skbuff.h 131.2 KB
Newer Older
1
/* SPDX-License-Identifier: GPL-2.0-or-later */
L
Linus Torvalds 已提交
2 3 4 5 6 7 8 9 10 11 12 13 14 15
/*
 *	Definitions for the 'struct sk_buff' memory handlers.
 *
 *	Authors:
 *		Alan Cox, <gw4pts@gw4pts.ampr.org>
 *		Florian La Roche, <rzsfl@rz.uni-sb.de>
 */

#ifndef _LINUX_SKBUFF_H
#define _LINUX_SKBUFF_H

#include <linux/kernel.h>
#include <linux/compiler.h>
#include <linux/time.h>
16
#include <linux/bug.h>
17
#include <linux/bvec.h>
L
Linus Torvalds 已提交
18
#include <linux/cache.h>
E
Eric Dumazet 已提交
19
#include <linux/rbtree.h>
20
#include <linux/socket.h>
21
#include <linux/refcount.h>
L
Linus Torvalds 已提交
22

A
Arun Sharma 已提交
23
#include <linux/atomic.h>
L
Linus Torvalds 已提交
24 25 26
#include <asm/types.h>
#include <linux/spinlock.h>
#include <linux/net.h>
27
#include <linux/textsearch.h>
L
Linus Torvalds 已提交
28
#include <net/checksum.h>
29
#include <linux/rcupdate.h>
30
#include <linux/hrtimer.h>
31
#include <linux/dma-mapping.h>
32
#include <linux/netdev_features.h>
33
#include <linux/sched.h>
34
#include <linux/sched/clock.h>
35
#include <net/flow_dissector.h>
36
#include <linux/splice.h>
37
#include <linux/in6.h>
38
#include <linux/if_packet.h>
39
#include <net/flow.h>
40 41 42
#if IS_ENABLED(CONFIG_NF_CONNTRACK)
#include <linux/netfilter/nf_conntrack_common.h>
#endif
L
Linus Torvalds 已提交
43

44 45 46 47 48 49
/* 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.
50 51
 * 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
52 53 54 55 56 57 58 59 60 61 62 63 64 65
 * 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
66
 *			  is TCP or UDP. The IPv4 header may contain IP options.
67 68 69 70 71 72 73
 *			  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
M
Miaohe Lin 已提交
74
 *			  IPv6|UDP where the Next Header field in the IPv6
75 76 77 78 79 80 81
 *			  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.
82
 *			 This flag is only used to disable the RX checksum
83 84 85 86 87
 *			 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
88
 *    verification is set in 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
 *
 *   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),
118
 *   GRE (checksum flag is set) and TCP, skb->csum_level would be set to
119
 *   two. If the device were only able to verify the UDP checksum and not
120
 *   GRE, either because it doesn't support GRE checksum or because GRE
121 122
 *   checksum is bad, skb->csum_level would be set to zero (TCP checksum is
 *   not considered in this case).
123 124 125 126
 *
 * CHECKSUM_COMPLETE:
 *
 *   This is the most generic way. The device supplied checksum of the _whole_
127
 *   packet as seen by netif_rx() and fills in skb->csum. This means the
128 129
 *   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
 *   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
156 157
 *   offset of the packet, but it should not attempt to validate that the
 *   checksum refers to a legitimate transport layer checksum -- it is the
158 159 160 161 162 163 164 165 166 167
 *   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
 *   This has the same meaning as CHECKSUM_NONE for checksum offload on
182
 *   output.
183
 *
184 185
 * CHECKSUM_COMPLETE:
 *   Not used in checksum output. If a driver observes a packet with this value
186
 *   set in skbuff, it should treat the packet as if CHECKSUM_NONE were set.
187 188 189 190 191
 *
 * 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
 *     will set csum_start and csum_offset accordingly, set ip_summed to
193 194 195 196 197 198
 *     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
 *
 *   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
203 204
 *     accordingly. Note that there is no indication in the skbuff that the
 *     CHECKSUM_PARTIAL refers to an FCOE checksum, so a driver that supports
205
 *     both IP checksum offload and FCOE CRC offload must verify which offload
206
 *     is configured for a packet, presumably by inspecting packet headers.
207 208 209 210 211 212 213
 *
 * 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
214 215 216
 * with GSO then ip_summed is CHECKSUM_PARTIAL, and both csum_start and
 * csum_offset are set to refer to the outermost checksum being offloaded
 * (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)))

241
struct ahash_request;
L
Linus Torvalds 已提交
242
struct net_device;
243
struct scatterlist;
J
Jens Axboe 已提交
244
struct pipe_inode_info;
H
Herbert Xu 已提交
245
struct iov_iter;
246
struct napi_struct;
247 248
struct bpf_prog;
union bpf_attr;
249
struct skb_ext;
L
Linus Torvalds 已提交
250

251
#if IS_ENABLED(CONFIG_BRIDGE_NETFILTER)
L
Linus Torvalds 已提交
252
struct nf_bridge_info {
253 254 255 256
	enum {
		BRNF_PROTO_UNCHANGED,
		BRNF_PROTO_8021Q,
		BRNF_PROTO_PPPOE
257
	} orig_proto:8;
258 259 260
	u8			pkt_otherhost:1;
	u8			in_prerouting:1;
	u8			bridged_dnat:1;
261
	__u16			frag_max_size;
262
	struct net_device	*physindev;
263 264 265

	/* always valid & non-NULL from FORWARD on, for physdev match */
	struct net_device	*physoutdev;
266
	union {
267
		/* prerouting: detect dnat in orig/reply direction */
268 269
		__be32          ipv4_daddr;
		struct in6_addr ipv6_daddr;
270 271 272 273 274 275

		/* 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];
276
	};
L
Linus Torvalds 已提交
277 278 279
};
#endif

280 281 282 283 284 285 286
#if IS_ENABLED(CONFIG_NET_TC_SKB_EXT)
/* Chain in tc_skb_ext will be used to share the tc chain with
 * ovs recirc_id. It will be set to the current chain by tc
 * and read by ovs to recirc_id.
 */
struct tc_skb_ext {
	__u32 chain;
287
	__u16 mru;
288 289 290
};
#endif

L
Linus Torvalds 已提交
291 292 293 294 295 296 297 298 299 300 301
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;

302 303 304 305 306 307
/* 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.
308
 */
309
#if (65536/PAGE_SIZE + 1) < 16
310
#define MAX_SKB_FRAGS 16UL
311
#else
312
#define MAX_SKB_FRAGS (65536/PAGE_SIZE + 1)
313
#endif
H
Hans Westgaard Ry 已提交
314
extern int sysctl_max_skb_frags;
L
Linus Torvalds 已提交
315

316 317 318 319 320
/* 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

321
typedef struct bio_vec skb_frag_t;
L
Linus Torvalds 已提交
322

323
/**
324
 * skb_frag_size() - Returns the size of a skb fragment
325 326
 * @frag: skb fragment
 */
E
Eric Dumazet 已提交
327 328
static inline unsigned int skb_frag_size(const skb_frag_t *frag)
{
329
	return frag->bv_len;
E
Eric Dumazet 已提交
330 331
}

332
/**
333
 * skb_frag_size_set() - Sets the size of a skb fragment
334 335 336
 * @frag: skb fragment
 * @size: size of fragment
 */
E
Eric Dumazet 已提交
337 338
static inline void skb_frag_size_set(skb_frag_t *frag, unsigned int size)
{
339
	frag->bv_len = size;
E
Eric Dumazet 已提交
340 341
}

342
/**
343
 * skb_frag_size_add() - Increments the size of a skb fragment by @delta
344 345 346
 * @frag: skb fragment
 * @delta: value to add
 */
E
Eric Dumazet 已提交
347 348
static inline void skb_frag_size_add(skb_frag_t *frag, int delta)
{
349
	frag->bv_len += delta;
E
Eric Dumazet 已提交
350 351
}

352
/**
353
 * skb_frag_size_sub() - Decrements the size of a skb fragment by @delta
354 355 356
 * @frag: skb fragment
 * @delta: value to subtract
 */
E
Eric Dumazet 已提交
357 358
static inline void skb_frag_size_sub(skb_frag_t *frag, int delta)
{
359
	frag->bv_len -= delta;
E
Eric Dumazet 已提交
360 361
}

362 363 364 365
/**
 * skb_frag_must_loop - Test if %p is a high memory page
 * @p: fragment's page
 */
366 367 368 369 370 371 372 373 374 375 376 377 378
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
379
 *	@f_off:		offset from start of f->bv_page
380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401
 *	@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))		\

402 403 404
#define HAVE_HW_TIME_STAMP

/**
405
 * struct skb_shared_hwtstamps - hardware time stamps
406 407 408 409
 * @hwtstamp:	hardware time stamp transformed into duration
 *		since arbitrary point in time
 *
 * Software time stamps generated by ktime_get_real() are stored in
410
 * skb->tstamp.
411 412 413 414 415 416 417 418 419 420 421
 *
 * 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;
};

422 423 424 425 426
/* Definitions for tx_flags in struct skb_shared_info */
enum {
	/* generate hardware time stamp */
	SKBTX_HW_TSTAMP = 1 << 0,

427
	/* generate software time stamp when queueing packet to NIC */
428 429 430 431 432
	SKBTX_SW_TSTAMP = 1 << 1,

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

433
	/* device driver supports TX zero-copy buffers */
E
Eric Dumazet 已提交
434
	SKBTX_DEV_ZEROCOPY = 1 << 3,
435 436

	/* generate wifi status information (where possible) */
E
Eric Dumazet 已提交
437
	SKBTX_WIFI_STATUS = 1 << 4,
438 439 440 441 442 443 444

	/* 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,
445 446 447

	/* generate software time stamp when entering packet scheduling */
	SKBTX_SCHED_TSTAMP = 1 << 6,
448 449
};

