netdevice.h 144.1 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
struct sfp_bus;
62 63
/* 802.11 specific */
struct wireless_dev;
64 65
/* 802.15.4 specific */
struct wpan_dev;
R
Robert Shearman 已提交
66
struct mpls_dev;
67 68
/* UDP Tunnel offloads */
struct udp_tunnel_info;
69
struct bpf_prog;
70
struct xdp_buff;
L
Linus Torvalds 已提交
71

72 73
void netdev_set_default_ethtool_ops(struct net_device *dev,
				    const struct ethtool_ops *ops);
74

75 76 77 78
/* Backlog congestion levels */
#define NET_RX_SUCCESS		0	/* keep 'em coming, baby */
#define NET_RX_DROP		1	/* packet dropped */

79 80 81 82 83 84 85 86 87 88 89 90
/*
 * 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
91 92
 * higher layers. Virtual network devices transmit synchronously; in this case
 * the driver transmit return codes are consumed by dev_queue_xmit(), and all
93 94 95 96 97
 * others are propagated to higher layers.
 */

/* qdisc ->enqueue() return codes. */
#define NET_XMIT_SUCCESS	0x00
98 99 100
#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
101

102 103 104
/* 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. */
105
#define net_xmit_eval(e)	((e) == NET_XMIT_CN ? 0 : (e))
L
Linus Torvalds 已提交
106 107
#define net_xmit_errno(e)	((e) != NET_XMIT_CN ? -ENOBUFS : 0)

108
/* Driver transmit return codes */
109
#define NETDEV_TX_MASK		0xf0
110

111
enum netdev_tx {
112
	__NETDEV_TX_MIN	 = INT_MIN,	/* make sure enum is signed */
113 114
	NETDEV_TX_OK	 = 0x00,	/* driver took care of packet */
	NETDEV_TX_BUSY	 = 0x10,	/* driver tx path was busy*/
115 116 117
};
typedef enum netdev_tx netdev_tx_t;

118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135
/*
 * 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
136
/*
B
Bjorn Helgaas 已提交
137
 *	Compute the worst-case header length according to the protocols
L
Linus Torvalds 已提交
138 139
 *	used.
 */
G
Graf Yang 已提交
140

141 142 143
#if defined(CONFIG_HYPERV_NET)
# define LL_MAX_HEADER 128
#elif defined(CONFIG_WLAN) || IS_ENABLED(CONFIG_AX25)
144 145 146 147 148
# if defined(CONFIG_MAC80211_MESH)
#  define LL_MAX_HEADER 128
# else
#  define LL_MAX_HEADER 96
# endif
L
Linus Torvalds 已提交
149
#else
150
# define LL_MAX_HEADER 32
L
Linus Torvalds 已提交
151 152
#endif

153 154
#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
155 156 157 158 159 160
#define MAX_HEADER LL_MAX_HEADER
#else
#define MAX_HEADER (LL_MAX_HEADER + 48)
#endif

/*
161 162
 *	Old network device statistics. Fields are native words
 *	(unsigned long) so they can be read and written atomically.
L
Linus Torvalds 已提交
163
 */
G
Graf Yang 已提交
164

E
Eric Dumazet 已提交
165
struct net_device_stats {
166 167 168 169 170 171 172 173 174
	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
175 176
	unsigned long	collisions;
	unsigned long	rx_length_errors;
177 178 179 180 181
	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
182 183 184 185 186 187 188 189 190 191 192 193 194
	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>

195
#ifdef CONFIG_RPS
196 197
#include <linux/static_key.h>
extern struct static_key rps_needed;
E
Eric Dumazet 已提交
198
extern struct static_key rfs_needed;
199 200
#endif

L
Linus Torvalds 已提交
201 202 203 204
struct neighbour;
struct neigh_parms;
struct sk_buff;

205 206 207 208
struct netdev_hw_addr {
	struct list_head	list;
	unsigned char		addr[MAX_ADDR_LEN];
	unsigned char		type;
J
Jiri Pirko 已提交
209 210 211 212
#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
213 214
#define NETDEV_HW_ADDR_T_MULTICAST	5
	bool			global_use;
215
	int			sync_cnt;
216
	int			refcount;
217
	int			synced;
218 219 220
	struct rcu_head		rcu_head;
};

221 222 223 224 225
struct netdev_hw_addr_list {
	struct list_head	list;
	int			count;
};

226 227 228 229
#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)
230

231 232 233 234
#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)
235

236 237
#define netdev_mc_count(dev) netdev_hw_addr_list_count(&(dev)->mc)
#define netdev_mc_empty(dev) netdev_hw_addr_list_empty(&(dev)->mc)
238
#define netdev_for_each_mc_addr(ha, dev) \
239
	netdev_hw_addr_list_for_each(ha, &(dev)->mc)
240

E
Eric Dumazet 已提交
241
struct hh_cache {
242
	unsigned int	hh_len;
243
	seqlock_t	hh_lock;
L
Linus Torvalds 已提交
244 245 246 247

	/* cached hardware header; allow for machine alignment needs.        */
#define HH_DATA_MOD	16
#define HH_DATA_OFF(__len) \
J
Jiri Benc 已提交
248
	(HH_DATA_MOD - (((__len - 1) & (HH_DATA_MOD - 1)) + 1))
L
Linus Torvalds 已提交
249 250 251 252 253
#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
254
/* Reserve HH_DATA_MOD byte-aligned hard_header_len, but at least that much.
L
Linus Torvalds 已提交
255 256 257 258 259 260 261 262
 * 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) \
263
	((((dev)->hard_header_len+(dev)->needed_headroom)&~(HH_DATA_MOD - 1)) + HH_DATA_MOD)
L
Linus Torvalds 已提交
264
#define LL_RESERVED_SPACE_EXTRA(dev,extra) \
265
	((((dev)->hard_header_len+(dev)->needed_headroom+(extra))&~(HH_DATA_MOD - 1)) + HH_DATA_MOD)
L
Linus Torvalds 已提交
266

267 268 269
struct header_ops {
	int	(*create) (struct sk_buff *skb, struct net_device *dev,
			   unsigned short type, const void *daddr,
270
			   const void *saddr, unsigned int len);
271
	int	(*parse)(const struct sk_buff *skb, unsigned char *haddr);
272
	int	(*cache)(const struct neighbour *neigh, struct hh_cache *hh, __be16 type);
273 274 275
	void	(*cache_update)(struct hh_cache *hh,
				const struct net_device *dev,
				const unsigned char *haddr);
276
	bool	(*validate)(const char *ll_header, unsigned int len);
277 278
};

L
Linus Torvalds 已提交
279
/* These flag bits are private to the generic network queueing
B
Bjorn Helgaas 已提交
280
 * layer; they may not be explicitly referenced by any other
L
Linus Torvalds 已提交
281 282 283
 * code.
 */

E
Eric Dumazet 已提交
284
enum netdev_state_t {
L
Linus Torvalds 已提交
285 286 287
	__LINK_STATE_START,
	__LINK_STATE_PRESENT,
	__LINK_STATE_NOCARRIER,
S
Stefan Rompf 已提交
288 289
	__LINK_STATE_LINKWATCH_PENDING,
	__LINK_STATE_DORMANT,
L
Linus Torvalds 已提交
290 291 292 293
};


/*
B
Bjorn Helgaas 已提交
294
 * This structure holds boot-time configured netdevice settings. They
G
Graf Yang 已提交
295
 * are then used in the device probing.
L
Linus Torvalds 已提交
296 297 298 299 300 301 302
 */
struct netdev_boot_setup {
	char name[IFNAMSIZ];
	struct ifmap map;
};
#define NETDEV_BOOT_SETUP_MAX 8

303
int __init netdev_boot_setup(char *str);
L
Linus Torvalds 已提交
304

305 306 307 308 309
struct gro_list {
	struct list_head	list;
	int			count;
};

310 311 312
/*
 * Structure for NAPI scheduling similar to tasklet but with weighting
 */
313
#define GRO_HASH_BUCKETS	8
314 315 316 317
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
318
	 * to the per-CPU poll_list, and whoever clears that bit
319 320 321 322 323 324
	 * can remove from the list right before clearing the bit.
	 */
	struct list_head	poll_list;

	unsigned long		state;
	int			weight;
325
	unsigned int		gro_count;
326 327 328 329
	int			(*poll)(struct napi_struct *, int);
#ifdef CONFIG_NETPOLL
	int			poll_owner;
#endif
H
Herbert Xu 已提交
330
	struct net_device	*dev;
331
	struct gro_list		gro_hash[GRO_HASH_BUCKETS];
H
Herbert Xu 已提交
332
	struct sk_buff		*skb;
333
	struct hrtimer		timer;
334
	struct list_head	dev_list;
E
Eliezer Tamir 已提交
335 336
	struct hlist_node	napi_hash_node;
	unsigned int		napi_id;
337 338
};

E
Eric Dumazet 已提交
339
enum {
340
	NAPI_STATE_SCHED,	/* Poll is scheduled */
E
Eric Dumazet 已提交
341
	NAPI_STATE_MISSED,	/* reschedule a napi */
D
David S. Miller 已提交
342
	NAPI_STATE_DISABLE,	/* Disable pending */
343
	NAPI_STATE_NPSVC,	/* Netpoll - don't dequeue from poll_list */
E
Eric Dumazet 已提交
344 345
	NAPI_STATE_HASHED,	/* In NAPI hash (busy polling possible) */
	NAPI_STATE_NO_BUSY_POLL,/* Do not add in napi_hash, no busy polling */
346 347 348 349
	NAPI_STATE_IN_BUSY_POLL,/* sk_busy_loop() owns this NAPI */
};

enum {
E
Eric Dumazet 已提交
350 351 352 353 354 355 356
	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),
357 358
};

359
enum gro_result {
360 361 362 363 364
	GRO_MERGED,
	GRO_MERGED_FREE,
	GRO_HELD,
	GRO_NORMAL,
	GRO_DROP,
365
	GRO_CONSUMED,
366
};
367
typedef enum gro_result gro_result_t;
368

369 370 371 372 373 374 375
/*
 * 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
376
 * @RX_HANDLER_PASS: Do nothing, pass the skb as if no rx_handler was called.
377 378 379 380 381 382 383 384 385 386 387 388 389 390
 *
 * 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
391
 * If the rx_handler consumed the skb in some way, it should return
392
 * RX_HANDLER_CONSUMED. This is appropriate when the rx_handler arranged for
B
Bjorn Helgaas 已提交
393
 * the skb to be delivered in some other way.
394 395 396 397 398
 *
 * 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
399
 * If the rx_handler decides the skb should be ignored, it should return
400
 * RX_HANDLER_EXACT. The skb will only be delivered to protocol handlers that
401
 * are registered on exact device (ptype->dev == skb->dev).
402
 *
B
Bjorn Helgaas 已提交
403
 * If the rx_handler didn't change skb->dev, but wants the skb to be normally
404 405 406 407 408 409 410 411 412 413 414 415 416 417
 * 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);
418

419
void __napi_schedule(struct napi_struct *n);
420
void __napi_schedule_irqoff(struct napi_struct *n);
421

422
static inline bool napi_disable_pending(struct napi_struct *n)
D
David S. Miller 已提交
423 424 425 426
{
	return test_bit(NAPI_STATE_DISABLE, &n->state);
}

E
Eric Dumazet 已提交
427
bool napi_schedule_prep(struct napi_struct *n);
428 429 430

/**
 *	napi_schedule - schedule NAPI poll
B
Bjorn Helgaas 已提交
431
 *	@n: NAPI context
432 433 434 435 436 437 438 439 440 441
 *
 * 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);
}

442 443
/**
 *	napi_schedule_irqoff - schedule NAPI poll
B
Bjorn Helgaas 已提交
444
 *	@n: NAPI context
445 446 447 448 449 450 451 452 453
 *
 * 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);
}

454
/* Try to reschedule poll. Called by dev->poll() after napi_complete().  */
455
static inline bool napi_reschedule(struct napi_struct *napi)
456 457 458
{
	if (napi_schedule_prep(napi)) {
		__napi_schedule(napi);
459
		return true;
460
	}
461
	return false;
462 463
}

464
bool napi_complete_done(struct napi_struct *n, int work_done);
465 466
/**
 *	napi_complete - NAPI processing complete
B
Bjorn Helgaas 已提交
467
 *	@n: NAPI context
468 469
 *
 * Mark NAPI processing as complete.
470
 * Consider using napi_complete_done() instead.
471
 * Return false if device should avoid rearming interrupts.
472
 */
473
static inline bool napi_complete(struct napi_struct *n)
474 475 476
{
	return napi_complete_done(n, 0);
}
477

E
Eliezer Tamir 已提交
478 479
/**
 *	napi_hash_del - remove a NAPI from global table
B
Bjorn Helgaas 已提交
480
 *	@napi: NAPI context
E
Eliezer Tamir 已提交
481
 *
B
Bjorn Helgaas 已提交
482
 * Warning: caller must observe RCU grace period
483 484
 * before freeing memory containing @napi, if
 * this function returns true.
485
 * Note: core networking stack automatically calls it
B
Bjorn Helgaas 已提交
486
 * from netif_napi_del().
487
 * Drivers might want to call this helper to combine all
B
Bjorn Helgaas 已提交
488
 * the needed RCU grace periods into a single one.
E
Eliezer Tamir 已提交
489
 */
490
bool napi_hash_del(struct napi_struct *napi);
E
Eliezer Tamir 已提交
491

492 493
/**
 *	napi_disable - prevent NAPI from scheduling
B
Bjorn Helgaas 已提交
494
 *	@n: NAPI context
495 496 497 498
 *
 * Stop NAPI from being scheduled on this context.
 * Waits till any outstanding processing completes.
 */
499
void napi_disable(struct napi_struct *n);
500 501 502

/**
 *	napi_enable - enable NAPI scheduling
B
Bjorn Helgaas 已提交
503
 *	@n: NAPI context
504 505 506 507 508 509 510
 *
 * 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));
511
	smp_mb__before_atomic();
512
	clear_bit(NAPI_STATE_SCHED, &n->state);
513
	clear_bit(NAPI_STATE_NPSVC, &n->state);
514 515
}

516 517
/**
 *	napi_synchronize - wait until NAPI is not running
B
Bjorn Helgaas 已提交
518
 *	@n: NAPI context
519 520 521 522 523 524 525
 *
 * 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)
{
526 527 528 529 530
	if (IS_ENABLED(CONFIG_SMP))
		while (test_bit(NAPI_STATE_SCHED, &n->state))
			msleep(1);
	else
		barrier();
531 532
}

E
Eric Dumazet 已提交
533
enum netdev_queue_state_t {
534 535
	__QUEUE_STATE_DRV_XOFF,
	__QUEUE_STATE_STACK_XOFF,
536
	__QUEUE_STATE_FROZEN,
537
};
538 539 540 541 542 543 544 545 546 547 548

#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)

549 550 551 552 553 554 555 556 557
/*
 * __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_*.
 */
558

559
struct netdev_queue {
560
/*
B
Bjorn Helgaas 已提交
561
 * read-mostly part
562
 */
563
	struct net_device	*dev;
564
	struct Qdisc __rcu	*qdisc;
565
	struct Qdisc		*qdisc_sleeping;
566
#ifdef CONFIG_SYSFS
T
Tom Herbert 已提交
567 568
	struct kobject		kobj;
#endif
569 570 571
#if defined(CONFIG_XPS) && defined(CONFIG_NUMA)
	int			numa_node;
#endif
572 573 574 575 576 577
	unsigned long		tx_maxrate;
	/*
	 * Number of TX timeouts for this queue
	 * (/sys/class/net/DEV/Q/trans_timeout)
	 */
	unsigned long		trans_timeout;
578 579 580

	/* Subordinate device that the queue has been assigned to */
	struct net_device	*sb_dev;
581
/*
B
Bjorn Helgaas 已提交
582
 * write-mostly part
583 584 585
 */
	spinlock_t		_xmit_lock ____cacheline_aligned_in_smp;
	int			xmit_lock_owner;
586
	/*
F
Florian Westphal 已提交
587
	 * Time (in jiffies) of last Tx
588 589
	 */
	unsigned long		trans_start;
590

T
Tom Herbert 已提交
591 592 593 594 595
	unsigned long		state;

#ifdef CONFIG_BQL
	struct dql		dql;
#endif
596
} ____cacheline_aligned_in_smp;
597

598 599 600 601
extern int sysctl_fb_tunnels_only_for_init_net;

static inline bool net_has_fallback_tunnels(const struct net *net)
{
602 603 604
	return net == &init_net ||
	       !IS_ENABLED(CONFIG_SYSCTL) ||
	       !sysctl_fb_tunnels_only_for_init_net;
605 606
}

607 608 609 610 611
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
612
	return NUMA_NO_NODE;
613 614 615 616 617 618 619 620 621 622
#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
623
#ifdef CONFIG_RPS
T
Tom Herbert 已提交
624 625 626 627 628 629 630 631 632
/*
 * 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];
};
633
#define RPS_MAP_SIZE(_num) (sizeof(struct rps_map) + ((_num) * sizeof(u16)))
T
Tom Herbert 已提交
634

T
Tom Herbert 已提交
635
/*
636 637 638
 * 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
639 640 641
 */
struct rps_dev_flow {
	u16 cpu;
642
	u16 filter;
T
Tom Herbert 已提交
643 644
	unsigned int last_qtail;
};
645
#define RPS_NO_FILTER 0xffff
T
Tom Herbert 已提交
646 647 648 649 650 651 652 653 654 655

