skbuff.h 112.4 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
/*
 *	Definitions for the 'struct sk_buff' memory handlers.
 *
 *	Authors:
 *		Alan Cox, <gw4pts@gw4pts.ampr.org>
 *		Florian La Roche, <rzsfl@rz.uni-sb.de>
 *
 *	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.
 */

#ifndef _LINUX_SKBUFF_H
#define _LINUX_SKBUFF_H

#include <linux/kernel.h>
18
#include <linux/kmemcheck.h>
L
Linus Torvalds 已提交
19 20
#include <linux/compiler.h>
#include <linux/time.h>
21
#include <linux/bug.h>
L
Linus Torvalds 已提交
22
#include <linux/cache.h>
E
Eric Dumazet 已提交
23
#include <linux/rbtree.h>
24
#include <linux/socket.h>
L
Linus Torvalds 已提交
25

A
Arun Sharma 已提交
26
#include <linux/atomic.h>
L
Linus Torvalds 已提交
27 28 29
#include <asm/types.h>
#include <linux/spinlock.h>
#include <linux/net.h>
30
#include <linux/textsearch.h>
L
Linus Torvalds 已提交
31
#include <net/checksum.h>
32
#include <linux/rcupdate.h>
33
#include <linux/hrtimer.h>
34
#include <linux/dma-mapping.h>
35
#include <linux/netdev_features.h>
36
#include <linux/sched.h>
37
#include <linux/sched/clock.h>
38
#include <net/flow_dissector.h>
39
#include <linux/splice.h>
40
#include <linux/in6.h>
41
#include <linux/if_packet.h>
42
#include <net/flow.h>
L
Linus Torvalds 已提交
43

44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88
/* The interface for checksum offload between the stack and networking drivers
 * is as follows...
 *
 * A. IP checksum related features
 *
 * Drivers advertise checksum offload capabilities in the features of a device.
 * From the stack's point of view these are capabilities offered by the driver,
 * a driver typically only advertises features that it is capable of offloading
 * to its device.
 *
 * The checksum related features are:
 *
 *	NETIF_F_HW_CSUM	- The driver (or its device) is able to compute one
 *			  IP (one's complement) checksum for any combination
 *			  of protocols or protocol layering. The checksum is
 *			  computed and set in a packet per the CHECKSUM_PARTIAL
 *			  interface (see below).
 *
 *	NETIF_F_IP_CSUM - Driver (device) is only able to checksum plain
 *			  TCP or UDP packets over IPv4. These are specifically
 *			  unencapsulated packets of the form IPv4|TCP or
 *			  IPv4|UDP where the Protocol field in the IPv4 header
 *			  is TCP or UDP. The IPv4 header may contain IP options
 *			  This feature cannot be set in features for a device
 *			  with NETIF_F_HW_CSUM also set. This feature is being
 *			  DEPRECATED (see below).
 *
 *	NETIF_F_IPV6_CSUM - Driver (device) is only able to checksum plain
 *			  TCP or UDP packets over IPv6. These are specifically
 *			  unencapsulated packets of the form IPv6|TCP or
 *			  IPv4|UDP where the Next Header field in the IPv6
 *			  header is either TCP or UDP. IPv6 extension headers
 *			  are not supported with this feature. This feature
 *			  cannot be set in features for a device with
 *			  NETIF_F_HW_CSUM also set. This feature is being
 *			  DEPRECATED (see below).
 *
 *	NETIF_F_RXCSUM - Driver (device) performs receive checksum offload.
 *			 This flag is used only used to disable the RX checksum
 *			 feature for a device. The stack will accept receive
 *			 checksum indication in packets received on a device
 *			 regardless of whether NETIF_F_RXCSUM is set.
 *
 * B. Checksumming of received packets by device. Indication of checksum
 *    verification is in set skb->ip_summed. Possible values are:
89 90 91
 *
 * CHECKSUM_NONE:
 *
92
 *   Device did not checksum this packet e.g. due to lack of capabilities.
93 94 95 96 97 98 99
 *   The packet contains full (though not verified) checksum in packet but
 *   not in skb->csum. Thus, skb->csum is undefined in this case.
 *
 * CHECKSUM_UNNECESSARY:
 *
 *   The hardware you're dealing with doesn't calculate the full checksum
 *   (as in CHECKSUM_COMPLETE), but it does parse headers and verify checksums
100 101
 *   for specific protocols. For such packets it will set CHECKSUM_UNNECESSARY
 *   if their checksums are okay. skb->csum is still undefined in this case
102 103
 *   though. A driver or device must never modify the checksum field in the
 *   packet even if checksum is verified.
104 105 106 107 108 109 110 111
 *
 *   CHECKSUM_UNNECESSARY is applicable to following protocols:
 *     TCP: IPv6 and IPv4.
 *     UDP: IPv4 and IPv6. A device may apply CHECKSUM_UNNECESSARY to a
 *       zero UDP checksum for either IPv4 or IPv6, the networking stack
 *       may perform further validation in this case.
 *     GRE: only if the checksum is present in the header.
 *     SCTP: indicates the CRC in SCTP header has been validated.
112
 *     FCOE: indicates the CRC in FC frame has been validated.
113 114 115 116 117 118 119 120 121 122
 *
 *   skb->csum_level indicates the number of consecutive checksums found in
 *   the packet minus one that have been verified as CHECKSUM_UNNECESSARY.
 *   For instance if a device receives an IPv6->UDP->GRE->IPv4->TCP packet
 *   and a device is able to verify the checksums for UDP (possibly zero),
 *   GRE (checksum flag is set), and TCP-- skb->csum_level would be set to
 *   two. If the device were only able to verify the UDP checksum and not
 *   GRE, either because it doesn't support GRE checksum of because GRE
 *   checksum is bad, skb->csum_level would be set to zero (TCP checksum is
 *   not considered in this case).
123 124 125 126 127 128 129
 *
 * CHECKSUM_COMPLETE:
 *
 *   This is the most generic way. The device supplied checksum of the _whole_
 *   packet as seen by netif_rx() and fills out in skb->csum. Meaning, the
 *   hardware doesn't need to parse L3/L4 headers to implement this.
 *
130 131 132 133
 *   Notes:
 *   - Even if device supports only some protocols, but is able to produce
 *     skb->csum, it MUST use CHECKSUM_COMPLETE, not CHECKSUM_UNNECESSARY.
 *   - CHECKSUM_COMPLETE is not applicable to SCTP and FCoE protocols.
134 135 136
 *
 * CHECKSUM_PARTIAL:
 *
137 138
 *   A checksum is set up to be offloaded to a device as described in the
 *   output description for CHECKSUM_PARTIAL. This may occur on a packet
139
 *   received directly from another Linux OS, e.g., a virtualized Linux kernel
140 141 142 143 144 145
 *   on the same host, or it may be set in the input path in GRO or remote
 *   checksum offload. For the purposes of checksum verification, the checksum
 *   referred to by skb->csum_start + skb->csum_offset and any preceding
 *   checksums in the packet are considered verified. Any checksums in the
 *   packet that are after the checksum being offloaded are not considered to
 *   be verified.
146
 *
147 148
 * C. Checksumming on transmit for non-GSO. The stack requests checksum offload
 *    in the skb->ip_summed for a packet. Values are:
149 150 151
 *
 * CHECKSUM_PARTIAL:
 *
152
 *   The driver is required to checksum the packet as seen by hard_start_xmit()
153
 *   from skb->csum_start up to the end, and to record/write the checksum at
154 155 156 157 158 159 160 161 162 163 164 165 166 167
 *   offset skb->csum_start + skb->csum_offset. A driver may verify that the
 *   csum_start and csum_offset values are valid values given the length and
 *   offset of the packet, however they should not attempt to validate that the
 *   checksum refers to a legitimate transport layer checksum-- it is the
 *   purview of the stack to validate that csum_start and csum_offset are set
 *   correctly.
 *
 *   When the stack requests checksum offload for a packet, the driver MUST
 *   ensure that the checksum is set correctly. A driver can either offload the
 *   checksum calculation to the device, or call skb_checksum_help (in the case
 *   that the device does not support offload for a particular checksum).
 *
 *   NETIF_F_IP_CSUM and NETIF_F_IPV6_CSUM are being deprecated in favor of
 *   NETIF_F_HW_CSUM. New devices should use NETIF_F_HW_CSUM to indicate
168 169 170 171 172
 *   checksum offload capability.
 *   skb_csum_hwoffload_help() can be called to resolve CHECKSUM_PARTIAL based
 *   on network device checksumming capabilities: if a packet does not match
 *   them, skb_checksum_help or skb_crc32c_help (depending on the value of
 *   csum_not_inet, see item D.) is called to resolve the checksum.
173
 *
174
 * CHECKSUM_NONE:
175
 *
176 177
 *   The skb was already checksummed by the protocol, or a checksum is not
 *   required.
178 179 180
 *
 * CHECKSUM_UNNECESSARY:
 *
181 182
 *   This has the same meaning on as CHECKSUM_NONE for checksum offload on
 *   output.
183
 *
184 185 186 187 188 189 190 191
 * CHECKSUM_COMPLETE:
 *   Not used in checksum output. If a driver observes a packet with this value
 *   set in skbuff, if should treat as CHECKSUM_NONE being set.
 *
 * D. Non-IP checksum (CRC) offloads
 *
 *   NETIF_F_SCTP_CRC - This feature indicates that a device is capable of
 *     offloading the SCTP CRC in a packet. To perform this offload the stack
192 193 194 195 196 197 198
 *     will set set csum_start and csum_offset accordingly, set ip_summed to
 *     CHECKSUM_PARTIAL and set csum_not_inet to 1, to provide an indication in
 *     the skbuff that the CHECKSUM_PARTIAL refers to CRC32c.
 *     A driver that supports both IP checksum offload and SCTP CRC32c offload
 *     must verify which offload is configured for a packet by testing the
 *     value of skb->csum_not_inet; skb_crc32c_csum_help is provided to resolve
 *     CHECKSUM_PARTIAL on skbs where csum_not_inet is set to 1.
199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216
 *
 *   NETIF_F_FCOE_CRC - This feature indicates that a device is capable of
 *     offloading the FCOE CRC in a packet. To perform this offload the stack
 *     will set ip_summed to CHECKSUM_PARTIAL and set csum_start and csum_offset
 *     accordingly. Note the there is no indication in the skbuff that the
 *     CHECKSUM_PARTIAL refers to an FCOE checksum, a driver that supports
 *     both IP checksum offload and FCOE CRC offload must verify which offload
 *     is configured for a packet presumably by inspecting packet headers.
 *
 * E. Checksumming on output with GSO.
 *
 * In the case of a GSO packet (skb_is_gso(skb) is true), checksum offload
 * is implied by the SKB_GSO_* flags in gso_type. Most obviously, if the
 * gso_type is SKB_GSO_TCPV4 or SKB_GSO_TCPV6, TCP checksum offload as
 * part of the GSO operation is implied. If a checksum is being offloaded
 * with GSO then ip_summed is CHECKSUM_PARTIAL, csum_start and csum_offset
 * are set to refer to the outermost checksum being offload (two offloaded
 * checksums are possible with UDP encapsulation).
217 218
 */

219
/* Don't change this without changing skb_csum_unnecessary! */
220 221 222 223
#define CHECKSUM_NONE		0
#define CHECKSUM_UNNECESSARY	1
#define CHECKSUM_COMPLETE	2
#define CHECKSUM_PARTIAL	3
L
Linus Torvalds 已提交
224

225 226 227
/* Maximum value in skb->csum_level */
#define SKB_MAX_CSUM_LEVEL	3

228
#define SKB_DATA_ALIGN(X)	ALIGN(X, SMP_CACHE_BYTES)
229
#define SKB_WITH_OVERHEAD(X)	\
230
	((X) - SKB_DATA_ALIGN(sizeof(struct skb_shared_info)))
231 232
#define SKB_MAX_ORDER(X, ORDER) \
	SKB_WITH_OVERHEAD((PAGE_SIZE << (ORDER)) - (X))
L
Linus Torvalds 已提交
233 234 235
#define SKB_MAX_HEAD(X)		(SKB_MAX_ORDER((X), 0))
#define SKB_MAX_ALLOC		(SKB_MAX_ORDER(0, 2))

E
Eric Dumazet 已提交
236 237 238 239 240
/* return minimum truesize of one skb containing X bytes of data */
#define SKB_TRUESIZE(X) ((X) +						\
			 SKB_DATA_ALIGN(sizeof(struct sk_buff)) +	\
			 SKB_DATA_ALIGN(sizeof(struct skb_shared_info)))

L
Linus Torvalds 已提交
241
struct net_device;
242
struct scatterlist;
J
Jens Axboe 已提交
243
struct pipe_inode_info;
H
Herbert Xu 已提交
244
struct iov_iter;
245
struct napi_struct;
L
Linus Torvalds 已提交
246

247
#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
L
Linus Torvalds 已提交
248 249 250
struct nf_conntrack {
	atomic_t use;
};
251
#endif
L
Linus Torvalds 已提交
252

253
#if IS_ENABLED(CONFIG_BRIDGE_NETFILTER)
L
Linus Torvalds 已提交
254
struct nf_bridge_info {
255
	refcount_t		use;
256 257 258 259
	enum {
		BRNF_PROTO_UNCHANGED,
		BRNF_PROTO_8021Q,
		BRNF_PROTO_PPPOE
260
	} orig_proto:8;
261 262 263
	u8			pkt_otherhost:1;
	u8			in_prerouting:1;
	u8			bridged_dnat:1;
264
	__u16			frag_max_size;
265
	struct net_device	*physindev;
266 267 268

	/* always valid & non-NULL from FORWARD on, for physdev match */
	struct net_device	*physoutdev;
269
	union {
270
		/* prerouting: detect dnat in orig/reply direction */
271 272
		__be32          ipv4_daddr;
		struct in6_addr ipv6_daddr;
273 274 275 276 277 278

		/* after prerouting + nat detected: store original source
		 * mac since neigh resolution overwrites it, only used while
		 * skb is out in neigh layer.
		 */
		char neigh_header[8];
279
	};
L
Linus Torvalds 已提交
280 281 282 283 284 285 286 287 288 289 290 291 292 293
};
#endif

struct sk_buff_head {
	/* These two members must be first. */
	struct sk_buff	*next;
	struct sk_buff	*prev;

	__u32		qlen;
	spinlock_t	lock;
};

struct sk_buff;

294 295 296 297 298 299
/* To allow 64K frame to be packed as single skb without frag_list we
 * require 64K/PAGE_SIZE pages plus 1 additional page to allow for
 * buffers which do not start on a page boundary.
 *
 * Since GRO uses frags we allocate at least 16 regardless of page
 * size.
300
 */
301
#if (65536/PAGE_SIZE + 1) < 16
302
#define MAX_SKB_FRAGS 16UL
303
#else
304
#define MAX_SKB_FRAGS (65536/PAGE_SIZE + 1)
305
#endif
H
Hans Westgaard Ry 已提交
306
extern int sysctl_max_skb_frags;
L
Linus Torvalds 已提交
307

308 309 310 311 312
/* Set skb_shinfo(skb)->gso_size to this in case you want skb_segment to
 * segment using its current segmentation instead.
 */
#define GSO_BY_FRAGS	0xFFFF

L
Linus Torvalds 已提交
313 314 315
typedef struct skb_frag_struct skb_frag_t;

struct skb_frag_struct {
316 317 318
	struct {
		struct page *p;
	} page;
319
#if (BITS_PER_LONG > 32) || (PAGE_SIZE >= 65536)
320 321
	__u32 page_offset;
	__u32 size;
322 323 324 325
#else
	__u16 page_offset;
	__u16 size;
#endif
L
Linus Torvalds 已提交
326 327
};

E
Eric Dumazet 已提交
328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347
static inline unsigned int skb_frag_size(const skb_frag_t *frag)
{
	return frag->size;
}

static inline void skb_frag_size_set(skb_frag_t *frag, unsigned int size)
{
	frag->size = size;
}

static inline void skb_frag_size_add(skb_frag_t *frag, int delta)
{
	frag->size += delta;
}

static inline void skb_frag_size_sub(skb_frag_t *frag, int delta)
{
	frag->size -= delta;
}

348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383
static inline bool skb_frag_must_loop(struct page *p)
{
#if defined(CONFIG_HIGHMEM)
	if (PageHighMem(p))
		return true;
#endif
	return false;
}

/**
 *	skb_frag_foreach_page - loop over pages in a fragment
 *
 *	@f:		skb frag to operate on
 *	@f_off:		offset from start of f->page.p
 *	@f_len:		length from f_off to loop over
 *	@p:		(temp var) current page
 *	@p_off:		(temp var) offset from start of current page,
 *	                           non-zero only on first page.
 *	@p_len:		(temp var) length in current page,
 *				   < PAGE_SIZE only on first and last page.
 *	@copied:	(temp var) length so far, excluding current p_len.
 *
 *	A fragment can hold a compound page, in which case per-page
 *	operations, notably kmap_atomic, must be called for each
 *	regular page.
 */
#define skb_frag_foreach_page(f, f_off, f_len, p, p_off, p_len, copied)	\
	for (p = skb_frag_page(f) + ((f_off) >> PAGE_SHIFT),		\
	     p_off = (f_off) & (PAGE_SIZE - 1),				\
	     p_len = skb_frag_must_loop(p) ?				\
	     min_t(u32, f_len, PAGE_SIZE - p_off) : f_len,		\
	     copied = 0;						\
	     copied < f_len;						\
	     copied += p_len, p++, p_off = 0,				\
	     p_len = min_t(u32, f_len - copied, PAGE_SIZE))		\

384 385 386
#define HAVE_HW_TIME_STAMP

/**
387
 * struct skb_shared_hwtstamps - hardware time stamps
388 389 390 391
 * @hwtstamp:	hardware time stamp transformed into duration
 *		since arbitrary point in time
 *
 * Software time stamps generated by ktime_get_real() are stored in
392
 * skb->tstamp.
393 394 395 396 397 398 399 400 401 402 403
 *
 * hwtstamps can only be compared against other hwtstamps from
 * the same device.
 *
 * This structure is attached to packets as part of the
 * &skb_shared_info. Use skb_hwtstamps() to get a pointer.
 */
struct skb_shared_hwtstamps {
	ktime_t	hwtstamp;
};

404 405 406 407 408
/* Definitions for tx_flags in struct skb_shared_info */
enum {
	/* generate hardware time stamp */
	SKBTX_HW_TSTAMP = 1 << 0,

409
	/* generate software time stamp when queueing packet to NIC */
410 411 412 413 414
	SKBTX_SW_TSTAMP = 1 << 1,

	/* device driver is going to provide hardware time stamp */
	SKBTX_IN_PROGRESS = 1 << 2,

415
	/* device driver supports TX zero-copy buffers */
E
Eric Dumazet 已提交
416
	SKBTX_DEV_ZEROCOPY = 1 << 3,
417 418

	/* generate wifi status information (where possible) */
E
Eric Dumazet 已提交
419
	SKBTX_WIFI_STATUS = 1 << 4,
420 421 422 423 424 425 426

	/* This indicates at least one fragment might be overwritten
	 * (as in vmsplice(), sendfile() ...)
	 * If we need to compute a TX checksum, we'll need to copy
	 * all frags to avoid possible bad checksum
	 */
	SKBTX_SHARED_FRAG = 1 << 5,
427 428 429

	/* generate software time stamp when entering packet scheduling */
	SKBTX_SCHED_TSTAMP = 1 << 6,
430 431
};

432
#define SKBTX_ANY_SW_TSTAMP	(SKBTX_SW_TSTAMP    | \
433
				 SKBTX_SCHED_TSTAMP)
434 435
#define SKBTX_ANY_TSTAMP	(SKBTX_HW_TSTAMP | SKBTX_ANY_SW_TSTAMP)