W
Willem de Bruijn 已提交
450
#define SKBTX_ZEROCOPY_FRAG	(SKBTX_DEV_ZEROCOPY | SKBTX_SHARED_FRAG)
451
#define SKBTX_ANY_SW_TSTAMP	(SKBTX_SW_TSTAMP    | \
452
				 SKBTX_SCHED_TSTAMP)
453 454
#define SKBTX_ANY_TSTAMP	(SKBTX_HW_TSTAMP | SKBTX_ANY_SW_TSTAMP)

455 456 457
/*
 * 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.
458 459
 * 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.
460 461
 * The ctx field is used to track device context.
 * The desc field is used to track userspace buffer index.
462 463
 */
struct ubuf_info {
464
	void (*callback)(struct ubuf_info *, bool zerocopy_success);
465 466 467 468 469 470 471 472 473 474 475 476
	union {
		struct {
			unsigned long desc;
			void *ctx;
		};
		struct {
			u32 id;
			u16 len;
			u16 zerocopy:1;
			u32 bytelen;
		};
	};
477
	refcount_t refcnt;
478 479 480 481 482

	struct mmpin {
		struct user_struct *user;
		unsigned int num_pg;
	} mmp;
483 484
};

W
Willem de Bruijn 已提交
485 486
#define skb_uarg(SKB)	((struct ubuf_info *)(skb_shinfo(SKB)->destructor_arg))

487 488 489
int mm_account_pinned_pages(struct mmpin *mmp, size_t size);
void mm_unaccount_pinned_pages(struct mmpin *mmp);

W
Willem de Bruijn 已提交
490
struct ubuf_info *sock_zerocopy_alloc(struct sock *sk, size_t size);
491 492
struct ubuf_info *sock_zerocopy_realloc(struct sock *sk, size_t size,
					struct ubuf_info *uarg);
W
Willem de Bruijn 已提交
493 494 495

static inline void sock_zerocopy_get(struct ubuf_info *uarg)
{
496
	refcount_inc(&uarg->refcnt);
W
Willem de Bruijn 已提交
497 498 499
}

void sock_zerocopy_put(struct ubuf_info *uarg);
500
void sock_zerocopy_put_abort(struct ubuf_info *uarg, bool have_uref);
W
Willem de Bruijn 已提交
501 502 503

void sock_zerocopy_callback(struct ubuf_info *uarg, bool success);

W
Willem de Bruijn 已提交
504
int skb_zerocopy_iter_dgram(struct sk_buff *skb, struct msghdr *msg, int len);
W
Willem de Bruijn 已提交
505 506 507 508
int skb_zerocopy_iter_stream(struct sock *sk, struct sk_buff *skb,
			     struct msghdr *msg, int len,
			     struct ubuf_info *uarg);

L
Linus Torvalds 已提交
509 510 511 512
/* This data is invariant across clones and lives at
 * the end of the header data, ie. at skb->end.
 */
struct skb_shared_info {
513 514 515
	__u8		__unused;
	__u8		meta_len;
	__u8		nr_frags;
516
	__u8		tx_flags;
517 518 519
	unsigned short	gso_size;
	/* Warning: this field is not always filled in (UFO)! */
	unsigned short	gso_segs;
L
Linus Torvalds 已提交
520
	struct sk_buff	*frag_list;
521
	struct skb_shared_hwtstamps hwtstamps;
522
	unsigned int	gso_type;
523
	u32		tskey;
E
Eric Dumazet 已提交
524 525 526 527 528 529

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

J
Johann Baudy 已提交
530 531 532
	/* Intermediate layers must ensure that destructor_arg
	 * remains valid until skb destructor */
	void *		destructor_arg;
533

534 535
	/* must be last field, see pskb_expand_head() */
	skb_frag_t	frags[MAX_SKB_FRAGS];
L
Linus Torvalds 已提交
536 537 538 539
};

/* 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
540 541
 * 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
542 543 544 545 546 547 548 549 550 551
 *
 * 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)

552 553

enum {
554 555 556
	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) */
557 558
};

559 560
enum {
	SKB_GSO_TCPV4 = 1 << 0,
561 562

	/* This indicates the skb is from an untrusted source. */
563
	SKB_GSO_DODGY = 1 << 1,
M
Michael Chan 已提交
564 565

	/* This indicates the tcp segment has CWR set. */
566
	SKB_GSO_TCP_ECN = 1 << 2,
H
Herbert Xu 已提交
567

568
	SKB_GSO_TCP_FIXEDID = 1 << 3,
569

570
	SKB_GSO_TCPV6 = 1 << 4,
571

572
	SKB_GSO_FCOE = 1 << 5,
573

574
	SKB_GSO_GRE = 1 << 6,
S
Simon Horman 已提交
575

576
	SKB_GSO_GRE_CSUM = 1 << 7,
E
Eric Dumazet 已提交
577

578
	SKB_GSO_IPXIP4 = 1 << 8,
E
Eric Dumazet 已提交
579

580
	SKB_GSO_IPXIP6 = 1 << 9,
581

582
	SKB_GSO_UDP_TUNNEL = 1 << 10,
T
Tom Herbert 已提交
583

584
	SKB_GSO_UDP_TUNNEL_CSUM = 1 << 11,
585

586
	SKB_GSO_PARTIAL = 1 << 12,
587

588
	SKB_GSO_TUNNEL_REMCSUM = 1 << 13,
M
Marcelo Ricardo Leitner 已提交
589

590
	SKB_GSO_SCTP = 1 << 14,
S
Steffen Klassert 已提交
591

592
	SKB_GSO_ESP = 1 << 15,
593 594

	SKB_GSO_UDP = 1 << 16,
W
Willem de Bruijn 已提交
595 596

	SKB_GSO_UDP_L4 = 1 << 17,
597 598

	SKB_GSO_FRAGLIST = 1 << 18,
599 600
};

601 602 603 604 605 606 607 608 609 610
#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

611
/**
L
Linus Torvalds 已提交
612 613 614
 *	struct sk_buff - socket buffer
 *	@next: Next buffer in list
 *	@prev: Previous buffer in list
615
 *	@tstamp: Time we arrived/left
616 617
 *	@skb_mstamp_ns: (aka @tstamp) earliest departure time; start point
 *		for retransmit timer
E
Eric Dumazet 已提交
618
 *	@rbnode: RB tree node, alternative to next/prev for netem/tcp
619
 *	@list: queue head
620
 *	@sk: Socket we are owned by
621 622
 *	@ip_defrag_offset: (aka @sk) alternate use of @sk, used in
 *		fragmentation management
L
Linus Torvalds 已提交
623
 *	@dev: Device we arrived on/are leaving by
624
 *	@dev_scratch: (aka @dev) alternate use of @dev when @dev would be %NULL
625
 *	@cb: Control buffer. Free for use by every layer. Put private vars here
E
Eric Dumazet 已提交
626
 *	@_skb_refdst: destination entry (with norefcount bit)
627
 *	@sp: the security path, used for xfrm
L
Linus Torvalds 已提交
628 629 630
 *	@len: Length of actual data
 *	@data_len: Data length
 *	@mac_len: Length of link layer header
631
 *	@hdr_len: writable header length of cloned skb
632 633 634
 *	@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
635
 *	@priority: Packet queueing priority
W
WANG Cong 已提交
636
 *	@ignore_df: allow local fragmentation
L
Linus Torvalds 已提交
637
 *	@cloned: Head may be cloned (check refcnt to be sure)
638
 *	@ip_summed: Driver fed us an IP checksum
L
Linus Torvalds 已提交
639 640
 *	@nohdr: Payload reference only, must not modify header
 *	@pkt_type: Packet class
641 642
 *	@fclone: skbuff clone status
 *	@ipvs_property: skbuff is owned by ipvs
643 644 645
 *	@inner_protocol_type: whether the inner protocol is
 *		ENCAP_TYPE_ETHER or ENCAP_TYPE_IPPROTO
 *	@remcsum_offload: remote checksum offload is enabled
646 647
 *	@offload_fwd_mark: Packet was L2-forwarded in hardware
 *	@offload_l3_fwd_mark: Packet was L3-forwarded in hardware
648
 *	@tc_skip_classify: do not classify packet. set by IFB device
649
 *	@tc_at_ingress: used within tc_classify to distinguish in/egress
650 651
 *	@redirected: packet was redirected by packet classifier
 *	@from_ingress: packet was redirected from the ingress path
652 653
 *	@peeked: this packet has been seen already, so stats have been
 *		done for it, don't do them again
654
 *	@nf_trace: netfilter packet trace flag
655 656
 *	@protocol: Packet protocol from driver
 *	@destructor: Destruct function
657
 *	@tcp_tsorted_anchor: list structure for TCP (tp->tsorted_sent_queue)
658
 *	@_nfct: Associated connection, if any (with nfctinfo bits)
L
Linus Torvalds 已提交
659
 *	@nf_bridge: Saved data about a bridged frame - see br_netfilter.c
660
 *	@skb_iif: ifindex of device we arrived on
L
Linus Torvalds 已提交
661
 *	@tc_index: Traffic control index
662
 *	@hash: the packet hash
663
 *	@queue_mapping: Queue mapping for multiqueue devices
664 665
 *	@head_frag: skb was allocated from page fragments,
 *		not allocated by kmalloc() or vmalloc().
666
 *	@pfmemalloc: skbuff was allocated from PFMEMALLOC reserves
667
 *	@active_extensions: active extensions (skb_ext_id types)
668
 *	@ndisc_nodetype: router type (from link layer)
669
 *	@ooo_okay: allow the mapping of a socket to a queue to be changed
670
 *	@l4_hash: indicate hash is a canonical 4-tuple hash over transport
671
 *		ports.
672
 *	@sw_hash: indicates hash was computed in software stack
673 674
 *	@wifi_acked_valid: wifi_acked was set
 *	@wifi_acked: whether frame was acked on wifi or not
675
 *	@no_fcs:  Request NIC to treat last 4 bytes as Ethernet FCS
676 677 678
 *	@encapsulation: indicates the inner headers in the skbuff are valid
 *	@encap_hdr_csum: software checksum is needed
 *	@csum_valid: checksum is already valid
679
 *	@csum_not_inet: use CRC32c to resolve CHECKSUM_PARTIAL
680 681 682 683
 *	@csum_complete_sw: checksum was completed by software
 *	@csum_level: indicates the number of consecutive checksums found in
 *		the packet minus one that have been verified as
 *		CHECKSUM_UNNECESSARY (max 3)
684
 *	@dst_pending_confirm: need to confirm neighbour
685
 *	@decrypted: Decrypted SKB
686
 *	@napi_id: id of the NAPI struct this skb came from
687
 *	@sender_cpu: (aka @napi_id) source CPU in XPS
688
 *	@secmark: security marking
689
 *	@mark: Generic packet mark
690 691 692
 *	@reserved_tailroom: (aka @mark) number of bytes of free space available
 *		at the tail of an sk_buff
 *	@vlan_present: VLAN tag is present
693
 *	@vlan_proto: vlan encapsulation protocol
694
 *	@vlan_tci: vlan tag control information
S
Simon Horman 已提交
695
 *	@inner_protocol: Protocol (encapsulation)
696 697
 *	@inner_ipproto: (aka @inner_protocol) stores ipproto when
 *		skb->inner_protocol_type == ENCAP_TYPE_IPPROTO;
698 699
 *	@inner_transport_header: Inner transport layer header (encapsulation)
 *	@inner_network_header: Network layer header (encapsulation)
700
 *	@inner_mac_header: Link layer header (encapsulation)
701 702 703 704 705 706 707 708 709
 *	@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
710
 *	@extensions: allocated extensions, valid if active_extensions is nonzero
L
Linus Torvalds 已提交
711 712 713
 */