/*
 * 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) + \
656
    ((_num) * sizeof(struct rps_dev_flow)))
T
Tom Herbert 已提交
657 658 659 660

/*
 * 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
661 662
 * Each entry is a 32bit value. Upper part is the high-order bits
 * of flow hash, lower part is CPU number.
663
 * rps_cpu_mask is used to partition the space, depending on number of
B
Bjorn Helgaas 已提交
664 665
 * possible CPUs : rps_cpu_mask = roundup_pow_of_two(nr_cpu_ids) - 1
 * For example, if 64 CPUs are possible, rps_cpu_mask = 0x3f,
666
 * meaning we use 32-6=26 bits for the hash.
T
Tom Herbert 已提交
667 668
 */
struct rps_sock_flow_table {
669
	u32	mask;
670 671

	u32	ents[0] ____cacheline_aligned_in_smp;
T
Tom Herbert 已提交
672
};
673
#define	RPS_SOCK_FLOW_TABLE_SIZE(_num) (offsetof(struct rps_sock_flow_table, ents[_num]))
T
Tom Herbert 已提交
674 675 676

#define RPS_NO_CPU 0xffff

677 678 679
extern u32 rps_cpu_mask;
extern struct rps_sock_flow_table __rcu *rps_sock_flow_table;

T
Tom Herbert 已提交
680 681 682 683
static inline void rps_record_sock_flow(struct rps_sock_flow_table *table,
					u32 hash)
{
	if (table && hash) {
684 685
		unsigned int index = hash & table->mask;
		u32 val = hash & ~rps_cpu_mask;
T
Tom Herbert 已提交
686

B
Bjorn Helgaas 已提交
687
		/* We only give a hint, preemption can change CPU under us */
688
		val |= raw_smp_processor_id();
T
Tom Herbert 已提交
689

690 691
		if (table->ents[index] != val)
			table->ents[index] = val;
T
Tom Herbert 已提交
692 693 694
	}
}

695
#ifdef CONFIG_RFS_ACCEL
696 697
bool rps_may_expire_flow(struct net_device *dev, u16 rxq_index, u32 flow_id,
			 u16 filter_id);
698
#endif
699
#endif /* CONFIG_RPS */
700

T
Tom Herbert 已提交
701 702
/* This structure contains an instance of an RX queue. */
struct netdev_rx_queue {
703
#ifdef CONFIG_RPS
E
Eric Dumazet 已提交
704 705
	struct rps_map __rcu		*rps_map;
	struct rps_dev_flow_table __rcu	*rps_flow_table;
706
#endif
E
Eric Dumazet 已提交
707
	struct kobject			kobj;
T
Tom Herbert 已提交
708
	struct net_device		*dev;
709
	struct xdp_rxq_info		xdp_rxq;
T
Tom Herbert 已提交
710
} ____cacheline_aligned_in_smp;
711 712 713 714 715 716

/*
 * RX queue sysfs structures and functions.
 */
struct rx_queue_attribute {
	struct attribute attr;
717
	ssize_t (*show)(struct netdev_rx_queue *queue, char *buf);
718
	ssize_t (*store)(struct netdev_rx_queue *queue,
719
			 const char *buf, size_t len);
720
};
721

T
Tom Herbert 已提交
722 723 724 725 726 727 728 729 730 731 732
#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];
};
733
#define XPS_MAP_SIZE(_num) (sizeof(struct xps_map) + ((_num) * sizeof(u16)))
734 735
#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
736 737 738 739 740 741

/*
 * This structure holds all XPS maps for device.  Maps are indexed by CPU.
 */
struct xps_dev_maps {
	struct rcu_head rcu;
742
	struct xps_map __rcu *attr_map[0]; /* Either CPUs map or RXQs map */
T
Tom Herbert 已提交
743
};
744 745

#define XPS_CPU_DEV_MAPS_SIZE(_tcs) (sizeof(struct xps_dev_maps) +	\
746
	(nr_cpu_ids * (_tcs) * sizeof(struct xps_map *)))
747 748 749 750

#define XPS_RXQ_DEV_MAPS_SIZE(_tcs, _rxqs) (sizeof(struct xps_dev_maps) +\
	(_rxqs * (_tcs) * sizeof(struct xps_map *)))

T
Tom Herbert 已提交
751 752
#endif /* CONFIG_XPS */

753 754 755 756 757 758 759 760
#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;
};

761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777
#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

778
#define MAX_PHYS_ITEM_ID_LEN 32
779

780 781
/* This structure holds a unique identifier to identify some
 * physical item (port for example) used by a netdevice.
782
 */
783 784
struct netdev_phys_item_id {
	unsigned char id[MAX_PHYS_ITEM_ID_LEN];
785 786 787
	unsigned char id_len;
};

788 789 790 791 792 793 794
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;
}

795 796 797
typedef u16 (*select_queue_fallback_t)(struct net_device *dev,
				       struct sk_buff *skb);

798
enum tc_setup_type {
799
	TC_SETUP_QDISC_MQPRIO,
800
	TC_SETUP_CLSU32,
801
	TC_SETUP_CLSFLOWER,
802
	TC_SETUP_CLSMATCHALL,
803
	TC_SETUP_CLSBPF,
804
	TC_SETUP_BLOCK,
805
	TC_SETUP_QDISC_CBS,
806
	TC_SETUP_QDISC_RED,
807
	TC_SETUP_QDISC_PRIO,
808
	TC_SETUP_QDISC_MQ,
809
	TC_SETUP_QDISC_ETF,
810 811
};

812 813
/* These structures hold the attributes of bpf state that are being passed
 * to the netdevice through the bpf op.
814
 */
815
enum bpf_netdev_command {
816 817 818 819 820 821 822 823
	/* 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,
824
	XDP_SETUP_PROG_HW,
825
	/* Check if a bpf program is set on the device.  The callee should
826 827
	 * set @prog_attached to one of XDP_ATTACHED_* values, note that "true"
	 * is equivalent to XDP_ATTACHED_DRV.
828 829
	 */
	XDP_QUERY_PROG,
830 831 832 833
	/* BPF program for offload callbacks, invoked at program load time. */
	BPF_OFFLOAD_VERIFIER_PREP,
	BPF_OFFLOAD_TRANSLATE,
	BPF_OFFLOAD_DESTROY,
834 835
	BPF_OFFLOAD_MAP_ALLOC,
	BPF_OFFLOAD_MAP_FREE,
836 837
	XDP_QUERY_XSK_UMEM,
	XDP_SETUP_XSK_UMEM,
838 839
};

840
struct bpf_prog_offload_ops;
841
struct netlink_ext_ack;
842
struct xdp_umem;
843

844 845
struct netdev_bpf {
	enum bpf_netdev_command command;
846 847
	union {
		/* XDP_SETUP_PROG */
848
		struct {
849
			u32 flags;
850 851 852
			struct bpf_prog *prog;
			struct netlink_ext_ack *extack;
		};
853
		/* XDP_QUERY_PROG */
M
Martin KaFai Lau 已提交
854
		struct {
855
			u8 prog_attached;
M
Martin KaFai Lau 已提交
856
			u32 prog_id;
857 858
			/* flags with which program was installed */
			u32 prog_flags;
M
Martin KaFai Lau 已提交
859
		};
860 861 862
		/* BPF_OFFLOAD_VERIFIER_PREP */
		struct {
			struct bpf_prog *prog;
863
			const struct bpf_prog_offload_ops *ops; /* callee set */
864 865 866 867 868
		} verifier;
		/* BPF_OFFLOAD_TRANSLATE, BPF_OFFLOAD_DESTROY */
		struct {
			struct bpf_prog *prog;
		} offload;
869 870 871 872
		/* BPF_OFFLOAD_MAP_ALLOC, BPF_OFFLOAD_MAP_FREE */
		struct {
			struct bpf_offloaded_map *offmap;
		};
873 874 875 876 877
		/* XDP_SETUP_XSK_UMEM */
		struct {
			struct xdp_umem *umem;
			u16 queue_id;
		} xsk;
878 879
	};
};
880

881 882 883 884 885 886 887
#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);
888
	void	(*xdo_dev_state_advance_esn) (struct xfrm_state *x);
889 890 891
};
#endif

I
Ilya Lesokhin 已提交
892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911
#if IS_ENABLED(CONFIG_TLS_DEVICE)
enum tls_offload_ctx_dir {
	TLS_OFFLOAD_CTX_DIR_RX,
	TLS_OFFLOAD_CTX_DIR_TX,
};

struct tls_crypto_info;
struct tls_context;

struct tlsdev_ops {
	int (*tls_dev_add)(struct net_device *netdev, struct sock *sk,
			   enum tls_offload_ctx_dir direction,
			   struct tls_crypto_info *crypto_info,
			   u32 start_offload_tcp_sn);
	void (*tls_dev_del)(struct net_device *netdev,
			    struct tls_context *ctx,
			    enum tls_offload_ctx_dir direction);
};
#endif

912 913 914 915 916
struct dev_ifalias {
	struct rcu_head rcuhead;
	char ifalias[];
};

917 918
/*
 * This structure defines the management hooks for network devices.
919 920
 * The following hooks can be defined; unless noted otherwise, they are
 * optional and can be filled with a null pointer.
921 922
 *
 * int (*ndo_init)(struct net_device *dev);
B
Bjorn Helgaas 已提交
923 924 925 926
 *     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.
927 928 929 930 931 932
 *
 * 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
933
 *     This function is called when a network device transitions to the up
934 935 936
 *     state.
 *
 * int (*ndo_stop)(struct net_device *dev);
B
Bjorn Helgaas 已提交
937
 *     This function is called when a network device transitions to the down
938 939
 *     state.
 *
940 941
 * netdev_tx_t (*ndo_start_xmit)(struct sk_buff *skb,
 *                               struct net_device *dev);
942
 *	Called when a packet needs to be transmitted.
943 944 945 946
 *	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
947
 *	Required; cannot be NULL.
948
 *
949 950 951 952 953 954 955 956 957
 * 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.
958
 *
959
 * u16 (*ndo_select_queue)(struct net_device *dev, struct sk_buff *skb,
960
 *                         void *accel_priv, select_queue_fallback_t fallback);
B
Bjorn Helgaas 已提交
961
 *	Called to decide which queue to use when device supports multiple
962 963
 *	transmit queues.
 *
964 965
 * 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
966
 *	changes to configuration when multicast or promiscuous is enabled.
967 968 969
 *
 * void (*ndo_set_rx_mode)(struct net_device *dev);
 *	This function is called device changes address list filtering.
970
 *	If driver handles unicast address filtering, it should set
B
Bjorn Helgaas 已提交
971
 *	IFF_UNICAST_FLT in its priv_flags.
972 973 974
 *
 * int (*ndo_set_mac_address)(struct net_device *dev, void *addr);
 *	This function  is called when the Media Access Control address
975
 *	needs to be changed. If this interface is not defined, the
B
Bjorn Helgaas 已提交
976
 *	MAC address can not be changed.
977 978 979 980 981
 *
 * 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
982 983
 *	Called when a user requests an ioctl which can't be handled by
 *	the generic interface code. If not defined ioctls return
984 985 986 987
 *	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
988
 *	is retained for legacy reasons; new devices should use the bus
989 990 991 992
 *	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
993
 *	of a device.
994
 *
995
 * void (*ndo_tx_timeout)(struct net_device *dev);
B
Bjorn Helgaas 已提交
996
 *	Callback used when the transmitter has not made any progress
997 998
 *	for dev->watchdog ticks.
 *
999 1000
 * void (*ndo_get_stats64)(struct net_device *dev,
 *                         struct rtnl_link_stats64 *storage);
1001
 * struct net_device_stats* (*ndo_get_stats)(struct net_device *dev);
1002
 *	Called when a user wants to get the network device usage
1003
 *	statistics. Drivers must do one of the following:
1004 1005
 *	1. Define @ndo_get_stats64 to fill in a zero-initialised
 *	   rtnl_link_stats64 structure passed by the caller.
1006
 *	2. Define @ndo_get_stats to update a net_device_stats structure
1007 1008 1009 1010 1011
 *	   (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.
1012
 *
1013
 * bool (*ndo_has_offload_stats)(const struct net_device *dev, int attr_id)
1014 1015 1016 1017 1018 1019 1020
 *	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.
 *
1021
 * int (*ndo_vlan_rx_add_vid)(struct net_device *dev, __be16 proto, u16 vid);
B
Bjorn Helgaas 已提交
1022
 *	If device supports VLAN filtering this function is called when a
1023
 *	VLAN id is registered.
1024
 *
1025
 * int (*ndo_vlan_rx_kill_vid)(struct net_device *dev, __be16 proto, u16 vid);
B
Bjorn Helgaas 已提交
1026
 *	If device supports VLAN filtering this function is called when a
1027
 *	VLAN id is unregistered.
1028 1029
 *
 * void (*ndo_poll_controller)(struct net_device *dev);
1030 1031 1032
 *
 *	SR-IOV management functions.
 * int (*ndo_set_vf_mac)(struct net_device *dev, int vf, u8* mac);
1033 1034
 * int (*ndo_set_vf_vlan)(struct net_device *dev, int vf, u16 vlan,
 *			  u8 qos, __be16 proto);
1035 1036
 * int (*ndo_set_vf_rate)(struct net_device *dev, int vf, int min_tx_rate,
 *			  int max_tx_rate);
1037
 * int (*ndo_set_vf_spoofchk)(struct net_device *dev, int vf, bool setting);
H
Hiroshi Shimamoto 已提交
1038
 * int (*ndo_set_vf_trust)(struct net_device *dev, int vf, bool setting);
1039 1040
 * int (*ndo_get_vf_config)(struct net_device *dev,
 *			    int vf, struct ifla_vf_info *ivf);
1041
 * int (*ndo_set_vf_link_state)(struct net_device *dev, int vf, int link_state);
1042 1043
 * int (*ndo_set_vf_port)(struct net_device *dev, int vf,
 *			  struct nlattr *port[]);
1044 1045 1046
 *
 *      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
1047
 *      with PF and querying it may introduce a theoretical security risk.
1048
 * int (*ndo_set_vf_rss_query_en)(struct net_device *dev, int vf, bool setting);
1049
 * int (*ndo_get_vf_port)(struct net_device *dev, int vf, struct sk_buff *skb);
1050
 * int (*ndo_setup_tc)(struct net_device *dev, enum tc_setup_type type,
1051
 *		       void *type_data);
F
Florian Fainelli 已提交
1052 1053 1054 1055
 *	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.
1056
 *
1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086
 *	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.
 *
1087 1088 1089 1090 1091 1092 1093
 * 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.
 *
1094 1095 1096 1097 1098 1099
 * 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.
 *
1100 1101 1102 1103 1104 1105
 *	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.
1106
 *
1107
 *	Slave management functions (for bridge, bonding, etc).
1108 1109 1110 1111 1112
 * 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.
1113 1114
 *
 *      Feature/offload setting functions.
1115 1116 1117 1118 1119 1120
 * 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.
 *
1121
 * int (*ndo_set_features)(struct net_device *dev, netdev_features_t features);
1122 1123 1124 1125
 *	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.
 *
1126 1127
 * int (*ndo_fdb_add)(struct ndmsg *ndm, struct nlattr *tb[],
 *		      struct net_device *dev,
1128
 *		      const unsigned char *addr, u16 vid, u16 flags)
1129
 *	Adds an FDB entry to dev for addr.
1130 1131
 * int (*ndo_fdb_del)(struct ndmsg *ndm, struct nlattr *tb[],
 *		      struct net_device *dev,
1132
 *		      const unsigned char *addr, u16 vid)
1133 1134
 *	Deletes the FDB entry from dev coresponding to addr.
 * int (*ndo_fdb_dump)(struct sk_buff *skb, struct netlink_callback *cb,
1135
 *		       struct net_device *dev, struct net_device *filter_dev,
1136
 *		       int *idx)
1137 1138
 *	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
1139
 *
1140 1141
 * int (*ndo_bridge_setlink)(struct net_device *dev, struct nlmsghdr *nlh,
 *			     u16 flags)
J
John Fastabend 已提交
1142
 * int (*ndo_bridge_getlink)(struct sk_buff *skb, u32 pid, u32 seq,
1143 1144
 *			     struct net_device *dev, u32 filter_mask,
 *			     int nlflags)
1145 1146
 * int (*ndo_bridge_dellink)(struct net_device *dev, struct nlmsghdr *nlh,
 *			     u16 flags);
J
Jiri Pirko 已提交
1147 1148 1149 1150 1151 1152 1153 1154
 *
 * 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.
1155 1156
 *
 * int (*ndo_get_phys_port_id)(struct net_device *dev,
1157
 *			       struct netdev_phys_item_id *ppid);
1158 1159 1160
 *	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.
1161
 *
1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174
 * 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.
 *
1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186
 * 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.
1187 1188 1189 1190
 * 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.
1191 1192
 * int (*ndo_get_iflink)(const struct net_device *dev);
 *	Called to get the iflink value of this device.
1193
 * void (*ndo_change_proto_down)(struct net_device *dev,
B
Bjorn Helgaas 已提交
1194
 *				 bool proto_down);
1195 1196 1197
 *	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.
1198 1199 1200 1201
 * 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.
1202 1203 1204 1205
 * 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
1206
 *	forward. Setting a negative value resets the rx headroom to the
1207
 *	default value.
1208
 * int (*ndo_bpf)(struct net_device *dev, struct netdev_bpf *bpf);
1209
 *	This function is used to set or query state related to XDP on the
1210 1211
 *	netdevice and manage BPF offload. See definition of
 *	enum bpf_netdev_command for details.
1212 1213
 * int (*ndo_xdp_xmit)(struct net_device *dev, int n, struct xdp_frame **xdp,
 *			u32 flags);
1214 1215 1216 1217 1218
 *	This function is used to submit @n XDP packets for transmit on a
 *	netdevice. Returns number of frames successfully transmitted, frames
 *	that got dropped are freed/returned via xdp_return_frame().
 *	Returns negative number, means general error invoking ndo, meaning
 *	no frames were xmit'ed and core-caller will free all frames.
1219 1220 1221 1222 1223 1224
 */
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);
1225 1226 1227 1228 1229
	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);
1230
	u16			(*ndo_select_queue)(struct net_device *dev,
1231
						    struct sk_buff *skb,
1232 1233
						    void *accel_priv,
						    select_queue_fallback_t fallback);
1234 1235 1236 1237 1238 1239 1240 1241 1242 1243
	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);
1244 1245 1246 1247
	int			(*ndo_change_mtu)(struct net_device *dev,
						  int new_mtu);
	int			(*ndo_neigh_setup)(struct net_device *dev,
						   struct neigh_parms *);
1248 1249
	void			(*ndo_tx_timeout) (struct net_device *dev);