436 437 438
/*
 * The callback notifies userspace to release buffers when skb DMA is done in
 * lower device, the skb last reference should be 0 when calling this.
439 440
 * The zerocopy_success argument is true if zero copy transmit occurred,
 * false on data copy or out of memory error caused by data copy attempt.
441 442
 * The ctx field is used to track device context.
 * The desc field is used to track userspace buffer index.
443 444
 */
struct ubuf_info {
445
	void (*callback)(struct ubuf_info *, bool zerocopy_success);
446
	void *ctx;
447
	unsigned long desc;
448 449
};

L
Linus Torvalds 已提交
450 451 452 453
/* This data is invariant across clones and lives at
 * the end of the header data, ie. at skb->end.
 */
struct skb_shared_info {
454
	unsigned short	_unused;
455 456
	unsigned char	nr_frags;
	__u8		tx_flags;
457 458 459
	unsigned short	gso_size;
	/* Warning: this field is not always filled in (UFO)! */
	unsigned short	gso_segs;
L
Linus Torvalds 已提交
460
	struct sk_buff	*frag_list;
461
	struct skb_shared_hwtstamps hwtstamps;
462
	unsigned int	gso_type;
463
	u32		tskey;
464
	__be32          ip6_frag_id;
E
Eric Dumazet 已提交
465 466 467 468 469 470

	/*
	 * Warning : all fields before dataref are cleared in __alloc_skb()
	 */
	atomic_t	dataref;

J
Johann Baudy 已提交
471 472 473
	/* Intermediate layers must ensure that destructor_arg
	 * remains valid until skb destructor */
	void *		destructor_arg;
474

475 476
	/* must be last field, see pskb_expand_head() */
	skb_frag_t	frags[MAX_SKB_FRAGS];
L
Linus Torvalds 已提交
477 478 479 480
};

/* We divide dataref into two halves.  The higher 16 bits hold references
 * to the payload part of skb->data.  The lower 16 bits hold references to
481 482
 * the entire skb->data.  A clone of a headerless skb holds the length of
 * the header in skb->hdr_len.
L
Linus Torvalds 已提交
483 484 485 486 487 488 489 490 491 492
 *
 * All users must obey the rule that the skb->data reference count must be
 * greater than or equal to the payload reference count.
 *
 * Holding a reference to the payload part means that the user does not
 * care about modifications to the header part of skb->data.
 */
#define SKB_DATAREF_SHIFT 16
#define SKB_DATAREF_MASK ((1 << SKB_DATAREF_SHIFT) - 1)

493 494

enum {
495 496 497
	SKB_FCLONE_UNAVAILABLE,	/* skb has no fclone (from head_cache) */
	SKB_FCLONE_ORIG,	/* orig skb (from fclone_cache) */
	SKB_FCLONE_CLONE,	/* companion fclone skb (from fclone_cache) */
498 499
};

500 501
enum {
	SKB_GSO_TCPV4 = 1 << 0,
502 503

	/* This indicates the skb is from an untrusted source. */
504
	SKB_GSO_DODGY = 1 << 1,
M
Michael Chan 已提交
505 506

	/* This indicates the tcp segment has CWR set. */
507
	SKB_GSO_TCP_ECN = 1 << 2,
H
Herbert Xu 已提交
508

509
	SKB_GSO_TCP_FIXEDID = 1 << 3,
510

511
	SKB_GSO_TCPV6 = 1 << 4,
512

513
	SKB_GSO_FCOE = 1 << 5,
514

515
	SKB_GSO_GRE = 1 << 6,
S
Simon Horman 已提交
516

517
	SKB_GSO_GRE_CSUM = 1 << 7,
E
Eric Dumazet 已提交
518

519
	SKB_GSO_IPXIP4 = 1 << 8,
E
Eric Dumazet 已提交
520

521
	SKB_GSO_IPXIP6 = 1 << 9,
522

523
	SKB_GSO_UDP_TUNNEL = 1 << 10,
T
Tom Herbert 已提交
524

525
	SKB_GSO_UDP_TUNNEL_CSUM = 1 << 11,
526

527
	SKB_GSO_PARTIAL = 1 << 12,
528

529
	SKB_GSO_TUNNEL_REMCSUM = 1 << 13,
M
Marcelo Ricardo Leitner 已提交
530

531
	SKB_GSO_SCTP = 1 << 14,
S
Steffen Klassert 已提交
532

533
	SKB_GSO_ESP = 1 << 15,
534 535
};

536 537 538 539 540 541 542 543 544 545
#if BITS_PER_LONG > 32
#define NET_SKBUFF_DATA_USES_OFFSET 1
#endif

#ifdef NET_SKBUFF_DATA_USES_OFFSET
typedef unsigned int sk_buff_data_t;
#else
typedef unsigned char *sk_buff_data_t;
#endif

L
Linus Torvalds 已提交
546 547 548 549
/** 
 *	struct sk_buff - socket buffer
 *	@next: Next buffer in list
 *	@prev: Previous buffer in list
550
 *	@tstamp: Time we arrived/left
E
Eric Dumazet 已提交
551
 *	@rbnode: RB tree node, alternative to next/prev for netem/tcp
552
 *	@sk: Socket we are owned by
L
Linus Torvalds 已提交
553
 *	@dev: Device we arrived on/are leaving by
554
 *	@cb: Control buffer. Free for use by every layer. Put private vars here
E
Eric Dumazet 已提交
555
 *	@_skb_refdst: destination entry (with norefcount bit)
556
 *	@sp: the security path, used for xfrm
L
Linus Torvalds 已提交
557 558 559
 *	@len: Length of actual data
 *	@data_len: Data length
 *	@mac_len: Length of link layer header
560
 *	@hdr_len: writable header length of cloned skb
561 562 563
 *	@csum: Checksum (must include start/offset pair)
 *	@csum_start: Offset from skb->head where checksumming should start
 *	@csum_offset: Offset from csum_start where checksum should be stored
564
 *	@priority: Packet queueing priority
W
WANG Cong 已提交
565
 *	@ignore_df: allow local fragmentation
L
Linus Torvalds 已提交
566
 *	@cloned: Head may be cloned (check refcnt to be sure)
567
 *	@ip_summed: Driver fed us an IP checksum
L
Linus Torvalds 已提交
568 569
 *	@nohdr: Payload reference only, must not modify header
 *	@pkt_type: Packet class
570 571
 *	@fclone: skbuff clone status
 *	@ipvs_property: skbuff is owned by ipvs
572
 *	@tc_skip_classify: do not classify packet. set by IFB device
573
 *	@tc_at_ingress: used within tc_classify to distinguish in/egress
574 575
 *	@tc_redirected: packet was redirected by a tc action
 *	@tc_from_ingress: if tc_redirected, tc_at_ingress at time of redirect
576 577
 *	@peeked: this packet has been seen already, so stats have been
 *		done for it, don't do them again
578
 *	@nf_trace: netfilter packet trace flag
579 580
 *	@protocol: Packet protocol from driver
 *	@destructor: Destruct function
581
 *	@_nfct: Associated connection, if any (with nfctinfo bits)
L
Linus Torvalds 已提交
582
 *	@nf_bridge: Saved data about a bridged frame - see br_netfilter.c
583
 *	@skb_iif: ifindex of device we arrived on
L
Linus Torvalds 已提交
584
 *	@tc_index: Traffic control index
585
 *	@hash: the packet hash
586
 *	@queue_mapping: Queue mapping for multiqueue devices
587
 *	@xmit_more: More SKBs are pending for this queue
588
 *	@ndisc_nodetype: router type (from link layer)
589
 *	@ooo_okay: allow the mapping of a socket to a queue to be changed
590
 *	@l4_hash: indicate hash is a canonical 4-tuple hash over transport
591
 *		ports.
592
 *	@sw_hash: indicates hash was computed in software stack
593 594
 *	@wifi_acked_valid: wifi_acked was set
 *	@wifi_acked: whether frame was acked on wifi or not
595
 *	@no_fcs:  Request NIC to treat last 4 bytes as Ethernet FCS
596
 *	@csum_not_inet: use CRC32c to resolve CHECKSUM_PARTIAL
597
 *	@dst_pending_confirm: need to confirm neighbour
E
Eliezer Tamir 已提交
598
  *	@napi_id: id of the NAPI struct this skb came from
599
 *	@secmark: security marking
600
 *	@mark: Generic packet mark
601
 *	@vlan_proto: vlan encapsulation protocol
602
 *	@vlan_tci: vlan tag control information
S
Simon Horman 已提交
603
 *	@inner_protocol: Protocol (encapsulation)
604 605
 *	@inner_transport_header: Inner transport layer header (encapsulation)
 *	@inner_network_header: Network layer header (encapsulation)
606
 *	@inner_mac_header: Link layer header (encapsulation)
607 608 609 610 611 612 613 614 615
 *	@transport_header: Transport layer header
 *	@network_header: Network layer header
 *	@mac_header: Link layer header
 *	@tail: Tail pointer
 *	@end: End pointer
 *	@head: Head of buffer
 *	@data: Data head pointer
 *	@truesize: Buffer size
 *	@users: User count - see {datagram,tcp}.c
L
Linus Torvalds 已提交
616 617 618
 */

struct sk_buff {
619
	union {
E
Eric Dumazet 已提交
620 621 622 623 624 625 626
		struct {
			/* These two members must be first. */
			struct sk_buff		*next;
			struct sk_buff		*prev;

			union {
				ktime_t		tstamp;
627
				u64		skb_mstamp;
E
Eric Dumazet 已提交
628 629 630
			};
		};
		struct rb_node	rbnode; /* used in netem & tcp stack */
631
	};
632
	struct sock		*sk;
L
Linus Torvalds 已提交
633

634 635 636 637 638 639 640 641
	union {
		struct net_device	*dev;
		/* Some protocols might use this space to store information,
		 * while device pointer would be NULL.
		 * UDP receive path is one user.
		 */
		unsigned long		dev_scratch;
	};
L
Linus Torvalds 已提交
642 643 644 645 646 647
	/*
	 * This is the control buffer. It is free to use for every
	 * layer. Please put your private variables there. If you
	 * want to keep them across layers you have to do a skb_clone()
	 * first. This is owned by whoever has the skb queued ATM.
	 */
648
	char			cb[48] __aligned(8);
L
Linus Torvalds 已提交
649

E
Eric Dumazet 已提交
650
	unsigned long		_skb_refdst;
651
	void			(*destructor)(struct sk_buff *skb);
652 653
#ifdef CONFIG_XFRM
	struct	sec_path	*sp;
654 655
#endif
#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
656
	unsigned long		 _nfct;
657
#endif
658
#if IS_ENABLED(CONFIG_BRIDGE_NETFILTER)
659
	struct nf_bridge_info	*nf_bridge;
660
#endif
L
Linus Torvalds 已提交
661
	unsigned int		len,
662 663 664
				data_len;
	__u16			mac_len,
				hdr_len;
665 666 667 668

	/* Following fields are _not_ copied in __copy_skb_header()
	 * Note that queue_mapping is here mostly to fill a hole.
	 */
669
	kmemcheck_bitfield_begin(flags1);
670
	__u16			queue_mapping;
671 672 673 674 675 676 677 678 679 680

/* if you move cloned around you also must adapt those constants */
#ifdef __BIG_ENDIAN_BITFIELD
#define CLONED_MASK	(1 << 7)
#else
#define CLONED_MASK	1
#endif
#define CLONED_OFFSET()		offsetof(struct sk_buff, __cloned_offset)

	__u8			__cloned_offset[0];
681
	__u8			cloned:1,
682
				nohdr:1,
683
				fclone:2,
684
				peeked:1,
685
				head_frag:1,
686 687
				xmit_more:1,
				__unused:1; /* one bit hole */
688
	kmemcheck_bitfield_end(flags1);
689

690 691 692
	/* fields enclosed in headers_start/headers_end are copied
	 * using a single memcpy() in __copy_skb_header()
	 */
693
	/* private: */
694
	__u32			headers_start[0];
695
	/* public: */
696

697 698 699 700 701
/* if you move pkt_type around you also must adapt those constants */
#ifdef __BIG_ENDIAN_BITFIELD
#define PKT_TYPE_MAX	(7 << 5)
#else
#define PKT_TYPE_MAX	7
L
Linus Torvalds 已提交
702
#endif
703
#define PKT_TYPE_OFFSET()	offsetof(struct sk_buff, __pkt_type_offset)
704

705
	__u8			__pkt_type_offset[0];
706
	__u8			pkt_type:3;
707
	__u8			pfmemalloc:1;
708 709 710 711
	__u8			ignore_df:1;

	__u8			nf_trace:1;
	__u8			ip_summed:2;
712
	__u8			ooo_okay:1;
713
	__u8			l4_hash:1;
714
	__u8			sw_hash:1;
715 716
	__u8			wifi_acked_valid:1;
	__u8			wifi_acked:1;
717

718
	__u8			no_fcs:1;
719
	/* Indicates the inner headers are valid in the skbuff. */
720
	__u8			encapsulation:1;
721
	__u8			encap_hdr_csum:1;
722
	__u8			csum_valid:1;
723
	__u8			csum_complete_sw:1;
724
	__u8			csum_level:2;
725
	__u8			csum_not_inet:1;
726

727
	__u8			dst_pending_confirm:1;
728 729 730 731
#ifdef CONFIG_IPV6_NDISC_NODETYPE
	__u8			ndisc_nodetype:2;
#endif
	__u8			ipvs_property:1;
T
Tom Herbert 已提交
732
	__u8			inner_protocol_type:1;
733
	__u8			remcsum_offload:1;
734 735 736
#ifdef CONFIG_NET_SWITCHDEV
	__u8			offload_fwd_mark:1;
#endif
737 738
#ifdef CONFIG_NET_CLS_ACT
	__u8			tc_skip_classify:1;
739
	__u8			tc_at_ingress:1;
740 741
	__u8			tc_redirected:1;
	__u8			tc_from_ingress:1;
742
#endif
743 744 745 746

#ifdef CONFIG_NET_SCHED
	__u16			tc_index;	/* traffic control index */
#endif
747

748 749 750 751 752 753 754 755 756 757 758 759
	union {
		__wsum		csum;
		struct {
			__u16	csum_start;
			__u16	csum_offset;
		};
	};
	__u32			priority;
	int			skb_iif;
	__u32			hash;
	__be16			vlan_proto;
	__u16			vlan_tci;
E
Eric Dumazet 已提交
760 761 762 763 764
#if defined(CONFIG_NET_RX_BUSY_POLL) || defined(CONFIG_XPS)
	union {
		unsigned int	napi_id;
		unsigned int	sender_cpu;
	};
765
#endif
766
#ifdef CONFIG_NETWORK_SECMARK
767
	__u32		secmark;
768 769
#endif

770 771
	union {
		__u32		mark;
E
Eric Dumazet 已提交
772
		__u32		reserved_tailroom;
773
	};
L
Linus Torvalds 已提交
774

T
Tom Herbert 已提交
775 776 777 778 779
	union {
		__be16		inner_protocol;
		__u8		inner_ipproto;
	};

780 781 782
	__u16			inner_transport_header;
	__u16			inner_network_header;
	__u16			inner_mac_header;
783 784

	__be16			protocol;
785 786 787
	__u16			transport_header;
	__u16			network_header;
	__u16			mac_header;
788

789
	/* private: */
790
	__u32			headers_end[0];
791
	/* public: */
792

L
Linus Torvalds 已提交
793
	/* These elements must be at the end, see alloc_skb() for details.  */
794
	sk_buff_data_t		tail;
795
	sk_buff_data_t		end;
L
Linus Torvalds 已提交
796
	unsigned char		*head,
797
				*data;
798
	unsigned int		truesize;
799
	refcount_t		users;
L
Linus Torvalds 已提交
800 801 802 803 804 805 806 807 808
};

#ifdef __KERNEL__
/*
 *	Handling routines are only of interest to the kernel
 */
#include <linux/slab.h>


809 810
#define SKB_ALLOC_FCLONE	0x01
#define SKB_ALLOC_RX		0x02
811
#define SKB_ALLOC_NAPI		0x04
812 813 814 815 816 817 818

/* Returns true if the skb was allocated from PFMEMALLOC reserves */
static inline bool skb_pfmemalloc(const struct sk_buff *skb)
{
	return unlikely(skb->pfmemalloc);
}

E
Eric Dumazet 已提交
819 820 821 822 823 824 825
/*
 * skb might have a dst pointer attached, refcounted or not.
 * _skb_refdst low order bit is set if refcount was _not_ taken
 */
#define SKB_DST_NOREF	1UL
#define SKB_DST_PTRMASK	~(SKB_DST_NOREF)

826
#define SKB_NFCT_PTRMASK	~(7UL)
E
Eric Dumazet 已提交
827 828 829 830 831 832
/**
 * skb_dst - returns skb dst_entry
 * @skb: buffer
 *
 * Returns skb dst_entry, regardless of reference taken or not.
 */
E
Eric Dumazet 已提交
833 834
static inline struct dst_entry *skb_dst(const struct sk_buff *skb)
{
E
Eric Dumazet 已提交
835 836 837 838 839 840 841
	/* If refdst was not refcounted, check we still are in a 
	 * rcu_read_lock section
	 */
	WARN_ON((skb->_skb_refdst & SKB_DST_NOREF) &&
		!rcu_read_lock_held() &&
		!rcu_read_lock_bh_held());
	return (struct dst_entry *)(skb->_skb_refdst & SKB_DST_PTRMASK);
E
Eric Dumazet 已提交
842 843
}

E
Eric Dumazet 已提交
844 845 846 847 848 849 850 851
/**
 * skb_dst_set - sets skb dst
 * @skb: buffer
 * @dst: dst entry
 *
 * Sets skb dst, assuming a reference was taken on dst and should
 * be released by skb_dst_drop()
 */
E
Eric Dumazet 已提交
852 853
static inline void skb_dst_set(struct sk_buff *skb, struct dst_entry *dst)
{
E
Eric Dumazet 已提交
854 855 856
	skb->_skb_refdst = (unsigned long)dst;
}

857 858 859 860 861 862 863 864 865 866 867 868
/**
 * skb_dst_set_noref - sets skb dst, hopefully, without taking reference
 * @skb: buffer
 * @dst: dst entry
 *
 * Sets skb dst, assuming a reference was not taken on dst.
 * If dst entry is cached, we do not take reference and dst_release
 * will be avoided by refdst_drop. If dst entry is not cached, we take
 * reference, so that last dst_release can destroy the dst immediately.
 */
static inline void skb_dst_set_noref(struct sk_buff *skb, struct dst_entry *dst)
{
869 870
	WARN_ON(!rcu_read_lock_held() && !rcu_read_lock_bh_held());
	skb->_skb_refdst = (unsigned long)dst | SKB_DST_NOREF;
871
}
E
Eric Dumazet 已提交
872 873

/**
L
Lucas De Marchi 已提交
874
 * skb_dst_is_noref - Test if skb dst isn't refcounted
E
Eric Dumazet 已提交
875 876 877 878 879
 * @skb: buffer
 */
static inline bool skb_dst_is_noref(const struct sk_buff *skb)
{
	return (skb->_skb_refdst & SKB_DST_NOREF) && skb_dst(skb);
E
Eric Dumazet 已提交
880 881
}

E
Eric Dumazet 已提交
882 883
static inline struct rtable *skb_rtable(const struct sk_buff *skb)
{
E
Eric Dumazet 已提交
884
	return (struct rtable *)skb_dst(skb);
E
Eric Dumazet 已提交
885 886
}

887 888 889 890 891 892 893 894 895
/* For mangling skb->pkt_type from user space side from applications
 * such as nft, tc, etc, we only allow a conservative subset of
 * possible pkt_types to be set.
*/
static inline bool skb_pkt_type_ok(u32 ptype)
{
	return ptype <= PACKET_OTHERHOST;
}

896 897 898 899 900 901 902 903 904
static inline unsigned int skb_napi_id(const struct sk_buff *skb)
{
#ifdef CONFIG_NET_RX_BUSY_POLL
	return skb->napi_id;
#else
	return 0;
#endif
}