struct sk_buff {
714
	union {
E
Eric Dumazet 已提交
715 716 717 718 719 720
		struct {
			/* These two members must be first. */
			struct sk_buff		*next;
			struct sk_buff		*prev;

			union {
E
Eric Dumazet 已提交
721 722 723 724 725 726
				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;
E
Eric Dumazet 已提交
727 728
			};
		};
729
		struct rb_node		rbnode; /* used in netem, ip4 defrag, and tcp stack */
730
		struct list_head	list;
731
	};
732 733 734 735 736

	union {
		struct sock		*sk;
		int			ip_defrag_offset;
	};
L
Linus Torvalds 已提交
737

738
	union {
E
Eric Dumazet 已提交
739
		ktime_t		tstamp;
740
		u64		skb_mstamp_ns; /* earliest departure time */
741
	};
L
Linus Torvalds 已提交
742 743 744 745 746 747
	/*
	 * 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.
	 */
748
	char			cb[48] __aligned(8);
L
Linus Torvalds 已提交
749

750 751 752 753 754 755 756 757
	union {
		struct {
			unsigned long	_skb_refdst;
			void		(*destructor)(struct sk_buff *skb);
		};
		struct list_head	tcp_tsorted_anchor;
	};

758
#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
759
	unsigned long		 _nfct;
760
#endif
L
Linus Torvalds 已提交
761
	unsigned int		len,
762 763 764
				data_len;
	__u16			mac_len,
				hdr_len;
765 766 767 768 769

	/* Following fields are _not_ copied in __copy_skb_header()
	 * Note that queue_mapping is here mostly to fill a hole.
	 */
	__u16			queue_mapping;
770 771 772 773 774 775 776 777 778

/* 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)

779
	/* private: */
780
	__u8			__cloned_offset[0];
781
	/* public: */
782
	__u8			cloned:1,
783
				nohdr:1,
784
				fclone:2,
785
				peeked:1,
786
				head_frag:1,
787
				pfmemalloc:1;
788 789 790
#ifdef CONFIG_SKB_EXTENSIONS
	__u8			active_extensions;
#endif
791 792 793
	/* fields enclosed in headers_start/headers_end are copied
	 * using a single memcpy() in __copy_skb_header()
	 */
794
	/* private: */
795
	__u32			headers_start[0];
796
	/* public: */
797

798 799 800 801 802
/* 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
803
#endif
804
#define PKT_TYPE_OFFSET()	offsetof(struct sk_buff, __pkt_type_offset)
805

806
	/* private: */
807
	__u8			__pkt_type_offset[0];
808
	/* public: */
809 810 811 812
	__u8			pkt_type:3;
	__u8			ignore_df:1;
	__u8			nf_trace:1;
	__u8			ip_summed:2;
813
	__u8			ooo_okay:1;
814

815
	__u8			l4_hash:1;
816
	__u8			sw_hash:1;
817 818
	__u8			wifi_acked_valid:1;
	__u8			wifi_acked:1;
819
	__u8			no_fcs:1;
820
	/* Indicates the inner headers are valid in the skbuff. */
821
	__u8			encapsulation:1;
822
	__u8			encap_hdr_csum:1;
823
	__u8			csum_valid:1;
824

M
Michał Mirosław 已提交
825 826 827 828 829 830
#ifdef __BIG_ENDIAN_BITFIELD
#define PKT_VLAN_PRESENT_BIT	7
#else
#define PKT_VLAN_PRESENT_BIT	0
#endif
#define PKT_VLAN_PRESENT_OFFSET()	offsetof(struct sk_buff, __pkt_vlan_present_offset)
831
	/* private: */
M
Michał Mirosław 已提交
832
	__u8			__pkt_vlan_present_offset[0];
833
	/* public: */
M
Michał Mirosław 已提交
834
	__u8			vlan_present:1;
835
	__u8			csum_complete_sw:1;
836
	__u8			csum_level:2;
837
	__u8			csum_not_inet:1;
838
	__u8			dst_pending_confirm:1;
839 840 841
#ifdef CONFIG_IPV6_NDISC_NODETYPE
	__u8			ndisc_nodetype:2;
#endif
842

M
Michał Mirosław 已提交
843
	__u8			ipvs_property:1;
T
Tom Herbert 已提交
844
	__u8			inner_protocol_type:1;
845
	__u8			remcsum_offload:1;
846 847
#ifdef CONFIG_NET_SWITCHDEV
	__u8			offload_fwd_mark:1;
848
	__u8			offload_l3_fwd_mark:1;
849
#endif
850 851
#ifdef CONFIG_NET_CLS_ACT
	__u8			tc_skip_classify:1;
852
	__u8			tc_at_ingress:1;
853 854 855 856
#endif
#ifdef CONFIG_NET_REDIRECT
	__u8			redirected:1;
	__u8			from_ingress:1;
857
#endif
858 859 860
#ifdef CONFIG_TLS_DEVICE
	__u8			decrypted:1;
#endif
861 862 863 864

#ifdef CONFIG_NET_SCHED
	__u16			tc_index;	/* traffic control index */
#endif
865

866 867 868 869 870 871 872 873 874 875 876 877
	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
878 879 880 881 882
#if defined(CONFIG_NET_RX_BUSY_POLL) || defined(CONFIG_XPS)
	union {
		unsigned int	napi_id;
		unsigned int	sender_cpu;
	};
883
#endif
884
#ifdef CONFIG_NETWORK_SECMARK
885
	__u32		secmark;
886 887
#endif

888 889
	union {
		__u32		mark;
E
Eric Dumazet 已提交
890
		__u32		reserved_tailroom;
891
	};
L
Linus Torvalds 已提交
892

T
Tom Herbert 已提交
893 894 895 896 897
	union {
		__be16		inner_protocol;
		__u8		inner_ipproto;
	};

898 899 900
	__u16			inner_transport_header;
	__u16			inner_network_header;
	__u16			inner_mac_header;
901 902

	__be16			protocol;
903 904 905
	__u16			transport_header;
	__u16			network_header;
	__u16			mac_header;
906

907
	/* private: */
908
	__u32			headers_end[0];
909
	/* public: */
910

L
Linus Torvalds 已提交
911
	/* These elements must be at the end, see alloc_skb() for details.  */
912
	sk_buff_data_t		tail;
913
	sk_buff_data_t		end;
L
Linus Torvalds 已提交
914
	unsigned char		*head,
915
				*data;
916
	unsigned int		truesize;
917
	refcount_t		users;
918 919 920 921 922

#ifdef CONFIG_SKB_EXTENSIONS
	/* only useable after checking ->active_extensions != 0 */
	struct skb_ext		*extensions;
#endif
L
Linus Torvalds 已提交
923 924 925 926 927 928 929
};

#ifdef __KERNEL__
/*
 *	Handling routines are only of interest to the kernel
 */

930 931
#define SKB_ALLOC_FCLONE	0x01
#define SKB_ALLOC_RX		0x02
932
#define SKB_ALLOC_NAPI		0x04
933

934 935 936 937
/**
 * skb_pfmemalloc - Test if the skb was allocated from PFMEMALLOC reserves
 * @skb: buffer
 */
938 939 940 941 942
static inline bool skb_pfmemalloc(const struct sk_buff *skb)
{
	return unlikely(skb->pfmemalloc);
}

E
Eric Dumazet 已提交
943 944 945 946 947 948 949 950 951 952 953 954 955
/*
 * 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)

/**
 * skb_dst - returns skb dst_entry
 * @skb: buffer
 *
 * Returns skb dst_entry, regardless of reference taken or not.
 */
E
Eric Dumazet 已提交
956 957
static inline struct dst_entry *skb_dst(const struct sk_buff *skb)
{
958
	/* If refdst was not refcounted, check we still are in a
E
Eric Dumazet 已提交
959 960 961 962 963 964
	 * 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
965 966
}

E
Eric Dumazet 已提交
967 968 969 970 971 972 973 974
/**
 * 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
975 976
static inline void skb_dst_set(struct sk_buff *skb, struct dst_entry *dst)
{
E
Eric Dumazet 已提交
977 978 979
	skb->_skb_refdst = (unsigned long)dst;
}

980 981 982 983 984 985 986 987 988 989 990 991
/**
 * 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)
{
992 993
	WARN_ON(!rcu_read_lock_held() && !rcu_read_lock_bh_held());
	skb->_skb_refdst = (unsigned long)dst | SKB_DST_NOREF;
994
}
E
Eric Dumazet 已提交
995 996