1250 1251
	void			(*ndo_get_stats64)(struct net_device *dev,
						   struct rtnl_link_stats64 *storage);
1252
	bool			(*ndo_has_offload_stats)(const struct net_device *dev, int attr_id);
1253 1254 1255
	int			(*ndo_get_offload_stats)(int attr_id,
							 const struct net_device *dev,
							 void *attr_data);
1256 1257
	struct net_device_stats* (*ndo_get_stats)(struct net_device *dev);

1258
	int			(*ndo_vlan_rx_add_vid)(struct net_device *dev,
1259
						       __be16 proto, u16 vid);
1260
	int			(*ndo_vlan_rx_kill_vid)(struct net_device *dev,
1261
						        __be16 proto, u16 vid);
1262 1263
#ifdef CONFIG_NET_POLL_CONTROLLER
	void                    (*ndo_poll_controller)(struct net_device *dev);
H
Herbert Xu 已提交
1264
	int			(*ndo_netpoll_setup)(struct net_device *dev,
1265
						     struct netpoll_info *info);
1266
	void			(*ndo_netpoll_cleanup)(struct net_device *dev);
1267
#endif
1268 1269 1270
	int			(*ndo_set_vf_mac)(struct net_device *dev,
						  int queue, u8 *mac);
	int			(*ndo_set_vf_vlan)(struct net_device *dev,
1271 1272
						   int queue, u16 vlan,
						   u8 qos, __be16 proto);
1273 1274 1275
	int			(*ndo_set_vf_rate)(struct net_device *dev,
						   int vf, int min_tx_rate,
						   int max_tx_rate);
1276 1277
	int			(*ndo_set_vf_spoofchk)(struct net_device *dev,
						       int vf, bool setting);
H
Hiroshi Shimamoto 已提交
1278 1279
	int			(*ndo_set_vf_trust)(struct net_device *dev,
						    int vf, bool setting);
1280 1281 1282
	int			(*ndo_get_vf_config)(struct net_device *dev,
						     int vf,
						     struct ifla_vf_info *ivf);
1283 1284
	int			(*ndo_set_vf_link_state)(struct net_device *dev,
							 int vf, int link_state);
1285 1286 1287 1288
	int			(*ndo_get_vf_stats)(struct net_device *dev,
						    int vf,
						    struct ifla_vf_stats
						    *vf_stats);
1289 1290 1291 1292 1293
	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);
1294 1295 1296
	int			(*ndo_set_vf_guid)(struct net_device *dev,
						   int vf, u64 guid,
						   int guid_type);
1297 1298 1299
	int			(*ndo_set_vf_rss_query_en)(
						   struct net_device *dev,
						   int vf, bool setting);
1300
	int			(*ndo_setup_tc)(struct net_device *dev,
1301
						enum tc_setup_type type,
1302
						void *type_data);
1303
#if IS_ENABLED(CONFIG_FCOE)
1304 1305
	int			(*ndo_fcoe_enable)(struct net_device *dev);
	int			(*ndo_fcoe_disable)(struct net_device *dev);
1306 1307 1308 1309 1310 1311
	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);
1312 1313 1314 1315
	int			(*ndo_fcoe_ddp_target)(struct net_device *dev,
						       u16 xid,
						       struct scatterlist *sgl,
						       unsigned int sgc);
1316 1317
	int			(*ndo_fcoe_get_hbainfo)(struct net_device *dev,
							struct netdev_fcoe_hbainfo *hbainfo);
1318 1319
#endif

1320
#if IS_ENABLED(CONFIG_LIBFCOE)
1321 1322 1323 1324
#define NETDEV_FCOE_WWNN 0
#define NETDEV_FCOE_WWPN 1
	int			(*ndo_fcoe_get_wwn)(struct net_device *dev,
						    u64 *wwn, int type);
1325
#endif
1326

1327 1328 1329 1330 1331 1332
#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
1333
	int			(*ndo_add_slave)(struct net_device *dev,
D
David Ahern 已提交
1334 1335
						 struct net_device *slave_dev,
						 struct netlink_ext_ack *extack);
1336 1337
	int			(*ndo_del_slave)(struct net_device *dev,
						 struct net_device *slave_dev);
1338 1339
	netdev_features_t	(*ndo_fix_features)(struct net_device *dev,
						    netdev_features_t features);
1340
	int			(*ndo_set_features)(struct net_device *dev,
1341
						    netdev_features_t features);
1342 1343 1344 1345
	int			(*ndo_neigh_construct)(struct net_device *dev,
						       struct neighbour *n);
	void			(*ndo_neigh_destroy)(struct net_device *dev,
						     struct neighbour *n);
1346 1347

	int			(*ndo_fdb_add)(struct ndmsg *ndm,
1348
					       struct nlattr *tb[],
1349
					       struct net_device *dev,
1350
					       const unsigned char *addr,
1351
					       u16 vid,
1352 1353
					       u16 flags);
	int			(*ndo_fdb_del)(struct ndmsg *ndm,
1354
					       struct nlattr *tb[],
1355
					       struct net_device *dev,
1356 1357
					       const unsigned char *addr,
					       u16 vid);
1358 1359 1360
	int			(*ndo_fdb_dump)(struct sk_buff *skb,
						struct netlink_callback *cb,
						struct net_device *dev,
1361
						struct net_device *filter_dev,
1362
						int *idx);
J
John Fastabend 已提交
1363 1364

	int			(*ndo_bridge_setlink)(struct net_device *dev,
1365 1366
						      struct nlmsghdr *nlh,
						      u16 flags);
J
John Fastabend 已提交
1367 1368
	int			(*ndo_bridge_getlink)(struct sk_buff *skb,
						      u32 pid, u32 seq,
1369
						      struct net_device *dev,
1370 1371
						      u32 filter_mask,
						      int nlflags);
1372
	int			(*ndo_bridge_dellink)(struct net_device *dev,
1373 1374
						      struct nlmsghdr *nlh,
						      u16 flags);
J
Jiri Pirko 已提交
1375 1376
	int			(*ndo_change_carrier)(struct net_device *dev,
						      bool new_carrier);
1377
	int			(*ndo_get_phys_port_id)(struct net_device *dev,
1378
							struct netdev_phys_item_id *ppid);
1379 1380
	int			(*ndo_get_phys_port_name)(struct net_device *dev,
							  char *name, size_t len);
1381 1382 1383 1384
	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);
1385 1386 1387 1388 1389
	void*			(*ndo_dfwd_add_station)(struct net_device *pdev,
							struct net_device *dev);
	void			(*ndo_dfwd_del_station)(struct net_device *pdev,
							void *priv);

1390
	int			(*ndo_get_lock_subclass)(struct net_device *dev);
1391 1392 1393
	int			(*ndo_set_tx_maxrate)(struct net_device *dev,
						      int queue_index,
						      u32 maxrate);
1394
	int			(*ndo_get_iflink)(const struct net_device *dev);
1395 1396
	int			(*ndo_change_proto_down)(struct net_device *dev,
							 bool proto_down);
1397 1398
	int			(*ndo_fill_metadata_dst)(struct net_device *dev,
						       struct sk_buff *skb);
1399 1400
	void			(*ndo_set_rx_headroom)(struct net_device *dev,
						       int needed_headroom);
1401 1402
	int			(*ndo_bpf)(struct net_device *dev,
					   struct netdev_bpf *bpf);
1403
	int			(*ndo_xdp_xmit)(struct net_device *dev, int n,
1404 1405
						struct xdp_frame **xdp,
						u32 flags);
1406 1407
	int			(*ndo_xsk_async_xmit)(struct net_device *dev,
						      u32 queue_id);
1408 1409
};

1410 1411 1412 1413 1414
/**
 * 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
1415
 * userspace; this means that the order of these flags can change
1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438
 * 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
1439 1440
 * @IFF_XMIT_DST_RELEASE_PERM: IFF_XMIT_DST_RELEASE not taking into account
 *	underlying stacked devices
1441
 * @IFF_L3MDEV_MASTER: device is an L3 master device
1442
 * @IFF_NO_QUEUE: device can run without qdisc attached
1443
 * @IFF_OPENVSWITCH: device is a Open vSwitch master
D
David Ahern 已提交
1444
 * @IFF_L3MDEV_SLAVE: device is enslaved to an L3 master device
J
Jiri Pirko 已提交
1445
 * @IFF_TEAM: device is a team device
1446
 * @IFF_RXFH_CONFIGURED: device has had Rx Flow indirection table configured
1447 1448
 * @IFF_PHONY_HEADROOM: the headroom value is controlled by an external
 *	entity (i.e. the master device for bridged veth)
1449
 * @IFF_MACSEC: device is a MACsec device
P
Paolo Abeni 已提交
1450
 * @IFF_NO_RX_HANDLER: device doesn't support the rx_handler hook
1451 1452
 * @IFF_FAILOVER: device is a failover master device
 * @IFF_FAILOVER_SLAVE: device is lower dev of a failover master device
1453 1454 1455 1456
 */
enum netdev_priv_flags {
	IFF_802_1Q_VLAN			= 1<<0,
	IFF_EBRIDGE			= 1<<1,
1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472
	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,
P
Paolo Abeni 已提交
1473 1474 1475 1476 1477 1478 1479 1480
	IFF_L3MDEV_MASTER		= 1<<18,
	IFF_NO_QUEUE			= 1<<19,
	IFF_OPENVSWITCH			= 1<<20,
	IFF_L3MDEV_SLAVE		= 1<<21,
	IFF_TEAM			= 1<<22,
	IFF_RXFH_CONFIGURED		= 1<<23,
	IFF_PHONY_HEADROOM		= 1<<24,
	IFF_MACSEC			= 1<<25,
P
Paolo Abeni 已提交
1481
	IFF_NO_RX_HANDLER		= 1<<26,
1482 1483
	IFF_FAILOVER			= 1<<27,
	IFF_FAILOVER_SLAVE		= 1<<28,
1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502
};

#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
1503
#define IFF_XMIT_DST_RELEASE_PERM	IFF_XMIT_DST_RELEASE_PERM
1504
#define IFF_L3MDEV_MASTER		IFF_L3MDEV_MASTER
1505
#define IFF_NO_QUEUE			IFF_NO_QUEUE
1506
#define IFF_OPENVSWITCH			IFF_OPENVSWITCH
1507
#define IFF_L3MDEV_SLAVE		IFF_L3MDEV_SLAVE
J
Jiri Pirko 已提交
1508
#define IFF_TEAM			IFF_TEAM
1509
#define IFF_RXFH_CONFIGURED		IFF_RXFH_CONFIGURED
1510
#define IFF_MACSEC			IFF_MACSEC
P
Paolo Abeni 已提交
1511
#define IFF_NO_RX_HANDLER		IFF_NO_RX_HANDLER
1512 1513
#define IFF_FAILOVER			IFF_FAILOVER
#define IFF_FAILOVER_SLAVE		IFF_FAILOVER_SLAVE
1514

1515 1516
/**
 *	struct net_device - The DEVICE structure.
1517 1518 1519 1520
 *
 *	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.
1521 1522 1523
 *
 *	@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
1524
 *		of the interface.
1525 1526 1527 1528 1529 1530 1531 1532 1533 1534
 *
 *	@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
1535 1536 1537 1538
 *	@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
1539 1540
 *	@ptype_all:     Device-specific packet handlers for all protocols
 *	@ptype_specific: Device-specific, protocol-specific packet handlers
1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556
 *
 *	@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
1557
 *	@group:		The group the device belongs to
1558 1559 1560 1561 1562 1563 1564 1565
 *
 *	@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
1566 1567
 *	@rx_nohandler:	nohandler dropped packets by core network on
 *			inactive devices, do not use this in drivers
1568 1569
 *	@carrier_up_count:	Number of times the carrier has been up
 *	@carrier_down_count:	Number of times the carrier has been down
1570 1571 1572 1573 1574 1575 1576 1577 1578
 *
 *	@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
1579 1580
 *	@ndisc_ops:	Includes callbacks for different IPv6 neighbour
 *			discovery handling. Necessary for e.g. 6LoWPAN.
E
Eric W. Biederman 已提交
1581
 *	@header_ops:	Includes callbacks for creating,parsing,caching,etc
1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593
 *			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
1594 1595
 *	@min_mtu:	Interface Minimum MTU value
 *	@max_mtu:	Interface Maximum MTU value
1596
 *	@type:		Interface hardware type
1597
 *	@hard_header_len: Maximum hardware header length.
1598
 *	@min_header_len:  Minimum hardware header length
1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610
 *
 *	@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
1611
 *	@neigh_priv_len:	Used in neigh_alloc()
1612 1613 1614 1615 1616
 * 	@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
1617
 *	@uc_promisc:		Counter that indicates promiscuous mode
1618 1619 1620
 *				has been enabled due to the need to listen to
 *				additional unicast addresses in a device that
 *				does not implement ndo_set_rx_mode()
1621 1622 1623 1624
 *	@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
1625 1626 1627
 *	@promiscuity:		Number of times the NIC is told to work in
 *				promiscuous mode; if it becomes 0 the NIC will
 *				exit promiscuous mode
1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649
 *	@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
1650 1651
 *	@miniq_ingress:		ingress/clsact qdisc specific data for
 *				ingress processing
1652 1653 1654
 *	@ingress_queue:		XXX: need comments on this one
 *	@broadcast:		hw bcast address
 *
1655 1656 1657 1658 1659 1660
 *	@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
 *
1661 1662 1663 1664 1665 1666 1667 1668
 *	@_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
1669 1670
 *	@miniq_egress:		clsact qdisc specific data for
 *				egress processing
1671
 *	@watchdog_timeo:	Represents the timeout that is used by
B
Bjorn Helgaas 已提交
1672
 *				the watchdog (see dev_watchdog())
1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683
 *	@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
 *
1684 1685
 *	@needs_free_netdev:	Should unregister perform free_netdev?
 *	@priv_destructor:	Called from unregister
1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711
 *	@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
1712
 *	@prio_tc_map:	XXX: need comments on this one
1713 1714 1715 1716 1717 1718
 *
 *	@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
1719
 *	@sfp_bus:	attached &struct sfp_bus structure.
1720
 *
1721 1722
 *	@qdisc_tx_busylock: lockdep class annotating Qdisc->busylock spinlock
 *	@qdisc_running_key: lockdep class annotating Qdisc->running seqcount
1723
 *
1724 1725 1726 1727
 *	@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
1728 1729 1730 1731
 *	FIXME: cleanup struct net_device such that network protocol info
 *	moves out.
 */

E
Eric Dumazet 已提交
1732
struct net_device {
L
Linus Torvalds 已提交
1733
	char			name[IFNAMSIZ];
1734
	struct hlist_node	name_hlist;
1735
	struct dev_ifalias	__rcu *ifalias;
L
Linus Torvalds 已提交
1736 1737 1738 1739
	/*
	 *	I/O specific fields
	 *	FIXME: Merge these and struct ifmap into one
	 */
1740 1741 1742 1743
	unsigned long		mem_end;
	unsigned long		mem_start;
	unsigned long		base_addr;
	int			irq;
L
Linus Torvalds 已提交
1744 1745

	/*
1746 1747
	 *	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
1748 1749 1750 1751 1752
	 *	part of the usual set specified in Space.c.
	 */

	unsigned long		state;

1753
	struct list_head	dev_list;
1754
	struct list_head	napi_list;
1755
	struct list_head	unreg_list;
1756
	struct list_head	close_list;
1757 1758
	struct list_head	ptype_all;
	struct list_head	ptype_specific;
1759 1760 1761 1762 1763 1764