905 906 907 908 909
/* decrement the reference count and return true if we can free the skb */
static inline bool skb_unref(struct sk_buff *skb)
{
	if (unlikely(!skb))
		return false;
910
	if (likely(refcount_read(&skb->users) == 1))
911
		smp_rmb();
912
	else if (likely(!refcount_dec_and_test(&skb->users)))
913 914 915 916 917
		return false;

	return true;
}

P
Paolo Abeni 已提交
918
void skb_release_head_state(struct sk_buff *skb);
919 920 921 922
void kfree_skb(struct sk_buff *skb);
void kfree_skb_list(struct sk_buff *segs);
void skb_tx_error(struct sk_buff *skb);
void consume_skb(struct sk_buff *skb);
P
Paolo Abeni 已提交
923
void consume_stateless_skb(struct sk_buff *skb);
924
void  __kfree_skb(struct sk_buff *skb);
925
extern struct kmem_cache *skbuff_head_cache;
E
Eric Dumazet 已提交
926

927 928 929
void kfree_skb_partial(struct sk_buff *skb, bool head_stolen);
bool skb_try_coalesce(struct sk_buff *to, struct sk_buff *from,
		      bool *fragstolen, int *delta_truesize);
E
Eric Dumazet 已提交
930

931 932
struct sk_buff *__alloc_skb(unsigned int size, gfp_t priority, int flags,
			    int node);
E
Eric Dumazet 已提交
933
struct sk_buff *__build_skb(void *data, unsigned int frag_size);
934
struct sk_buff *build_skb(void *data, unsigned int frag_size);
935
static inline struct sk_buff *alloc_skb(unsigned int size,
A
Al Viro 已提交
936
					gfp_t priority)
937
{
E
Eric Dumazet 已提交
938
	return __alloc_skb(size, priority, 0, NUMA_NO_NODE);
939 940
}

941 942 943 944 945 946
struct sk_buff *alloc_skb_with_frags(unsigned long header_len,
				     unsigned long data_len,
				     int max_page_order,
				     int *errcode,
				     gfp_t gfp_mask);

947 948 949 950 951 952
/* Layout of fast clones : [skb1][skb2][fclone_ref] */
struct sk_buff_fclones {
	struct sk_buff	skb1;

	struct sk_buff	skb2;

953
	refcount_t	fclone_ref;
954 955 956 957
};

/**
 *	skb_fclone_busy - check if fclone is busy
958
 *	@sk: socket
959 960
 *	@skb: buffer
 *
M
Masanari Iida 已提交
961
 * Returns true if skb is a fast clone, and its clone is not freed.
962 963
 * Some drivers call skb_orphan() in their ndo_start_xmit(),
 * so we also check that this didnt happen.
964
 */
965 966
static inline bool skb_fclone_busy(const struct sock *sk,
				   const struct sk_buff *skb)
967 968 969 970 971 972
{
	const struct sk_buff_fclones *fclones;

	fclones = container_of(skb, struct sk_buff_fclones, skb1);

	return skb->fclone == SKB_FCLONE_ORIG &&
973
	       refcount_read(&fclones->fclone_ref) > 1 &&
974
	       fclones->skb2.sk == sk;
975 976
}

977
static inline struct sk_buff *alloc_skb_fclone(unsigned int size,
A
Al Viro 已提交
978
					       gfp_t priority)
979
{
980
	return __alloc_skb(size, priority, SKB_ALLOC_FCLONE, NUMA_NO_NODE);
981 982
}

983 984 985 986
struct sk_buff *skb_morph(struct sk_buff *dst, struct sk_buff *src);
int skb_copy_ubufs(struct sk_buff *skb, gfp_t gfp_mask);
struct sk_buff *skb_clone(struct sk_buff *skb, gfp_t priority);
struct sk_buff *skb_copy(const struct sk_buff *skb, gfp_t priority);
987 988 989 990 991 992 993
struct sk_buff *__pskb_copy_fclone(struct sk_buff *skb, int headroom,
				   gfp_t gfp_mask, bool fclone);
static inline struct sk_buff *__pskb_copy(struct sk_buff *skb, int headroom,
					  gfp_t gfp_mask)
{
	return __pskb_copy_fclone(skb, headroom, gfp_mask, false);
}
994 995 996 997 998 999

int pskb_expand_head(struct sk_buff *skb, int nhead, int ntail, gfp_t gfp_mask);
struct sk_buff *skb_realloc_headroom(struct sk_buff *skb,
				     unsigned int headroom);
struct sk_buff *skb_copy_expand(const struct sk_buff *skb, int newheadroom,
				int newtailroom, gfp_t priority);
1000 1001 1002 1003
int __must_check skb_to_sgvec_nomark(struct sk_buff *skb, struct scatterlist *sg,
				     int offset, int len);
int __must_check skb_to_sgvec(struct sk_buff *skb, struct scatterlist *sg,
			      int offset, int len);
1004 1005
int skb_cow_data(struct sk_buff *skb, int tailbits, struct sk_buff **trailer);
int skb_pad(struct sk_buff *skb, int pad);
1006
#define dev_kfree_skb(a)	consume_skb(a)
L
Linus Torvalds 已提交
1007

1008 1009 1010 1011
int skb_append_datato_frags(struct sock *sk, struct sk_buff *skb,
			    int getfrag(void *from, char *to, int offset,
					int len, int odd, struct sk_buff *skb),
			    void *from, int length);
1012

1013 1014 1015
int skb_append_pagefrags(struct sk_buff *skb, struct page *page,
			 int offset, size_t size);

E
Eric Dumazet 已提交
1016
struct skb_seq_state {
1017 1018 1019 1020 1021 1022 1023 1024 1025
	__u32		lower_offset;
	__u32		upper_offset;
	__u32		frag_idx;
	__u32		stepped_offset;
	struct sk_buff	*root_skb;
	struct sk_buff	*cur_skb;
	__u8		*frag_data;
};

1026 1027 1028 1029 1030
void skb_prepare_seq_read(struct sk_buff *skb, unsigned int from,
			  unsigned int to, struct skb_seq_state *st);
unsigned int skb_seq_read(unsigned int consumed, const u8 **data,
			  struct skb_seq_state *st);
void skb_abort_seq_read(struct skb_seq_state *st);
1031

1032
unsigned int skb_find_text(struct sk_buff *skb, unsigned int from,
1033
			   unsigned int to, struct ts_config *config);
1034

T
Tom Herbert 已提交
1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067
/*
 * Packet hash types specify the type of hash in skb_set_hash.
 *
 * Hash types refer to the protocol layer addresses which are used to
 * construct a packet's hash. The hashes are used to differentiate or identify
 * flows of the protocol layer for the hash type. Hash types are either
 * layer-2 (L2), layer-3 (L3), or layer-4 (L4).
 *
 * Properties of hashes:
 *
 * 1) Two packets in different flows have different hash values
 * 2) Two packets in the same flow should have the same hash value
 *
 * A hash at a higher layer is considered to be more specific. A driver should
 * set the most specific hash possible.
 *
 * A driver cannot indicate a more specific hash than the layer at which a hash
 * was computed. For instance an L3 hash cannot be set as an L4 hash.
 *
 * A driver may indicate a hash level which is less specific than the
 * actual layer the hash was computed on. For instance, a hash computed
 * at L4 may be considered an L3 hash. This should only be done if the
 * driver can't unambiguously determine that the HW computed the hash at
 * the higher layer. Note that the "should" in the second property above
 * permits this.
 */
enum pkt_hash_types {
	PKT_HASH_TYPE_NONE,	/* Undefined type */
	PKT_HASH_TYPE_L2,	/* Input: src_MAC, dest_MAC */
	PKT_HASH_TYPE_L3,	/* Input: src_IP, dst_IP */
	PKT_HASH_TYPE_L4,	/* Input: src_IP, dst_IP, src_port, dst_port */
};

1068
static inline void skb_clear_hash(struct sk_buff *skb)
T
Tom Herbert 已提交
1069
{
1070
	skb->hash = 0;
1071
	skb->sw_hash = 0;
1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085
	skb->l4_hash = 0;
}

static inline void skb_clear_hash_if_not_l4(struct sk_buff *skb)
{
	if (!skb->l4_hash)
		skb_clear_hash(skb);
}

static inline void
__skb_set_hash(struct sk_buff *skb, __u32 hash, bool is_sw, bool is_l4)
{
	skb->l4_hash = is_l4;
	skb->sw_hash = is_sw;
1086
	skb->hash = hash;
T
Tom Herbert 已提交
1087 1088
}

1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101
static inline void
skb_set_hash(struct sk_buff *skb, __u32 hash, enum pkt_hash_types type)
{
	/* Used by drivers to set hash from HW */
	__skb_set_hash(skb, hash, false, type == PKT_HASH_TYPE_L4);
}

static inline void
__skb_set_sw_hash(struct sk_buff *skb, __u32 hash, bool is_l4)
{
	__skb_set_hash(skb, hash, true, is_l4);
}

1102
void __skb_get_hash(struct sk_buff *skb);
1103
u32 __skb_get_hash_symmetric(const struct sk_buff *skb);
1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122
u32 skb_get_poff(const struct sk_buff *skb);
u32 __skb_get_poff(const struct sk_buff *skb, void *data,
		   const struct flow_keys *keys, int hlen);
__be32 __skb_flow_get_ports(const struct sk_buff *skb, int thoff, u8 ip_proto,
			    void *data, int hlen_proto);

static inline __be32 skb_flow_get_ports(const struct sk_buff *skb,
					int thoff, u8 ip_proto)
{
	return __skb_flow_get_ports(skb, thoff, ip_proto, NULL, 0);
}

void skb_flow_dissector_init(struct flow_dissector *flow_dissector,
			     const struct flow_dissector_key *key,
			     unsigned int key_count);

bool __skb_flow_dissect(const struct sk_buff *skb,
			struct flow_dissector *flow_dissector,
			void *target_container,
1123 1124
			void *data, __be16 proto, int nhoff, int hlen,
			unsigned int flags);
1125 1126 1127

static inline bool skb_flow_dissect(const struct sk_buff *skb,
				    struct flow_dissector *flow_dissector,
1128
				    void *target_container, unsigned int flags)
1129 1130
{
	return __skb_flow_dissect(skb, flow_dissector, target_container,
1131
				  NULL, 0, 0, 0, flags);
1132 1133 1134
}

static inline bool skb_flow_dissect_flow_keys(const struct sk_buff *skb,
1135 1136
					      struct flow_keys *flow,
					      unsigned int flags)
1137 1138 1139
{
	memset(flow, 0, sizeof(*flow));
	return __skb_flow_dissect(skb, &flow_keys_dissector, flow,
1140
				  NULL, 0, 0, 0, flags);
1141 1142 1143 1144
}

static inline bool skb_flow_dissect_flow_keys_buf(struct flow_keys *flow,
						  void *data, __be16 proto,
1145 1146
						  int nhoff, int hlen,
						  unsigned int flags)
1147 1148 1149
{
	memset(flow, 0, sizeof(*flow));
	return __skb_flow_dissect(NULL, &flow_keys_buf_dissector, flow,
1150
				  data, proto, nhoff, hlen, flags);
1151 1152
}

1153
static inline __u32 skb_get_hash(struct sk_buff *skb)
1154
{
1155
	if (!skb->l4_hash && !skb->sw_hash)
1156
		__skb_get_hash(skb);
1157

1158
	return skb->hash;
1159 1160
}

1161
static inline __u32 skb_get_hash_flowi6(struct sk_buff *skb, const struct flowi6 *fl6)
1162
{
1163 1164
	if (!skb->l4_hash && !skb->sw_hash) {
		struct flow_keys keys;
1165
		__u32 hash = __get_hash_from_flowi6(fl6, &keys);
1166

1167
		__skb_set_sw_hash(skb, hash, flow_keys_have_l4(&keys));
1168
	}
1169 1170 1171 1172

	return skb->hash;
}

T
Tom Herbert 已提交
1173 1174
__u32 skb_get_hash_perturb(const struct sk_buff *skb, u32 perturb);

T
Tom Herbert 已提交
1175 1176
static inline __u32 skb_get_hash_raw(const struct sk_buff *skb)
{
1177
	return skb->hash;
T
Tom Herbert 已提交
1178 1179
}

1180 1181
static inline void skb_copy_hash(struct sk_buff *to, const struct sk_buff *from)
{
1182
	to->hash = from->hash;
1183
	to->sw_hash = from->sw_hash;
1184
	to->l4_hash = from->l4_hash;
1185 1186
};

1187 1188 1189 1190 1191
#ifdef NET_SKBUFF_DATA_USES_OFFSET
static inline unsigned char *skb_end_pointer(const struct sk_buff *skb)
{
	return skb->head + skb->end;
}
1192 1193 1194 1195 1196

static inline unsigned int skb_end_offset(const struct sk_buff *skb)
{
	return skb->end;
}
1197 1198 1199 1200 1201
#else
static inline unsigned char *skb_end_pointer(const struct sk_buff *skb)
{
	return skb->end;
}
1202 1203 1204 1205 1206

static inline unsigned int skb_end_offset(const struct sk_buff *skb)
{
	return skb->end - skb->head;
}
1207 1208
#endif

L
Linus Torvalds 已提交
1209
/* Internal */
1210
#define skb_shinfo(SKB)	((struct skb_shared_info *)(skb_end_pointer(SKB)))
L
Linus Torvalds 已提交
1211

1212 1213 1214 1215 1216
static inline struct skb_shared_hwtstamps *skb_hwtstamps(struct sk_buff *skb)
{
	return &skb_shinfo(skb)->hwtstamps;
}

L
Linus Torvalds 已提交
1217 1218 1219 1220 1221 1222 1223 1224
/**
 *	skb_queue_empty - check if a queue is empty
 *	@list: queue head
 *
 *	Returns true if the queue is empty, false otherwise.
 */
static inline int skb_queue_empty(const struct sk_buff_head *list)
{
1225
	return list->next == (const struct sk_buff *) list;
L
Linus Torvalds 已提交
1226 1227
}

D
David S. Miller 已提交
1228 1229 1230 1231 1232 1233 1234 1235 1236 1237
/**
 *	skb_queue_is_last - check if skb is the last entry in the queue
 *	@list: queue head
 *	@skb: buffer
 *
 *	Returns true if @skb is the last buffer on the list.
 */
static inline bool skb_queue_is_last(const struct sk_buff_head *list,
				     const struct sk_buff *skb)
{
1238
	return skb->next == (const struct sk_buff *) list;
D
David S. Miller 已提交
1239 1240
}

1241 1242 1243 1244 1245 1246 1247 1248 1249 1250
/**
 *	skb_queue_is_first - check if skb is the first entry in the queue
 *	@list: queue head
 *	@skb: buffer
 *
 *	Returns true if @skb is the first buffer on the list.
 */
static inline bool skb_queue_is_first(const struct sk_buff_head *list,
				      const struct sk_buff *skb)
{
1251
	return skb->prev == (const struct sk_buff *) list;
1252 1253
}

D
David S. Miller 已提交
1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271
/**
 *	skb_queue_next - return the next packet in the queue
 *	@list: queue head
 *	@skb: current buffer
 *
 *	Return the next packet in @list after @skb.  It is only valid to
 *	call this if skb_queue_is_last() evaluates to false.
 */
static inline struct sk_buff *skb_queue_next(const struct sk_buff_head *list,
					     const struct sk_buff *skb)
{
	/* This BUG_ON may seem severe, but if we just return then we
	 * are going to dereference garbage.
	 */
	BUG_ON(skb_queue_is_last(list, skb));
	return skb->next;
}

1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289
/**
 *	skb_queue_prev - return the prev packet in the queue
 *	@list: queue head
 *	@skb: current buffer
 *
 *	Return the prev packet in @list before @skb.  It is only valid to
 *	call this if skb_queue_is_first() evaluates to false.
 */
static inline struct sk_buff *skb_queue_prev(const struct sk_buff_head *list,
					     const struct sk_buff *skb)
{
	/* This BUG_ON may seem severe, but if we just return then we
	 * are going to dereference garbage.
	 */
	BUG_ON(skb_queue_is_first(list, skb));
	return skb->prev;
}

L
Linus Torvalds 已提交
1290 1291 1292 1293 1294 1295 1296 1297 1298
/**
 *	skb_get - reference buffer
 *	@skb: buffer to reference
 *
 *	Makes another reference to a socket buffer and returns a pointer
 *	to the buffer.
 */
static inline struct sk_buff *skb_get(struct sk_buff *skb)
{
1299
	refcount_inc(&skb->users);
L
Linus Torvalds 已提交
1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321
	return skb;
}

/*
 * If users == 1, we are the only owner and are can avoid redundant
 * atomic change.
 */

/**
 *	skb_cloned - is the buffer a clone
 *	@skb: buffer to check
 *
 *	Returns true if the buffer was generated with skb_clone() and is
 *	one of multiple shared copies of the buffer. Cloned buffers are
 *	shared data so must not be written to under normal circumstances.
 */
static inline int skb_cloned(const struct sk_buff *skb)
{
	return skb->cloned &&
	       (atomic_read(&skb_shinfo(skb)->dataref) & SKB_DATAREF_MASK) != 1;
}

1322 1323
static inline int skb_unclone(struct sk_buff *skb, gfp_t pri)
{
1324
	might_sleep_if(gfpflags_allow_blocking(pri));
1325 1326 1327 1328 1329 1330 1331

	if (skb_cloned(skb))
		return pskb_expand_head(skb, 0, 0, pri);

	return 0;
}

L
Linus Torvalds 已提交
1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350
/**
 *	skb_header_cloned - is the header a clone
 *	@skb: buffer to check
 *
 *	Returns true if modifying the header part of the buffer requires
 *	the data to be copied.
 */
static inline int skb_header_cloned(const struct sk_buff *skb)
{
	int dataref;

	if (!skb->cloned)
		return 0;

	dataref = atomic_read(&skb_shinfo(skb)->dataref);
	dataref = (dataref & SKB_DATAREF_MASK) - (dataref >> SKB_DATAREF_SHIFT);
	return dataref != 1;
}

1351 1352 1353 1354 1355 1356 1357 1358 1359 1360
static inline int skb_header_unclone(struct sk_buff *skb, gfp_t pri)
{
	might_sleep_if(gfpflags_allow_blocking(pri));

	if (skb_header_cloned(skb))
		return pskb_expand_head(skb, 0, 0, pri);

	return 0;
}

L
Linus Torvalds 已提交
1361 1362 1363 1364 1365 1366 1367
/**
 *	skb_header_release - release reference to header
 *	@skb: buffer to operate on
 *
 *	Drop a reference to the header part of the buffer.  This is done
 *	by acquiring a payload reference.  You must not read from the header
 *	part of skb->data after this.
1368
 *	Note : Check if you can use __skb_header_release() instead.
L
Linus Torvalds 已提交
1369 1370 1371 1372 1373 1374 1375 1376
 */
static inline void skb_header_release(struct sk_buff *skb)
{
	BUG_ON(skb->nohdr);
	skb->nohdr = 1;
	atomic_add(1 << SKB_DATAREF_SHIFT, &skb_shinfo(skb)->dataref);
}

1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390
/**
 *	__skb_header_release - release reference to header
 *	@skb: buffer to operate on
 *
 *	Variant of skb_header_release() assuming skb is private to caller.
 *	We can avoid one atomic operation.
 */
static inline void __skb_header_release(struct sk_buff *skb)
{
	skb->nohdr = 1;
	atomic_set(&skb_shinfo(skb)->dataref, 1 + (1 << SKB_DATAREF_SHIFT));
}


L
Linus Torvalds 已提交
1391 1392 1393 1394 1395 1396 1397 1398 1399
/**
 *	skb_shared - is the buffer shared
 *	@skb: buffer to check
 *
 *	Returns true if more than one person has a reference to this
 *	buffer.
 */
static inline int skb_shared(const struct sk_buff *skb)
{
1400
	return refcount_read(&skb->users) != 1;
L
Linus Torvalds 已提交
1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415
}