/**
L
Lucas De Marchi 已提交
997
 * skb_dst_is_noref - Test if skb dst isn't refcounted
E
Eric Dumazet 已提交
998 999 1000 1001 1002
 * @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
1003 1004
}

1005 1006 1007 1008
/**
 * skb_rtable - Returns the skb &rtable
 * @skb: buffer
 */
E
Eric Dumazet 已提交
1009 1010
static inline struct rtable *skb_rtable(const struct sk_buff *skb)
{
E
Eric Dumazet 已提交
1011
	return (struct rtable *)skb_dst(skb);
E
Eric Dumazet 已提交
1012 1013
}

1014 1015 1016 1017 1018 1019 1020 1021 1022
/* 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;
}

1023 1024 1025 1026
/**
 * skb_napi_id - Returns the skb's NAPI id
 * @skb: buffer
 */
1027 1028 1029 1030 1031 1032 1033 1034 1035
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
}

1036 1037 1038 1039 1040 1041
/**
 * skb_unref - decrement the skb's reference count
 * @skb: buffer
 *
 * Returns true if we can free the skb.
 */
1042 1043 1044 1045
static inline bool skb_unref(struct sk_buff *skb)
{
	if (unlikely(!skb))
		return false;
1046
	if (likely(refcount_read(&skb->users) == 1))
1047
		smp_rmb();
1048
	else if (likely(!refcount_dec_and_test(&skb->users)))
1049 1050 1051 1052 1053
		return false;

	return true;
}

P
Paolo Abeni 已提交
1054
void skb_release_head_state(struct sk_buff *skb);
1055 1056
void kfree_skb(struct sk_buff *skb);
void kfree_skb_list(struct sk_buff *segs);
1057
void skb_dump(const char *level, const struct sk_buff *skb, bool full_pkt);
1058
void skb_tx_error(struct sk_buff *skb);
1059 1060

#ifdef CONFIG_TRACEPOINTS
1061
void consume_skb(struct sk_buff *skb);
1062 1063 1064 1065 1066 1067 1068
#else
static inline void consume_skb(struct sk_buff *skb)
{
	return kfree_skb(skb);
}
#endif

1069
void __consume_stateless_skb(struct sk_buff *skb);
1070
void  __kfree_skb(struct sk_buff *skb);
1071
extern struct kmem_cache *skbuff_head_cache;
E
Eric Dumazet 已提交
1072

1073 1074 1075
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
1076

1077 1078
struct sk_buff *__alloc_skb(unsigned int size, gfp_t priority, int flags,
			    int node);
E
Eric Dumazet 已提交
1079
struct sk_buff *__build_skb(void *data, unsigned int frag_size);
1080
struct sk_buff *build_skb(void *data, unsigned int frag_size);
1081 1082
struct sk_buff *build_skb_around(struct sk_buff *skb,
				 void *data, unsigned int frag_size);
1083 1084 1085 1086 1087 1088 1089 1090

/**
 * alloc_skb - allocate a network buffer
 * @size: size to allocate
 * @priority: allocation mask
 *
 * This function is a convenient wrapper around __alloc_skb().
 */
1091
static inline struct sk_buff *alloc_skb(unsigned int size,
A
Al Viro 已提交
1092
					gfp_t priority)
1093
{
E
Eric Dumazet 已提交
1094
	return __alloc_skb(size, priority, 0, NUMA_NO_NODE);
1095 1096
}

1097 1098 1099 1100 1101
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);
1102
struct sk_buff *alloc_skb_for_msg(struct sk_buff *first);
1103

1104 1105 1106 1107 1108 1109
/* Layout of fast clones : [skb1][skb2][fclone_ref] */
struct sk_buff_fclones {
	struct sk_buff	skb1;

	struct sk_buff	skb2;

1110
	refcount_t	fclone_ref;
1111 1112 1113 1114
};

/**
 *	skb_fclone_busy - check if fclone is busy
1115
 *	@sk: socket
1116 1117
 *	@skb: buffer
 *
M
Masanari Iida 已提交
1118
 * Returns true if skb is a fast clone, and its clone is not freed.
1119 1120
 * Some drivers call skb_orphan() in their ndo_start_xmit(),
 * so we also check that this didnt happen.
1121
 */
1122 1123
static inline bool skb_fclone_busy(const struct sock *sk,
				   const struct sk_buff *skb)
1124 1125 1126 1127 1128 1129
{
	const struct sk_buff_fclones *fclones;

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

	return skb->fclone == SKB_FCLONE_ORIG &&
1130
	       refcount_read(&fclones->fclone_ref) > 1 &&
1131
	       fclones->skb2.sk == sk;
1132 1133
}

1134 1135 1136 1137 1138 1139 1140
/**
 * alloc_skb_fclone - allocate a network buffer from fclone cache
 * @size: size to allocate
 * @priority: allocation mask
 *
 * This function is a convenient wrapper around __alloc_skb().
 */
1141
static inline struct sk_buff *alloc_skb_fclone(unsigned int size,
A
Al Viro 已提交
1142
					       gfp_t priority)
1143
{
1144
	return __alloc_skb(size, priority, SKB_ALLOC_FCLONE, NUMA_NO_NODE);
1145 1146
}

1147
struct sk_buff *skb_morph(struct sk_buff *dst, struct sk_buff *src);
1148
void skb_headers_offset_update(struct sk_buff *skb, int off);
1149 1150
int skb_copy_ubufs(struct sk_buff *skb, gfp_t gfp_mask);
struct sk_buff *skb_clone(struct sk_buff *skb, gfp_t priority);
1151
void skb_copy_header(struct sk_buff *new, const struct sk_buff *old);
1152
struct sk_buff *skb_copy(const struct sk_buff *skb, gfp_t priority);
1153 1154 1155 1156 1157 1158 1159
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);
}
1160 1161 1162 1163 1164 1165

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);
1166 1167 1168 1169
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);
1170
int skb_cow_data(struct sk_buff *skb, int tailbits, struct sk_buff **trailer);
1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187
int __skb_pad(struct sk_buff *skb, int pad, bool free_on_error);

/**
 *	skb_pad			-	zero pad the tail of an skb
 *	@skb: buffer to pad
 *	@pad: space to pad
 *
 *	Ensure that a buffer is followed by a padding area that is zero
 *	filled. Used by network drivers which may DMA or transfer data
 *	beyond the buffer end onto the wire.
 *
 *	May return error in out of memory cases. The skb is freed on error.
 */
static inline int skb_pad(struct sk_buff *skb, int pad)
{
	return __skb_pad(skb, pad, true);
}
1188
#define dev_kfree_skb(a)	consume_skb(a)
L
Linus Torvalds 已提交
1189

1190 1191 1192
int skb_append_pagefrags(struct sk_buff *skb, struct page *page,
			 int offset, size_t size);

E
Eric Dumazet 已提交
1193
struct skb_seq_state {
1194 1195 1196 1197 1198 1199 1200 1201 1202
	__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;
};

1203 1204 1205 1206 1207
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);
1208

1209
unsigned int skb_find_text(struct sk_buff *skb, unsigned int from,
1210
			   unsigned int to, struct ts_config *config);
1211

T
Tom Herbert 已提交
1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244
/*
 * 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 */
};

1245
static inline void skb_clear_hash(struct sk_buff *skb)
T
Tom Herbert 已提交
1246
{
1247
	skb->hash = 0;
1248
	skb->sw_hash = 0;
1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262
	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;
1263
	skb->hash = hash;
T
Tom Herbert 已提交
1264 1265
}

1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278
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);
}

1279
void __skb_get_hash(struct sk_buff *skb);
1280
u32 __skb_get_hash_symmetric(const struct sk_buff *skb);
1281 1282
u32 skb_get_poff(const struct sk_buff *skb);
u32 __skb_get_poff(const struct sk_buff *skb, void *data,
1283
		   const struct flow_keys_basic *keys, int hlen);
1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296
__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);

1297 1298
struct bpf_flow_dissector;
bool bpf_flow_dissect(struct bpf_prog *prog, struct bpf_flow_dissector *ctx,
1299
		      __be16 proto, int nhoff, int hlen, unsigned int flags);
1300

1301 1302
bool __skb_flow_dissect(const struct net *net,
			const struct sk_buff *skb,
1303 1304
			struct flow_dissector *flow_dissector,
			void *target_container,
1305 1306
			void *data, __be16 proto, int nhoff, int hlen,
			unsigned int flags);
1307 1308 1309

static inline bool skb_flow_dissect(const struct sk_buff *skb,
				    struct flow_dissector *flow_dissector,
1310
				    void *target_container, unsigned int flags)
1311
{
1312 1313
	return __skb_flow_dissect(NULL, skb, flow_dissector,
				  target_container, NULL, 0, 0, 0, flags);
1314 1315 1316
}

static inline bool skb_flow_dissect_flow_keys(const struct sk_buff *skb,
1317 1318
					      struct flow_keys *flow,
					      unsigned int flags)
1319 1320
{
	memset(flow, 0, sizeof(*flow));
1321 1322
	return __skb_flow_dissect(NULL, skb, &flow_keys_dissector,
				  flow, NULL, 0, 0, 0, flags);
1323 1324
}

1325
static inline bool
1326 1327
skb_flow_dissect_flow_keys_basic(const struct net *net,
				 const struct sk_buff *skb,
1328 1329 1330
				 struct flow_keys_basic *flow, void *data,
				 __be16 proto, int nhoff, int hlen,
				 unsigned int flags)
1331 1332
{
	memset(flow, 0, sizeof(*flow));
1333
	return __skb_flow_dissect(net, skb, &flow_keys_basic_dissector, flow,
1334
				  data, proto, nhoff, hlen, flags);
1335 1336
}

1337 1338 1339 1340
void skb_flow_dissect_meta(const struct sk_buff *skb,
			   struct flow_dissector *flow_dissector,
			   void *target_container);

