G
Gal Pressman 已提交
4433
#define netdev_WARN_ONCE(dev, format, args...)				\
4434
	WARN_ONCE(1, "netdevice: %s%s: " format, netdev_name(dev),	\
4435 4436
		  netdev_reg_state(dev), ##args)

4437 4438 4439 4440 4441 4442 4443 4444
/* netif printk helpers, similar to netdev_printk */

#define netif_printk(priv, type, level, dev, fmt, args...)	\
do {					  			\
	if (netif_msg_##type(priv))				\
		netdev_printk(level, (dev), fmt, ##args);	\
} while (0)

4445 4446 4447 4448 4449 4450
#define netif_level(level, priv, type, dev, fmt, args...)	\
do {								\
	if (netif_msg_##type(priv))				\
		netdev_##level(dev, fmt, ##args);		\
} while (0)

4451
#define netif_emerg(priv, type, dev, fmt, args...)		\
4452
	netif_level(emerg, priv, type, dev, fmt, ##args)
4453
#define netif_alert(priv, type, dev, fmt, args...)		\
4454
	netif_level(alert, priv, type, dev, fmt, ##args)
4455
#define netif_crit(priv, type, dev, fmt, args...)		\
4456
	netif_level(crit, priv, type, dev, fmt, ##args)
4457
#define netif_err(priv, type, dev, fmt, args...)		\
4458
	netif_level(err, priv, type, dev, fmt, ##args)
4459
#define netif_warn(priv, type, dev, fmt, args...)		\
4460
	netif_level(warn, priv, type, dev, fmt, ##args)
4461
#define netif_notice(priv, type, dev, fmt, args...)		\
4462
	netif_level(notice, priv, type, dev, fmt, ##args)
4463
#define netif_info(priv, type, dev, fmt, args...)		\
4464
	netif_level(info, priv, type, dev, fmt, ##args)
4465

4466
#if defined(CONFIG_DYNAMIC_DEBUG)
4467 4468 4469
#define netif_dbg(priv, type, netdev, format, args...)		\
do {								\
	if (netif_msg_##type(priv))				\
4470
		dynamic_netdev_dbg(netdev, format, ##args);	\
4471
} while (0)
4472 4473 4474
#elif defined(DEBUG)
#define netif_dbg(priv, type, dev, format, args...)		\
	netif_printk(priv, type, KERN_DEBUG, dev, format, ##args)
4475 4476 4477 4478 4479 4480 4481 4482 4483
#else
#define netif_dbg(priv, type, dev, format, args...)			\
({									\
	if (0)								\
		netif_printk(priv, type, KERN_DEBUG, dev, format, ##args); \
	0;								\
})
#endif

4484 4485 4486 4487 4488 4489 4490 4491 4492
/* if @cond then downgrade to debug, else print at @level */
#define netif_cond_dbg(priv, type, netdev, cond, level, fmt, args...)     \
	do {                                                              \
		if (cond)                                                 \
			netif_dbg(priv, type, netdev, fmt, ##args);       \
		else                                                      \
			netif_ ## level(priv, type, netdev, fmt, ##args); \
	} while (0)

4493
#if defined(VERBOSE_DEBUG)
4494
#define netif_vdbg	netif_dbg
4495 4496 4497 4498
#else
#define netif_vdbg(priv, type, dev, format, args...)		\
({								\
	if (0)							\
4499
		netif_printk(priv, type, KERN_DEBUG, dev, format, ##args); \
4500 4501 4502
	0;							\
})
#endif
4503

4504 4505 4506 4507 4508 4509 4510 4511 4512 4513 4514 4515 4516 4517 4518 4519 4520 4521 4522 4523 4524 4525
/*
 *	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.
 *
 *		0800	IP
 *		0001	802.3
 *		0002	AX.25
 *		0004	802.2
 *		8035	RARP
 *		0005	SNAP
 *		0805	X.25
 *		0806	ARP
 *		8137	IPX
 *		0009	Localtalk
 *		86DD	IPv6
 */
#define PTYPE_HASH_SIZE	(16)
#define PTYPE_HASH_MASK	(PTYPE_HASH_SIZE - 1)

4526
#endif	/* _LINUX_NETDEVICE_H */