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

1765 1766 1767 1768
	netdev_features_t	features;
	netdev_features_t	hw_features;
	netdev_features_t	wanted_features;
	netdev_features_t	vlan_features;
1769
	netdev_features_t	hw_enc_features;
S
Simon Horman 已提交
1770
	netdev_features_t	mpls_features;
1771
	netdev_features_t	gso_partial_features;
1772

L
Linus Torvalds 已提交
1773
	int			ifindex;
1774
	int			group;
L
Linus Torvalds 已提交
1775

R
Rusty Russell 已提交
1776
	struct net_device_stats	stats;
1777 1778 1779

	atomic_long_t		rx_dropped;
	atomic_long_t		tx_dropped;
1780
	atomic_long_t		rx_nohandler;
L
Linus Torvalds 已提交
1781

1782 1783 1784 1785
	/* Stats to monitor link on/off, flapping */
	atomic_t		carrier_up_count;
	atomic_t		carrier_down_count;

1786
#ifdef CONFIG_WIRELESS_EXT
B
Bjorn Helgaas 已提交
1787 1788
	const struct iw_handler_def *wireless_handlers;
	struct iw_public_data	*wireless_data;
1789
#endif
1790
	const struct net_device_ops *netdev_ops;
1791
	const struct ethtool_ops *ethtool_ops;
S
Scott Feldman 已提交
1792
#ifdef CONFIG_NET_SWITCHDEV
J
Jiri Pirko 已提交
1793
	const struct switchdev_ops *switchdev_ops;
S
Scott Feldman 已提交
1794
#endif
1795 1796 1797
#ifdef CONFIG_NET_L3_MASTER_DEV
	const struct l3mdev_ops	*l3mdev_ops;
#endif
1798 1799 1800
#if IS_ENABLED(CONFIG_IPV6)
	const struct ndisc_ops *ndisc_ops;
#endif
L
Linus Torvalds 已提交
1801

1802
#ifdef CONFIG_XFRM_OFFLOAD
1803 1804 1805
	const struct xfrmdev_ops *xfrmdev_ops;
#endif

I
Ilya Lesokhin 已提交
1806 1807 1808 1809
#if IS_ENABLED(CONFIG_TLS_DEVICE)
	const struct tlsdev_ops *tlsdev_ops;
#endif

1810 1811
	const struct header_ops *header_ops;

1812 1813 1814
	unsigned int		flags;
	unsigned int		priv_flags;

L
Linus Torvalds 已提交
1815
	unsigned short		gflags;
1816
	unsigned short		padded;
L
Linus Torvalds 已提交
1817

1818 1819
	unsigned char		operstate;
	unsigned char		link_mode;
S
Stefan Rompf 已提交
1820

1821 1822
	unsigned char		if_port;
	unsigned char		dma;
1823

1824
	unsigned int		mtu;
1825 1826
	unsigned int		min_mtu;
	unsigned int		max_mtu;
1827 1828
	unsigned short		type;
	unsigned short		hard_header_len;
1829
	unsigned char		min_header_len;
L
Linus Torvalds 已提交
1830

1831 1832 1833
	unsigned short		needed_headroom;
	unsigned short		needed_tailroom;

L
Linus Torvalds 已提交
1834
	/* Interface address info. */
1835 1836 1837
	unsigned char		perm_addr[MAX_ADDR_LEN];
	unsigned char		addr_assign_type;
	unsigned char		addr_len;
1838
	unsigned short		neigh_priv_len;
1839 1840
	unsigned short          dev_id;
	unsigned short          dev_port;
J
Jiri Pirko 已提交
1841
	spinlock_t		addr_list_lock;
1842 1843
	unsigned char		name_assign_type;
	bool			uc_promisc;
1844 1845 1846 1847
	struct netdev_hw_addr_list	uc;
	struct netdev_hw_addr_list	mc;
	struct netdev_hw_addr_list	dev_addrs;

1848 1849 1850
#ifdef CONFIG_SYSFS
	struct kset		*queues_kset;
#endif
1851 1852
	unsigned int		promiscuity;
	unsigned int		allmulti;
L
Linus Torvalds 已提交
1853 1854


B
Bjorn Helgaas 已提交
1855
	/* Protocol-specific pointers */
1856

1857
#if IS_ENABLED(CONFIG_VLAN_8021Q)
1858
	struct vlan_info __rcu	*vlan_info;
1859
#endif
1860
#if IS_ENABLED(CONFIG_NET_DSA)
1861
	struct dsa_port		*dsa_ptr;
1862 1863
#endif
#if IS_ENABLED(CONFIG_TIPC)
1864
	struct tipc_bearer __rcu *tipc_ptr;
1865
#endif
1866
#if IS_ENABLED(CONFIG_IRDA) || IS_ENABLED(CONFIG_ATALK)
1867
	void 			*atalk_ptr;
1868
#endif
1869
	struct in_device __rcu	*ip_ptr;
1870
#if IS_ENABLED(CONFIG_DECNET)
1871
	struct dn_dev __rcu     *dn_ptr;
1872
#endif
1873
	struct inet6_dev __rcu	*ip6_ptr;
1874
#if IS_ENABLED(CONFIG_AX25)
1875
	void			*ax25_ptr;
1876
#endif
1877
	struct wireless_dev	*ieee80211_ptr;
1878
	struct wpan_dev		*ieee802154_ptr;
R
Robert Shearman 已提交
1879 1880 1881
#if IS_ENABLED(CONFIG_MPLS_ROUTING)
	struct mpls_dev __rcu	*mpls_ptr;
#endif
L
Linus Torvalds 已提交
1882

1883
/*
E
Eric Dumazet 已提交
1884
 * Cache lines mostly used on receive path (including eth_type_trans())
1885 1886
 */
	/* Interface address info used in eth_type_trans() */
1887
	unsigned char		*dev_addr;
1888

T
Tom Herbert 已提交
1889 1890
	struct netdev_rx_queue	*_rx;
	unsigned int		num_rx_queues;
1891
	unsigned int		real_num_rx_queues;
T
Tom Herbert 已提交
1892

1893
	struct bpf_prog __rcu	*xdp_prog;
1894
	unsigned long		gro_flush_timeout;
1895 1896
	rx_handler_func_t __rcu	*rx_handler;
	void __rcu		*rx_handler_data;
1897

1898
#ifdef CONFIG_NET_CLS_ACT
1899
	struct mini_Qdisc __rcu	*miniq_ingress;
1900
#endif
1901
	struct netdev_queue __rcu *ingress_queue;
1902
#ifdef CONFIG_NETFILTER_INGRESS
1903
	struct nf_hook_entries __rcu *nf_hooks_ingress;
1904
#endif
1905

1906
	unsigned char		broadcast[MAX_ADDR_LEN];
1907 1908 1909 1910
#ifdef CONFIG_RFS_ACCEL
	struct cpu_rmap		*rx_cpu_rmap;
#endif
	struct hlist_node	index_hlist;
E
Eric Dumazet 已提交
1911 1912 1913 1914

/*
 * Cache lines mostly used on transmit path
 */
1915 1916
	struct netdev_queue	*_tx ____cacheline_aligned_in_smp;
	unsigned int		num_tx_queues;
1917
	unsigned int		real_num_tx_queues;
1918
	struct Qdisc		*qdisc;
1919 1920 1921
#ifdef CONFIG_NET_SCHED
	DECLARE_HASHTABLE	(qdisc_hash, 4);
#endif
1922
	unsigned int		tx_queue_len;
1923
	spinlock_t		tx_global_lock;
1924
	int			watchdog_timeo;
E
Eric Dumazet 已提交
1925

T
Tom Herbert 已提交
1926
#ifdef CONFIG_XPS
1927 1928
	struct xps_dev_maps __rcu *xps_cpus_map;
	struct xps_dev_maps __rcu *xps_rxqs_map;
T
Tom Herbert 已提交
1929
#endif
D
Daniel Borkmann 已提交
1930
#ifdef CONFIG_NET_CLS_ACT
1931
	struct mini_Qdisc __rcu	*miniq_egress;
D
Daniel Borkmann 已提交
1932
#endif
1933

1934 1935 1936
	/* These may be needed for future network-power-down code. */
	struct timer_list	watchdog_timer;

E
Eric Dumazet 已提交
1937
	int __percpu		*pcpu_refcnt;
L
Linus Torvalds 已提交
1938 1939
	struct list_head	todo_list;

1940
	struct list_head	link_watch_list;
1941

L
Linus Torvalds 已提交
1942
	enum { NETREG_UNINITIALIZED=0,
1943
	       NETREG_REGISTERED,	/* completed register_netdevice */
L
Linus Torvalds 已提交
1944 1945 1946
	       NETREG_UNREGISTERING,	/* called unregister_netdevice */
	       NETREG_UNREGISTERED,	/* completed unregister todo */
	       NETREG_RELEASED,		/* called free_netdev */
1947
	       NETREG_DUMMY,		/* dummy device for NAPI poll */
1948 1949
	} reg_state:8;

1950
	bool dismantle;
1951 1952 1953 1954 1955

	enum {
		RTNL_LINK_INITIALIZED,
		RTNL_LINK_INITIALIZING,
	} rtnl_link_state:16;
L
Linus Torvalds 已提交
1956

1957 1958
	bool needs_free_netdev;
	void (*priv_destructor)(struct net_device *dev);
L
Linus Torvalds 已提交
1959 1960

#ifdef CONFIG_NETPOLL
1961
	struct netpoll_info __rcu	*npinfo;
L
Linus Torvalds 已提交
1962
#endif
1963

1964
	possible_net_t			nd_net;
1965

D
David S. Miller 已提交
1966
	/* mid-layer private */
E
Eric Dumazet 已提交
1967
	union {
1968 1969
		void					*ml_priv;
		struct pcpu_lstats __percpu		*lstats;
1970
		struct pcpu_sw_netstats __percpu	*tstats;
1971 1972
		struct pcpu_dstats __percpu		*dstats;
		struct pcpu_vstats __percpu		*vstats;
E
Eric Dumazet 已提交
1973
	};
1974

1975
#if IS_ENABLED(CONFIG_GARP)
E
Eric Dumazet 已提交
1976
	struct garp_port __rcu	*garp_port;
1977 1978
#endif
#if IS_ENABLED(CONFIG_MRP)
1979
	struct mrp_port __rcu	*mrp_port;
1980
#endif
L
Linus Torvalds 已提交
1981

B
Bjorn Helgaas 已提交
1982
	struct device		dev;
1983
	const struct attribute_group *sysfs_groups[4];
1984
	const struct attribute_group *sysfs_rx_queue_group;
P
Patrick McHardy 已提交
1985 1986

	const struct rtnl_link_ops *rtnl_link_ops;
1987

1988 1989 1990
	/* for setting kernel sock attribute on TCP connection setup */
#define GSO_MAX_SIZE		65536
	unsigned int		gso_max_size;
1991 1992
#define GSO_MAX_SEGS		65535
	u16			gso_max_segs;
1993

J
Jeff Kirsher 已提交
1994
#ifdef CONFIG_DCB
1995
	const struct dcbnl_rtnl_ops *dcbnl_ops;
1996
#endif
1997
	s16			num_tc;
B
Bjorn Helgaas 已提交
1998 1999
	struct netdev_tc_txq	tc_to_txq[TC_MAX_QUEUE];
	u8			prio_tc_map[TC_BITMASK + 1];
2000

2001
#if IS_ENABLED(CONFIG_FCOE)
2002
	unsigned int		fcoe_ddp_xid;
2003
#endif
2004
#if IS_ENABLED(CONFIG_CGROUP_NET_PRIO)
2005
	struct netprio_map __rcu *priomap;
2006
#endif
B
Bjorn Helgaas 已提交
2007
	struct phy_device	*phydev;
2008
	struct sfp_bus		*sfp_bus;
B
Bjorn Helgaas 已提交
2009
	struct lock_class_key	*qdisc_tx_busylock;
2010
	struct lock_class_key	*qdisc_running_key;
B
Bjorn Helgaas 已提交
2011
	bool			proto_down;
L
Linus Torvalds 已提交
2012
};
2013
#define to_net_dev(d) container_of(d, struct net_device, dev)
L
Linus Torvalds 已提交
2014

D
David S. Miller 已提交
2015 2016 2017 2018 2019 2020 2021
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
2022 2023
#define	NETDEV_ALIGN		32

2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039
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;
}

2040
int netdev_txq_to_tc(struct net_device *dev, unsigned int txq);
2041 2042 2043
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);
2044 2045 2046 2047 2048 2049 2050

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

2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061
void netdev_unbind_sb_channel(struct net_device *dev,
			      struct net_device *sb_dev);
int netdev_bind_sb_channel_queue(struct net_device *dev,
				 struct net_device *sb_dev,
				 u8 tc, u16 count, u16 offset);
int netdev_set_sb_channel(struct net_device *dev, u16 channel);
static inline int netdev_get_sb_channel(struct net_device *dev)
{
	return max_t(int, -dev->num_tc, 0);
}

2062 2063 2064 2065 2066 2067 2068
static inline
struct netdev_queue *netdev_get_tx_queue(const struct net_device *dev,
					 unsigned int index)
{
	return &dev->_tx[index];
}

2069 2070 2071 2072 2073 2074
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));
}

2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086
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);
}

2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103
#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);	\
}

2104
struct netdev_queue *netdev_pick_tx(struct net_device *dev,
2105 2106
				    struct sk_buff *skb,
				    void *accel_priv);
2107

2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127
/* 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);
}

2128 2129 2130 2131 2132 2133
/*
 * Net namespace inlines
 */
static inline
struct net *dev_net(const struct net_device *dev)
{
E
Eric Dumazet 已提交
2134
	return read_pnet(&dev->nd_net);
2135 2136 2137
}

static inline
2138
void dev_net_set(struct net_device *dev, struct net *net)
2139
{
2140
	write_pnet(&dev->nd_net, net);
2141 2142
}

2143 2144 2145 2146 2147 2148
/**
 *	netdev_priv - access network device private data
 *	@dev: network device
 *
 * Get network device private data
 */
2149
static inline void *netdev_priv(const struct net_device *dev)
L
Linus Torvalds 已提交
2150
{
2151
	return (char *)dev + ALIGN(sizeof(struct net_device), NETDEV_ALIGN);
L
Linus Torvalds 已提交
2152 2153 2154 2155 2156
}

/* Set the sysfs physical device reference for the network logical device
 * if set prior to registration will cause a symlink during initialization.
 */
2157
#define SET_NETDEV_DEV(net, pdev)	((net)->dev.parent = (pdev))
L
Linus Torvalds 已提交
2158

2159
/* Set the sysfs device type for the network logical device to allow
2160
 * fine-grained identification of different network device types. For
B
Bjorn Helgaas 已提交
2161
 * example Ethernet, Wireless LAN, Bluetooth, WiMAX etc.
2162 2163 2164
 */
#define SET_NETDEV_DEVTYPE(net, devtype)	((net)->dev.type = (devtype))

E
Eric Dumazet 已提交
2165 2166 2167 2168 2169
/* Default NAPI poll() weight
 * Device drivers are strongly advised to not use bigger value
 */
#define NAPI_POLL_WEIGHT 64

2170
/**
B
Bjorn Helgaas 已提交
2171
 *	netif_napi_add - initialize a NAPI context
2172
 *	@dev:  network device
B
Bjorn Helgaas 已提交
2173
 *	@napi: NAPI context
2174 2175 2176
 *	@poll: polling function
 *	@weight: default weight
 *
B
Bjorn Helgaas 已提交
2177 2178
 * netif_napi_add() must be used to initialize a NAPI context prior to calling
 * *any* of the other NAPI-related functions.
2179
 */
2180 2181
void netif_napi_add(struct net_device *dev, struct napi_struct *napi,
		    int (*poll)(struct napi_struct *, int), int weight);
2182

E
Eric Dumazet 已提交
2183
/**
B
Bjorn Helgaas 已提交
2184
 *	netif_tx_napi_add - initialize a NAPI context
E
Eric Dumazet 已提交
2185
 *	@dev:  network device
B
Bjorn Helgaas 已提交
2186
 *	@napi: NAPI context
E
Eric Dumazet 已提交
2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202
 *	@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);
}

2203
/**
B
Bjorn Helgaas 已提交
2204 2205
 *  netif_napi_del - remove a NAPI context
 *  @napi: NAPI context
2206
 *
B
Bjorn Helgaas 已提交
2207
 *  netif_napi_del() removes a NAPI context from the network device NAPI list
2208
 */
2209 2210 2211
void netif_napi_del(struct napi_struct *napi);

struct napi_gro_cb {
2212
	/* Virtual address of skb_shinfo(skb)->frags[0].page + offset. */
B
Bjorn Helgaas 已提交
2213
	void	*frag0;
2214

2215 2216 2217
	/* Length of frag0. */
	unsigned int frag0_len;

2218
	/* This indicates where we are processing relative to skb->data. */
B
Bjorn Helgaas 已提交
2219
	int	data_offset;
2220

2221
	/* This is non-zero if the packet cannot be merged with the new skb. */
2222 2223 2224 2225
	u16	flush;

	/* Save the IP ID here and check when we get to the transport layer */
	u16	flush_id;
2226 2227

	/* Number of segments aggregated. */
2228 2229
	u16	count;

2230 2231 2232
	/* Start offset for remote checksum offload */
	u16	gro_remcsum_start;

2233 2234
	/* jiffies when first packet was created/queued */
	unsigned long age;
2235