1341
/* Gets a skb connection tracking info, ctinfo map should be a
1342
 * map of mapsize to translate enum ip_conntrack_info states
1343 1344 1345 1346 1347 1348 1349 1350
 * to user states.
 */
void
skb_flow_dissect_ct(const struct sk_buff *skb,
		    struct flow_dissector *flow_dissector,
		    void *target_container,
		    u16 *ctinfo_map,
		    size_t mapsize);
1351 1352 1353 1354 1355
void
skb_flow_dissect_tunnel_info(const struct sk_buff *skb,
			     struct flow_dissector *flow_dissector,
			     void *target_container);

1356 1357 1358 1359
void skb_flow_dissect_hash(const struct sk_buff *skb,
			   struct flow_dissector *flow_dissector,
			   void *target_container);

1360
static inline __u32 skb_get_hash(struct sk_buff *skb)
1361
{
1362
	if (!skb->l4_hash && !skb->sw_hash)
1363
		__skb_get_hash(skb);
1364

1365
	return skb->hash;
1366 1367
}

1368
static inline __u32 skb_get_hash_flowi6(struct sk_buff *skb, const struct flowi6 *fl6)
1369
{
1370 1371
	if (!skb->l4_hash && !skb->sw_hash) {
		struct flow_keys keys;
1372
		__u32 hash = __get_hash_from_flowi6(fl6, &keys);
1373

1374
		__skb_set_sw_hash(skb, hash, flow_keys_have_l4(&keys));
1375
	}
1376 1377 1378 1379

	return skb->hash;
}

1380 1381
__u32 skb_get_hash_perturb(const struct sk_buff *skb,
			   const siphash_key_t *perturb);
T
Tom Herbert 已提交
1382

T
Tom Herbert 已提交
1383 1384
static inline __u32 skb_get_hash_raw(const struct sk_buff *skb)
{
1385
	return skb->hash;
T
Tom Herbert 已提交
1386 1387
}

1388 1389
static inline void skb_copy_hash(struct sk_buff *to, const struct sk_buff *from)
{
1390
	to->hash = from->hash;
1391
	to->sw_hash = from->sw_hash;
1392
	to->l4_hash = from->l4_hash;
1393 1394
};

1395 1396 1397 1398 1399 1400 1401 1402
static inline void skb_copy_decrypted(struct sk_buff *to,
				      const struct sk_buff *from)
{
#ifdef CONFIG_TLS_DEVICE
	to->decrypted = from->decrypted;
#endif
}

1403 1404 1405 1406 1407
#ifdef NET_SKBUFF_DATA_USES_OFFSET
static inline unsigned char *skb_end_pointer(const struct sk_buff *skb)
{
	return skb->head + skb->end;
}
1408 1409 1410 1411 1412

static inline unsigned int skb_end_offset(const struct sk_buff *skb)
{
	return skb->end;
}
1413 1414 1415 1416 1417
#else
static inline unsigned char *skb_end_pointer(const struct sk_buff *skb)
{
	return skb->end;
}
1418 1419 1420 1421 1422

static inline unsigned int skb_end_offset(const struct sk_buff *skb)
{
	return skb->end - skb->head;
}
1423 1424
#endif

L
Linus Torvalds 已提交
1425
/* Internal */
1426
#define skb_shinfo(SKB)	((struct skb_shared_info *)(skb_end_pointer(SKB)))
L
Linus Torvalds 已提交
1427

1428 1429 1430 1431 1432
static inline struct skb_shared_hwtstamps *skb_hwtstamps(struct sk_buff *skb)
{
	return &skb_shinfo(skb)->hwtstamps;
}

W
Willem de Bruijn 已提交
1433 1434 1435 1436 1437 1438 1439
static inline struct ubuf_info *skb_zcopy(struct sk_buff *skb)
{
	bool is_zcopy = skb && skb_shinfo(skb)->tx_flags & SKBTX_DEV_ZEROCOPY;

	return is_zcopy ? skb_uarg(skb) : NULL;
}

1440 1441
static inline void skb_zcopy_set(struct sk_buff *skb, struct ubuf_info *uarg,
				 bool *have_ref)
W
Willem de Bruijn 已提交
1442 1443
{
	if (skb && uarg && !skb_zcopy(skb)) {
1444 1445 1446 1447
		if (unlikely(have_ref && *have_ref))
			*have_ref = false;
		else
			sock_zerocopy_get(uarg);
W
Willem de Bruijn 已提交
1448 1449 1450 1451 1452
		skb_shinfo(skb)->destructor_arg = uarg;
		skb_shinfo(skb)->tx_flags |= SKBTX_ZEROCOPY_FRAG;
	}
}

1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468
static inline void skb_zcopy_set_nouarg(struct sk_buff *skb, void *val)
{
	skb_shinfo(skb)->destructor_arg = (void *)((uintptr_t) val | 0x1UL);
	skb_shinfo(skb)->tx_flags |= SKBTX_ZEROCOPY_FRAG;
}

static inline bool skb_zcopy_is_nouarg(struct sk_buff *skb)
{
	return (uintptr_t) skb_shinfo(skb)->destructor_arg & 0x1UL;
}

static inline void *skb_zcopy_get_nouarg(struct sk_buff *skb)
{
	return (void *)((uintptr_t) skb_shinfo(skb)->destructor_arg & ~0x1UL);
}

W
Willem de Bruijn 已提交
1469 1470 1471 1472 1473 1474
/* Release a reference on a zerocopy structure */
static inline void skb_zcopy_clear(struct sk_buff *skb, bool zerocopy)
{
	struct ubuf_info *uarg = skb_zcopy(skb);

	if (uarg) {
1475 1476 1477
		if (skb_zcopy_is_nouarg(skb)) {
			/* no notification callback */
		} else if (uarg->callback == sock_zerocopy_callback) {
1478 1479
			uarg->zerocopy = uarg->zerocopy && zerocopy;
			sock_zerocopy_put(uarg);
1480
		} else {
1481 1482 1483
			uarg->callback(uarg, zerocopy);
		}

W
Willem de Bruijn 已提交
1484 1485 1486 1487 1488 1489 1490 1491 1492 1493
		skb_shinfo(skb)->tx_flags &= ~SKBTX_ZEROCOPY_FRAG;
	}
}

/* Abort a zerocopy operation and revert zckey on error in send syscall */
static inline void skb_zcopy_abort(struct sk_buff *skb)
{
	struct ubuf_info *uarg = skb_zcopy(skb);

	if (uarg) {
1494
		sock_zerocopy_put_abort(uarg, false);
W
Willem de Bruijn 已提交
1495 1496 1497 1498
		skb_shinfo(skb)->tx_flags &= ~SKBTX_ZEROCOPY_FRAG;
	}
}

1499 1500 1501 1502 1503
static inline void skb_mark_not_on_list(struct sk_buff *skb)
{
	skb->next = NULL;
}

1504
/* Iterate through singly-linked GSO fragments of an skb. */
1505 1506 1507
#define skb_list_walk_safe(first, skb, next_skb)                               \
	for ((skb) = (first), (next_skb) = (skb) ? (skb)->next : NULL; (skb);  \
	     (skb) = (next_skb), (next_skb) = (skb) ? (skb)->next : NULL)
1508

1509 1510 1511 1512 1513 1514
static inline void skb_list_del_init(struct sk_buff *skb)
{
	__list_del_entry(&skb->list);
	skb_mark_not_on_list(skb);
}

L
Linus Torvalds 已提交
1515 1516 1517 1518 1519 1520 1521 1522
/**
 *	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)
{
1523
	return list->next == (const struct sk_buff *) list;
L
Linus Torvalds 已提交
1524 1525
}

E
Eric Dumazet 已提交
1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538
/**
 *	skb_queue_empty_lockless - check if a queue is empty
 *	@list: queue head
 *
 *	Returns true if the queue is empty, false otherwise.
 *	This variant can be used in lockless contexts.
 */
static inline bool skb_queue_empty_lockless(const struct sk_buff_head *list)
{
	return READ_ONCE(list->next) == (const struct sk_buff *) list;
}


D
David S. Miller 已提交
1539 1540 1541 1542 1543 1544 1545 1546 1547 1548
/**
 *	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)
{
1549
	return skb->next == (const struct sk_buff *) list;
D
David S. Miller 已提交
1550 1551
}

1552 1553 1554 1555 1556 1557 1558 1559 1560 1561
/**
 *	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)
{
1562
	return skb->prev == (const struct sk_buff *) list;
1563 1564
}

D
David S. Miller 已提交
1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582
/**
 *	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;
}

1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600
/**
 *	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
1601 1602 1603 1604 1605 1606 1607 1608 1609
/**
 *	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)
{
1610
	refcount_inc(&skb->users);
L
Linus Torvalds 已提交
1611 1612 1613 1614
	return skb;
}

/*
1615
 * If users == 1, we are the only owner and can avoid redundant atomic changes.
L
Linus Torvalds 已提交
1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631
 */

/**
 *	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;
}

1632 1633
static inline int skb_unclone(struct sk_buff *skb, gfp_t pri)
{
1634
	might_sleep_if(gfpflags_allow_blocking(pri));
1635 1636 1637 1638 1639 1640 1641

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

	return 0;
}

L
Linus Torvalds 已提交
1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660
/**
 *	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;
}

1661 1662 1663 1664 1665 1666 1667 1668 1669 1670
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;
}

1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681
/**
 *	__skb_header_release - release reference to header
 *	@skb: buffer to operate on
 */
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
1682 1683 1684 1685 1686 1687 1688 1689 1690
/**
 *	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)
{
1691
	return refcount_read(&skb->users) != 1;
L
Linus Torvalds 已提交
1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706
}