/**
 *	skb_share_check - check if buffer is shared and if so clone it
 *	@skb: buffer to check
 *	@pri: priority for memory allocation
 *
 *	If the buffer is shared the buffer is cloned and the old copy
 *	drops a reference. A new clone with a single reference is returned.
 *	If the buffer is not shared the original buffer is returned. When
 *	being called from interrupt status or with spinlocks held pri must
 *	be GFP_ATOMIC.
 *
 *	NULL is returned on a memory allocation failure.
 */
1416
static inline struct sk_buff *skb_share_check(struct sk_buff *skb, gfp_t pri)
L
Linus Torvalds 已提交
1417
{
1418
	might_sleep_if(gfpflags_allow_blocking(pri));
L
Linus Torvalds 已提交
1419 1420
	if (skb_shared(skb)) {
		struct sk_buff *nskb = skb_clone(skb, pri);
1421 1422 1423 1424 1425

		if (likely(nskb))
			consume_skb(skb);
		else
			kfree_skb(skb);
L
Linus Torvalds 已提交
1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450
		skb = nskb;
	}
	return skb;
}

/*
 *	Copy shared buffers into a new sk_buff. We effectively do COW on
 *	packets to handle cases where we have a local reader and forward
 *	and a couple of other messy ones. The normal one is tcpdumping
 *	a packet thats being forwarded.
 */

/**
 *	skb_unshare - make a copy of a shared buffer
 *	@skb: buffer to check
 *	@pri: priority for memory allocation
 *
 *	If the socket buffer is a clone then this function creates a new
 *	copy of the data, drops a reference count on the old copy and returns
 *	the new copy with the reference count at 1. If the buffer is not a clone
 *	the original buffer is returned. When called with a spinlock held or
 *	from interrupt state @pri must be %GFP_ATOMIC
 *
 *	%NULL is returned on a memory allocation failure.
 */
1451
static inline struct sk_buff *skb_unshare(struct sk_buff *skb,
A
Al Viro 已提交
1452
					  gfp_t pri)
L
Linus Torvalds 已提交
1453
{
1454
	might_sleep_if(gfpflags_allow_blocking(pri));
L
Linus Torvalds 已提交
1455 1456
	if (skb_cloned(skb)) {
		struct sk_buff *nskb = skb_copy(skb, pri);
1457 1458 1459 1460 1461 1462

		/* Free our shared copy */
		if (likely(nskb))
			consume_skb(skb);
		else
			kfree_skb(skb);
L
Linus Torvalds 已提交
1463 1464 1465 1466 1467 1468
		skb = nskb;
	}
	return skb;
}

/**
1469
 *	skb_peek - peek at the head of an &sk_buff_head
L
Linus Torvalds 已提交
1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480
 *	@list_: list to peek at
 *
 *	Peek an &sk_buff. Unlike most other operations you _MUST_
 *	be careful with this one. A peek leaves the buffer on the
 *	list and someone else may run off with it. You must hold
 *	the appropriate locks or have a private queue to do this.
 *
 *	Returns %NULL for an empty list or a pointer to the head element.
 *	The reference count is not incremented and the reference is therefore
 *	volatile. Use with caution.
 */
1481
static inline struct sk_buff *skb_peek(const struct sk_buff_head *list_)
L
Linus Torvalds 已提交
1482
{
1483 1484 1485 1486 1487
	struct sk_buff *skb = list_->next;

	if (skb == (struct sk_buff *)list_)
		skb = NULL;
	return skb;
L
Linus Torvalds 已提交
1488 1489
}

P
Pavel Emelyanov 已提交
1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502
/**
 *	skb_peek_next - peek skb following the given one from a queue
 *	@skb: skb to start from
 *	@list_: list to peek at
 *
 *	Returns %NULL when the end of the list is met or a pointer to the
 *	next element. The reference count is not incremented and the
 *	reference is therefore volatile. Use with caution.
 */
static inline struct sk_buff *skb_peek_next(struct sk_buff *skb,
		const struct sk_buff_head *list_)
{
	struct sk_buff *next = skb->next;
1503

P
Pavel Emelyanov 已提交
1504 1505 1506 1507 1508
	if (next == (struct sk_buff *)list_)
		next = NULL;
	return next;
}

L
Linus Torvalds 已提交
1509
/**
1510
 *	skb_peek_tail - peek at the tail of an &sk_buff_head
L
Linus Torvalds 已提交
1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521
 *	@list_: list to peek at
 *
 *	Peek an &sk_buff. Unlike most other operations you _MUST_
 *	be careful with this one. A peek leaves the buffer on the
 *	list and someone else may run off with it. You must hold
 *	the appropriate locks or have a private queue to do this.
 *
 *	Returns %NULL for an empty list or a pointer to the tail element.
 *	The reference count is not incremented and the reference is therefore
 *	volatile. Use with caution.
 */
1522
static inline struct sk_buff *skb_peek_tail(const struct sk_buff_head *list_)
L
Linus Torvalds 已提交
1523
{
1524 1525 1526 1527 1528 1529
	struct sk_buff *skb = list_->prev;

	if (skb == (struct sk_buff *)list_)
		skb = NULL;
	return skb;

L
Linus Torvalds 已提交
1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542
}

/**
 *	skb_queue_len	- get queue length
 *	@list_: list to measure
 *
 *	Return the length of an &sk_buff queue.
 */
static inline __u32 skb_queue_len(const struct sk_buff_head *list_)
{
	return list_->qlen;
}

1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558
/**
 *	__skb_queue_head_init - initialize non-spinlock portions of sk_buff_head
 *	@list: queue to initialize
 *
 *	This initializes only the list and queue length aspects of
 *	an sk_buff_head object.  This allows to initialize the list
 *	aspects of an sk_buff_head without reinitializing things like
 *	the spinlock.  It can also be used for on-stack sk_buff_head
 *	objects where the spinlock is known to not be used.
 */
static inline void __skb_queue_head_init(struct sk_buff_head *list)
{
	list->prev = list->next = (struct sk_buff *)list;
	list->qlen = 0;
}

1559 1560 1561 1562 1563 1564 1565 1566
/*
 * This function creates a split out lock class for each invocation;
 * this is needed for now since a whole lot of users of the skb-queue
 * infrastructure in drivers have different locking usage (in hardirq)
 * than the networking core (in softirq only). In the long run either the
 * network layer or drivers should need annotation to consolidate the
 * main types of usage into 3 classes.
 */
L
Linus Torvalds 已提交
1567 1568 1569
static inline void skb_queue_head_init(struct sk_buff_head *list)
{
	spin_lock_init(&list->lock);
1570
	__skb_queue_head_init(list);
L
Linus Torvalds 已提交
1571 1572
}

1573 1574 1575 1576 1577 1578 1579
static inline void skb_queue_head_init_class(struct sk_buff_head *list,
		struct lock_class_key *class)
{
	skb_queue_head_init(list);
	lockdep_set_class(&list->lock, class);
}

L
Linus Torvalds 已提交
1580
/*
1581
 *	Insert an sk_buff on a list.
L
Linus Torvalds 已提交
1582 1583 1584 1585
 *
 *	The "__skb_xxxx()" functions are the non-atomic ones that
 *	can only be called with interrupts disabled.
 */
1586 1587
void skb_insert(struct sk_buff *old, struct sk_buff *newsk,
		struct sk_buff_head *list);
1588 1589 1590 1591 1592 1593 1594 1595 1596
static inline void __skb_insert(struct sk_buff *newsk,
				struct sk_buff *prev, struct sk_buff *next,
				struct sk_buff_head *list)
{
	newsk->next = next;
	newsk->prev = prev;
	next->prev  = prev->next = newsk;
	list->qlen++;
}
L
Linus Torvalds 已提交
1597

1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621
static inline void __skb_queue_splice(const struct sk_buff_head *list,
				      struct sk_buff *prev,
				      struct sk_buff *next)
{
	struct sk_buff *first = list->next;
	struct sk_buff *last = list->prev;

	first->prev = prev;
	prev->next = first;

	last->next = next;
	next->prev = last;
}

/**
 *	skb_queue_splice - join two skb lists, this is designed for stacks
 *	@list: the new list to add
 *	@head: the place to add it in the first list
 */
static inline void skb_queue_splice(const struct sk_buff_head *list,
				    struct sk_buff_head *head)
{
	if (!skb_queue_empty(list)) {
		__skb_queue_splice(list, (struct sk_buff *) head, head->next);
1622
		head->qlen += list->qlen;
1623 1624 1625 1626
	}
}

/**
E
Eric Dumazet 已提交
1627
 *	skb_queue_splice_init - join two skb lists and reinitialise the emptied list
1628 1629 1630 1631 1632 1633 1634 1635 1636 1637
 *	@list: the new list to add
 *	@head: the place to add it in the first list
 *
 *	The list at @list is reinitialised
 */
static inline void skb_queue_splice_init(struct sk_buff_head *list,
					 struct sk_buff_head *head)
{
	if (!skb_queue_empty(list)) {
		__skb_queue_splice(list, (struct sk_buff *) head, head->next);
1638
		head->qlen += list->qlen;
1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652
		__skb_queue_head_init(list);
	}
}

/**
 *	skb_queue_splice_tail - join two skb lists, each list being a queue
 *	@list: the new list to add
 *	@head: the place to add it in the first list
 */
static inline void skb_queue_splice_tail(const struct sk_buff_head *list,
					 struct sk_buff_head *head)
{
	if (!skb_queue_empty(list)) {
		__skb_queue_splice(list, head->prev, (struct sk_buff *) head);
1653
		head->qlen += list->qlen;
1654 1655 1656 1657
	}
}

/**
E
Eric Dumazet 已提交
1658
 *	skb_queue_splice_tail_init - join two skb lists and reinitialise the emptied list
1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669
 *	@list: the new list to add
 *	@head: the place to add it in the first list
 *
 *	Each of the lists is a queue.
 *	The list at @list is reinitialised
 */
static inline void skb_queue_splice_tail_init(struct sk_buff_head *list,
					      struct sk_buff_head *head)
{
	if (!skb_queue_empty(list)) {
		__skb_queue_splice(list, head->prev, (struct sk_buff *) head);
1670
		head->qlen += list->qlen;
1671 1672 1673 1674
		__skb_queue_head_init(list);
	}
}

L
Linus Torvalds 已提交
1675
/**
1676
 *	__skb_queue_after - queue a buffer at the list head
L
Linus Torvalds 已提交
1677
 *	@list: list to use
1678
 *	@prev: place after this buffer
L
Linus Torvalds 已提交
1679 1680
 *	@newsk: buffer to queue
 *
1681
 *	Queue a buffer int the middle of a list. This function takes no locks
L
Linus Torvalds 已提交
1682 1683 1684 1685
 *	and you must therefore hold required locks before calling it.
 *
 *	A buffer cannot be placed on two lists at the same time.
 */
1686 1687 1688
static inline void __skb_queue_after(struct sk_buff_head *list,
				     struct sk_buff *prev,
				     struct sk_buff *newsk)
L
Linus Torvalds 已提交
1689
{
1690
	__skb_insert(newsk, prev, prev->next, list);
L
Linus Torvalds 已提交
1691 1692
}

1693 1694
void skb_append(struct sk_buff *old, struct sk_buff *newsk,
		struct sk_buff_head *list);
1695

1696 1697 1698 1699 1700 1701 1702
static inline void __skb_queue_before(struct sk_buff_head *list,
				      struct sk_buff *next,
				      struct sk_buff *newsk)
{
	__skb_insert(newsk, next->prev, next, list);
}

1703 1704 1705 1706 1707 1708 1709 1710 1711 1712
/**
 *	__skb_queue_head - queue a buffer at the list head
 *	@list: list to use
 *	@newsk: buffer to queue
 *
 *	Queue a buffer at the start of a list. This function takes no locks
 *	and you must therefore hold required locks before calling it.
 *
 *	A buffer cannot be placed on two lists at the same time.
 */
1713
void skb_queue_head(struct sk_buff_head *list, struct sk_buff *newsk);
1714 1715 1716 1717 1718 1719
static inline void __skb_queue_head(struct sk_buff_head *list,
				    struct sk_buff *newsk)
{
	__skb_queue_after(list, (struct sk_buff *)list, newsk);
}

L
Linus Torvalds 已提交
1720 1721 1722 1723 1724 1725 1726 1727 1728 1729
/**
 *	__skb_queue_tail - queue a buffer at the list tail
 *	@list: list to use
 *	@newsk: buffer to queue
 *
 *	Queue a buffer at the end of a list. This function takes no locks
 *	and you must therefore hold required locks before calling it.
 *
 *	A buffer cannot be placed on two lists at the same time.
 */
1730
void skb_queue_tail(struct sk_buff_head *list, struct sk_buff *newsk);
L
Linus Torvalds 已提交
1731 1732 1733
static inline void __skb_queue_tail(struct sk_buff_head *list,
				   struct sk_buff *newsk)
{
1734
	__skb_queue_before(list, (struct sk_buff *)list, newsk);
L
Linus Torvalds 已提交
1735 1736 1737 1738 1739 1740
}

/*
 * remove sk_buff from list. _Must_ be called atomically, and with
 * the list known..
 */
1741
void skb_unlink(struct sk_buff *skb, struct sk_buff_head *list);
L
Linus Torvalds 已提交
1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753
static inline void __skb_unlink(struct sk_buff *skb, struct sk_buff_head *list)
{
	struct sk_buff *next, *prev;

	list->qlen--;
	next	   = skb->next;
	prev	   = skb->prev;
	skb->next  = skb->prev = NULL;
	next->prev = prev;
	prev->next = next;
}

1754 1755 1756 1757 1758 1759 1760 1761
/**
 *	__skb_dequeue - remove from the head of the queue
 *	@list: list to dequeue from
 *
 *	Remove the head of the list. This function does not take any locks
 *	so must be used with appropriate locks held only. The head item is
 *	returned or %NULL if the list is empty.
 */
1762
struct sk_buff *skb_dequeue(struct sk_buff_head *list);
1763 1764 1765 1766 1767 1768 1769
static inline struct sk_buff *__skb_dequeue(struct sk_buff_head *list)
{
	struct sk_buff *skb = skb_peek(list);
	if (skb)
		__skb_unlink(skb, list);
	return skb;
}
L
Linus Torvalds 已提交
1770 1771 1772 1773 1774 1775 1776 1777 1778

/**
 *	__skb_dequeue_tail - remove from the tail of the queue
 *	@list: list to dequeue from
 *
 *	Remove the tail of the list. This function does not take any locks
 *	so must be used with appropriate locks held only. The tail item is
 *	returned or %NULL if the list is empty.
 */
1779
struct sk_buff *skb_dequeue_tail(struct sk_buff_head *list);
L
Linus Torvalds 已提交
1780 1781 1782 1783 1784 1785 1786 1787 1788
static inline struct sk_buff *__skb_dequeue_tail(struct sk_buff_head *list)
{
	struct sk_buff *skb = skb_peek_tail(list);
	if (skb)
		__skb_unlink(skb, list);
	return skb;
}


1789
static inline bool skb_is_nonlinear(const struct sk_buff *skb)
L
Linus Torvalds 已提交
1790 1791 1792 1793 1794 1795 1796 1797 1798
{
	return skb->data_len;
}

static inline unsigned int skb_headlen(const struct sk_buff *skb)
{
	return skb->len - skb->data_len;
}

1799
static inline unsigned int skb_pagelen(const struct sk_buff *skb)
L
Linus Torvalds 已提交
1800
{
1801
	unsigned int i, len = 0;
L
Linus Torvalds 已提交
1802

1803
	for (i = skb_shinfo(skb)->nr_frags - 1; (int)i >= 0; i--)
E
Eric Dumazet 已提交
1804
		len += skb_frag_size(&skb_shinfo(skb)->frags[i]);
L
Linus Torvalds 已提交
1805 1806 1807
	return len + skb_headlen(skb);
}

1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822
/**
 * __skb_fill_page_desc - initialise a paged fragment in an skb
 * @skb: buffer containing fragment to be initialised
 * @i: paged fragment index to initialise
 * @page: the page to use for this fragment
 * @off: the offset to the data with @page
 * @size: the length of the data
 *
 * Initialises the @i'th fragment of @skb to point to &size bytes at
 * offset @off within @page.
 *
 * Does not take any additional reference on the fragment.
 */
static inline void __skb_fill_page_desc(struct sk_buff *skb, int i,
					struct page *page, int off, int size)
L
Linus Torvalds 已提交
1823 1824 1825
{
	skb_frag_t *frag = &skb_shinfo(skb)->frags[i];

1826
	/*
1827 1828 1829
	 * Propagate page pfmemalloc to the skb if we can. The problem is
	 * that not all callers have unique ownership of the page but rely
	 * on page_is_pfmemalloc doing the right thing(tm).
1830
	 */
1831
	frag->page.p		  = page;
L
Linus Torvalds 已提交
1832
	frag->page_offset	  = off;
E
Eric Dumazet 已提交
1833
	skb_frag_size_set(frag, size);
1834 1835

	page = compound_head(page);
1836
	if (page_is_pfmemalloc(page))
1837
		skb->pfmemalloc	= true;
1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848
}

/**
 * skb_fill_page_desc - initialise a paged fragment in an skb
 * @skb: buffer containing fragment to be initialised
 * @i: paged fragment index to initialise
 * @page: the page to use for this fragment
 * @off: the offset to the data with @page
 * @size: the length of the data
 *
 * As per __skb_fill_page_desc() -- initialises the @i'th fragment of
M
Mathias Krause 已提交
1849
 * @skb to point to @size bytes at offset @off within @page. In
1850 1851 1852 1853 1854 1855 1856 1857
 * addition updates @skb such that @i is the last fragment.
 *
 * Does not take any additional reference on the fragment.
 */
static inline void skb_fill_page_desc(struct sk_buff *skb, int i,
				      struct page *page, int off, int size)
{
	__skb_fill_page_desc(skb, i, page, off, size);
L
Linus Torvalds 已提交
1858 1859 1860
	skb_shinfo(skb)->nr_frags = i + 1;
}

1861 1862
void skb_add_rx_frag(struct sk_buff *skb, int i, struct page *page, int off,
		     int size, unsigned int truesize);
P
Peter Zijlstra 已提交
1863

J
Jason Wang 已提交
1864 1865 1866
void skb_coalesce_rx_frag(struct sk_buff *skb, int i, int size,
			  unsigned int truesize);

L
Linus Torvalds 已提交
1867
#define SKB_PAGE_ASSERT(skb) 	BUG_ON(skb_shinfo(skb)->nr_frags)
1868
#define SKB_FRAG_ASSERT(skb) 	BUG_ON(skb_has_frag_list(skb))
L
Linus Torvalds 已提交
1869 1870
#define SKB_LINEAR_ASSERT(skb)  BUG_ON(skb_is_nonlinear(skb))

1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886
#ifdef NET_SKBUFF_DATA_USES_OFFSET
static inline unsigned char *skb_tail_pointer(const struct sk_buff *skb)
{
	return skb->head + skb->tail;
}

static inline void skb_reset_tail_pointer(struct sk_buff *skb)
{
	skb->tail = skb->data - skb->head;
}

static inline void skb_set_tail_pointer(struct sk_buff *skb, const int offset)
{
	skb_reset_tail_pointer(skb);
	skb->tail += offset;
}
1887

1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902
#else /* NET_SKBUFF_DATA_USES_OFFSET */
static inline unsigned char *skb_tail_pointer(const struct sk_buff *skb)
{
	return skb->tail;
}

static inline void skb_reset_tail_pointer(struct sk_buff *skb)
{
	skb->tail = skb->data;
}

static inline void skb_set_tail_pointer(struct sk_buff *skb, const int offset)
{
	skb->tail = skb->data + offset;
}
1903

1904 1905
#endif /* NET_SKBUFF_DATA_USES_OFFSET */

L
Linus Torvalds 已提交
1906 1907 1908
/*
 *	Add data to an sk_buff
 */
1909 1910 1911
void *pskb_put(struct sk_buff *skb, struct sk_buff *tail, int len);
void *skb_put(struct sk_buff *skb, unsigned int len);
static inline void *__skb_put(struct sk_buff *skb, unsigned int len)
L
Linus Torvalds 已提交
1912
{
1913
	void *tmp = skb_tail_pointer(skb);
L
Linus Torvalds 已提交
1914 1915 1916 1917 1918 1919
	SKB_LINEAR_ASSERT(skb);
	skb->tail += len;
	skb->len  += len;
	return tmp;
}