T
Tom Herbert 已提交
2236
	/* Used in ipv6_gro_receive() and foo-over-udp */
2237 2238
	u16	proto;

2239 2240 2241
	/* This is non-zero if the packet may be of the same flow. */
	u8	same_flow:1;

2242 2243
	/* Used in tunnel GRO receive */
	u8	encap_mark:1;
T
Tom Herbert 已提交
2244 2245 2246 2247

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

2248 2249
	/* Number of checksums via CHECKSUM_UNNECESSARY */
	u8	csum_cnt:3;
2250

2251 2252 2253 2254 2255
	/* Free the skb? */
	u8	free:2;
#define NAPI_GRO_FREE		  1
#define NAPI_GRO_FREE_STOLEN_HEAD 2

2256 2257 2258
	/* Used in foo-over-udp, set in udp[46]_gro_receive */
	u8	is_ipv6:1;

2259 2260 2261
	/* Used in GRE, set in fou/gue_gro_receive */
	u8	is_fou:1;

2262 2263 2264
	/* Used to determine if flush_id can be ignored */
	u8	is_atomic:1;

S
Sabrina Dubroca 已提交
2265 2266 2267 2268
	/* Number of gro_receive callbacks this packet already went through */
	u8 recursion_counter:4;

	/* 1 bit hole */
2269

2270 2271 2272
	/* used to support CHECKSUM_COMPLETE for tunneling protocols */
	__wsum	csum;

2273 2274
	/* used in skb_gro_receive() slow path */
	struct sk_buff *last;
2275 2276 2277
};

#define NAPI_GRO_CB(skb) ((struct napi_gro_cb *)(skb)->cb)
2278

S
Sabrina Dubroca 已提交
2279 2280 2281 2282 2283 2284
#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;
}

2285 2286 2287 2288
typedef struct sk_buff *(*gro_receive_t)(struct list_head *, struct sk_buff *);
static inline struct sk_buff *call_gro_receive(gro_receive_t cb,
					       struct list_head *head,
					       struct sk_buff *skb)
S
Sabrina Dubroca 已提交
2289 2290 2291 2292 2293 2294 2295 2296 2297
{
	if (unlikely(gro_recursion_inc_test(skb))) {
		NAPI_GRO_CB(skb)->flush |= 1;
		return NULL;
	}

	return cb(head, skb);
}

2298 2299 2300 2301 2302 2303
typedef struct sk_buff *(*gro_receive_sk_t)(struct sock *, struct list_head *,
					    struct sk_buff *);
static inline struct sk_buff *call_gro_receive_sk(gro_receive_sk_t cb,
						  struct sock *sk,
						  struct list_head *head,
						  struct sk_buff *skb)
S
Sabrina Dubroca 已提交
2304 2305 2306 2307 2308 2309 2310 2311 2312
{
	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
2313
struct packet_type {
D
David S. Miller 已提交
2314 2315 2316 2317 2318 2319
	__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 *);
2320 2321 2322
	void			(*list_func) (struct list_head *,
					      struct packet_type *,
					      struct net_device *);
2323 2324
	bool			(*id_match)(struct packet_type *ptype,
					    struct sock *sk);
L
Linus Torvalds 已提交
2325 2326 2327 2328
	void			*af_packet_priv;
	struct list_head	list;
};

2329
struct offload_callbacks {
2330
	struct sk_buff		*(*gso_segment)(struct sk_buff *skb,
2331
						netdev_features_t features);
2332 2333
	struct sk_buff		*(*gro_receive)(struct list_head *head,
						struct sk_buff *skb);
2334
	int			(*gro_complete)(struct sk_buff *skb, int nhoff);
2335 2336 2337 2338
};

struct packet_offload {
	__be16			 type;	/* This is really htons(ether_type). */
2339
	u16			 priority;
2340 2341
	struct offload_callbacks callbacks;
	struct list_head	 list;
L
Linus Torvalds 已提交
2342 2343
};

B
Bjorn Helgaas 已提交
2344
/* often modified stats are per-CPU, other are shared (netdev->stats) */
2345 2346 2347 2348 2349 2350 2351 2352
struct pcpu_sw_netstats {
	u64     rx_packets;
	u64     rx_bytes;
	u64     tx_packets;
	u64     tx_bytes;
	struct u64_stats_sync   syncp;
};

2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364
#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;							\
2365 2366
})

2367
#define netdev_alloc_pcpu_stats(type)					\
2368
	__netdev_alloc_pcpu_stats(type, GFP_KERNEL)
2369

2370 2371 2372 2373 2374 2375 2376 2377 2378
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,
};

2379 2380 2381 2382 2383 2384 2385 2386 2387 2388
enum netdev_lag_hash {
	NETDEV_LAG_HASH_NONE,
	NETDEV_LAG_HASH_L2,
	NETDEV_LAG_HASH_L34,
	NETDEV_LAG_HASH_L23,
	NETDEV_LAG_HASH_E23,
	NETDEV_LAG_HASH_E34,
	NETDEV_LAG_HASH_UNKNOWN,
};

2389 2390
struct netdev_lag_upper_info {
	enum netdev_lag_tx_type tx_type;
2391
	enum netdev_lag_hash hash_type;
2392 2393
};

2394 2395 2396 2397 2398
struct netdev_lag_lower_state_info {
	u8 link_up : 1,
	   tx_enabled : 1;
};

L
Linus Torvalds 已提交
2399 2400
#include <linux/notifier.h>

2401 2402 2403
/* netdevice notifier chain. Please remember to update netdev_cmd_to_name()
 * and the rtnetlink notification exclusion list in rtnetlink_event() when
 * adding new types.
2404
 */
2405 2406 2407 2408
enum netdev_cmd {
	NETDEV_UP	= 1,	/* For now you can't veto a device up/down */
	NETDEV_DOWN,
	NETDEV_REBOOT,		/* Tell a protocol stack a network interface
2409 2410 2411
				   detected a hardware crash and restarted
				   - we can use this eg to kick tcp sessions
				   once done */
2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437
	NETDEV_CHANGE,		/* Notify device state change */
	NETDEV_REGISTER,
	NETDEV_UNREGISTER,
	NETDEV_CHANGEMTU,	/* notify after mtu change happened */
	NETDEV_CHANGEADDR,
	NETDEV_GOING_DOWN,
	NETDEV_CHANGENAME,
	NETDEV_FEAT_CHANGE,
	NETDEV_BONDING_FAILOVER,
	NETDEV_PRE_UP,
	NETDEV_PRE_TYPE_CHANGE,
	NETDEV_POST_TYPE_CHANGE,
	NETDEV_POST_INIT,
	NETDEV_RELEASE,
	NETDEV_NOTIFY_PEERS,
	NETDEV_JOIN,
	NETDEV_CHANGEUPPER,
	NETDEV_RESEND_IGMP,
	NETDEV_PRECHANGEMTU,	/* notify before mtu change happened */
	NETDEV_CHANGEINFODATA,
	NETDEV_BONDING_INFO,
	NETDEV_PRECHANGEUPPER,
	NETDEV_CHANGELOWERSTATE,
	NETDEV_UDP_TUNNEL_PUSH_INFO,
	NETDEV_UDP_TUNNEL_DROP_INFO,
	NETDEV_CHANGE_TX_QUEUE_LEN,
2438 2439 2440 2441
	NETDEV_CVLAN_FILTER_PUSH_INFO,
	NETDEV_CVLAN_FILTER_DROP_INFO,
	NETDEV_SVLAN_FILTER_PUSH_INFO,
	NETDEV_SVLAN_FILTER_DROP_INFO,
2442 2443
};
const char *netdev_cmd_to_name(enum netdev_cmd cmd);
2444

2445 2446
int register_netdevice_notifier(struct notifier_block *nb);
int unregister_netdevice_notifier(struct notifier_block *nb);
2447 2448

struct netdev_notifier_info {
2449 2450
	struct net_device	*dev;
	struct netlink_ext_ack	*extack;
2451 2452
};

2453 2454 2455 2456 2457
struct netdev_notifier_change_info {
	struct netdev_notifier_info info; /* must be first */
	unsigned int flags_changed;
};

2458 2459 2460 2461
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
2462
	bool linking; /* is the notification for link or unlink */
2463
	void *upper_info; /* upper dev info */
2464 2465
};

2466 2467 2468 2469 2470
struct netdev_notifier_changelowerstate_info {
	struct netdev_notifier_info info; /* must be first */
	void *lower_state_info; /* is lower dev state */
};

2471 2472 2473 2474
static inline void netdev_notifier_info_init(struct netdev_notifier_info *info,
					     struct net_device *dev)
{
	info->dev = dev;
2475
	info->extack = NULL;
2476 2477
}

2478 2479 2480 2481 2482 2483
static inline struct net_device *
netdev_notifier_info_to_dev(const struct netdev_notifier_info *info)
{
	return info->dev;
}

2484 2485 2486 2487 2488 2489
static inline struct netlink_ext_ack *
netdev_notifier_info_to_extack(const struct netdev_notifier_info *info)
{
	return info->extack;
}

2490
int call_netdevice_notifiers(unsigned long val, struct net_device *dev);
2491 2492


L
Linus Torvalds 已提交
2493 2494
extern rwlock_t				dev_base_lock;		/* Device list lock */

2495 2496
#define for_each_netdev(net, d)		\
		list_for_each_entry(d, &(net)->dev_base_head, dev_list)
2497 2498
#define for_each_netdev_reverse(net, d)	\
		list_for_each_entry_reverse(d, &(net)->dev_base_head, dev_list)
2499 2500
#define for_each_netdev_rcu(net, d)		\
		list_for_each_entry_rcu(d, &(net)->dev_base_head, dev_list)
2501 2502 2503 2504
#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)
2505 2506
#define for_each_netdev_continue_rcu(net, d)		\
	list_for_each_entry_continue_rcu(d, &(net)->dev_base_head, dev_list)
2507 2508
#define for_each_netdev_in_bond_rcu(bond, slave)	\
		for_each_netdev_rcu(&init_net, slave)	\
2509
			if (netdev_master_upper_dev_get_rcu(slave) == (bond))
2510
#define net_device_entry(lh)	list_entry(lh, struct net_device, dev_list)
2511

2512 2513 2514 2515 2516
static inline struct net_device *next_net_device(struct net_device *dev)
{
	struct list_head *lh;
	struct net *net;

2517
	net = dev_net(dev);
2518 2519 2520 2521
	lh = dev->dev_list.next;
	return lh == &net->dev_base_head ? NULL : net_device_entry(lh);
}

2522 2523 2524 2525 2526 2527
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
2528
	lh = rcu_dereference(list_next_rcu(&dev->dev_list));
2529 2530 2531
	return lh == &net->dev_base_head ? NULL : net_device_entry(lh);
}

2532 2533 2534 2535 2536
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);
}
2537

E
Eric Dumazet 已提交
2538 2539 2540 2541 2542 2543 2544
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);
}

2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556
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);

2557
int dev_get_iflink(const struct net_device *dev);
2558
int dev_fill_metadata_dst(struct net_device *dev, struct sk_buff *skb);
2559 2560
struct net_device *__dev_get_by_flags(struct net *net, unsigned short flags,
				      unsigned short mask);
2561 2562 2563 2564 2565
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);
2566 2567
void dev_close(struct net_device *dev);
void dev_close_many(struct list_head *head, bool unlink);
2568
void dev_disable_lro(struct net_device *dev);
2569
int dev_loopback_xmit(struct net *net, struct sock *sk, struct sk_buff *newskb);
2570
int dev_queue_xmit(struct sk_buff *skb);
2571
int dev_queue_xmit_accel(struct sk_buff *skb, void *accel_priv);
2572
int dev_direct_xmit(struct sk_buff *skb, u16 queue_id);
2573 2574 2575
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);
2576 2577 2578 2579 2580
static inline void unregister_netdevice(struct net_device *dev)
{
	unregister_netdevice_queue(dev, NULL);
}

2581 2582
int netdev_refcnt_read(const struct net_device *dev);
void free_netdev(struct net_device *dev);
2583
void netdev_freemem(struct net_device *dev);
2584 2585
void synchronize_net(void);
int init_dummy_netdev(struct net_device *dev);
2586

2587
DECLARE_PER_CPU(int, xmit_recursion);
2588 2589
#define XMIT_RECURSION_LIMIT	10

2590 2591 2592 2593 2594
static inline int dev_recursion_level(void)
{
	return this_cpu_read(xmit_recursion);
}

2595 2596 2597
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);
2598
struct net_device *dev_get_by_napi_id(unsigned int napi_id);
2599 2600
int netdev_get_name(struct net *net, char *name, int ifindex);
int dev_restart(struct net_device *dev);
2601
int skb_gro_receive(struct sk_buff *p, struct sk_buff *skb);
2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617

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;
}

2618 2619
static inline void *skb_gro_header_fast(struct sk_buff *skb,
					unsigned int offset)
2620
{
2621 2622
	return NAPI_GRO_CB(skb)->frag0 + offset;
}
2623

2624 2625 2626 2627
static inline int skb_gro_header_hard(struct sk_buff *skb, unsigned int hlen)
{
	return NAPI_GRO_CB(skb)->frag0_len < hlen;
}
2628

2629 2630 2631 2632 2633 2634
static inline void skb_gro_frag0_invalidate(struct sk_buff *skb)
{
	NAPI_GRO_CB(skb)->frag0 = NULL;
	NAPI_GRO_CB(skb)->frag0_len = 0;
}

2635 2636 2637
static inline void *skb_gro_header_slow(struct sk_buff *skb, unsigned int hlen,
					unsigned int offset)
{
2638 2639 2640
	if (!pskb_may_pull(skb, hlen))
		return NULL;

2641
	skb_gro_frag0_invalidate(skb);
2642
	return skb->data + offset;
2643
}
L
Linus Torvalds 已提交
2644

H
Herbert Xu 已提交
2645 2646
static inline void *skb_gro_network_header(struct sk_buff *skb)
{
2647 2648
	return (NAPI_GRO_CB(skb)->frag0 ?: skb->data) +
	       skb_network_offset(skb);
H
Herbert Xu 已提交
2649 2650
}

2651 2652 2653
static inline void skb_gro_postpull_rcsum(struct sk_buff *skb,
					const void *start, unsigned int len)
{
T
Tom Herbert 已提交
2654
	if (NAPI_GRO_CB(skb)->csum_valid)
2655 2656 2657 2658
		NAPI_GRO_CB(skb)->csum = csum_sub(NAPI_GRO_CB(skb)->csum,
						  csum_partial(start, len, 0));
}

T
Tom Herbert 已提交
2659 2660 2661 2662 2663 2664 2665
/* 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);

2666 2667
static inline bool skb_at_gro_remcsum_start(struct sk_buff *skb)
{
2668
	return (NAPI_GRO_CB(skb)->gro_remcsum_start == skb_gro_offset(skb));
2669 2670
}

T
Tom Herbert 已提交
2671 2672 2673 2674
static inline bool __skb_gro_checksum_validate_needed(struct sk_buff *skb,
						      bool zero_okay,
						      __sum16 check)
{
2675 2676 2677
	return ((skb->ip_summed != CHECKSUM_PARTIAL ||
		skb_checksum_start_offset(skb) <
		 skb_gro_offset(skb)) &&
2678
		!skb_at_gro_remcsum_start(skb) &&
2679
		NAPI_GRO_CB(skb)->csum_cnt == 0 &&
T
Tom Herbert 已提交
2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696
		(!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)
{
2697 2698 2699 2700 2701 2702 2703 2704 2705
	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
2706 2707 2708 2709 2710 2711 2712 2713 2714 2715
	}
}

#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));		\
2716
	if (!__ret)							\
T
Tom Herbert 已提交
2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730
		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)

2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750
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)

2751 2752 2753 2754 2755 2756 2757
struct gro_remcsum {
	int offset;
	__wsum delta;
};

static inline void skb_gro_remcsum_init(struct gro_remcsum *grc)
{
2758
	grc->offset = 0;
2759 2760 2761
	grc->delta = 0;
}

2762 2763 2764 2765 2766
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)
2767 2768
{
	__wsum delta;
2769
	size_t plen = hdrlen + max_t(size_t, offset + sizeof(u16), start);
2770 2771 2772

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

2773
	if (!nopartial) {
2774 2775 2776 2777 2778 2779 2780 2781 2782
		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;
2783 2784
	}

2785 2786
	delta = remcsum_adjust(ptr + hdrlen, NAPI_GRO_CB(skb)->csum,
			       start, offset);
2787 2788 2789

	/* Adjust skb->csum since we changed the packet */
	NAPI_GRO_CB(skb)->csum = csum_add(NAPI_GRO_CB(skb)->csum, delta);
2790

2791
	grc->offset = off + hdrlen + offset;
2792
	grc->delta = delta;
2793 2794

	return ptr;
2795 2796
}

2797 2798 2799
static inline void skb_gro_remcsum_cleanup(struct sk_buff *skb,
					   struct gro_remcsum *grc)
{
2800 2801 2802
	void *ptr;
	size_t plen = grc->offset + sizeof(u16);

2803 2804 2805
	if (!grc->delta)
		return;

2806 2807 2808 2809 2810 2811 2812 2813
	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);
2814
}
2815