/**
 *	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.
 */
1707
static inline struct sk_buff *skb_share_check(struct sk_buff *skb, gfp_t pri)
L
Linus Torvalds 已提交
1708
{
1709
	might_sleep_if(gfpflags_allow_blocking(pri));
L
Linus Torvalds 已提交
1710 1711
	if (skb_shared(skb)) {
		struct sk_buff *nskb = skb_clone(skb, pri);
1712 1713 1714 1715 1716

		if (likely(nskb))
			consume_skb(skb);
		else
			kfree_skb(skb);
L
Linus Torvalds 已提交
1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741
		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.
 */
1742
static inline struct sk_buff *skb_unshare(struct sk_buff *skb,
A
Al Viro 已提交
1743
					  gfp_t pri)
L
Linus Torvalds 已提交
1744
{
1745
	might_sleep_if(gfpflags_allow_blocking(pri));
L
Linus Torvalds 已提交
1746 1747
	if (skb_cloned(skb)) {
		struct sk_buff *nskb = skb_copy(skb, pri);
1748 1749 1750 1751 1752 1753

		/* Free our shared copy */
		if (likely(nskb))
			consume_skb(skb);
		else
			kfree_skb(skb);
L
Linus Torvalds 已提交
1754 1755 1756 1757 1758 1759
		skb = nskb;
	}
	return skb;
}

/**
1760
 *	skb_peek - peek at the head of an &sk_buff_head
L
Linus Torvalds 已提交
1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771
 *	@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.
 */
1772
static inline struct sk_buff *skb_peek(const struct sk_buff_head *list_)
L
Linus Torvalds 已提交
1773
{
1774 1775 1776 1777 1778
	struct sk_buff *skb = list_->next;

	if (skb == (struct sk_buff *)list_)
		skb = NULL;
	return skb;
L
Linus Torvalds 已提交
1779 1780
}

1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791
/**
 *	__skb_peek - peek at the head of a non-empty &sk_buff_head
 *	@list_: list to peek at
 *
 *	Like skb_peek(), but the caller knows that the list is not empty.
 */
static inline struct sk_buff *__skb_peek(const struct sk_buff_head *list_)
{
	return list_->next;
}

P
Pavel Emelyanov 已提交
1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804
/**
 *	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;
1805

P
Pavel Emelyanov 已提交
1806 1807 1808 1809 1810
	if (next == (struct sk_buff *)list_)
		next = NULL;
	return next;
}

L
Linus Torvalds 已提交
1811
/**
1812
 *	skb_peek_tail - peek at the tail of an &sk_buff_head
L
Linus Torvalds 已提交
1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823
 *	@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.
 */
1824
static inline struct sk_buff *skb_peek_tail(const struct sk_buff_head *list_)
L
Linus Torvalds 已提交
1825
{
1826
	struct sk_buff *skb = READ_ONCE(list_->prev);
1827 1828 1829 1830 1831

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

L
Linus Torvalds 已提交
1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844
}

/**
 *	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;
}

1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856
/**
 *	skb_queue_len_lockless	- get queue length
 *	@list_: list to measure
 *
 *	Return the length of an &sk_buff queue.
 *	This variant can be used in lockless contexts.
 */
static inline __u32 skb_queue_len_lockless(const struct sk_buff_head *list_)
{
	return READ_ONCE(list_->qlen);
}

1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872
/**
 *	__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;
}

1873 1874 1875 1876 1877 1878 1879 1880
/*
 * 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
1881 1882 1883
static inline void skb_queue_head_init(struct sk_buff_head *list)
{
	spin_lock_init(&list->lock);
1884
	__skb_queue_head_init(list);
L
Linus Torvalds 已提交
1885 1886
}

1887 1888 1889 1890 1891 1892 1893
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
1894
/*
1895
 *	Insert an sk_buff on a list.
L
Linus Torvalds 已提交
1896 1897 1898 1899
 *
 *	The "__skb_xxxx()" functions are the non-atomic ones that
 *	can only be called with interrupts disabled.
 */
1900 1901 1902 1903
static inline void __skb_insert(struct sk_buff *newsk,
				struct sk_buff *prev, struct sk_buff *next,
				struct sk_buff_head *list)
{
1904 1905 1906
	/* See skb_queue_empty_lockless() and skb_peek_tail()
	 * for the opposite READ_ONCE()
	 */
E
Eric Dumazet 已提交
1907 1908 1909 1910
	WRITE_ONCE(newsk->next, next);
	WRITE_ONCE(newsk->prev, prev);
	WRITE_ONCE(next->prev, newsk);
	WRITE_ONCE(prev->next, newsk);
1911 1912
	list->qlen++;
}
L
Linus Torvalds 已提交
1913

1914 1915 1916 1917 1918 1919 1920
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;

E
Eric Dumazet 已提交
1921 1922
	WRITE_ONCE(first->prev, prev);
	WRITE_ONCE(prev->next, first);
1923

E
Eric Dumazet 已提交
1924 1925
	WRITE_ONCE(last->next, next);
	WRITE_ONCE(next->prev, last);
1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937
}

/**
 *	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);
1938
		head->qlen += list->qlen;
1939 1940 1941 1942
	}
}

/**
E
Eric Dumazet 已提交
1943
 *	skb_queue_splice_init - join two skb lists and reinitialise the emptied list
1944 1945 1946 1947 1948 1949 1950 1951 1952 1953
 *	@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);
1954
		head->qlen += list->qlen;
1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968
		__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);
1969
		head->qlen += list->qlen;
1970 1971 1972 1973
	}
}

/**
E
Eric Dumazet 已提交
1974
 *	skb_queue_splice_tail_init - join two skb lists and reinitialise the emptied list
1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985
 *	@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);
1986
		head->qlen += list->qlen;
1987 1988 1989 1990
		__skb_queue_head_init(list);
	}
}

L
Linus Torvalds 已提交
1991
/**
1992
 *	__skb_queue_after - queue a buffer at the list head
L
Linus Torvalds 已提交
1993
 *	@list: list to use
1994
 *	@prev: place after this buffer
L
Linus Torvalds 已提交
1995 1996
 *	@newsk: buffer to queue
 *
1997
 *	Queue a buffer int the middle of a list. This function takes no locks
L
Linus Torvalds 已提交
1998 1999 2000 2001
 *	and you must therefore hold required locks before calling it.
 *
 *	A buffer cannot be placed on two lists at the same time.
 */
2002 2003 2004
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
2005
{
2006
	__skb_insert(newsk, prev, prev->next, list);
L
Linus Torvalds 已提交
2007 2008
}

2009 2010
void skb_append(struct sk_buff *old, struct sk_buff *newsk,
		struct sk_buff_head *list);
2011

2012 2013 2014 2015 2016 2017 2018
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);
}

2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033
/**
 *	__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.
 */
static inline void __skb_queue_head(struct sk_buff_head *list,
				    struct sk_buff *newsk)
{
	__skb_queue_after(list, (struct sk_buff *)list, newsk);
}
2034
void skb_queue_head(struct sk_buff_head *list, struct sk_buff *newsk);
2035

L
Linus Torvalds 已提交
2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048
/**
 *	__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.
 */
static inline void __skb_queue_tail(struct sk_buff_head *list,
				   struct sk_buff *newsk)
{
2049
	__skb_queue_before(list, (struct sk_buff *)list, newsk);
L
Linus Torvalds 已提交
2050
}
2051
void skb_queue_tail(struct sk_buff_head *list, struct sk_buff *newsk);
L
Linus Torvalds 已提交
2052 2053 2054 2055 2056

/*
 * remove sk_buff from list. _Must_ be called atomically, and with
 * the list known..
 */
2057
void skb_unlink(struct sk_buff *skb, struct sk_buff_head *list);
L
Linus Torvalds 已提交
2058 2059 2060 2061
static inline void __skb_unlink(struct sk_buff *skb, struct sk_buff_head *list)
{
	struct sk_buff *next, *prev;

2062
	WRITE_ONCE(list->qlen, list->qlen - 1);
L
Linus Torvalds 已提交
2063 2064 2065
	next	   = skb->next;
	prev	   = skb->prev;
	skb->next  = skb->prev = NULL;
E
Eric Dumazet 已提交
2066 2067
	WRITE_ONCE(next->prev, prev);
	WRITE_ONCE(prev->next, next);
L
Linus Torvalds 已提交
2068 2069
}

2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084
/**
 *	__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.
 */
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;
}
2085
struct sk_buff *skb_dequeue(struct sk_buff_head *list);
L
Linus Torvalds 已提交
2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101

/**
 *	__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.
 */
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;
}
2102
struct sk_buff *skb_dequeue_tail(struct sk_buff_head *list);
L
Linus Torvalds 已提交
2103 2104


2105
static inline bool skb_is_nonlinear(const struct sk_buff *skb)
L
Linus Torvalds 已提交
2106 2107 2108 2109 2110 2111 2112 2113 2114
{
	return skb->data_len;
}

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

2115
static inline unsigned int __skb_pagelen(const struct sk_buff *skb)
L
Linus Torvalds 已提交
2116
{
2117
	unsigned int i, len = 0;
L
Linus Torvalds 已提交
2118

2119
	for (i = skb_shinfo(skb)->nr_frags - 1; (int)i >= 0; i--)
E
Eric Dumazet 已提交
2120
		len += skb_frag_size(&skb_shinfo(skb)->frags[i]);
2121 2122 2123 2124 2125 2126
	return len;
}

static inline unsigned int skb_pagelen(const struct sk_buff *skb)
{
	return skb_headlen(skb) + __skb_pagelen(skb);
L
Linus Torvalds 已提交
2127 2128
}

2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143
/**
 * __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
2144 2145 2146
{
	skb_frag_t *frag = &skb_shinfo(skb)->frags[i];

2147
	/*
2148 2149 2150
	 * 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).
2151
	 */
2152
	frag->bv_page		  = page;
2153
	frag->bv_offset		  = off;
E
Eric Dumazet 已提交
2154
	skb_frag_size_set(frag, size);
2155 2156

	page = compound_head(page);
2157
	if (page_is_pfmemalloc(page))
2158
		skb->pfmemalloc	= true;
2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169
}

/**
 * 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
2170
 * @skb to point to @size bytes at offset @off within @page. In
2171 2172 2173 2174 2175 2176 2177 2178
 * 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
2179 2180 2181
	skb_shinfo(skb)->nr_frags = i + 1;
}