1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941
static inline void *__skb_put_zero(struct sk_buff *skb, unsigned int len)
{
	void *tmp = __skb_put(skb, len);

	memset(tmp, 0, len);
	return tmp;
}

static inline void *__skb_put_data(struct sk_buff *skb, const void *data,
				   unsigned int len)
{
	void *tmp = __skb_put(skb, len);

	memcpy(tmp, data, len);
	return tmp;
}

static inline void __skb_put_u8(struct sk_buff *skb, u8 val)
{
	*(u8 *)__skb_put(skb, 1) = val;
}

1942
static inline void *skb_put_zero(struct sk_buff *skb, unsigned int len)
1943
{
1944
	void *tmp = skb_put(skb, len);
1945 1946 1947 1948 1949 1950

	memset(tmp, 0, len);

	return tmp;
}

1951 1952 1953 1954 1955 1956 1957 1958 1959 1960
static inline void *skb_put_data(struct sk_buff *skb, const void *data,
				 unsigned int len)
{
	void *tmp = skb_put(skb, len);

	memcpy(tmp, data, len);

	return tmp;
}

1961 1962 1963 1964 1965
static inline void skb_put_u8(struct sk_buff *skb, u8 val)
{
	*(u8 *)skb_put(skb, 1) = val;
}

1966 1967
void *skb_push(struct sk_buff *skb, unsigned int len);
static inline void *__skb_push(struct sk_buff *skb, unsigned int len)
L
Linus Torvalds 已提交
1968 1969 1970 1971 1972 1973
{
	skb->data -= len;
	skb->len  += len;
	return skb->data;
}

1974 1975
void *skb_pull(struct sk_buff *skb, unsigned int len);
static inline void *__skb_pull(struct sk_buff *skb, unsigned int len)
L
Linus Torvalds 已提交
1976 1977 1978 1979 1980 1981
{
	skb->len -= len;
	BUG_ON(skb->len < skb->data_len);
	return skb->data += len;
}

1982
static inline void *skb_pull_inline(struct sk_buff *skb, unsigned int len)
1983 1984 1985 1986
{
	return unlikely(len > skb->len) ? NULL : __skb_pull(skb, len);
}

1987
void *__pskb_pull_tail(struct sk_buff *skb, int delta);
L
Linus Torvalds 已提交
1988

1989
static inline void *__pskb_pull(struct sk_buff *skb, unsigned int len)
L
Linus Torvalds 已提交
1990 1991
{
	if (len > skb_headlen(skb) &&
G
Gerrit Renker 已提交
1992
	    !__pskb_pull_tail(skb, len - skb_headlen(skb)))
L
Linus Torvalds 已提交
1993 1994 1995 1996 1997
		return NULL;
	skb->len -= len;
	return skb->data += len;
}

1998
static inline void *pskb_pull(struct sk_buff *skb, unsigned int len)
L
Linus Torvalds 已提交
1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
{
	return unlikely(len > skb->len) ? NULL : __pskb_pull(skb, len);
}

static inline int pskb_may_pull(struct sk_buff *skb, unsigned int len)
{
	if (likely(len <= skb_headlen(skb)))
		return 1;
	if (unlikely(len > skb->len))
		return 0;
G
Gerrit Renker 已提交
2009
	return __pskb_pull_tail(skb, len - skb_headlen(skb)) != NULL;
L
Linus Torvalds 已提交
2010 2011
}

2012 2013
void skb_condense(struct sk_buff *skb);

L
Linus Torvalds 已提交
2014 2015 2016 2017 2018 2019
/**
 *	skb_headroom - bytes at buffer head
 *	@skb: buffer to check
 *
 *	Return the number of bytes of free space at the head of an &sk_buff.
 */
2020
static inline unsigned int skb_headroom(const struct sk_buff *skb)
L
Linus Torvalds 已提交
2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032
{
	return skb->data - skb->head;
}

/**
 *	skb_tailroom - bytes at buffer end
 *	@skb: buffer to check
 *
 *	Return the number of bytes of free space at the tail of an sk_buff
 */
static inline int skb_tailroom(const struct sk_buff *skb)
{
2033
	return skb_is_nonlinear(skb) ? 0 : skb->end - skb->tail;
L
Linus Torvalds 已提交
2034 2035
}

2036 2037 2038 2039 2040 2041 2042 2043 2044
/**
 *	skb_availroom - bytes at buffer end
 *	@skb: buffer to check
 *
 *	Return the number of bytes of free space at the tail of an sk_buff
 *	allocated by sk_stream_alloc()
 */
static inline int skb_availroom(const struct sk_buff *skb)
{
E
Eric Dumazet 已提交
2045 2046 2047 2048
	if (skb_is_nonlinear(skb))
		return 0;

	return skb->end - skb->tail - skb->reserved_tailroom;
2049 2050
}

L
Linus Torvalds 已提交
2051 2052 2053 2054 2055 2056 2057 2058
/**
 *	skb_reserve - adjust headroom
 *	@skb: buffer to alter
 *	@len: bytes to move
 *
 *	Increase the headroom of an empty &sk_buff by reducing the tail
 *	room. This is only allowed for an empty buffer.
 */
2059
static inline void skb_reserve(struct sk_buff *skb, int len)
L
Linus Torvalds 已提交
2060 2061 2062 2063 2064
{
	skb->data += len;
	skb->tail += len;
}

2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088
/**
 *	skb_tailroom_reserve - adjust reserved_tailroom
 *	@skb: buffer to alter
 *	@mtu: maximum amount of headlen permitted
 *	@needed_tailroom: minimum amount of reserved_tailroom
 *
 *	Set reserved_tailroom so that headlen can be as large as possible but
 *	not larger than mtu and tailroom cannot be smaller than
 *	needed_tailroom.
 *	The required headroom should already have been reserved before using
 *	this function.
 */
static inline void skb_tailroom_reserve(struct sk_buff *skb, unsigned int mtu,
					unsigned int needed_tailroom)
{
	SKB_LINEAR_ASSERT(skb);
	if (mtu < skb_tailroom(skb) - needed_tailroom)
		/* use at most mtu */
		skb->reserved_tailroom = skb_tailroom(skb) - mtu;
	else
		/* use up to all available space */
		skb->reserved_tailroom = needed_tailroom;
}

T
Tom Herbert 已提交
2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105
#define ENCAP_TYPE_ETHER	0
#define ENCAP_TYPE_IPPROTO	1

static inline void skb_set_inner_protocol(struct sk_buff *skb,
					  __be16 protocol)
{
	skb->inner_protocol = protocol;
	skb->inner_protocol_type = ENCAP_TYPE_ETHER;
}

static inline void skb_set_inner_ipproto(struct sk_buff *skb,
					 __u8 ipproto)
{
	skb->inner_ipproto = ipproto;
	skb->inner_protocol_type = ENCAP_TYPE_IPPROTO;
}

2106 2107
static inline void skb_reset_inner_headers(struct sk_buff *skb)
{
2108
	skb->inner_mac_header = skb->mac_header;
2109 2110 2111 2112
	skb->inner_network_header = skb->network_header;
	skb->inner_transport_header = skb->transport_header;
}

2113 2114 2115 2116 2117
static inline void skb_reset_mac_len(struct sk_buff *skb)
{
	skb->mac_len = skb->network_header - skb->mac_header;
}

2118 2119 2120 2121 2122 2123
static inline unsigned char *skb_inner_transport_header(const struct sk_buff
							*skb)
{
	return skb->head + skb->inner_transport_header;
}

2124 2125 2126 2127 2128
static inline int skb_inner_transport_offset(const struct sk_buff *skb)
{
	return skb_inner_transport_header(skb) - skb->data;
}

2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157
static inline void skb_reset_inner_transport_header(struct sk_buff *skb)
{
	skb->inner_transport_header = skb->data - skb->head;
}

static inline void skb_set_inner_transport_header(struct sk_buff *skb,
						   const int offset)
{
	skb_reset_inner_transport_header(skb);
	skb->inner_transport_header += offset;
}

static inline unsigned char *skb_inner_network_header(const struct sk_buff *skb)
{
	return skb->head + skb->inner_network_header;
}

static inline void skb_reset_inner_network_header(struct sk_buff *skb)
{
	skb->inner_network_header = skb->data - skb->head;
}

static inline void skb_set_inner_network_header(struct sk_buff *skb,
						const int offset)
{
	skb_reset_inner_network_header(skb);
	skb->inner_network_header += offset;
}

2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173
static inline unsigned char *skb_inner_mac_header(const struct sk_buff *skb)
{
	return skb->head + skb->inner_mac_header;
}

static inline void skb_reset_inner_mac_header(struct sk_buff *skb)
{
	skb->inner_mac_header = skb->data - skb->head;
}

static inline void skb_set_inner_mac_header(struct sk_buff *skb,
					    const int offset)
{
	skb_reset_inner_mac_header(skb);
	skb->inner_mac_header += offset;
}
2174 2175
static inline bool skb_transport_header_was_set(const struct sk_buff *skb)
{
C
Cong Wang 已提交
2176
	return skb->transport_header != (typeof(skb->transport_header))~0U;
2177 2178
}

2179 2180
static inline unsigned char *skb_transport_header(const struct sk_buff *skb)
{
2181
	return skb->head + skb->transport_header;
2182 2183
}

2184 2185
static inline void skb_reset_transport_header(struct sk_buff *skb)
{
2186
	skb->transport_header = skb->data - skb->head;
2187 2188
}

2189 2190 2191
static inline void skb_set_transport_header(struct sk_buff *skb,
					    const int offset)
{
2192 2193
	skb_reset_transport_header(skb);
	skb->transport_header += offset;
2194 2195
}

2196 2197
static inline unsigned char *skb_network_header(const struct sk_buff *skb)
{
2198
	return skb->head + skb->network_header;
2199 2200
}

2201 2202
static inline void skb_reset_network_header(struct sk_buff *skb)
{
2203
	skb->network_header = skb->data - skb->head;
2204 2205
}

2206 2207
static inline void skb_set_network_header(struct sk_buff *skb, const int offset)
{
2208 2209
	skb_reset_network_header(skb);
	skb->network_header += offset;
2210 2211
}

2212
static inline unsigned char *skb_mac_header(const struct sk_buff *skb)
2213
{
2214
	return skb->head + skb->mac_header;
2215 2216
}

2217 2218 2219 2220 2221
static inline int skb_mac_offset(const struct sk_buff *skb)
{
	return skb_mac_header(skb) - skb->data;
}

2222 2223 2224 2225 2226
static inline u32 skb_mac_header_len(const struct sk_buff *skb)
{
	return skb->network_header - skb->mac_header;
}

2227
static inline int skb_mac_header_was_set(const struct sk_buff *skb)
2228
{
C
Cong Wang 已提交
2229
	return skb->mac_header != (typeof(skb->mac_header))~0U;
2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242
}

static inline void skb_reset_mac_header(struct sk_buff *skb)
{
	skb->mac_header = skb->data - skb->head;
}

static inline void skb_set_mac_header(struct sk_buff *skb, const int offset)
{
	skb_reset_mac_header(skb);
	skb->mac_header += offset;
}

2243 2244 2245 2246 2247
static inline void skb_pop_mac_header(struct sk_buff *skb)
{
	skb->mac_header = skb->network_header;
}

2248 2249 2250 2251 2252 2253 2254
static inline void skb_probe_transport_header(struct sk_buff *skb,
					      const int offset_hint)
{
	struct flow_keys keys;

	if (skb_transport_header_was_set(skb))
		return;
2255
	else if (skb_flow_dissect_flow_keys(skb, &keys, 0))
2256
		skb_set_transport_header(skb, keys.control.thoff);
2257 2258 2259 2260
	else
		skb_set_transport_header(skb, offset_hint);
}

2261 2262 2263 2264 2265 2266 2267 2268 2269 2270
static inline void skb_mac_header_rebuild(struct sk_buff *skb)
{
	if (skb_mac_header_was_set(skb)) {
		const unsigned char *old_mac = skb_mac_header(skb);

		skb_set_mac_header(skb, -skb->mac_len);
		memmove(skb_mac_header(skb), old_mac, skb->mac_len);
	}
}

2271 2272 2273 2274 2275
static inline int skb_checksum_start_offset(const struct sk_buff *skb)
{
	return skb->csum_start - skb_headroom(skb);
}

2276 2277 2278 2279 2280
static inline unsigned char *skb_checksum_start(const struct sk_buff *skb)
{
	return skb->head + skb->csum_start;
}

2281 2282 2283 2284 2285 2286 2287 2288 2289 2290
static inline int skb_transport_offset(const struct sk_buff *skb)
{
	return skb_transport_header(skb) - skb->data;
}

static inline u32 skb_network_header_len(const struct sk_buff *skb)
{
	return skb->transport_header - skb->network_header;
}

2291 2292 2293 2294 2295
static inline u32 skb_inner_network_header_len(const struct sk_buff *skb)
{
	return skb->inner_transport_header - skb->inner_network_header;
}

2296 2297 2298 2299
static inline int skb_network_offset(const struct sk_buff *skb)
{
	return skb_network_header(skb) - skb->data;
}
2300

2301 2302 2303 2304 2305
static inline int skb_inner_network_offset(const struct sk_buff *skb)
{
	return skb_inner_network_header(skb) - skb->data;
}

2306 2307 2308 2309 2310
static inline int pskb_network_may_pull(struct sk_buff *skb, unsigned int len)
{
	return pskb_may_pull(skb, skb_network_offset(skb) + len);
}

L
Linus Torvalds 已提交
2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321
/*
 * CPUs often take a performance hit when accessing unaligned memory
 * locations. The actual performance hit varies, it can be small if the
 * hardware handles it or large if we have to take an exception and fix it
 * in software.
 *
 * Since an ethernet header is 14 bytes network drivers often end up with
 * the IP header at an unaligned offset. The IP header can be aligned by
 * shifting the start of the packet by 2 bytes. Drivers should do this
 * with:
 *
2322
 * skb_reserve(skb, NET_IP_ALIGN);
L
Linus Torvalds 已提交
2323 2324 2325 2326
 *
 * The downside to this alignment of the IP header is that the DMA is now
 * unaligned. On some architectures the cost of an unaligned DMA is high
 * and this cost outweighs the gains made by aligning the IP header.
2327
 *
L
Linus Torvalds 已提交
2328 2329 2330 2331 2332 2333 2334
 * Since this trade off varies between architectures, we allow NET_IP_ALIGN
 * to be overridden.
 */
#ifndef NET_IP_ALIGN
#define NET_IP_ALIGN	2
#endif

2335 2336 2337 2338
/*
 * The networking layer reserves some headroom in skb data (via
 * dev_alloc_skb). This is used to avoid having to reallocate skb data when
 * the header has to grow. In the default case, if the header has to grow
2339
 * 32 bytes or less we avoid the reallocation.
2340 2341 2342 2343 2344 2345 2346
 *
 * Unfortunately this headroom changes the DMA alignment of the resulting
 * network packet. As for NET_IP_ALIGN, this unaligned DMA is expensive
 * on some architectures. An architecture can override this value,
 * perhaps setting it to a cacheline in size (since that will maintain
 * cacheline alignment of the DMA). It must be a power of 2.
 *
2347
 * Various parts of the networking layer expect at least 32 bytes of
2348
 * headroom, you should not reduce this.
2349 2350 2351 2352
 *
 * Using max(32, L1_CACHE_BYTES) makes sense (especially with RPS)
 * to reduce average number of cache lines per packet.
 * get_rps_cpus() for example only access one 64 bytes aligned block :
2353
 * NET_IP_ALIGN(2) + ethernet_header(14) + IP_header(20/40) + ports(8)
2354 2355
 */
#ifndef NET_SKB_PAD
2356
#define NET_SKB_PAD	max(32, L1_CACHE_BYTES)
2357 2358
#endif

2359
int ___pskb_trim(struct sk_buff *skb, unsigned int len);
L
Linus Torvalds 已提交
2360

2361
static inline void __skb_set_length(struct sk_buff *skb, unsigned int len)
L
Linus Torvalds 已提交
2362
{
2363
	if (unlikely(skb_is_nonlinear(skb))) {
2364 2365 2366
		WARN_ON(1);
		return;
	}
2367 2368
	skb->len = len;
	skb_set_tail_pointer(skb, len);
L
Linus Torvalds 已提交
2369 2370
}

2371 2372 2373 2374 2375
static inline void __skb_trim(struct sk_buff *skb, unsigned int len)
{
	__skb_set_length(skb, len);
}

2376
void skb_trim(struct sk_buff *skb, unsigned int len);
L
Linus Torvalds 已提交
2377 2378 2379

static inline int __pskb_trim(struct sk_buff *skb, unsigned int len)
{
2380 2381 2382 2383
	if (skb->data_len)
		return ___pskb_trim(skb, len);
	__skb_trim(skb, len);
	return 0;
L
Linus Torvalds 已提交
2384 2385 2386 2387 2388 2389 2390
}

static inline int pskb_trim(struct sk_buff *skb, unsigned int len)
{
	return (len < skb->len) ? __pskb_trim(skb, len) : 0;
}

2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405
/**
 *	pskb_trim_unique - remove end from a paged unique (not cloned) buffer
 *	@skb: buffer to alter
 *	@len: new length
 *
 *	This is identical to pskb_trim except that the caller knows that
 *	the skb is not cloned so we should never get an error due to out-
 *	of-memory.
 */
static inline void pskb_trim_unique(struct sk_buff *skb, unsigned int len)
{
	int err = pskb_trim(skb, len);
	BUG_ON(err);
}

2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419
static inline int __skb_grow(struct sk_buff *skb, unsigned int len)
{
	unsigned int diff = len - skb->len;

	if (skb_tailroom(skb) < diff) {
		int ret = pskb_expand_head(skb, 0, diff - skb_tailroom(skb),
					   GFP_ATOMIC);
		if (ret)
			return ret;
	}
	__skb_set_length(skb, len);
	return 0;
}

L
Linus Torvalds 已提交
2420 2421 2422 2423 2424 2425 2426 2427 2428 2429
/**
 *	skb_orphan - orphan a buffer
 *	@skb: buffer to orphan
 *
 *	If a buffer currently has an owner then we call the owner's
 *	destructor function and make the @skb unowned. The buffer continues
 *	to exist but is no longer charged to its former owner.
 */
static inline void skb_orphan(struct sk_buff *skb)
{
E
Eric Dumazet 已提交
2430
	if (skb->destructor) {
L
Linus Torvalds 已提交
2431
		skb->destructor(skb);
E
Eric Dumazet 已提交
2432 2433
		skb->destructor = NULL;
		skb->sk		= NULL;
2434 2435
	} else {
		BUG_ON(skb->sk);
E
Eric Dumazet 已提交
2436
	}
L
Linus Torvalds 已提交
2437 2438
}

2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454
/**
 *	skb_orphan_frags - orphan the frags contained in a buffer
 *	@skb: buffer to orphan frags from
 *	@gfp_mask: allocation mask for replacement pages
 *
 *	For each frag in the SKB which needs a destructor (i.e. has an
 *	owner) create a copy of that frag and release the original
 *	page by calling the destructor.
 */
static inline int skb_orphan_frags(struct sk_buff *skb, gfp_t gfp_mask)
{
	if (likely(!(skb_shinfo(skb)->tx_flags & SKBTX_DEV_ZEROCOPY)))
		return 0;
	return skb_copy_ubufs(skb, gfp_mask);
}