2816
#ifdef CONFIG_XFRM_OFFLOAD
2817
static inline void skb_gro_flush_final(struct sk_buff *skb, struct sk_buff *pp, int flush)
2818 2819 2820 2821
{
	if (PTR_ERR(pp) != -EINPROGRESS)
		NAPI_GRO_CB(skb)->flush |= flush;
}
2822
static inline void skb_gro_flush_final_remcsum(struct sk_buff *skb,
2823
					       struct sk_buff *pp,
2824 2825 2826 2827 2828 2829 2830 2831 2832
					       int flush,
					       struct gro_remcsum *grc)
{
	if (PTR_ERR(pp) != -EINPROGRESS) {
		NAPI_GRO_CB(skb)->flush |= flush;
		skb_gro_remcsum_cleanup(skb, grc);
		skb->remcsum_offload = 0;
	}
}
2833
#else
2834
static inline void skb_gro_flush_final(struct sk_buff *skb, struct sk_buff *pp, int flush)
2835 2836 2837
{
	NAPI_GRO_CB(skb)->flush |= flush;
}
2838
static inline void skb_gro_flush_final_remcsum(struct sk_buff *skb,
2839
					       struct sk_buff *pp,
2840 2841 2842 2843 2844 2845 2846
					       int flush,
					       struct gro_remcsum *grc)
{
	NAPI_GRO_CB(skb)->flush |= flush;
	skb_gro_remcsum_cleanup(skb, grc);
	skb->remcsum_offload = 0;
}
2847
#endif
2848

2849 2850
static inline int dev_hard_header(struct sk_buff *skb, struct net_device *dev,
				  unsigned short type,
2851
				  const void *daddr, const void *saddr,
2852
				  unsigned int len)
2853
{
2854
	if (!dev->header_ops || !dev->header_ops->create)
2855
		return 0;
2856 2857

	return dev->header_ops->create(skb, dev, type, daddr, saddr, len);
2858 2859
}

S
Stephen Hemminger 已提交
2860 2861 2862 2863 2864
static inline int dev_parse_header(const struct sk_buff *skb,
				   unsigned char *haddr)
{
	const struct net_device *dev = skb->dev;

2865
	if (!dev->header_ops || !dev->header_ops->parse)
S
Stephen Hemminger 已提交
2866
		return 0;
2867
	return dev->header_ops->parse(skb, haddr);
S
Stephen Hemminger 已提交
2868 2869
}

2870 2871 2872 2873 2874 2875
/* 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;
2876 2877
	if (len < dev->min_header_len)
		return false;
2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889

	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;
}

2890 2891
typedef int gifconf_func_t(struct net_device * dev, char __user * bufptr,
			   int len, int size);
2892
int register_gifconf(unsigned int family, gifconf_func_t *gifconf);
L
Linus Torvalds 已提交
2893 2894 2895 2896 2897
static inline int unregister_gifconf(unsigned int family)
{
	return register_gifconf(family, NULL);
}

2898
#ifdef CONFIG_NET_FLOW_LIMIT
2899
#define FLOW_LIMIT_HISTORY	(1 << 7)  /* must be ^2 and !overflow buckets */
2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910
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
2911
/*
B
Bjorn Helgaas 已提交
2912
 * Incoming packets are placed on per-CPU queues
L
Linus Torvalds 已提交
2913
 */
E
Eric Dumazet 已提交
2914
struct softnet_data {
L
Linus Torvalds 已提交
2915
	struct list_head	poll_list;
2916
	struct sk_buff_head	process_queue;
L
Linus Torvalds 已提交
2917

C
Changli Gao 已提交
2918
	/* stats */
2919 2920 2921
	unsigned int		processed;
	unsigned int		time_squeeze;
	unsigned int		received_rps;
2922
#ifdef CONFIG_RPS
E
Eric Dumazet 已提交
2923
	struct softnet_data	*rps_ipi_list;
E
Eric Dumazet 已提交
2924 2925 2926 2927 2928 2929 2930
#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;
2931 2932 2933
#ifdef CONFIG_XFRM_OFFLOAD
	struct sk_buff_head	xfrm_backlog;
#endif
E
Eric Dumazet 已提交
2934
#ifdef CONFIG_RPS
2935 2936 2937 2938 2939 2940
	/* 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 */
2941
	call_single_data_t	csd ____cacheline_aligned_in_smp;
E
Eric Dumazet 已提交
2942 2943
	struct softnet_data	*rps_ipi_next;
	unsigned int		cpu;
2944
	unsigned int		input_queue_tail;
2945
#endif
2946
	unsigned int		dropped;
T
Tom Herbert 已提交
2947
	struct sk_buff_head	input_pkt_queue;
2948
	struct napi_struct	backlog;
2949

L
Linus Torvalds 已提交
2950 2951
};

2952
static inline void input_queue_head_incr(struct softnet_data *sd)
T
Tom Herbert 已提交
2953 2954
{
#ifdef CONFIG_RPS
2955 2956 2957 2958 2959 2960 2961 2962 2963
	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
2964 2965 2966
#endif
}

T
Tom Herbert 已提交
2967
DECLARE_PER_CPU_ALIGNED(struct softnet_data, softnet_data);
L
Linus Torvalds 已提交
2968

2969
void __netif_schedule(struct Qdisc *q);
2970
void netif_schedule_queue(struct netdev_queue *txq);
2971

2972 2973 2974 2975 2976 2977 2978 2979
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));
}

2980
static __always_inline void netif_tx_start_queue(struct netdev_queue *dev_queue)
2981
{
2982
	clear_bit(__QUEUE_STATE_DRV_XOFF, &dev_queue->state);
2983 2984
}

2985 2986 2987 2988 2989 2990
/**
 *	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
2991 2992
static inline void netif_start_queue(struct net_device *dev)
{
2993
	netif_tx_start_queue(netdev_get_tx_queue(dev, 0));
L
Linus Torvalds 已提交
2994 2995
}

2996 2997 2998 2999 3000 3001 3002 3003 3004 3005
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);
	}
}

3006
void netif_tx_wake_queue(struct netdev_queue *dev_queue);
3007

3008 3009 3010 3011 3012 3013 3014
/**
 *	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.
 */
3015 3016
static inline void netif_wake_queue(struct net_device *dev)
{
3017
	netif_tx_wake_queue(netdev_get_tx_queue(dev, 0));
L
Linus Torvalds 已提交
3018 3019
}

3020 3021 3022 3023 3024 3025 3026 3027 3028 3029
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);
	}
}

3030
static __always_inline void netif_tx_stop_queue(struct netdev_queue *dev_queue)
3031
{
3032
	set_bit(__QUEUE_STATE_DRV_XOFF, &dev_queue->state);
3033 3034
}

3035 3036 3037 3038 3039 3040 3041
/**
 *	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
3042 3043
static inline void netif_stop_queue(struct net_device *dev)
{
3044
	netif_tx_stop_queue(netdev_get_tx_queue(dev, 0));
L
Linus Torvalds 已提交
3045 3046
}

3047
void netif_tx_stop_all_queues(struct net_device *dev);
3048

3049
static inline bool netif_tx_queue_stopped(const struct netdev_queue *dev_queue)
3050
{
3051
	return test_bit(__QUEUE_STATE_DRV_XOFF, &dev_queue->state);
3052 3053
}

3054 3055 3056 3057 3058 3059
/**
 *	netif_queue_stopped - test if transmit queue is flowblocked
 *	@dev: network device
 *
 *	Test if transmit queue on device is currently unable to send.
 */
3060
static inline bool netif_queue_stopped(const struct net_device *dev)
L
Linus Torvalds 已提交
3061
{
3062
	return netif_tx_queue_stopped(netdev_get_tx_queue(dev, 0));
L
Linus Torvalds 已提交
3063 3064
}

3065
static inline bool netif_xmit_stopped(const struct netdev_queue *dev_queue)
3066
{
3067 3068 3069
	return dev_queue->state & QUEUE_STATE_ANY_XOFF;
}

3070 3071
static inline bool
netif_xmit_frozen_or_stopped(const struct netdev_queue *dev_queue)
3072 3073 3074 3075
{
	return dev_queue->state & QUEUE_STATE_ANY_XOFF_OR_FROZEN;
}

3076 3077 3078 3079 3080 3081
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;
}

3082 3083 3084 3085 3086
/**
 *	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
3087
 * to give appropriate hint to the CPU.
3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100
 */
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
3101
 * to give appropriate hint to the CPU.
3102 3103 3104 3105 3106 3107 3108 3109
 */
static inline void netdev_txq_bql_complete_prefetchw(struct netdev_queue *dev_queue)
{
#ifdef CONFIG_BQL
	prefetchw(&dev_queue->dql.limit);
#endif
}

3110 3111 3112
static inline void netdev_tx_sent_queue(struct netdev_queue *dev_queue,
					unsigned int bytes)
{
T
Tom Herbert 已提交
3113 3114
#ifdef CONFIG_BQL
	dql_queued(&dev_queue->dql, bytes);
3115 3116 3117 3118 3119 3120 3121 3122 3123 3124 3125 3126 3127 3128 3129 3130

	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
3131
#endif
3132 3133
}

3134 3135 3136 3137 3138 3139 3140 3141 3142
/**
 * 	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
 */
3143 3144 3145 3146 3147 3148
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,
3149
					     unsigned int pkts, unsigned int bytes)
3150
{
T
Tom Herbert 已提交
3151
#ifdef CONFIG_BQL
3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168
	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
3169
#endif
3170 3171
}

3172 3173 3174 3175 3176 3177 3178 3179 3180 3181
/**
 * 	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()
 */
3182
static inline void netdev_completed_queue(struct net_device *dev,
3183
					  unsigned int pkts, unsigned int bytes)
3184 3185 3186 3187 3188 3189
{
	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
3190
#ifdef CONFIG_BQL
3191
	clear_bit(__QUEUE_STATE_STACK_XOFF, &q->state);
T
Tom Herbert 已提交
3192 3193
	dql_reset(&q->dql);
#endif
3194 3195
}

3196 3197 3198 3199 3200 3201 3202
/**
 * 	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
 */
3203 3204 3205
static inline void netdev_reset_queue(struct net_device *dev_queue)
{
	netdev_tx_reset_queue(netdev_get_tx_queue(dev_queue, 0));
3206 3207
}

3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227
/**
 * 	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;
}

3228 3229 3230 3231 3232 3233
/**
 *	netif_running - test if up
 *	@dev: network device
 *
 *	Test if the device has been brought up.
 */
3234
static inline bool netif_running(const struct net_device *dev)
L
Linus Torvalds 已提交
3235 3236 3237 3238
{
	return test_bit(__LINK_STATE_START, &dev->state);
}

3239
/*
B
Bjorn Helgaas 已提交
3240
 * Routines to manage the subqueues on a device.  We only need start,
3241 3242 3243 3244
 * 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.
 */
3245 3246 3247 3248 3249 3250 3251 3252

/**
 *	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.
 */
3253 3254
static inline void netif_start_subqueue(struct net_device *dev, u16 queue_index)
{
3255
	struct netdev_queue *txq = netdev_get_tx_queue(dev, queue_index);
3256 3257

	netif_tx_start_queue(txq);
3258 3259
}

3260 3261 3262 3263 3264 3265 3266
/**
 *	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.
 */
3267 3268
static inline void netif_stop_subqueue(struct net_device *dev, u16 queue_index)
{
3269
	struct netdev_queue *txq = netdev_get_tx_queue(dev, queue_index);
3270
	netif_tx_stop_queue(txq);
3271 3272
}

3273 3274 3275 3276 3277 3278 3279
/**
 *	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.
 */
3280 3281
static inline bool __netif_subqueue_stopped(const struct net_device *dev,
					    u16 queue_index)
3282
{
3283
	struct netdev_queue *txq = netdev_get_tx_queue(dev, queue_index);
3284 3285

	return netif_tx_queue_stopped(txq);
3286 3287
}

3288 3289
static inline bool netif_subqueue_stopped(const struct net_device *dev,
					  struct sk_buff *skb)
3290 3291 3292
{
	return __netif_subqueue_stopped(dev, skb_get_queue_mapping(skb));
}
3293

3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306
/**
 *	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);
}
3307

3308
#ifdef CONFIG_XPS
3309
int netif_set_xps_queue(struct net_device *dev, const struct cpumask *mask,
3310
			u16 index);
3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396
int __netif_set_xps_queue(struct net_device *dev, const unsigned long *mask,
			  u16 index, bool is_rxqs_map);

/**
 *	netif_attr_test_mask - Test a CPU or Rx queue set in a mask
 *	@j: CPU/Rx queue index
 *	@mask: bitmask of all cpus/rx queues
 *	@nr_bits: number of bits in the bitmask
 *
 * Test if a CPU or Rx queue index is set in a mask of all CPU/Rx queues.
 */
static inline bool netif_attr_test_mask(unsigned long j,
					const unsigned long *mask,
					unsigned int nr_bits)
{
	cpu_max_bits_warn(j, nr_bits);
	return test_bit(j, mask);
}

/**
 *	netif_attr_test_online - Test for online CPU/Rx queue
 *	@j: CPU/Rx queue index
 *	@online_mask: bitmask for CPUs/Rx queues that are online
 *	@nr_bits: number of bits in the bitmask
 *
 * Returns true if a CPU/Rx queue is online.
 */
static inline bool netif_attr_test_online(unsigned long j,
					  const unsigned long *online_mask,
					  unsigned int nr_bits)
{
	cpu_max_bits_warn(j, nr_bits);

	if (online_mask)
		return test_bit(j, online_mask);

	return (j < nr_bits);
}

/**
 *	netif_attrmask_next - get the next CPU/Rx queue in a cpu/Rx queues mask
 *	@n: CPU/Rx queue index
 *	@srcp: the cpumask/Rx queue mask pointer
 *	@nr_bits: number of bits in the bitmask
 *
 * Returns >= nr_bits if no further CPUs/Rx queues set.
 */
static inline unsigned int netif_attrmask_next(int n, const unsigned long *srcp,
					       unsigned int nr_bits)
{
	/* -1 is a legal arg here. */
	if (n != -1)
		cpu_max_bits_warn(n, nr_bits);

	if (srcp)
		return find_next_bit(srcp, nr_bits, n + 1);

	return n + 1;
}

/**
 *	netif_attrmask_next_and - get the next CPU/Rx queue in *src1p & *src2p
 *	@n: CPU/Rx queue index
 *	@src1p: the first CPUs/Rx queues mask pointer
 *	@src2p: the second CPUs/Rx queues mask pointer
 *	@nr_bits: number of bits in the bitmask
 *
 * Returns >= nr_bits if no further CPUs/Rx queues set in both.
 */
static inline int netif_attrmask_next_and(int n, const unsigned long *src1p,
					  const unsigned long *src2p,
					  unsigned int nr_bits)
{
	/* -1 is a legal arg here. */
	if (n != -1)
		cpu_max_bits_warn(n, nr_bits);

	if (src1p && src2p)
		return find_next_and_bit(src1p, src2p, nr_bits, n + 1);
	else if (src1p)
		return find_next_bit(src1p, nr_bits, n + 1);
	else if (src2p)
		return find_next_bit(src2p, nr_bits, n + 1);

	return n + 1;
}
3397 3398
#else
static inline int netif_set_xps_queue(struct net_device *dev,
3399
				      const struct cpumask *mask,
3400 3401 3402 3403 3404 3405
				      u16 index)
{
	return 0;
}
#endif

3406 3407 3408 3409 3410 3411
/**
 *	netif_is_multiqueue - test if device has multiple transmit queues
 *	@dev: network device
 *
 * Check if device has multiple transmit queues
 */
3412
static inline bool netif_is_multiqueue(const struct net_device *dev)
3413
{
E
Eric Dumazet 已提交
3414
	return dev->num_tx_queues > 1;
3415
}
L
Linus Torvalds 已提交
3416

3417
int netif_set_real_num_tx_queues(struct net_device *dev, unsigned int txq);
3418

3419
#ifdef CONFIG_SYSFS
3420
int netif_set_real_num_rx_queues(struct net_device *dev, unsigned int rxq);
3421 3422 3423 3424 3425 3426 3427 3428
#else
static inline int netif_set_real_num_rx_queues(struct net_device *dev,
						unsigned int rxq)
{
	return 0;
}
#endif

3429 3430 3431 3432 3433 3434
static inline struct netdev_rx_queue *
__netif_get_rx_queue(struct net_device *dev, unsigned int rxq)
{
	return dev->_rx + rxq;
}

3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446
#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

3447
#define DEFAULT_MAX_NUM_RSS_QUEUES	(8)
3448
int netif_get_num_default_rss_queues(void);
3449

3450 3451 3452 3453 3454 3455 3456
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
3457

3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475
/*
 * 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
3476
 */
3477 3478 3479 3480 3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495
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
3496

3497 3498
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);
3499 3500
int netif_rx(struct sk_buff *skb);
int netif_rx_ni(struct sk_buff *skb);
3501
int netif_receive_skb(struct sk_buff *skb);
3502
int netif_receive_skb_core(struct sk_buff *skb);
3503
void netif_receive_skb_list(struct list_head *head);
3504 3505 3506 3507
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);
3508 3509
struct packet_offload *gro_find_receive_by_type(__be16 type);
struct packet_offload *gro_find_complete_by_type(__be16 type);
3510 3511 3512 3513 3514 3515 3516