2182 2183
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
2184

J
Jason Wang 已提交
2185 2186 2187
void skb_coalesce_rx_frag(struct sk_buff *skb, int i, int size,
			  unsigned int truesize);

L
Linus Torvalds 已提交
2188 2189
#define SKB_LINEAR_ASSERT(skb)  BUG_ON(skb_is_nonlinear(skb))

2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205
#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;
}
2206

2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221
#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;
}
2222

2223 2224
#endif /* NET_SKBUFF_DATA_USES_OFFSET */

L
Linus Torvalds 已提交
2225 2226 2227
/*
 *	Add data to an sk_buff
 */
2228 2229 2230
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
2231
{
2232
	void *tmp = skb_tail_pointer(skb);
L
Linus Torvalds 已提交
2233 2234 2235 2236 2237 2238
	SKB_LINEAR_ASSERT(skb);
	skb->tail += len;
	skb->len  += len;
	return tmp;
}

2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260
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;
}

2261
static inline void *skb_put_zero(struct sk_buff *skb, unsigned int len)
2262
{
2263
	void *tmp = skb_put(skb, len);
2264 2265 2266 2267 2268 2269

	memset(tmp, 0, len);

	return tmp;
}

2270 2271 2272 2273 2274 2275 2276 2277 2278 2279
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;
}

2280 2281 2282 2283 2284
static inline void skb_put_u8(struct sk_buff *skb, u8 val)
{
	*(u8 *)skb_put(skb, 1) = val;
}

2285 2286
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
2287 2288 2289 2290 2291 2292
{
	skb->data -= len;
	skb->len  += len;
	return skb->data;
}

2293 2294
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
2295 2296 2297 2298 2299 2300
{
	skb->len -= len;
	BUG_ON(skb->len < skb->data_len);
	return skb->data += len;
}

2301
static inline void *skb_pull_inline(struct sk_buff *skb, unsigned int len)
2302 2303 2304 2305
{
	return unlikely(len > skb->len) ? NULL : __skb_pull(skb, len);
}

2306
void *__pskb_pull_tail(struct sk_buff *skb, int delta);
L
Linus Torvalds 已提交
2307

2308
static inline void *__pskb_pull(struct sk_buff *skb, unsigned int len)
L
Linus Torvalds 已提交
2309 2310
{
	if (len > skb_headlen(skb) &&
G
Gerrit Renker 已提交
2311
	    !__pskb_pull_tail(skb, len - skb_headlen(skb)))
L
Linus Torvalds 已提交
2312 2313 2314 2315 2316
		return NULL;
	skb->len -= len;
	return skb->data += len;
}

2317
static inline void *pskb_pull(struct sk_buff *skb, unsigned int len)
L
Linus Torvalds 已提交
2318 2319 2320 2321
{
	return unlikely(len > skb->len) ? NULL : __pskb_pull(skb, len);
}

2322
static inline bool pskb_may_pull(struct sk_buff *skb, unsigned int len)
L
Linus Torvalds 已提交
2323 2324
{
	if (likely(len <= skb_headlen(skb)))
2325
		return true;
L
Linus Torvalds 已提交
2326
	if (unlikely(len > skb->len))
2327
		return false;
G
Gerrit Renker 已提交
2328
	return __pskb_pull_tail(skb, len - skb_headlen(skb)) != NULL;
L
Linus Torvalds 已提交
2329 2330
}

2331 2332
void skb_condense(struct sk_buff *skb);

L
Linus Torvalds 已提交
2333 2334 2335 2336 2337 2338
/**
 *	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.
 */
2339
static inline unsigned int skb_headroom(const struct sk_buff *skb)
L
Linus Torvalds 已提交
2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351
{
	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)
{
2352
	return skb_is_nonlinear(skb) ? 0 : skb->end - skb->tail;
L
Linus Torvalds 已提交
2353 2354
}

2355 2356 2357 2358 2359 2360 2361 2362 2363
/**
 *	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
2364 2365 2366 2367
	if (skb_is_nonlinear(skb))
		return 0;

	return skb->end - skb->tail - skb->reserved_tailroom;
2368 2369
}

L
Linus Torvalds 已提交
2370 2371 2372 2373 2374 2375 2376 2377
/**
 *	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.
 */
2378
static inline void skb_reserve(struct sk_buff *skb, int len)
L
Linus Torvalds 已提交
2379 2380 2381 2382 2383
{
	skb->data += len;
	skb->tail += len;
}

2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407
/**
 *	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
2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424
#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;
}

2425 2426
static inline void skb_reset_inner_headers(struct sk_buff *skb)
{
2427
	skb->inner_mac_header = skb->mac_header;
2428 2429 2430 2431
	skb->inner_network_header = skb->network_header;
	skb->inner_transport_header = skb->transport_header;
}

2432 2433 2434 2435 2436
static inline void skb_reset_mac_len(struct sk_buff *skb)
{
	skb->mac_len = skb->network_header - skb->mac_header;
}

2437 2438 2439 2440 2441 2442
static inline unsigned char *skb_inner_transport_header(const struct sk_buff
							*skb)
{
	return skb->head + skb->inner_transport_header;
}

2443 2444 2445 2446 2447
static inline int skb_inner_transport_offset(const struct sk_buff *skb)
{
	return skb_inner_transport_header(skb) - skb->data;
}

2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476
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;
}

2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492
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;
}
2493 2494
static inline bool skb_transport_header_was_set(const struct sk_buff *skb)
{
C
Cong Wang 已提交
2495
	return skb->transport_header != (typeof(skb->transport_header))~0U;
2496 2497
}

2498 2499
static inline unsigned char *skb_transport_header(const struct sk_buff *skb)
{
2500
	return skb->head + skb->transport_header;
2501 2502
}

2503 2504
static inline void skb_reset_transport_header(struct sk_buff *skb)
{
2505
	skb->transport_header = skb->data - skb->head;
2506 2507
}

2508 2509 2510
static inline void skb_set_transport_header(struct sk_buff *skb,
					    const int offset)
{
2511 2512
	skb_reset_transport_header(skb);
	skb->transport_header += offset;
2513 2514
}

2515 2516
static inline unsigned char *skb_network_header(const struct sk_buff *skb)
{
2517
	return skb->head + skb->network_header;
2518 2519
}

2520 2521
static inline void skb_reset_network_header(struct sk_buff *skb)
{
2522
	skb->network_header = skb->data - skb->head;
2523 2524
}

2525 2526
static inline void skb_set_network_header(struct sk_buff *skb, const int offset)
{
2527 2528
	skb_reset_network_header(skb);
	skb->network_header += offset;
2529 2530
}

2531
static inline unsigned char *skb_mac_header(const struct sk_buff *skb)
2532
{
2533
	return skb->head + skb->mac_header;
2534 2535
}

2536 2537 2538 2539 2540
static inline int skb_mac_offset(const struct sk_buff *skb)
{
	return skb_mac_header(skb) - skb->data;
}

2541 2542 2543 2544 2545
static inline u32 skb_mac_header_len(const struct sk_buff *skb)
{
	return skb->network_header - skb->mac_header;
}

2546
static inline int skb_mac_header_was_set(const struct sk_buff *skb)
2547
{
C
Cong Wang 已提交
2548
	return skb->mac_header != (typeof(skb->mac_header))~0U;
2549 2550
}

2551 2552 2553 2554 2555
static inline void skb_unset_mac_header(struct sk_buff *skb)
{
	skb->mac_header = (typeof(skb->mac_header))~0U;
}

2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566
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;
}

2567 2568 2569 2570 2571
static inline void skb_pop_mac_header(struct sk_buff *skb)
{
	skb->mac_header = skb->network_header;
}

2572
static inline void skb_probe_transport_header(struct sk_buff *skb)
2573
{
2574
	struct flow_keys_basic keys;
2575 2576 2577

	if (skb_transport_header_was_set(skb))
		return;
2578

2579 2580
	if (skb_flow_dissect_flow_keys_basic(NULL, skb, &keys,
					     NULL, 0, 0, 0, 0))
2581
		skb_set_transport_header(skb, keys.control.thoff);
2582 2583
}

2584 2585 2586 2587 2588 2589 2590 2591 2592 2593
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);
	}
}

2594 2595 2596 2597 2598
static inline int skb_checksum_start_offset(const struct sk_buff *skb)
{
	return skb->csum_start - skb_headroom(skb);
}

2599 2600 2601 2602 2603
static inline unsigned char *skb_checksum_start(const struct sk_buff *skb)
{
	return skb->head + skb->csum_start;
}

2604 2605 2606 2607 2608 2609 2610 2611 2612 2613
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;
}

2614 2615 2616 2617 2618
static inline u32 skb_inner_network_header_len(const struct sk_buff *skb)
{
	return skb->inner_transport_header - skb->inner_network_header;
}

2619 2620 2621 2622
static inline int skb_network_offset(const struct sk_buff *skb)
{
	return skb_network_header(skb) - skb->data;
}
2623

2624 2625 2626 2627 2628
static inline int skb_inner_network_offset(const struct sk_buff *skb)
{
	return skb_inner_network_header(skb) - skb->data;
}

2629 2630 2631 2632 2633
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
2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644
/*
 * 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:
 *
2645
 * skb_reserve(skb, NET_IP_ALIGN);
L
Linus Torvalds 已提交
2646 2647 2648 2649
 *
 * 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.
2650
 *
L
Linus Torvalds 已提交
2651 2652 2653 2654 2655 2656 2657
 * 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

2658 2659 2660 2661
/*
 * 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
2662
 * 32 bytes or less we avoid the reallocation.
2663 2664 2665 2666 2667 2668 2669
 *
 * 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.
 *
2670
 * Various parts of the networking layer expect at least 32 bytes of
2671
 * headroom, you should not reduce this.
2672 2673 2674
 *
 * Using max(32, L1_CACHE_BYTES) makes sense (especially with RPS)
 * to reduce average number of cache lines per packet.
M
Miaohe Lin 已提交
2675
 * get_rps_cpu() for example only access one 64 bytes aligned block :
2676
 * NET_IP_ALIGN(2) + ethernet_header(14) + IP_header(20/40) + ports(8)
2677 2678
 */
#ifndef NET_SKB_PAD
2679
#define NET_SKB_PAD	max(32, L1_CACHE_BYTES)
2680 2681
#endif