L
Linus Torvalds 已提交
2455 2456 2457 2458 2459 2460 2461 2462
/**
 *	__skb_queue_purge - empty a list
 *	@list: list to empty
 *
 *	Delete all buffers on an &sk_buff list. Each buffer is removed from
 *	the list and one reference dropped. This function does not take the
 *	list lock and the caller must hold the relevant locks to use it.
 */
2463
void skb_queue_purge(struct sk_buff_head *list);
L
Linus Torvalds 已提交
2464 2465 2466 2467 2468 2469 2470
static inline void __skb_queue_purge(struct sk_buff_head *list)
{
	struct sk_buff *skb;
	while ((skb = __skb_dequeue(list)) != NULL)
		kfree_skb(skb);
}

2471 2472
void skb_rbtree_purge(struct rb_root *root);

2473
void *netdev_alloc_frag(unsigned int fragsz);
L
Linus Torvalds 已提交
2474

2475 2476
struct sk_buff *__netdev_alloc_skb(struct net_device *dev, unsigned int length,
				   gfp_t gfp_mask);
2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491

/**
 *	netdev_alloc_skb - allocate an skbuff for rx on a specific device
 *	@dev: network device to receive on
 *	@length: length to allocate
 *
 *	Allocate a new &sk_buff and assign it a usage count of one. The
 *	buffer has unspecified headroom built in. Users should allocate
 *	the headroom they think they need without accounting for the
 *	built in space. The built in space is used for optimisations.
 *
 *	%NULL is returned if there is no free memory. Although this function
 *	allocates memory it can be called from an interrupt.
 */
static inline struct sk_buff *netdev_alloc_skb(struct net_device *dev,
2492
					       unsigned int length)
2493 2494 2495 2496
{
	return __netdev_alloc_skb(dev, length, GFP_ATOMIC);
}

2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510
/* legacy helper around __netdev_alloc_skb() */
static inline struct sk_buff *__dev_alloc_skb(unsigned int length,
					      gfp_t gfp_mask)
{
	return __netdev_alloc_skb(NULL, length, gfp_mask);
}

/* legacy helper around netdev_alloc_skb() */
static inline struct sk_buff *dev_alloc_skb(unsigned int length)
{
	return netdev_alloc_skb(NULL, length);
}


2511 2512
static inline struct sk_buff *__netdev_alloc_skb_ip_align(struct net_device *dev,
		unsigned int length, gfp_t gfp)
2513
{
2514
	struct sk_buff *skb = __netdev_alloc_skb(dev, length + NET_IP_ALIGN, gfp);
2515 2516 2517 2518 2519 2520

	if (NET_IP_ALIGN && skb)
		skb_reserve(skb, NET_IP_ALIGN);
	return skb;
}

2521 2522 2523 2524 2525 2526
static inline struct sk_buff *netdev_alloc_skb_ip_align(struct net_device *dev,
		unsigned int length)
{
	return __netdev_alloc_skb_ip_align(dev, length, GFP_ATOMIC);
}

2527 2528
static inline void skb_free_frag(void *addr)
{
2529
	page_frag_free(addr);
2530 2531
}

2532
void *napi_alloc_frag(unsigned int fragsz);
2533 2534 2535 2536 2537 2538 2539
struct sk_buff *__napi_alloc_skb(struct napi_struct *napi,
				 unsigned int length, gfp_t gfp_mask);
static inline struct sk_buff *napi_alloc_skb(struct napi_struct *napi,
					     unsigned int length)
{
	return __napi_alloc_skb(napi, length, GFP_ATOMIC);
}
2540 2541 2542
void napi_consume_skb(struct sk_buff *skb, int budget);

void __kfree_skb_flush(void);
2543
void __kfree_skb_defer(struct sk_buff *skb);
2544

2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571
/**
 * __dev_alloc_pages - allocate page for network Rx
 * @gfp_mask: allocation priority. Set __GFP_NOMEMALLOC if not for network Rx
 * @order: size of the allocation
 *
 * Allocate a new page.
 *
 * %NULL is returned if there is no free memory.
*/
static inline struct page *__dev_alloc_pages(gfp_t gfp_mask,
					     unsigned int order)
{
	/* This piece of code contains several assumptions.
	 * 1.  This is for device Rx, therefor a cold page is preferred.
	 * 2.  The expectation is the user wants a compound page.
	 * 3.  If requesting a order 0 page it will not be compound
	 *     due to the check to see if order has a value in prep_new_page
	 * 4.  __GFP_MEMALLOC is ignored if __GFP_NOMEMALLOC is set due to
	 *     code in gfp_to_alloc_flags that should be enforcing this.
	 */
	gfp_mask |= __GFP_COLD | __GFP_COMP | __GFP_MEMALLOC;

	return alloc_pages_node(NUMA_NO_NODE, gfp_mask, order);
}

static inline struct page *dev_alloc_pages(unsigned int order)
{
2572
	return __dev_alloc_pages(GFP_ATOMIC | __GFP_NOWARN, order);
2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589
}

/**
 * __dev_alloc_page - allocate a page for network Rx
 * @gfp_mask: allocation priority. Set __GFP_NOMEMALLOC if not for network Rx
 *
 * Allocate a new page.
 *
 * %NULL is returned if there is no free memory.
 */
static inline struct page *__dev_alloc_page(gfp_t gfp_mask)
{
	return __dev_alloc_pages(gfp_mask, 0);
}

static inline struct page *dev_alloc_page(void)
{
2590
	return dev_alloc_pages(0);
2591 2592
}

2593 2594 2595 2596 2597 2598 2599 2600
/**
 *	skb_propagate_pfmemalloc - Propagate pfmemalloc if skb is allocated after RX page
 *	@page: The page that was allocated from skb_alloc_page
 *	@skb: The skb that may need pfmemalloc set
 */
static inline void skb_propagate_pfmemalloc(struct page *page,
					     struct sk_buff *skb)
{
2601
	if (page_is_pfmemalloc(page))
2602 2603 2604
		skb->pfmemalloc = true;
}

2605
/**
2606
 * skb_frag_page - retrieve the page referred to by a paged fragment
2607 2608 2609 2610 2611 2612
 * @frag: the paged fragment
 *
 * Returns the &struct page associated with @frag.
 */
static inline struct page *skb_frag_page(const skb_frag_t *frag)
{
2613
	return frag->page.p;
2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698
}

/**
 * __skb_frag_ref - take an addition reference on a paged fragment.
 * @frag: the paged fragment
 *
 * Takes an additional reference on the paged fragment @frag.
 */
static inline void __skb_frag_ref(skb_frag_t *frag)
{
	get_page(skb_frag_page(frag));
}

/**
 * skb_frag_ref - take an addition reference on a paged fragment of an skb.
 * @skb: the buffer
 * @f: the fragment offset.
 *
 * Takes an additional reference on the @f'th paged fragment of @skb.
 */
static inline void skb_frag_ref(struct sk_buff *skb, int f)
{
	__skb_frag_ref(&skb_shinfo(skb)->frags[f]);
}

/**
 * __skb_frag_unref - release a reference on a paged fragment.
 * @frag: the paged fragment
 *
 * Releases a reference on the paged fragment @frag.
 */
static inline void __skb_frag_unref(skb_frag_t *frag)
{
	put_page(skb_frag_page(frag));
}

/**
 * skb_frag_unref - release a reference on a paged fragment of an skb.
 * @skb: the buffer
 * @f: the fragment offset
 *
 * Releases a reference on the @f'th paged fragment of @skb.
 */
static inline void skb_frag_unref(struct sk_buff *skb, int f)
{
	__skb_frag_unref(&skb_shinfo(skb)->frags[f]);
}

/**
 * skb_frag_address - gets the address of the data contained in a paged fragment
 * @frag: the paged fragment buffer
 *
 * Returns the address of the data within @frag. The page must already
 * be mapped.
 */
static inline void *skb_frag_address(const skb_frag_t *frag)
{
	return page_address(skb_frag_page(frag)) + frag->page_offset;
}

/**
 * skb_frag_address_safe - gets the address of the data contained in a paged fragment
 * @frag: the paged fragment buffer
 *
 * Returns the address of the data within @frag. Checks that the page
 * is mapped and returns %NULL otherwise.
 */
static inline void *skb_frag_address_safe(const skb_frag_t *frag)
{
	void *ptr = page_address(skb_frag_page(frag));
	if (unlikely(!ptr))
		return NULL;

	return ptr + frag->page_offset;
}

/**
 * __skb_frag_set_page - sets the page contained in a paged fragment
 * @frag: the paged fragment
 * @page: the page to set
 *
 * Sets the fragment @frag to contain @page.
 */
static inline void __skb_frag_set_page(skb_frag_t *frag, struct page *page)
{
2699
	frag->page.p = page;
2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715
}

/**
 * skb_frag_set_page - sets the page contained in a paged fragment of an skb
 * @skb: the buffer
 * @f: the fragment offset
 * @page: the page to set
 *
 * Sets the @f'th fragment of @skb to contain @page.
 */
static inline void skb_frag_set_page(struct sk_buff *skb, int f,
				     struct page *page)
{
	__skb_frag_set_page(&skb_shinfo(skb)->frags[f], page);
}

E
Eric Dumazet 已提交
2716 2717
bool skb_page_frag_refill(unsigned int sz, struct page_frag *pfrag, gfp_t prio);

2718 2719
/**
 * skb_frag_dma_map - maps a paged fragment via the DMA API
2720
 * @dev: the device to map the fragment to
2721 2722 2723 2724
 * @frag: the paged fragment to map
 * @offset: the offset within the fragment (starting at the
 *          fragment's own offset)
 * @size: the number of bytes to map
2725
 * @dir: the direction of the mapping (``PCI_DMA_*``)
2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737
 *
 * Maps the page associated with @frag to @device.
 */
static inline dma_addr_t skb_frag_dma_map(struct device *dev,
					  const skb_frag_t *frag,
					  size_t offset, size_t size,
					  enum dma_data_direction dir)
{
	return dma_map_page(dev, skb_frag_page(frag),
			    frag->page_offset + offset, size, dir);
}

E
Eric Dumazet 已提交
2738 2739 2740 2741 2742 2743
static inline struct sk_buff *pskb_copy(struct sk_buff *skb,
					gfp_t gfp_mask)
{
	return __pskb_copy(skb, skb_headroom(skb), gfp_mask);
}

2744 2745 2746 2747 2748 2749 2750 2751

static inline struct sk_buff *pskb_copy_for_clone(struct sk_buff *skb,
						  gfp_t gfp_mask)
{
	return __pskb_copy_fclone(skb, skb_headroom(skb), gfp_mask, true);
}


2752 2753 2754 2755 2756 2757 2758 2759
/**
 *	skb_clone_writable - is the header of a clone writable
 *	@skb: buffer to check
 *	@len: length up to which to write
 *
 *	Returns true if modifying the header part of the cloned buffer
 *	does not requires the data to be copied.
 */
2760
static inline int skb_clone_writable(const struct sk_buff *skb, unsigned int len)
2761 2762 2763 2764 2765
{
	return !skb_header_cloned(skb) &&
	       skb_headroom(skb) + len <= skb->hdr_len;
}

2766 2767 2768 2769 2770 2771 2772
static inline int skb_try_make_writable(struct sk_buff *skb,
					unsigned int write_len)
{
	return skb_cloned(skb) && !skb_clone_writable(skb, write_len) &&
	       pskb_expand_head(skb, 0, 0, GFP_ATOMIC);
}

H
Herbert Xu 已提交
2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786
static inline int __skb_cow(struct sk_buff *skb, unsigned int headroom,
			    int cloned)
{
	int delta = 0;

	if (headroom > skb_headroom(skb))
		delta = headroom - skb_headroom(skb);

	if (delta || cloned)
		return pskb_expand_head(skb, ALIGN(delta, NET_SKB_PAD), 0,
					GFP_ATOMIC);
	return 0;
}

L
Linus Torvalds 已提交
2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800
/**
 *	skb_cow - copy header of skb when it is required
 *	@skb: buffer to cow
 *	@headroom: needed headroom
 *
 *	If the skb passed lacks sufficient headroom or its data part
 *	is shared, data is reallocated. If reallocation fails, an error
 *	is returned and original skb is not changed.
 *
 *	The result is skb with writable area skb->head...skb->tail
 *	and at least @headroom of space at head.
 */
static inline int skb_cow(struct sk_buff *skb, unsigned int headroom)
{
H
Herbert Xu 已提交
2801 2802
	return __skb_cow(skb, headroom, skb_cloned(skb));
}
L
Linus Torvalds 已提交
2803

H
Herbert Xu 已提交
2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816
/**
 *	skb_cow_head - skb_cow but only making the head writable
 *	@skb: buffer to cow
 *	@headroom: needed headroom
 *
 *	This function is identical to skb_cow except that we replace the
 *	skb_cloned check by skb_header_cloned.  It should be used when
 *	you only need to push on some header and do not need to modify
 *	the data.
 */
static inline int skb_cow_head(struct sk_buff *skb, unsigned int headroom)
{
	return __skb_cow(skb, headroom, skb_header_cloned(skb));
L
Linus Torvalds 已提交
2817 2818 2819 2820 2821 2822 2823 2824 2825
}

/**
 *	skb_padto	- pad an skbuff up to a minimal size
 *	@skb: buffer to pad
 *	@len: minimal length
 *
 *	Pads up a buffer to ensure the trailing bytes exist and are
 *	blanked. If the buffer already contains sufficient data it
2826 2827
 *	is untouched. Otherwise it is extended. Returns zero on
 *	success. The skb is freed on error.
L
Linus Torvalds 已提交
2828
 */
2829
static inline int skb_padto(struct sk_buff *skb, unsigned int len)
L
Linus Torvalds 已提交
2830 2831 2832
{
	unsigned int size = skb->len;
	if (likely(size >= len))
2833
		return 0;
G
Gerrit Renker 已提交
2834
	return skb_pad(skb, len - size);
L
Linus Torvalds 已提交
2835 2836
}

2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858 2859
/**
 *	skb_put_padto - increase size and pad an skbuff up to a minimal size
 *	@skb: buffer to pad
 *	@len: minimal length
 *
 *	Pads up a buffer to ensure the trailing bytes exist and are
 *	blanked. If the buffer already contains sufficient data it
 *	is untouched. Otherwise it is extended. Returns zero on
 *	success. The skb is freed on error.
 */
static inline int skb_put_padto(struct sk_buff *skb, unsigned int len)
{
	unsigned int size = skb->len;

	if (unlikely(size < len)) {
		len -= size;
		if (skb_pad(skb, len))
			return -ENOMEM;
		__skb_put(skb, len);
	}
	return 0;
}

L
Linus Torvalds 已提交
2860
static inline int skb_add_data(struct sk_buff *skb,
2861
			       struct iov_iter *from, int copy)
L
Linus Torvalds 已提交
2862 2863 2864 2865
{
	const int off = skb->len;

	if (skb->ip_summed == CHECKSUM_NONE) {
2866
		__wsum csum = 0;
2867 2868
		if (csum_and_copy_from_iter_full(skb_put(skb, copy), copy,
					         &csum, from)) {
L
Linus Torvalds 已提交
2869 2870 2871
			skb->csum = csum_block_add(skb->csum, csum, off);
			return 0;
		}
2872
	} else if (copy_from_iter_full(skb_put(skb, copy), copy, from))
L
Linus Torvalds 已提交
2873 2874 2875 2876 2877 2878
		return 0;

	__skb_trim(skb, off);
	return -EFAULT;
}

2879 2880
static inline bool skb_can_coalesce(struct sk_buff *skb, int i,
				    const struct page *page, int off)
L
Linus Torvalds 已提交
2881 2882
{
	if (i) {
E
Eric Dumazet 已提交
2883
		const struct skb_frag_struct *frag = &skb_shinfo(skb)->frags[i - 1];
L
Linus Torvalds 已提交
2884

2885
		return page == skb_frag_page(frag) &&
E
Eric Dumazet 已提交
2886
		       off == frag->page_offset + skb_frag_size(frag);
L
Linus Torvalds 已提交
2887
	}
2888
	return false;
L
Linus Torvalds 已提交
2889 2890
}

H
Herbert Xu 已提交
2891 2892 2893 2894 2895
static inline int __skb_linearize(struct sk_buff *skb)
{
	return __pskb_pull_tail(skb, skb->data_len) ? 0 : -ENOMEM;
}

L
Linus Torvalds 已提交
2896 2897 2898 2899 2900 2901 2902
/**
 *	skb_linearize - convert paged skb to linear one
 *	@skb: buffer to linarize
 *
 *	If there is no free memory -ENOMEM is returned, otherwise zero
 *	is returned and the old skb data released.
 */
H
Herbert Xu 已提交
2903 2904 2905 2906 2907
static inline int skb_linearize(struct sk_buff *skb)
{
	return skb_is_nonlinear(skb) ? __skb_linearize(skb) : 0;
}

2908 2909 2910 2911 2912 2913 2914 2915 2916
/**
 * skb_has_shared_frag - can any frag be overwritten
 * @skb: buffer to test
 *
 * Return true if the skb has at least one frag that might be modified
 * by an external entity (as in vmsplice()/sendfile())
 */
static inline bool skb_has_shared_frag(const struct sk_buff *skb)
{
2917 2918
	return skb_is_nonlinear(skb) &&
	       skb_shinfo(skb)->tx_flags & SKBTX_SHARED_FRAG;
2919 2920
}

H
Herbert Xu 已提交
2921 2922 2923 2924 2925 2926 2927 2928
/**
 *	skb_linearize_cow - make sure skb is linear and writable
 *	@skb: buffer to process
 *
 *	If there is no free memory -ENOMEM is returned, otherwise zero
 *	is returned and the old skb data released.
 */
static inline int skb_linearize_cow(struct sk_buff *skb)
L
Linus Torvalds 已提交
2929
{
H
Herbert Xu 已提交
2930 2931
	return skb_is_nonlinear(skb) || skb_cloned(skb) ?
	       __skb_linearize(skb) : 0;
L
Linus Torvalds 已提交
2932 2933
}

2934 2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945
static __always_inline void
__skb_postpull_rcsum(struct sk_buff *skb, const void *start, unsigned int len,
		     unsigned int off)
{
	if (skb->ip_summed == CHECKSUM_COMPLETE)
		skb->csum = csum_block_sub(skb->csum,
					   csum_partial(start, len, 0), off);
	else if (skb->ip_summed == CHECKSUM_PARTIAL &&
		 skb_checksum_start_offset(skb) < 0)
		skb->ip_summed = CHECKSUM_NONE;
}

L
Linus Torvalds 已提交
2946 2947 2948 2949 2950 2951 2952
/**
 *	skb_postpull_rcsum - update checksum for received skb after pull
 *	@skb: buffer to update
 *	@start: start of data before pull
 *	@len: length of data pulled
 *
 *	After doing a pull on a received packet, you need to call this to
2953 2954
 *	update the CHECKSUM_COMPLETE checksum, or set ip_summed to
 *	CHECKSUM_NONE so that it can be recomputed from scratch.
L
Linus Torvalds 已提交
2955 2956
 */
static inline void skb_postpull_rcsum(struct sk_buff *skb,
2957
				      const void *start, unsigned int len)
L
Linus Torvalds 已提交
2958
{
2959
	__skb_postpull_rcsum(skb, start, len, 0);
L
Linus Torvalds 已提交
2960 2961
}

2962 2963 2964 2965 2966 2967 2968 2969
static __always_inline void
__skb_postpush_rcsum(struct sk_buff *skb, const void *start, unsigned int len,
		     unsigned int off)
{
	if (skb->ip_summed == CHECKSUM_COMPLETE)
		skb->csum = csum_block_add(skb->csum,
					   csum_partial(start, len, 0), off);
}
2970

2971 2972 2973 2974 2975 2976 2977 2978 2979
/**
 *	skb_postpush_rcsum - update checksum for received skb after push
 *	@skb: buffer to update
 *	@start: start of data after push
 *	@len: length of data pushed
 *
 *	After doing a push on a received packet, you need to call this to
 *	update the CHECKSUM_COMPLETE checksum.
 */
2980 2981 2982
static inline void skb_postpush_rcsum(struct sk_buff *skb,
				      const void *start, unsigned int len)
{
2983
	__skb_postpush_rcsum(skb, start, len, 0);
2984 2985
}