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

M
Mahesh Bandewar 已提交
3517
bool netdev_is_rx_handler_busy(struct net_device *dev);
3518 3519 3520 3521 3522 3523
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);
3524 3525
int dev_ioctl(struct net *net, unsigned int cmd, struct ifreq *ifr,
		bool *need_copyout);
3526
int dev_ifconf(struct net *net, struct ifconf *, int);
3527 3528 3529 3530
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);
3531 3532
void __dev_notify_flags(struct net_device *, unsigned int old_flags,
			unsigned int gchanges);
3533 3534
int dev_change_name(struct net_device *, const char *);
int dev_set_alias(struct net_device *, const char *, size_t);
3535
int dev_get_alias(const struct net_device *, char *, size_t);
3536
int dev_change_net_namespace(struct net_device *, struct net *, const char *);
3537
int __dev_set_mtu(struct net_device *, int);
3538
int dev_set_mtu(struct net_device *, int);
3539
int dev_change_tx_queue_len(struct net_device *, unsigned long);
3540 3541 3542 3543
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,
3544
			 struct netdev_phys_item_id *ppid);
3545 3546
int dev_get_phys_port_name(struct net_device *dev,
			   char *name, size_t len);
3547
int dev_change_proto_down(struct net_device *dev, bool proto_down);
3548
struct sk_buff *validate_xmit_skb_list(struct sk_buff *skb, struct net_device *dev, bool *again);
3549 3550
struct sk_buff *dev_hard_start_xmit(struct sk_buff *skb, struct net_device *dev,
				    struct netdev_queue *txq, int *ret);
3551

3552
typedef int (*bpf_op_t)(struct net_device *dev, struct netdev_bpf *bpf);
3553 3554
int dev_change_xdp_fd(struct net_device *dev, struct netlink_ext_ack *extack,
		      int fd, u32 flags);
3555 3556
void __dev_xdp_query(struct net_device *dev, bpf_op_t xdp_op,
		     struct netdev_bpf *xdp);
3557

H
Herbert Xu 已提交
3558
int __dev_forward_skb(struct net_device *dev, struct sk_buff *skb);
3559
int dev_forward_skb(struct net_device *dev, struct sk_buff *skb);
3560 3561
bool is_skb_forwardable(const struct net_device *dev,
			const struct sk_buff *skb);
L
Linus Torvalds 已提交
3562

3563 3564 3565 3566 3567 3568 3569 3570 3571 3572 3573 3574 3575 3576 3577
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;
}

3578 3579
void dev_queue_xmit_nit(struct sk_buff *skb, struct net_device *dev);

3580
extern int		netdev_budget;
3581
extern unsigned int	netdev_budget_usecs;
L
Linus Torvalds 已提交
3582 3583

/* Called by rtnetlink.c:rtnl_unlock() */
3584
void netdev_run_todo(void);
L
Linus Torvalds 已提交
3585

3586 3587 3588 3589
/**
 *	dev_put - release reference to device
 *	@dev: network device
 *
3590
 * Release reference to device to allow it to be freed.
3591
 */
L
Linus Torvalds 已提交
3592 3593
static inline void dev_put(struct net_device *dev)
{
3594
	this_cpu_dec(*dev->pcpu_refcnt);
L
Linus Torvalds 已提交
3595 3596
}

3597 3598 3599 3600
/**
 *	dev_hold - get reference to device
 *	@dev: network device
 *
3601
 * Hold reference to device to keep it from being freed.
3602
 */
3603 3604
static inline void dev_hold(struct net_device *dev)
{
3605
	this_cpu_inc(*dev->pcpu_refcnt);
3606
}
L
Linus Torvalds 已提交
3607 3608 3609 3610

/* 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
3611 3612 3613 3614
 *
 * 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
3615 3616
 */

3617 3618 3619
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
3620

3621 3622 3623 3624 3625 3626
/**
 *	netif_carrier_ok - test if carrier present
 *	@dev: network device
 *
 * Check if carrier is present on device
 */
3627
static inline bool netif_carrier_ok(const struct net_device *dev)
L
Linus Torvalds 已提交
3628 3629 3630 3631
{
	return !test_bit(__LINK_STATE_NOCARRIER, &dev->state);
}

3632
unsigned long dev_trans_start(struct net_device *dev);
3633

3634
void __netdev_watchdog_up(struct net_device *dev);
L
Linus Torvalds 已提交
3635

3636
void netif_carrier_on(struct net_device *dev);
L
Linus Torvalds 已提交
3637

3638
void netif_carrier_off(struct net_device *dev);
L
Linus Torvalds 已提交
3639

3640 3641 3642 3643 3644 3645 3646 3647 3648 3649 3650 3651
/**
 *	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
3652 3653 3654 3655 3656 3657
static inline void netif_dormant_on(struct net_device *dev)
{
	if (!test_and_set_bit(__LINK_STATE_DORMANT, &dev->state))
		linkwatch_fire_event(dev);
}

3658 3659 3660 3661 3662 3663
/**
 *	netif_dormant_off - set device as not dormant.
 *	@dev: network device
 *
 * Device is not in dormant state.
 */
S
Stefan Rompf 已提交
3664 3665 3666 3667 3668 3669
static inline void netif_dormant_off(struct net_device *dev)
{
	if (test_and_clear_bit(__LINK_STATE_DORMANT, &dev->state))
		linkwatch_fire_event(dev);
}

3670
/**
3671
 *	netif_dormant - test if device is dormant
3672 3673
 *	@dev: network device
 *
3674
 * Check if device is dormant.
3675
 */
3676
static inline bool netif_dormant(const struct net_device *dev)
S
Stefan Rompf 已提交
3677 3678 3679 3680 3681
{
	return test_bit(__LINK_STATE_DORMANT, &dev->state);
}


3682 3683 3684 3685 3686 3687
/**
 *	netif_oper_up - test if device is operational
 *	@dev: network device
 *
 * Check if carrier is operational
 */
3688
static inline bool netif_oper_up(const struct net_device *dev)
E
Eric Dumazet 已提交
3689
{
S
Stefan Rompf 已提交
3690 3691 3692 3693
	return (dev->operstate == IF_OPER_UP ||
		dev->operstate == IF_OPER_UNKNOWN /* backward compat */);
}

3694 3695 3696 3697 3698 3699
/**
 *	netif_device_present - is device available or removed
 *	@dev: network device
 *
 * Check if device has not been removed from system.
 */
3700
static inline bool netif_device_present(struct net_device *dev)
L
Linus Torvalds 已提交
3701 3702 3703 3704
{
	return test_bit(__LINK_STATE_PRESENT, &dev->state);
}

3705
void netif_device_detach(struct net_device *dev);
L
Linus Torvalds 已提交
3706

3707
void netif_device_attach(struct net_device *dev);
L
Linus Torvalds 已提交
3708 3709 3710 3711 3712 3713 3714 3715 3716 3717 3718 3719 3720 3721 3722 3723 3724 3725 3726 3727 3728 3729 3730 3731 3732 3733 3734 3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750 3751 3752 3753 3754 3755 3756 3757

/*
 * 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;
}

3758
static inline void __netif_tx_lock(struct netdev_queue *txq, int cpu)
H
Herbert Xu 已提交
3759
{
3760 3761
	spin_lock(&txq->_xmit_lock);
	txq->xmit_lock_owner = cpu;
3762 3763
}

3764 3765 3766 3767 3768 3769 3770 3771 3772 3773 3774
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);
}

3775 3776 3777 3778 3779 3780
static inline void __netif_tx_lock_bh(struct netdev_queue *txq)
{
	spin_lock_bh(&txq->_xmit_lock);
	txq->xmit_lock_owner = smp_processor_id();
}

3781
static inline bool __netif_tx_trylock(struct netdev_queue *txq)
3782
{
3783
	bool ok = spin_trylock(&txq->_xmit_lock);
3784 3785 3786 3787 3788 3789 3790 3791 3792 3793 3794 3795 3796 3797 3798 3799 3800
	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
3801 3802 3803 3804 3805 3806
static inline void txq_trans_update(struct netdev_queue *txq)
{
	if (txq->xmit_lock_owner != -1)
		txq->trans_start = jiffies;
}

3807 3808 3809
/* 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
3810 3811 3812 3813
	struct netdev_queue *txq = netdev_get_tx_queue(dev, 0);

	if (txq->trans_start != jiffies)
		txq->trans_start = jiffies;
3814 3815
}

3816 3817 3818 3819 3820 3821
/**
 *	netif_tx_lock - grab network device transmit lock
 *	@dev: network device
 *
 * Get network device transmit lock
 */
3822 3823
static inline void netif_tx_lock(struct net_device *dev)
{
3824
	unsigned int i;
3825
	int cpu;
3826

3827 3828
	spin_lock(&dev->tx_global_lock);
	cpu = smp_processor_id();
3829 3830
	for (i = 0; i < dev->num_tx_queues; i++) {
		struct netdev_queue *txq = netdev_get_tx_queue(dev, i);
3831 3832 3833 3834 3835 3836 3837

		/* 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.
		 */
3838
		__netif_tx_lock(txq, cpu);
3839 3840
		set_bit(__QUEUE_STATE_FROZEN, &txq->state);
		__netif_tx_unlock(txq);
3841
	}
H
Herbert Xu 已提交
3842 3843 3844 3845
}

static inline void netif_tx_lock_bh(struct net_device *dev)
{
3846 3847
	local_bh_disable();
	netif_tx_lock(dev);
H
Herbert Xu 已提交
3848 3849 3850 3851
}

static inline void netif_tx_unlock(struct net_device *dev)
{
3852 3853 3854 3855
	unsigned int i;

	for (i = 0; i < dev->num_tx_queues; i++) {
		struct netdev_queue *txq = netdev_get_tx_queue(dev, i);
3856

3857 3858 3859 3860 3861
		/* 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);
3862
		netif_schedule_queue(txq);
3863 3864
	}
	spin_unlock(&dev->tx_global_lock);
H
Herbert Xu 已提交
3865 3866 3867 3868
}

static inline void netif_tx_unlock_bh(struct net_device *dev)
{
3869 3870
	netif_tx_unlock(dev);
	local_bh_enable();
H
Herbert Xu 已提交
3871 3872
}

3873
#define HARD_TX_LOCK(dev, txq, cpu) {			\
3874
	if ((dev->features & NETIF_F_LLTX) == 0) {	\
3875
		__netif_tx_lock(txq, cpu);		\
3876 3877
	} else {					\
		__netif_tx_acquire(txq);		\
3878 3879 3880
	}						\
}

3881 3882 3883
#define HARD_TX_TRYLOCK(dev, txq)			\
	(((dev->features & NETIF_F_LLTX) == 0) ?	\
		__netif_tx_trylock(txq) :		\
3884
		__netif_tx_acquire(txq))
3885

3886
#define HARD_TX_UNLOCK(dev, txq) {			\
3887
	if ((dev->features & NETIF_F_LLTX) == 0) {	\
3888
		__netif_tx_unlock(txq);			\
3889 3890
	} else {					\
		__netif_tx_release(txq);		\
3891 3892 3893
	}						\
}

L
Linus Torvalds 已提交
3894 3895
static inline void netif_tx_disable(struct net_device *dev)
{
3896
	unsigned int i;
3897
	int cpu;
3898

3899 3900
	local_bh_disable();
	cpu = smp_processor_id();
3901 3902
	for (i = 0; i < dev->num_tx_queues; i++) {
		struct netdev_queue *txq = netdev_get_tx_queue(dev, i);
3903 3904

		__netif_tx_lock(txq, cpu);
3905
		netif_tx_stop_queue(txq);
3906
		__netif_tx_unlock(txq);
3907
	}
3908
	local_bh_enable();
L
Linus Torvalds 已提交
3909 3910
}

3911 3912 3913 3914 3915
static inline void netif_addr_lock(struct net_device *dev)
{
	spin_lock(&dev->addr_list_lock);
}

3916 3917
static inline void netif_addr_lock_nested(struct net_device *dev)
{
3918 3919 3920 3921 3922 3923
	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);
3924 3925
}

3926 3927 3928 3929 3930 3931 3932 3933 3934 3935 3936 3937 3938 3939 3940
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);
}

3941
/*
3942
 * dev_addrs walker. Should be used only for read access. Call with
3943 3944 3945
 * rcu_read_lock held.
 */
#define for_each_dev_addr(dev, ha) \
3946
		list_for_each_entry_rcu(ha, &dev->dev_addrs.list, list)
3947

L
Linus Torvalds 已提交
3948 3949
/* These functions live elsewhere (drivers/net/net_init.c, but related) */

3950
void ether_setup(struct net_device *dev);
L
Linus Torvalds 已提交
3951 3952

/* Support for loadable net-drivers */
3953
struct net_device *alloc_netdev_mqs(int sizeof_priv, const char *name,
3954
				    unsigned char name_assign_type,
3955 3956
				    void (*setup)(struct net_device *),
				    unsigned int txqs, unsigned int rxqs);
3957 3958 3959
int dev_get_valid_name(struct net *net, struct net_device *dev,
		       const char *name);

3960 3961
#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
3962

3963 3964 3965
#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
3966

3967 3968
int register_netdev(struct net_device *dev);
void unregister_netdev(struct net_device *dev);
3969

3970
/* General hardware address lists handling functions */
3971 3972 3973 3974
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);
3975 3976 3977 3978 3979 3980 3981 3982 3983
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 *));
3984
void __hw_addr_init(struct netdev_hw_addr_list *list);
3985

3986
/* Functions used for device addresses handling */
3987 3988 3989 3990 3991 3992
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);
3993 3994

/* Functions used for unicast addresses handling */
3995 3996 3997 3998 3999 4000 4001 4002
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);
4003

4004 4005 4006 4007 4008 4009 4010 4011
/**
 *  __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
4012
 */
4013 4014 4015 4016 4017 4018 4019 4020 4021 4022
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);
}

/**
4023
 *  __dev_uc_unsync - Remove synchronized addresses from device
4024 4025 4026 4027
 *  @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
4028
 */
4029 4030 4031 4032 4033 4034 4035
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);
}

4036
/* Functions used for multicast addresses handling */
4037 4038 4039 4040 4041 4042 4043 4044 4045 4046
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);
4047

4048 4049 4050 4051 4052 4053 4054 4055
/**
 *  __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
4056
 */
4057 4058 4059 4060 4061 4062 4063 4064 4065 4066
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);
}

/**
4067
 *  __dev_mc_unsync - Remove synchronized addresses from device
4068 4069 4070 4071
 *  @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
4072
 */
4073 4074 4075 4076 4077 4078 4079
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);
}

4080
/* Functions used for secondary unicast and multicast support */
4081 4082 4083 4084 4085 4086 4087
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
4088
/* Load a device via the kmod */
4089 4090 4091 4092 4093
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);
4094

L
Linus Torvalds 已提交
4095
extern int		netdev_max_backlog;
E
Eric Dumazet 已提交
4096
extern int		netdev_tstamp_prequeue;
L
Linus Torvalds 已提交
4097
extern int		weight_p;
4098 4099 4100 4101
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
4102

4103
bool netdev_has_upper_dev(struct net_device *dev, struct net_device *upper_dev);
4104 4105
struct net_device *netdev_upper_get_next_dev_rcu(struct net_device *dev,
						     struct list_head **iter);
4106 4107
struct net_device *netdev_all_upper_get_next_dev_rcu(struct net_device *dev,
						     struct list_head **iter);
4108

4109 4110 4111 4112 4113 4114 4115
/* 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)))

4116 4117 4118 4119 4120 4121 4122 4123
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);

4124 4125
bool netdev_has_any_upper_dev(struct net_device *dev);

4126 4127 4128 4129
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);
4130 4131 4132 4133 4134 4135 4136 4137 4138 4139 4140 4141 4142

#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)))

4143 4144
void *netdev_lower_get_next(struct net_device *dev,
				struct list_head **iter);
4145

4146
#define netdev_for_each_lower_dev(dev, ldev, iter) \
4147
	for (iter = (dev)->adj_list.lower.next, \
4148 4149 4150 4151
	     ldev = netdev_lower_get_next(dev, &(iter)); \
	     ldev; \
	     ldev = netdev_lower_get_next(dev, &(iter)))

4152 4153 4154 4155 4156
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);

4157 4158 4159 4160 4161 4162 4163 4164 4165
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);

4166
void *netdev_adjacent_get_private(struct list_head *adj_list);
4167
void *netdev_lower_get_first_private_rcu(struct net_device *dev);
4168 4169
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);
4170 4171
int netdev_upper_dev_link(struct net_device *dev, struct net_device *upper_dev,
			  struct netlink_ext_ack *extack);
4172
int netdev_master_upper_dev_link(struct net_device *dev,
4173
				 struct net_device *upper_dev,
4174 4175
				 void *upper_priv, void *upper_info,
				 struct netlink_ext_ack *extack);
4176 4177
void netdev_upper_dev_unlink(struct net_device *dev,
			     struct net_device *upper_dev);
4178
void netdev_adjacent_rename_links(struct net_device *dev, char *oldname);
4179 4180
void *netdev_lower_dev_get_private(struct net_device *dev,
				   struct net_device *lower_dev);
4181 4182
void netdev_lower_state_changed(struct net_device *lower_dev,
				void *lower_state_info);
4183 4184 4185

/* RSS keys are 40 or 52 bytes long */
#define NETDEV_RSS_KEY_LEN 52
4186
extern u8 netdev_rss_key[NETDEV_RSS_KEY_LEN] __read_mostly;
4187 4188
void netdev_rss_key_fill(void *buffer, size_t len);