2986
void *skb_pull_rcsum(struct sk_buff *skb, unsigned int len);
2987

2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998
/**
 *	skb_push_rcsum - push skb and update receive checksum
 *	@skb: buffer to update
 *	@len: length of data pulled
 *
 *	This function performs an skb_push on the packet and updates
 *	the CHECKSUM_COMPLETE checksum.  It should be used on
 *	receive path processing instead of skb_push unless you know
 *	that the checksum difference is zero (e.g., a valid IP header)
 *	or you are setting ip_summed to CHECKSUM_NONE.
 */
2999
static inline void *skb_push_rcsum(struct sk_buff *skb, unsigned int len)
3000 3001 3002 3003 3004 3005
{
	skb_push(skb, len);
	skb_postpush_rcsum(skb, skb->data, len);
	return skb->data;
}

3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023
/**
 *	pskb_trim_rcsum - trim received skb and update checksum
 *	@skb: buffer to trim
 *	@len: new length
 *
 *	This is exactly the same as pskb_trim except that it ensures the
 *	checksum of received packets are still valid after the operation.
 */

static inline int pskb_trim_rcsum(struct sk_buff *skb, unsigned int len)
{
	if (likely(len >= skb->len))
		return 0;
	if (skb->ip_summed == CHECKSUM_COMPLETE)
		skb->ip_summed = CHECKSUM_NONE;
	return __pskb_trim(skb, len);
}

3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038
static inline int __skb_trim_rcsum(struct sk_buff *skb, unsigned int len)
{
	if (skb->ip_summed == CHECKSUM_COMPLETE)
		skb->ip_summed = CHECKSUM_NONE;
	__skb_trim(skb, len);
	return 0;
}

static inline int __skb_grow_rcsum(struct sk_buff *skb, unsigned int len)
{
	if (skb->ip_summed == CHECKSUM_COMPLETE)
		skb->ip_summed = CHECKSUM_NONE;
	return __skb_grow(skb, len);
}

L
Linus Torvalds 已提交
3039 3040
#define skb_queue_walk(queue, skb) \
		for (skb = (queue)->next;					\
3041
		     skb != (struct sk_buff *)(queue);				\
L
Linus Torvalds 已提交
3042 3043
		     skb = skb->next)

3044 3045 3046 3047 3048
#define skb_queue_walk_safe(queue, skb, tmp)					\
		for (skb = (queue)->next, tmp = skb->next;			\
		     skb != (struct sk_buff *)(queue);				\
		     skb = tmp, tmp = skb->next)

3049
#define skb_queue_walk_from(queue, skb)						\
3050
		for (; skb != (struct sk_buff *)(queue);			\
3051 3052 3053 3054 3055 3056 3057
		     skb = skb->next)

#define skb_queue_walk_from_safe(queue, skb, tmp)				\
		for (tmp = skb->next;						\
		     skb != (struct sk_buff *)(queue);				\
		     skb = tmp, tmp = skb->next)

3058 3059
#define skb_queue_reverse_walk(queue, skb) \
		for (skb = (queue)->prev;					\
3060
		     skb != (struct sk_buff *)(queue);				\
3061 3062
		     skb = skb->prev)

3063 3064 3065 3066 3067 3068 3069 3070 3071
#define skb_queue_reverse_walk_safe(queue, skb, tmp)				\
		for (skb = (queue)->prev, tmp = skb->prev;			\
		     skb != (struct sk_buff *)(queue);				\
		     skb = tmp, tmp = skb->prev)

#define skb_queue_reverse_walk_from_safe(queue, skb, tmp)			\
		for (tmp = skb->prev;						\
		     skb != (struct sk_buff *)(queue);				\
		     skb = tmp, tmp = skb->prev)
L
Linus Torvalds 已提交
3072

3073
static inline bool skb_has_frag_list(const struct sk_buff *skb)
3074 3075 3076 3077 3078 3079 3080 3081 3082 3083 3084 3085
{
	return skb_shinfo(skb)->frag_list != NULL;
}

static inline void skb_frag_list_init(struct sk_buff *skb)
{
	skb_shinfo(skb)->frag_list = NULL;
}

#define skb_walk_frags(skb, iter)	\
	for (iter = skb_shinfo(skb)->frag_list; iter; iter = iter->next)

3086 3087 3088

int __skb_wait_for_more_packets(struct sock *sk, int *err, long *timeo_p,
				const struct sk_buff *skb);
3089 3090 3091 3092 3093 3094 3095
struct sk_buff *__skb_try_recv_from_queue(struct sock *sk,
					  struct sk_buff_head *queue,
					  unsigned int flags,
					  void (*destructor)(struct sock *sk,
							   struct sk_buff *skb),
					  int *peeked, int *off, int *err,
					  struct sk_buff **last);
3096
struct sk_buff *__skb_try_recv_datagram(struct sock *sk, unsigned flags,
3097 3098
					void (*destructor)(struct sock *sk,
							   struct sk_buff *skb),
3099 3100
					int *peeked, int *off, int *err,
					struct sk_buff **last);
3101
struct sk_buff *__skb_recv_datagram(struct sock *sk, unsigned flags,
3102 3103
				    void (*destructor)(struct sock *sk,
						       struct sk_buff *skb),
3104 3105 3106 3107 3108
				    int *peeked, int *off, int *err);
struct sk_buff *skb_recv_datagram(struct sock *sk, unsigned flags, int noblock,
				  int *err);
unsigned int datagram_poll(struct file *file, struct socket *sock,
			   struct poll_table_struct *wait);
A
Al Viro 已提交
3109 3110
int skb_copy_datagram_iter(const struct sk_buff *from, int offset,
			   struct iov_iter *to, int size);
3111 3112 3113
static inline int skb_copy_datagram_msg(const struct sk_buff *from, int offset,
					struct msghdr *msg, int size)
{
3114
	return skb_copy_datagram_iter(from, offset, &msg->msg_iter, size);
3115
}
3116 3117
int skb_copy_and_csum_datagram_msg(struct sk_buff *skb, int hlen,
				   struct msghdr *msg);
3118 3119 3120
int skb_copy_datagram_from_iter(struct sk_buff *skb, int offset,
				 struct iov_iter *from, int len);
int zerocopy_sg_from_iter(struct sk_buff *skb, struct iov_iter *frm);
3121
void skb_free_datagram(struct sock *sk, struct sk_buff *skb);
3122 3123 3124 3125 3126 3127
void __skb_free_datagram_locked(struct sock *sk, struct sk_buff *skb, int len);
static inline void skb_free_datagram_locked(struct sock *sk,
					    struct sk_buff *skb)
{
	__skb_free_datagram_locked(sk, skb, 0);
}
3128 3129 3130 3131 3132
int skb_kill_datagram(struct sock *sk, struct sk_buff *skb, unsigned int flags);
int skb_copy_bits(const struct sk_buff *skb, int offset, void *to, int len);
int skb_store_bits(struct sk_buff *skb, int offset, const void *from, int len);
__wsum skb_copy_and_csum_bits(const struct sk_buff *skb, int offset, u8 *to,
			      int len, __wsum csum);
3133
int skb_splice_bits(struct sk_buff *skb, struct sock *sk, unsigned int offset,
3134
		    struct pipe_inode_info *pipe, unsigned int len,
A
Al Viro 已提交
3135
		    unsigned int flags);
3136 3137 3138
int skb_send_sock_locked(struct sock *sk, struct sk_buff *skb, int offset,
			 int len);
int skb_send_sock(struct sock *sk, struct sk_buff *skb, int offset, int len);
3139
void skb_copy_and_csum_dev(const struct sk_buff *skb, u8 *to);
3140
unsigned int skb_zerocopy_headlen(const struct sk_buff *from);
3141 3142
int skb_zerocopy(struct sk_buff *to, struct sk_buff *from,
		 int len, int hlen);
3143 3144 3145
void skb_split(struct sk_buff *skb, struct sk_buff *skb1, const u32 len);
int skb_shift(struct sk_buff *tgt, struct sk_buff *skb, int shiftlen);
void skb_scrub_packet(struct sk_buff *skb, bool xnet);
3146
unsigned int skb_gso_transport_seglen(const struct sk_buff *skb);
3147
bool skb_gso_validate_mtu(const struct sk_buff *skb, unsigned int mtu);
3148
struct sk_buff *skb_segment(struct sk_buff *skb, netdev_features_t features);
3149
struct sk_buff *skb_vlan_untag(struct sk_buff *skb);
3150
int skb_ensure_writable(struct sk_buff *skb, int write_len);
3151
int __skb_vlan_pop(struct sk_buff *skb, u16 *vlan_tci);
3152 3153
int skb_vlan_pop(struct sk_buff *skb);
int skb_vlan_push(struct sk_buff *skb, __be16 vlan_proto, u16 vlan_tci);
3154 3155
struct sk_buff *pskb_extract(struct sk_buff *skb, int off, int to_copy,
			     gfp_t gfp);
3156

A
Al Viro 已提交
3157 3158
static inline int memcpy_from_msg(void *data, struct msghdr *msg, int len)
{
3159
	return copy_from_iter_full(data, len, &msg->msg_iter) ? 0 : -EFAULT;
A
Al Viro 已提交
3160 3161
}

A
Al Viro 已提交
3162 3163
static inline int memcpy_to_msg(struct msghdr *msg, void *data, int len)
{
3164
	return copy_to_iter(data, len, &msg->msg_iter) == len ? 0 : -EFAULT;
A
Al Viro 已提交
3165 3166
}

3167 3168 3169 3170 3171
struct skb_checksum_ops {
	__wsum (*update)(const void *mem, int len, __wsum wsum);
	__wsum (*combine)(__wsum csum, __wsum csum2, int offset, int len);
};

3172 3173
extern const struct skb_checksum_ops *crc32c_csum_stub __read_mostly;

3174 3175 3176 3177 3178
__wsum __skb_checksum(const struct sk_buff *skb, int offset, int len,
		      __wsum csum, const struct skb_checksum_ops *ops);
__wsum skb_checksum(const struct sk_buff *skb, int offset, int len,
		    __wsum csum);

3179 3180 3181
static inline void * __must_check
__skb_header_pointer(const struct sk_buff *skb, int offset,
		     int len, void *data, int hlen, void *buffer)
L
Linus Torvalds 已提交
3182
{
3183
	if (hlen - offset >= len)
3184
		return data + offset;
L
Linus Torvalds 已提交
3185

3186 3187
	if (!skb ||
	    skb_copy_bits(skb, offset, buffer, len) < 0)
L
Linus Torvalds 已提交
3188 3189 3190 3191 3192
		return NULL;

	return buffer;
}

3193 3194
static inline void * __must_check
skb_header_pointer(const struct sk_buff *skb, int offset, int len, void *buffer)
3195 3196 3197 3198 3199
{
	return __skb_header_pointer(skb, offset, len, skb->data,
				    skb_headlen(skb), buffer);
}

3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217
/**
 *	skb_needs_linearize - check if we need to linearize a given skb
 *			      depending on the given device features.
 *	@skb: socket buffer to check
 *	@features: net device features
 *
 *	Returns true if either:
 *	1. skb has frag_list and the device doesn't support FRAGLIST, or
 *	2. skb is fragmented and the device does not support SG.
 */
static inline bool skb_needs_linearize(struct sk_buff *skb,
				       netdev_features_t features)
{
	return skb_is_nonlinear(skb) &&
	       ((skb_has_frag_list(skb) && !(features & NETIF_F_FRAGLIST)) ||
		(skb_shinfo(skb)->nr_frags && !(features & NETIF_F_SG)));
}

3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231
static inline void skb_copy_from_linear_data(const struct sk_buff *skb,
					     void *to,
					     const unsigned int len)
{
	memcpy(to, skb->data, len);
}

static inline void skb_copy_from_linear_data_offset(const struct sk_buff *skb,
						    const int offset, void *to,
						    const unsigned int len)
{
	memcpy(to, skb->data + offset, len);
}

3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246
static inline void skb_copy_to_linear_data(struct sk_buff *skb,
					   const void *from,
					   const unsigned int len)
{
	memcpy(skb->data, from, len);
}

static inline void skb_copy_to_linear_data_offset(struct sk_buff *skb,
						  const int offset,
						  const void *from,
						  const unsigned int len)
{
	memcpy(skb->data + offset, from, len);
}

3247
void skb_init(void);
L
Linus Torvalds 已提交
3248

3249 3250 3251 3252 3253
static inline ktime_t skb_get_ktime(const struct sk_buff *skb)
{
	return skb->tstamp;
}

3254 3255 3256 3257 3258 3259 3260 3261 3262
/**
 *	skb_get_timestamp - get timestamp from a skb
 *	@skb: skb to get stamp from
 *	@stamp: pointer to struct timeval to store stamp in
 *
 *	Timestamps are stored in the skb as offsets to a base timestamp.
 *	This function converts the offset back to a struct timeval and stores
 *	it in stamp.
 */
3263 3264
static inline void skb_get_timestamp(const struct sk_buff *skb,
				     struct timeval *stamp)
3265
{
3266
	*stamp = ktime_to_timeval(skb->tstamp);
3267 3268
}

3269 3270 3271 3272 3273 3274
static inline void skb_get_timestampns(const struct sk_buff *skb,
				       struct timespec *stamp)
{
	*stamp = ktime_to_timespec(skb->tstamp);
}

3275
static inline void __net_timestamp(struct sk_buff *skb)
3276
{
3277
	skb->tstamp = ktime_get_real();
3278 3279
}

3280 3281 3282 3283 3284
static inline ktime_t net_timedelta(ktime_t t)
{
	return ktime_sub(ktime_get_real(), t);
}

3285 3286
static inline ktime_t net_invalid_timestamp(void)
{
T
Thomas Gleixner 已提交
3287
	return 0;
3288
}
3289

3290 3291
struct sk_buff *skb_clone_sk(struct sk_buff *skb);

3292 3293
#ifdef CONFIG_NETWORK_PHY_TIMESTAMPING

3294 3295
void skb_clone_tx_timestamp(struct sk_buff *skb);
bool skb_defer_rx_timestamp(struct sk_buff *skb);
3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312

#else /* CONFIG_NETWORK_PHY_TIMESTAMPING */

static inline void skb_clone_tx_timestamp(struct sk_buff *skb)
{
}

static inline bool skb_defer_rx_timestamp(struct sk_buff *skb)
{
	return false;
}

#endif /* !CONFIG_NETWORK_PHY_TIMESTAMPING */

/**
 * skb_complete_tx_timestamp() - deliver cloned skb with tx timestamps
 *
3313 3314
 * PHY drivers may accept clones of transmitted packets for
 * timestamping via their phy_driver.txtstamp method. These drivers
3315 3316
 * must call this function to return the skb back to the stack with a
 * timestamp.
3317
 *
3318
 * @skb: clone of the the original outgoing packet
3319
 * @hwtstamps: hardware time stamps
3320 3321 3322 3323 3324
 *
 */
void skb_complete_tx_timestamp(struct sk_buff *skb,
			       struct skb_shared_hwtstamps *hwtstamps);

3325 3326 3327 3328
void __skb_tstamp_tx(struct sk_buff *orig_skb,
		     struct skb_shared_hwtstamps *hwtstamps,
		     struct sock *sk, int tstype);

3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 3339
/**
 * skb_tstamp_tx - queue clone of skb with send time stamps
 * @orig_skb:	the original outgoing packet
 * @hwtstamps:	hardware time stamps, may be NULL if not available
 *
 * If the skb has a socket associated, then this function clones the
 * skb (thus sharing the actual data and optional structures), stores
 * the optional hardware time stamping information (if non NULL) or
 * generates a software time stamp (otherwise), then queues the clone
 * to the error queue of the socket.  Errors are silently ignored.
 */
3340 3341
void skb_tstamp_tx(struct sk_buff *orig_skb,
		   struct skb_shared_hwtstamps *hwtstamps);
3342

3343 3344 3345 3346
/**
 * skb_tx_timestamp() - Driver hook for transmit timestamping
 *
 * Ethernet MAC Drivers should call this function in their hard_xmit()
3347
 * function immediately before giving the sk_buff to the MAC hardware.
3348
 *
3349 3350 3351 3352
 * Specifically, one should make absolutely sure that this function is
 * called before TX completion of this packet can trigger.  Otherwise
 * the packet could potentially already be freed.
 *
3353 3354 3355 3356
 * @skb: A socket buffer.
 */
static inline void skb_tx_timestamp(struct sk_buff *skb)
{
3357
	skb_clone_tx_timestamp(skb);
3358 3359
	if (skb_shinfo(skb)->tx_flags & SKBTX_SW_TSTAMP)
		skb_tstamp_tx(skb, NULL);
3360 3361
}

3362 3363 3364 3365 3366 3367 3368 3369 3370
/**
 * skb_complete_wifi_ack - deliver skb with wifi status
 *
 * @skb: the original outgoing packet
 * @acked: ack status
 *
 */
void skb_complete_wifi_ack(struct sk_buff *skb, bool acked);

3371 3372
__sum16 __skb_checksum_complete_head(struct sk_buff *skb, int len);
__sum16 __skb_checksum_complete(struct sk_buff *skb);
3373

3374 3375
static inline int skb_csum_unnecessary(const struct sk_buff *skb)
{
3376 3377 3378 3379
	return ((skb->ip_summed == CHECKSUM_UNNECESSARY) ||
		skb->csum_valid ||
		(skb->ip_summed == CHECKSUM_PARTIAL &&
		 skb_checksum_start_offset(skb) >= 0));
3380 3381
}

3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397
/**
 *	skb_checksum_complete - Calculate checksum of an entire packet
 *	@skb: packet to process
 *
 *	This function calculates the checksum over the entire packet plus
 *	the value of skb->csum.  The latter can be used to supply the
 *	checksum of a pseudo header as used by TCP/UDP.  It returns the
 *	checksum.
 *
 *	For protocols that contain complete checksums such as ICMP/TCP/UDP,
 *	this function can be used to verify that checksum on received
 *	packets.  In that case the function should return zero if the
 *	checksum is correct.  In particular, this function will return zero
 *	if skb->ip_summed is CHECKSUM_UNNECESSARY which indicates that the
 *	hardware has already verified the correctness of the checksum.
 */
3398
static inline __sum16 skb_checksum_complete(struct sk_buff *skb)
3399
{
3400 3401
	return skb_csum_unnecessary(skb) ?
	       0 : __skb_checksum_complete(skb);
3402 3403
}

3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424
static inline void __skb_decr_checksum_unnecessary(struct sk_buff *skb)
{
	if (skb->ip_summed == CHECKSUM_UNNECESSARY) {
		if (skb->csum_level == 0)
			skb->ip_summed = CHECKSUM_NONE;
		else
			skb->csum_level--;
	}
}

static inline void __skb_incr_checksum_unnecessary(struct sk_buff *skb)
{
	if (skb->ip_summed == CHECKSUM_UNNECESSARY) {
		if (skb->csum_level < SKB_MAX_CSUM_LEVEL)
			skb->csum_level++;
	} else if (skb->ip_summed == CHECKSUM_NONE) {
		skb->ip_summed = CHECKSUM_UNNECESSARY;
		skb->csum_level = 0;
	}
}

3425 3426 3427 3428 3429 3430 3431 3432 3433
/* Check if we need to perform checksum complete validation.
 *
 * Returns true if checksum complete is needed, false otherwise
 * (either checksum is unnecessary or zero checksum is allowed).
 */
static inline bool __skb_checksum_validate_needed(struct sk_buff *skb,
						  bool zero_okay,
						  __sum16 check)
{
3434 3435
	if (skb_csum_unnecessary(skb) || (zero_okay && !check)) {
		skb->csum_valid = 1;
3436
		__skb_decr_checksum_unnecessary(skb);
3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447
		return false;
	}

	return true;
}

/* For small packets <= CHECKSUM_BREAK peform checksum complete directly
 * in checksum_init.
 */
#define CHECKSUM_BREAK 76