4189
int dev_get_nest_level(struct net_device *dev);
4190
int skb_checksum_help(struct sk_buff *skb);
4191
int skb_crc32c_csum_help(struct sk_buff *skb);
4192 4193 4194
int skb_csum_hwoffload_help(struct sk_buff *skb,
			    const netdev_features_t features);

4195 4196 4197 4198
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);
4199

4200 4201 4202 4203 4204 4205 4206 4207 4208 4209 4210 4211 4212
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);

4213 4214 4215 4216 4217
static inline
struct sk_buff *skb_gso_segment(struct sk_buff *skb, netdev_features_t features)
{
	return __skb_gso_segment(skb, features, true);
}
4218
__be16 skb_network_protocol(struct sk_buff *skb, int *depth);
4219 4220 4221 4222

static inline bool can_checksum_protocol(netdev_features_t features,
					 __be16 protocol)
{
4223 4224 4225 4226 4227 4228 4229 4230 4231 4232 4233 4234 4235 4236 4237 4238 4239 4240
	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;
	}
4241
}
4242

4243
#ifdef CONFIG_BUG
4244
void netdev_rx_csum_fault(struct net_device *dev);
4245 4246 4247 4248 4249
#else
static inline void netdev_rx_csum_fault(struct net_device *dev)
{
}
#endif
L
Linus Torvalds 已提交
4250
/* rx skb timestamps */
4251 4252
void net_enable_timestamp(void);
void net_disable_timestamp(void);
L
Linus Torvalds 已提交
4253

4254
#ifdef CONFIG_PROC_FS
4255
int __init dev_proc_init(void);
4256 4257
#else
#define dev_proc_init() 0
4258 4259
#endif

D
David S. Miller 已提交
4260
static inline netdev_tx_t __netdev_start_xmit(const struct net_device_ops *ops,
4261 4262
					      struct sk_buff *skb, struct net_device *dev,
					      bool more)
D
David S. Miller 已提交
4263
{
4264
	skb->xmit_more = more ? 1 : 0;
4265
	return ops->ndo_start_xmit(skb, dev);
D
David S. Miller 已提交
4266 4267
}

4268
static inline netdev_tx_t netdev_start_xmit(struct sk_buff *skb, struct net_device *dev,
4269
					    struct netdev_queue *txq, bool more)
D
David S. Miller 已提交
4270 4271
{
	const struct net_device_ops *ops = dev->netdev_ops;
4272
	int rc;
D
David S. Miller 已提交
4273

4274
	rc = __netdev_start_xmit(ops, skb, dev, more);
4275 4276 4277 4278
	if (rc == NETDEV_TX_OK)
		txq_trans_update(txq);

	return rc;
D
David S. Miller 已提交
4279 4280
}

4281
int netdev_class_create_file_ns(const struct class_attribute *class_attr,
4282
				const void *ns);
4283
void netdev_class_remove_file_ns(const struct class_attribute *class_attr,
4284
				 const void *ns);
4285

4286
static inline int netdev_class_create_file(const struct class_attribute *class_attr)
4287 4288 4289 4290
{
	return netdev_class_create_file_ns(class_attr, NULL);
}

4291
static inline void netdev_class_remove_file(const struct class_attribute *class_attr)
4292 4293 4294
{
	netdev_class_remove_file_ns(class_attr, NULL);
}
4295

4296
extern const struct kobj_ns_type_operations net_ns_type_operations;
4297

4298
const char *netdev_drivername(const struct net_device *dev);
4299

4300
void linkwatch_run_queue(void);
4301

4302 4303 4304
static inline netdev_features_t netdev_intersect_features(netdev_features_t f1,
							  netdev_features_t f2)
{
4305 4306
	if ((f1 ^ f2) & NETIF_F_HW_CSUM) {
		if (f1 & NETIF_F_HW_CSUM)
4307
			f1 |= (NETIF_F_IP_CSUM|NETIF_F_IPV6_CSUM);
4308
		else
4309
			f2 |= (NETIF_F_IP_CSUM|NETIF_F_IPV6_CSUM);
4310
	}
4311

4312
	return f1 & f2;
4313 4314
}

4315 4316
static inline netdev_features_t netdev_get_wanted_features(
	struct net_device *dev)
4317 4318 4319
{
	return (dev->features & ~dev->hw_features) | dev->wanted_features;
}
4320 4321
netdev_features_t netdev_increment_features(netdev_features_t all,
	netdev_features_t one, netdev_features_t mask);
4322 4323 4324 4325 4326 4327 4328 4329 4330 4331 4332

/* 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);
}

4333
int __netdev_update_features(struct net_device *dev);
4334
void netdev_update_features(struct net_device *dev);
4335
void netdev_change_features(struct net_device *dev);
4336

4337 4338 4339
void netif_stacked_transfer_operstate(const struct net_device *rootdev,
					struct net_device *dev);

4340 4341 4342
netdev_features_t passthru_features_check(struct sk_buff *skb,
					  struct net_device *dev,
					  netdev_features_t features);
4343
netdev_features_t netif_skb_features(struct sk_buff *skb);
4344

4345
static inline bool net_gso_ok(netdev_features_t features, int gso_type)
4346
{
4347
	netdev_features_t feature = (netdev_features_t)gso_type << NETIF_F_GSO_SHIFT;
4348 4349 4350 4351 4352

	/* 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));
4353
	BUILD_BUG_ON(SKB_GSO_TCP_FIXEDID != (NETIF_F_TSO_MANGLEID >> NETIF_F_GSO_SHIFT));
4354 4355
	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));
4356 4357
	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));
4358 4359
	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));
4360 4361
	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));
4362
	BUILD_BUG_ON(SKB_GSO_PARTIAL != (NETIF_F_GSO_PARTIAL >> NETIF_F_GSO_SHIFT));
4363
	BUILD_BUG_ON(SKB_GSO_TUNNEL_REMCSUM != (NETIF_F_GSO_TUNNEL_REMCSUM >> NETIF_F_GSO_SHIFT));
M
Marcelo Ricardo Leitner 已提交
4364
	BUILD_BUG_ON(SKB_GSO_SCTP    != (NETIF_F_GSO_SCTP >> NETIF_F_GSO_SHIFT));
S
Steffen Klassert 已提交
4365
	BUILD_BUG_ON(SKB_GSO_ESP != (NETIF_F_GSO_ESP >> NETIF_F_GSO_SHIFT));
4366
	BUILD_BUG_ON(SKB_GSO_UDP != (NETIF_F_GSO_UDP >> NETIF_F_GSO_SHIFT));
4367
	BUILD_BUG_ON(SKB_GSO_UDP_L4 != (NETIF_F_GSO_UDP_L4 >> NETIF_F_GSO_SHIFT));
4368

4369
	return (features & feature) == feature;
4370 4371
}

4372
static inline bool skb_gso_ok(struct sk_buff *skb, netdev_features_t features)
4373
{
H
Herbert Xu 已提交
4374
	return net_gso_ok(features, skb_shinfo(skb)->gso_type) &&
4375
	       (!skb_has_frag_list(skb) || (features & NETIF_F_FRAGLIST));
4376 4377
}

4378
static inline bool netif_needs_gso(struct sk_buff *skb,
4379
				   netdev_features_t features)
4380
{
4381
	return skb_is_gso(skb) && (!skb_gso_ok(skb, features) ||
4382 4383
		unlikely((skb->ip_summed != CHECKSUM_PARTIAL) &&
			 (skb->ip_summed != CHECKSUM_UNNECESSARY)));
4384 4385
}

4386 4387 4388 4389 4390 4391
static inline void netif_set_gso_max_size(struct net_device *dev,
					  unsigned int size)
{
	dev->gso_max_size = size;
}

4392 4393 4394 4395 4396 4397 4398 4399 4400 4401 4402 4403 4404
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;
}

4405 4406 4407 4408 4409
static inline bool netif_is_macsec(const struct net_device *dev)
{
	return dev->priv_flags & IFF_MACSEC;
}

4410
static inline bool netif_is_macvlan(const struct net_device *dev)
4411 4412 4413 4414
{
	return dev->priv_flags & IFF_MACVLAN;
}

4415
static inline bool netif_is_macvlan_port(const struct net_device *dev)
4416 4417 4418 4419
{
	return dev->priv_flags & IFF_MACVLAN_PORT;
}

4420
static inline bool netif_is_bond_master(const struct net_device *dev)
4421 4422 4423 4424
{
	return dev->flags & IFF_MASTER && dev->priv_flags & IFF_BONDING;
}

4425
static inline bool netif_is_bond_slave(const struct net_device *dev)
J
Jiri Pirko 已提交
4426 4427 4428 4429
{
	return dev->flags & IFF_SLAVE && dev->priv_flags & IFF_BONDING;
}

4430 4431 4432 4433 4434
static inline bool netif_supports_nofcs(struct net_device *dev)
{
	return dev->priv_flags & IFF_SUPP_NOFCS;
}

4435
static inline bool netif_is_l3_master(const struct net_device *dev)
4436
{
4437
	return dev->priv_flags & IFF_L3MDEV_MASTER;
4438 4439
}

D
David Ahern 已提交
4440 4441 4442 4443 4444
static inline bool netif_is_l3_slave(const struct net_device *dev)
{
	return dev->priv_flags & IFF_L3MDEV_SLAVE;
}

4445 4446 4447 4448 4449
static inline bool netif_is_bridge_master(const struct net_device *dev)
{
	return dev->priv_flags & IFF_EBRIDGE;
}

4450 4451 4452 4453 4454
static inline bool netif_is_bridge_port(const struct net_device *dev)
{
	return dev->priv_flags & IFF_BRIDGE_PORT;
}

4455 4456 4457 4458 4459
static inline bool netif_is_ovs_master(const struct net_device *dev)
{
	return dev->priv_flags & IFF_OPENVSWITCH;
}

J
Jiri Pirko 已提交
4460 4461 4462 4463 4464
static inline bool netif_is_ovs_port(const struct net_device *dev)
{
	return dev->priv_flags & IFF_OVS_DATAPATH;
}

4465
static inline bool netif_is_team_master(const struct net_device *dev)
J
Jiri Pirko 已提交
4466 4467 4468 4469
{
	return dev->priv_flags & IFF_TEAM;
}

4470
static inline bool netif_is_team_port(const struct net_device *dev)
J
Jiri Pirko 已提交
4471 4472 4473 4474
{
	return dev->priv_flags & IFF_TEAM_PORT;
}

4475
static inline bool netif_is_lag_master(const struct net_device *dev)
J
Jiri Pirko 已提交
4476 4477 4478 4479
{
	return netif_is_bond_master(dev) || netif_is_team_master(dev);
}

4480
static inline bool netif_is_lag_port(const struct net_device *dev)
J
Jiri Pirko 已提交
4481 4482 4483 4484
{
	return netif_is_bond_slave(dev) || netif_is_team_port(dev);
}

4485 4486 4487 4488 4489
static inline bool netif_is_rxfh_configured(const struct net_device *dev)
{
	return dev->priv_flags & IFF_RXFH_CONFIGURED;
}

4490 4491 4492 4493 4494 4495 4496 4497 4498 4499
static inline bool netif_is_failover(const struct net_device *dev)
{
	return dev->priv_flags & IFF_FAILOVER;
}

static inline bool netif_is_failover_slave(const struct net_device *dev)
{
	return dev->priv_flags & IFF_FAILOVER_SLAVE;
}

4500 4501 4502 4503 4504 4505
/* 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);
}

4506 4507 4508 4509 4510 4511 4512
/* 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;
}

4513
extern struct pernet_operations __net_initdata loopback_net_ops;
4514

4515 4516 4517 4518 4519 4520
/* Logging, debugging and troubleshooting/diagnostic helpers. */

/* netdev_printk helpers, similar to dev_printk */

static inline const char *netdev_name(const struct net_device *dev)
{
4521 4522
	if (!dev->name[0] || strchr(dev->name, '%'))
		return "(unnamed net_device)";
4523 4524 4525
	return dev->name;
}

4526 4527 4528 4529 4530
static inline bool netdev_unregistering(const struct net_device *dev)
{
	return dev->reg_state == NETREG_UNREGISTERING;
}

4531 4532 4533 4534 4535 4536 4537 4538 4539 4540 4541 4542 4543 4544 4545
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)";
}

4546
__printf(3, 4)
4547 4548
void netdev_printk(const char *level, const struct net_device *dev,
		   const char *format, ...);
4549
__printf(2, 3)
4550
void netdev_emerg(const struct net_device *dev, const char *format, ...);
4551
__printf(2, 3)
4552
void netdev_alert(const struct net_device *dev, const char *format, ...);
4553
__printf(2, 3)
4554
void netdev_crit(const struct net_device *dev, const char *format, ...);
4555
__printf(2, 3)
4556
void netdev_err(const struct net_device *dev, const char *format, ...);
4557
__printf(2, 3)
4558
void netdev_warn(const struct net_device *dev, const char *format, ...);
4559
__printf(2, 3)
4560
void netdev_notice(const struct net_device *dev, const char *format, ...);
4561
__printf(2, 3)
4562
void netdev_info(const struct net_device *dev, const char *format, ...);
4563

4564 4565 4566 4567 4568 4569 4570 4571 4572 4573 4574 4575 4576 4577 4578 4579 4580 4581 4582 4583 4584 4585 4586 4587 4588
#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__)

4589 4590 4591
#define MODULE_ALIAS_NETDEV(device) \
	MODULE_ALIAS("netdev-" device)

4592
#if defined(CONFIG_DYNAMIC_DEBUG)
4593 4594
#define netdev_dbg(__dev, format, args...)			\
do {								\
4595
	dynamic_netdev_dbg(__dev, format, ##args);		\
4596
} while (0)
4597 4598 4599
#elif defined(DEBUG)
#define netdev_dbg(__dev, format, args...)			\
	netdev_printk(KERN_DEBUG, __dev, format, ##args)
4600 4601 4602 4603 4604 4605 4606 4607 4608 4609 4610 4611 4612 4613 4614 4615 4616 4617 4618 4619 4620 4621 4622 4623 4624 4625
#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...)			\
4626
	WARN(1, "netdevice: %s%s: " format, netdev_name(dev),	\
4627
	     netdev_reg_state(dev), ##args)
4628

G
Gal Pressman 已提交
4629
#define netdev_WARN_ONCE(dev, format, args...)				\
4630
	WARN_ONCE(1, "netdevice: %s%s: " format, netdev_name(dev),	\
4631 4632
		  netdev_reg_state(dev), ##args)

4633 4634 4635 4636 4637 4638 4639 4640
/* 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)

4641 4642 4643 4644 4645 4646
#define netif_level(level, priv, type, dev, fmt, args...)	\
do {								\
	if (netif_msg_##type(priv))				\
		netdev_##level(dev, fmt, ##args);		\
} while (0)

4647
#define netif_emerg(priv, type, dev, fmt, args...)		\
4648
	netif_level(emerg, priv, type, dev, fmt, ##args)
4649
#define netif_alert(priv, type, dev, fmt, args...)		\
4650
	netif_level(alert, priv, type, dev, fmt, ##args)
4651
#define netif_crit(priv, type, dev, fmt, args...)		\
4652
	netif_level(crit, priv, type, dev, fmt, ##args)
4653
#define netif_err(priv, type, dev, fmt, args...)		\
4654
	netif_level(err, priv, type, dev, fmt, ##args)
4655
#define netif_warn(priv, type, dev, fmt, args...)		\
4656
	netif_level(warn, priv, type, dev, fmt, ##args)
4657
#define netif_notice(priv, type, dev, fmt, args...)		\
4658
	netif_level(notice, priv, type, dev, fmt, ##args)
4659
#define netif_info(priv, type, dev, fmt, args...)		\
4660
	netif_level(info, priv, type, dev, fmt, ##args)
4661

4662
#if defined(CONFIG_DYNAMIC_DEBUG)
4663 4664 4665
#define netif_dbg(priv, type, netdev, format, args...)		\
do {								\
	if (netif_msg_##type(priv))				\
4666
		dynamic_netdev_dbg(netdev, format, ##args);	\
4667
} while (0)
4668 4669 4670
#elif defined(DEBUG)
#define netif_dbg(priv, type, dev, format, args...)		\
	netif_printk(priv, type, KERN_DEBUG, dev, format, ##args)
4671 4672 4673 4674 4675 4676 4677 4678 4679
#else
#define netif_dbg(priv, type, dev, format, args...)			\
({									\
	if (0)								\
		netif_printk(priv, type, KERN_DEBUG, dev, format, ##args); \
	0;								\
})
#endif

4680 4681 4682 4683 4684 4685 4686 4687 4688
/* 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)

4689
#if defined(VERBOSE_DEBUG)
4690
#define netif_vdbg	netif_dbg
4691 4692 4693 4694
#else
#define netif_vdbg(priv, type, dev, format, args...)		\
({								\
	if (0)							\
4695
		netif_printk(priv, type, KERN_DEBUG, dev, format, ##args); \
4696 4697 4698
	0;							\
})
#endif
4699

4700 4701 4702 4703 4704 4705 4706 4707 4708 4709 4710 4711 4712 4713 4714 4715 4716 4717 4718 4719 4720 4721
/*
 *	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)

4722
#endif	/* _LINUX_NETDEVICE_H */