3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459
/* Unset checksum-complete
 *
 * Unset checksum complete can be done when packet is being modified
 * (uncompressed for instance) and checksum-complete value is
 * invalidated.
 */
static inline void skb_checksum_complete_unset(struct sk_buff *skb)
{
	if (skb->ip_summed == CHECKSUM_COMPLETE)
		skb->ip_summed = CHECKSUM_NONE;
}

3460 3461 3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474
/* Validate (init) checksum based on checksum complete.
 *
 * Return values:
 *   0: checksum is validated or try to in skb_checksum_complete. In the latter
 *	case the ip_summed will not be CHECKSUM_UNNECESSARY and the pseudo
 *	checksum is stored in skb->csum for use in __skb_checksum_complete
 *   non-zero: value of invalid checksum
 *
 */
static inline __sum16 __skb_checksum_validate_complete(struct sk_buff *skb,
						       bool complete,
						       __wsum psum)
{
	if (skb->ip_summed == CHECKSUM_COMPLETE) {
		if (!csum_fold(csum_add(psum, skb->csum))) {
3475
			skb->csum_valid = 1;
3476 3477 3478 3479 3480 3481
			return 0;
		}
	}

	skb->csum = psum;

3482 3483 3484 3485 3486 3487 3488
	if (complete || skb->len <= CHECKSUM_BREAK) {
		__sum16 csum;

		csum = __skb_checksum_complete(skb);
		skb->csum_valid = !csum;
		return csum;
	}
3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502 3503 3504 3505 3506 3507 3508 3509 3510 3511

	return 0;
}

static inline __wsum null_compute_pseudo(struct sk_buff *skb, int proto)
{
	return 0;
}

/* Perform checksum validate (init). Note that this is a macro since we only
 * want to calculate the pseudo header which is an input function if necessary.
 * First we try to validate without any computation (checksum unnecessary) and
 * then calculate based on checksum complete calling the function to compute
 * pseudo header.
 *
 * Return values:
 *   0: checksum is validated or try to in skb_checksum_complete
 *   non-zero: value of invalid checksum
 */
#define __skb_checksum_validate(skb, proto, complete,			\
				zero_okay, check, compute_pseudo)	\
({									\
	__sum16 __ret = 0;						\
3512
	skb->csum_valid = 0;						\
3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523 3524 3525 3526 3527 3528 3529
	if (__skb_checksum_validate_needed(skb, zero_okay, check))	\
		__ret = __skb_checksum_validate_complete(skb,		\
				complete, compute_pseudo(skb, proto));	\
	__ret;								\
})

#define skb_checksum_init(skb, proto, compute_pseudo)			\
	__skb_checksum_validate(skb, proto, false, false, 0, compute_pseudo)

#define skb_checksum_init_zero_check(skb, proto, check, compute_pseudo)	\
	__skb_checksum_validate(skb, proto, false, true, check, compute_pseudo)

#define skb_checksum_validate(skb, proto, compute_pseudo)		\
	__skb_checksum_validate(skb, proto, true, false, 0, compute_pseudo)

#define skb_checksum_validate_zero_check(skb, proto, check,		\
					 compute_pseudo)		\
3530
	__skb_checksum_validate(skb, proto, true, true, check, compute_pseudo)
3531 3532 3533 3534

#define skb_checksum_simple_validate(skb)				\
	__skb_checksum_validate(skb, 0, true, false, 0, null_compute_pseudo)

3535 3536
static inline bool __skb_checksum_convert_check(struct sk_buff *skb)
{
3537
	return (skb->ip_summed == CHECKSUM_NONE && skb->csum_valid);
3538 3539 3540 3541 3542 3543 3544 3545 3546 3547 3548 3549 3550 3551 3552 3553
}

static inline void __skb_checksum_convert(struct sk_buff *skb,
					  __sum16 check, __wsum pseudo)
{
	skb->csum = ~pseudo;
	skb->ip_summed = CHECKSUM_COMPLETE;
}

#define skb_checksum_try_convert(skb, proto, check, compute_pseudo)	\
do {									\
	if (__skb_checksum_convert_check(skb))				\
		__skb_checksum_convert(skb, check,			\
				       compute_pseudo(skb, proto));	\
} while (0)

3554 3555 3556 3557 3558 3559 3560 3561
static inline void skb_remcsum_adjust_partial(struct sk_buff *skb, void *ptr,
					      u16 start, u16 offset)
{
	skb->ip_summed = CHECKSUM_PARTIAL;
	skb->csum_start = ((unsigned char *)ptr + start) - skb->head;
	skb->csum_offset = offset - start;
}

3562 3563 3564 3565 3566 3567
/* Update skbuf and packet to reflect the remote checksum offload operation.
 * When called, ptr indicates the starting point for skb->csum when
 * ip_summed is CHECKSUM_COMPLETE. If we need create checksum complete
 * here, skb_postpull_rcsum is done so skb->csum start is ptr.
 */
static inline void skb_remcsum_process(struct sk_buff *skb, void *ptr,
3568
				       int start, int offset, bool nopartial)
3569 3570 3571
{
	__wsum delta;

3572 3573 3574 3575 3576
	if (!nopartial) {
		skb_remcsum_adjust_partial(skb, ptr, start, offset);
		return;
	}

3577 3578 3579 3580 3581 3582 3583 3584 3585 3586 3587
	 if (unlikely(skb->ip_summed != CHECKSUM_COMPLETE)) {
		__skb_checksum_complete(skb);
		skb_postpull_rcsum(skb, skb->data, ptr - (void *)skb->data);
	}

	delta = remcsum_adjust(ptr, skb->csum, start, offset);

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

3588 3589 3590
static inline struct nf_conntrack *skb_nfct(const struct sk_buff *skb)
{
#if IS_ENABLED(CONFIG_NF_CONNTRACK)
3591
	return (void *)(skb->_nfct & SKB_NFCT_PTRMASK);
3592 3593 3594 3595 3596
#else
	return NULL;
#endif
}

3597
#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
3598
void nf_conntrack_destroy(struct nf_conntrack *nfct);
L
Linus Torvalds 已提交
3599 3600 3601
static inline void nf_conntrack_put(struct nf_conntrack *nfct)
{
	if (nfct && atomic_dec_and_test(&nfct->use))
3602
		nf_conntrack_destroy(nfct);
L
Linus Torvalds 已提交
3603 3604 3605 3606 3607 3608
}
static inline void nf_conntrack_get(struct nf_conntrack *nfct)
{
	if (nfct)
		atomic_inc(&nfct->use);
}
3609
#endif
3610
#if IS_ENABLED(CONFIG_BRIDGE_NETFILTER)
L
Linus Torvalds 已提交
3611 3612
static inline void nf_bridge_put(struct nf_bridge_info *nf_bridge)
{
3613
	if (nf_bridge && refcount_dec_and_test(&nf_bridge->use))
L
Linus Torvalds 已提交
3614 3615 3616 3617 3618
		kfree(nf_bridge);
}
static inline void nf_bridge_get(struct nf_bridge_info *nf_bridge)
{
	if (nf_bridge)
3619
		refcount_inc(&nf_bridge->use);
L
Linus Torvalds 已提交
3620 3621
}
#endif /* CONFIG_BRIDGE_NETFILTER */
3622 3623
static inline void nf_reset(struct sk_buff *skb)
{
3624
#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
3625 3626
	nf_conntrack_put(skb_nfct(skb));
	skb->_nfct = 0;
3627
#endif
3628
#if IS_ENABLED(CONFIG_BRIDGE_NETFILTER)
3629 3630 3631 3632 3633
	nf_bridge_put(skb->nf_bridge);
	skb->nf_bridge = NULL;
#endif
}

3634 3635
static inline void nf_reset_trace(struct sk_buff *skb)
{
3636
#if IS_ENABLED(CONFIG_NETFILTER_XT_TARGET_TRACE) || defined(CONFIG_NF_TABLES)
G
Gao feng 已提交
3637 3638
	skb->nf_trace = 0;
#endif
3639 3640
}

3641
/* Note: This doesn't put any conntrack and bridge info in dst. */
3642 3643
static inline void __nf_copy(struct sk_buff *dst, const struct sk_buff *src,
			     bool copy)
3644
{
3645
#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
3646 3647
	dst->_nfct = src->_nfct;
	nf_conntrack_get(skb_nfct(src));
3648
#endif
3649
#if IS_ENABLED(CONFIG_BRIDGE_NETFILTER)
3650 3651 3652
	dst->nf_bridge  = src->nf_bridge;
	nf_bridge_get(src->nf_bridge);
#endif
3653
#if IS_ENABLED(CONFIG_NETFILTER_XT_TARGET_TRACE) || defined(CONFIG_NF_TABLES)
3654 3655
	if (copy)
		dst->nf_trace = src->nf_trace;
3656
#endif
3657 3658
}

3659 3660 3661
static inline void nf_copy(struct sk_buff *dst, const struct sk_buff *src)
{
#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
3662
	nf_conntrack_put(skb_nfct(dst));
3663
#endif
3664
#if IS_ENABLED(CONFIG_BRIDGE_NETFILTER)
3665 3666
	nf_bridge_put(dst->nf_bridge);
#endif
3667
	__nf_copy(dst, src, true);
3668 3669
}

3670 3671 3672 3673 3674 3675 3676 3677 3678 3679 3680 3681 3682 3683 3684 3685 3686 3687
#ifdef CONFIG_NETWORK_SECMARK
static inline void skb_copy_secmark(struct sk_buff *to, const struct sk_buff *from)
{
	to->secmark = from->secmark;
}

static inline void skb_init_secmark(struct sk_buff *skb)
{
	skb->secmark = 0;
}
#else
static inline void skb_copy_secmark(struct sk_buff *to, const struct sk_buff *from)
{ }

static inline void skb_init_secmark(struct sk_buff *skb)
{ }
#endif

3688 3689 3690 3691 3692 3693
static inline bool skb_irq_freeable(const struct sk_buff *skb)
{
	return !skb->destructor &&
#if IS_ENABLED(CONFIG_XFRM)
		!skb->sp &&
#endif
3694
		!skb_nfct(skb) &&
3695 3696 3697 3698
		!skb->_skb_refdst &&
		!skb_has_frag_list(skb);
}

3699 3700 3701 3702 3703
static inline void skb_set_queue_mapping(struct sk_buff *skb, u16 queue_mapping)
{
	skb->queue_mapping = queue_mapping;
}

3704
static inline u16 skb_get_queue_mapping(const struct sk_buff *skb)
3705 3706 3707 3708
{
	return skb->queue_mapping;
}

3709 3710 3711 3712 3713
static inline void skb_copy_queue_mapping(struct sk_buff *to, const struct sk_buff *from)
{
	to->queue_mapping = from->queue_mapping;
}

3714 3715 3716 3717 3718
static inline void skb_record_rx_queue(struct sk_buff *skb, u16 rx_queue)
{
	skb->queue_mapping = rx_queue + 1;
}

3719
static inline u16 skb_get_rx_queue(const struct sk_buff *skb)
3720 3721 3722 3723
{
	return skb->queue_mapping - 1;
}

3724
static inline bool skb_rx_queue_recorded(const struct sk_buff *skb)
3725
{
E
Eric Dumazet 已提交
3726
	return skb->queue_mapping != 0;
3727 3728
}

3729 3730 3731 3732 3733 3734 3735 3736 3737 3738
static inline void skb_set_dst_pending_confirm(struct sk_buff *skb, u32 val)
{
	skb->dst_pending_confirm = val;
}

static inline bool skb_get_dst_pending_confirm(const struct sk_buff *skb)
{
	return skb->dst_pending_confirm != 0;
}

3739 3740
static inline struct sec_path *skb_sec_path(struct sk_buff *skb)
{
3741
#ifdef CONFIG_XFRM
3742 3743 3744 3745
	return skb->sp;
#else
	return NULL;
#endif
3746
}
3747

3748 3749 3750
/* Keeps track of mac header offset relative to skb->head.
 * It is useful for TSO of Tunneling protocol. e.g. GRE.
 * For non-tunnel skb it points to skb_mac_header() and for
3751 3752 3753
 * tunnel skb it points to outer mac header.
 * Keeps track of level of encapsulation of network headers.
 */
3754
struct skb_gso_cb {
3755 3756 3757 3758
	union {
		int	mac_offset;
		int	data_offset;
	};
3759
	int	encap_level;
3760
	__wsum	csum;
3761
	__u16	csum_start;
3762
};
3763 3764
#define SKB_SGO_CB_OFFSET	32
#define SKB_GSO_CB(skb) ((struct skb_gso_cb *)((skb)->cb + SKB_SGO_CB_OFFSET))
3765 3766 3767 3768 3769 3770 3771

static inline int skb_tnl_header_len(const struct sk_buff *inner_skb)
{
	return (skb_mac_header(inner_skb) - inner_skb->head) -
		SKB_GSO_CB(inner_skb)->mac_offset;
}

3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785 3786
static inline int gso_pskb_expand_head(struct sk_buff *skb, int extra)
{
	int new_headroom, headroom;
	int ret;

	headroom = skb_headroom(skb);
	ret = pskb_expand_head(skb, extra, 0, GFP_ATOMIC);
	if (ret)
		return ret;

	new_headroom = skb_headroom(skb);
	SKB_GSO_CB(skb)->mac_offset += (new_headroom - headroom);
	return 0;
}

3787 3788 3789 3790 3791 3792 3793 3794 3795 3796
static inline void gso_reset_checksum(struct sk_buff *skb, __wsum res)
{
	/* Do not update partial checksums if remote checksum is enabled. */
	if (skb->remcsum_offload)
		return;

	SKB_GSO_CB(skb)->csum = res;
	SKB_GSO_CB(skb)->csum_start = skb_checksum_start(skb) - skb->head;
}

3797 3798 3799 3800 3801 3802 3803 3804 3805 3806
/* Compute the checksum for a gso segment. First compute the checksum value
 * from the start of transport header to SKB_GSO_CB(skb)->csum_start, and
 * then add in skb->csum (checksum from csum_start to end of packet).
 * skb->csum and csum_start are then updated to reflect the checksum of the
 * resultant packet starting from the transport header-- the resultant checksum
 * is in the res argument (i.e. normally zero or ~ of checksum of a pseudo
 * header.
 */
static inline __sum16 gso_make_checksum(struct sk_buff *skb, __wsum res)
{
3807 3808 3809
	unsigned char *csum_start = skb_transport_header(skb);
	int plen = (skb->head + SKB_GSO_CB(skb)->csum_start) - csum_start;
	__wsum partial = SKB_GSO_CB(skb)->csum;
3810

3811 3812
	SKB_GSO_CB(skb)->csum = res;
	SKB_GSO_CB(skb)->csum_start = csum_start - skb->head;
3813

3814
	return csum_fold(csum_partial(csum_start, plen, partial));
3815 3816
}

3817
static inline bool skb_is_gso(const struct sk_buff *skb)
H
Herbert Xu 已提交
3818 3819 3820 3821
{
	return skb_shinfo(skb)->gso_size;
}

3822
/* Note: Should be called only if skb_is_gso(skb) is true */
3823
static inline bool skb_is_gso_v6(const struct sk_buff *skb)
B
Brice Goglin 已提交
3824 3825 3826 3827
{
	return skb_shinfo(skb)->gso_type & SKB_GSO_TCPV6;
}

3828 3829 3830 3831 3832 3833 3834
static inline void skb_gso_reset(struct sk_buff *skb)
{
	skb_shinfo(skb)->gso_size = 0;
	skb_shinfo(skb)->gso_segs = 0;
	skb_shinfo(skb)->gso_type = 0;
}

3835
void __skb_warn_lro_forwarding(const struct sk_buff *skb);
3836 3837 3838 3839 3840

static inline bool skb_warn_if_lro(const struct sk_buff *skb)
{
	/* LRO sets gso_size but not gso_type, whereas if GSO is really
	 * wanted then gso_type will be set. */
3841 3842
	const struct skb_shared_info *shinfo = skb_shinfo(skb);

3843 3844
	if (skb_is_nonlinear(skb) && shinfo->gso_size != 0 &&
	    unlikely(shinfo->gso_type == 0)) {
3845 3846 3847 3848 3849 3850
		__skb_warn_lro_forwarding(skb);
		return true;
	}
	return false;
}

3851 3852 3853 3854 3855 3856 3857
static inline void skb_forward_csum(struct sk_buff *skb)
{
	/* Unfortunately we don't support this one.  Any brave souls? */
	if (skb->ip_summed == CHECKSUM_COMPLETE)
		skb->ip_summed = CHECKSUM_NONE;
}

3858 3859 3860 3861 3862 3863 3864 3865
/**
 * skb_checksum_none_assert - make sure skb ip_summed is CHECKSUM_NONE
 * @skb: skb to check
 *
 * fresh skbs have their ip_summed set to CHECKSUM_NONE.
 * Instead of forcing ip_summed to CHECKSUM_NONE, we can
 * use this helper, to document places where we make this assertion.
 */
3866
static inline void skb_checksum_none_assert(const struct sk_buff *skb)
3867 3868 3869 3870 3871 3872
{
#ifdef DEBUG
	BUG_ON(skb->ip_summed != CHECKSUM_NONE);
#endif
}

3873
bool skb_partial_csum_set(struct sk_buff *skb, u16 start, u16 off);
3874

P
Paul Durrant 已提交
3875
int skb_checksum_setup(struct sk_buff *skb, bool recalculate);
3876 3877 3878
struct sk_buff *skb_checksum_trimmed(struct sk_buff *skb,
				     unsigned int transport_len,
				     __sum16(*skb_chkf)(struct sk_buff *skb));
P
Paul Durrant 已提交
3879

3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891 3892
/**
 * skb_head_is_locked - Determine if the skb->head is locked down
 * @skb: skb to check
 *
 * The head on skbs build around a head frag can be removed if they are
 * not cloned.  This function returns true if the skb head is locked down
 * due to either being allocated via kmalloc, or by being a clone with
 * multiple references to the head.
 */
static inline bool skb_head_is_locked(const struct sk_buff *skb)
{
	return !skb->head_frag || skb_cloned(skb);
}
3893 3894 3895 3896 3897 3898 3899 3900 3901 3902 3903 3904 3905 3906 3907 3908 3909

/**
 * skb_gso_network_seglen - Return length of individual segments of a gso packet
 *
 * @skb: GSO skb
 *
 * skb_gso_network_seglen is used to determine the real size of the
 * individual segments, including Layer3 (IP, IPv6) and L4 headers (TCP/UDP).
 *
 * The MAC/L2 header is not accounted for.
 */
static inline unsigned int skb_gso_network_seglen(const struct sk_buff *skb)
{
	unsigned int hdr_len = skb_transport_header(skb) -
			       skb_network_header(skb);
	return hdr_len + skb_gso_transport_seglen(skb);
}
T
Thomas Graf 已提交
3910

3911 3912 3913
/* Local Checksum Offload.
 * Compute outer checksum based on the assumption that the
 * inner checksum will be offloaded later.
3914 3915
 * See Documentation/networking/checksum-offloads.txt for
 * explanation of how this works.
3916 3917 3918 3919 3920 3921
 * Fill in outer checksum adjustment (e.g. with sum of outer
 * pseudo-header) before calling.
 * Also ensure that inner checksum is in linear data area.
 */
static inline __wsum lco_csum(struct sk_buff *skb)
{
3922 3923 3924
	unsigned char *csum_start = skb_checksum_start(skb);
	unsigned char *l4_hdr = skb_transport_header(skb);
	__wsum partial;
3925 3926

	/* Start with complement of inner checksum adjustment */
3927 3928 3929
	partial = ~csum_unfold(*(__force __sum16 *)(csum_start +
						    skb->csum_offset));

3930
	/* Add in checksum of our headers (incl. outer checksum
3931
	 * adjustment filled in by caller) and return result.
3932
	 */
3933
	return csum_partial(l4_hdr, csum_start - l4_hdr, partial);
3934 3935
}

L
Linus Torvalds 已提交
3936 3937
#endif	/* __KERNEL__ */
#endif	/* _LINUX_SKBUFF_H */