2682
int ___pskb_trim(struct sk_buff *skb, unsigned int len);
L
Linus Torvalds 已提交
2683

2684
static inline void __skb_set_length(struct sk_buff *skb, unsigned int len)
L
Linus Torvalds 已提交
2685
{
2686
	if (WARN_ON(skb_is_nonlinear(skb)))
2687
		return;
2688 2689
	skb->len = len;
	skb_set_tail_pointer(skb, len);
L
Linus Torvalds 已提交
2690 2691
}

2692 2693 2694 2695 2696
static inline void __skb_trim(struct sk_buff *skb, unsigned int len)
{
	__skb_set_length(skb, len);
}

2697
void skb_trim(struct sk_buff *skb, unsigned int len);
L
Linus Torvalds 已提交
2698 2699 2700

static inline int __pskb_trim(struct sk_buff *skb, unsigned int len)
{
2701 2702 2703 2704
	if (skb->data_len)
		return ___pskb_trim(skb, len);
	__skb_trim(skb, len);
	return 0;
L
Linus Torvalds 已提交
2705 2706 2707 2708 2709 2710 2711
}

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

2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726
/**
 *	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);
}

2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740
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
2741 2742 2743 2744 2745 2746 2747 2748 2749 2750
/**
 *	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
2751
	if (skb->destructor) {
L
Linus Torvalds 已提交
2752
		skb->destructor(skb);
E
Eric Dumazet 已提交
2753 2754
		skb->destructor = NULL;
		skb->sk		= NULL;
2755 2756
	} else {
		BUG_ON(skb->sk);
E
Eric Dumazet 已提交
2757
	}
L
Linus Torvalds 已提交
2758 2759
}

2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770
/**
 *	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)
{
W
Willem de Bruijn 已提交
2771 2772
	if (likely(!skb_zcopy(skb)))
		return 0;
2773 2774
	if (!skb_zcopy_is_nouarg(skb) &&
	    skb_uarg(skb)->callback == sock_zerocopy_callback)
W
Willem de Bruijn 已提交
2775 2776 2777 2778 2779 2780 2781 2782
		return 0;
	return skb_copy_ubufs(skb, gfp_mask);
}

/* Frags must be orphaned, even if refcounted, if skb might loop to rx path */
static inline int skb_orphan_frags_rx(struct sk_buff *skb, gfp_t gfp_mask)
{
	if (likely(!skb_zcopy(skb)))
2783 2784 2785 2786
		return 0;
	return skb_copy_ubufs(skb, gfp_mask);
}

L
Linus Torvalds 已提交
2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800
/**
 *	__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.
 */
static inline void __skb_queue_purge(struct sk_buff_head *list)
{
	struct sk_buff *skb;
	while ((skb = __skb_dequeue(list)) != NULL)
		kfree_skb(skb);
}
2801
void skb_queue_purge(struct sk_buff_head *list);
L
Linus Torvalds 已提交
2802

2803
unsigned int skb_rbtree_purge(struct rb_root *root);
2804

2805
void *netdev_alloc_frag(unsigned int fragsz);
L
Linus Torvalds 已提交
2806

2807 2808
struct sk_buff *__netdev_alloc_skb(struct net_device *dev, unsigned int length,
				   gfp_t gfp_mask);
2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823

/**
 *	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,
2824
					       unsigned int length)
2825 2826 2827 2828
{
	return __netdev_alloc_skb(dev, length, GFP_ATOMIC);
}

2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842
/* 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);
}


2843 2844
static inline struct sk_buff *__netdev_alloc_skb_ip_align(struct net_device *dev,
		unsigned int length, gfp_t gfp)
2845
{
2846
	struct sk_buff *skb = __netdev_alloc_skb(dev, length + NET_IP_ALIGN, gfp);
2847 2848 2849 2850 2851 2852

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

2853 2854 2855 2856 2857 2858
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);
}

2859 2860
static inline void skb_free_frag(void *addr)
{
2861
	page_frag_free(addr);
2862 2863
}

2864
void *napi_alloc_frag(unsigned int fragsz);
2865 2866 2867 2868 2869 2870 2871
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);
}
2872 2873 2874
void napi_consume_skb(struct sk_buff *skb, int budget);

void __kfree_skb_flush(void);
2875
void __kfree_skb_defer(struct sk_buff *skb);
2876

2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896
/**
 * __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.
	 */
M
Mel Gorman 已提交
2897
	gfp_mask |= __GFP_COMP | __GFP_MEMALLOC;
2898 2899 2900 2901 2902 2903

	return alloc_pages_node(NUMA_NO_NODE, gfp_mask, order);
}

static inline struct page *dev_alloc_pages(unsigned int order)
{
2904
	return __dev_alloc_pages(GFP_ATOMIC | __GFP_NOWARN, order);
2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921
}

/**
 * __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)
{
2922
	return dev_alloc_pages(0);
2923 2924
}

2925 2926 2927 2928 2929 2930 2931 2932
/**
 *	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)
{
2933
	if (page_is_pfmemalloc(page))
2934 2935 2936
		skb->pfmemalloc = true;
}

2937 2938 2939 2940 2941 2942
/**
 * skb_frag_off() - Returns the offset of a skb fragment
 * @frag: the paged fragment
 */
static inline unsigned int skb_frag_off(const skb_frag_t *frag)
{
2943
	return frag->bv_offset;
2944 2945 2946 2947 2948 2949 2950 2951 2952
}

/**
 * skb_frag_off_add() - Increments the offset of a skb fragment by @delta
 * @frag: skb fragment
 * @delta: value to add
 */
static inline void skb_frag_off_add(skb_frag_t *frag, int delta)
{
2953
	frag->bv_offset += delta;
2954 2955 2956 2957 2958 2959 2960 2961 2962
}

/**
 * skb_frag_off_set() - Sets the offset of a skb fragment
 * @frag: skb fragment
 * @offset: offset of fragment
 */
static inline void skb_frag_off_set(skb_frag_t *frag, unsigned int offset)
{
2963
	frag->bv_offset = offset;
2964 2965 2966 2967 2968 2969 2970 2971 2972 2973
}

/**
 * skb_frag_off_copy() - Sets the offset of a skb fragment from another fragment
 * @fragto: skb fragment where offset is set
 * @fragfrom: skb fragment offset is copied from
 */
static inline void skb_frag_off_copy(skb_frag_t *fragto,
				     const skb_frag_t *fragfrom)
{
2974
	fragto->bv_offset = fragfrom->bv_offset;
2975 2976
}

2977
/**
2978
 * skb_frag_page - retrieve the page referred to by a paged fragment
2979 2980 2981 2982 2983 2984
 * @frag: the paged fragment
 *
 * Returns the &struct page associated with @frag.
 */
static inline struct page *skb_frag_page(const skb_frag_t *frag)
{
2985
	return frag->bv_page;
2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042
}

/**
 * __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)
{
3043
	return page_address(skb_frag_page(frag)) + skb_frag_off(frag);
3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058
}

/**
 * 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;

3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070
	return ptr + skb_frag_off(frag);
}

/**
 * skb_frag_page_copy() - sets the page in a fragment from another fragment
 * @fragto: skb fragment where page is set
 * @fragfrom: skb fragment page is copied from
 */
static inline void skb_frag_page_copy(skb_frag_t *fragto,
				      const skb_frag_t *fragfrom)
{
	fragto->bv_page = fragfrom->bv_page;
3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081
}

/**
 * __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)
{
3082
	frag->bv_page = page;
3083 3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098
}

/**
 * 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
3099 3100
bool skb_page_frag_refill(unsigned int sz, struct page_frag *pfrag, gfp_t prio);

3101 3102
/**
 * skb_frag_dma_map - maps a paged fragment via the DMA API
3103
 * @dev: the device to map the fragment to
3104 3105 3106 3107
 * @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
3108
 * @dir: the direction of the mapping (``PCI_DMA_*``)
3109 3110 3111 3112 3113 3114 3115 3116 3117
 *
 * 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),
3118
			    skb_frag_off(frag) + offset, size, dir);
3119 3120
}

E
Eric Dumazet 已提交
3121 3122 3123 3124 3125 3126
static inline struct sk_buff *pskb_copy(struct sk_buff *skb,
					gfp_t gfp_mask)
{
	return __pskb_copy(skb, skb_headroom(skb), gfp_mask);
}

3127 3128 3129 3130 3131 3132 3133 3134

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);
}


3135 3136 3137 3138 3139 3140 3141 3142
/**
 *	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.
 */
3143
static inline int skb_clone_writable(const struct sk_buff *skb, unsigned int len)
3144 3145 3146 3147 3148
{
	return !skb_header_cloned(skb) &&
	       skb_headroom(skb) + len <= skb->hdr_len;
}

3149 3150 3151 3152 3153 3154 3155
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
3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169
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
3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183
/**
 *	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
3184 3185
	return __skb_cow(skb, headroom, skb_cloned(skb));
}
L
Linus Torvalds 已提交
3186

H
Herbert Xu 已提交
3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199
/**
 *	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
3200 3201 3202 3203 3204 3205 3206 3207 3208
}

/**
 *	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
3209 3210
 *	is untouched. Otherwise it is extended. Returns zero on
 *	success. The skb is freed on error.
L
Linus Torvalds 已提交
3211
 */
3212
static inline int skb_padto(struct sk_buff *skb, unsigned int len)
L
Linus Torvalds 已提交
3213 3214 3215
{
	unsigned int size = skb->len;
	if (likely(size >= len))
3216
		return 0;
G
Gerrit Renker 已提交
3217
	return skb_pad(skb, len - size);
L
Linus Torvalds 已提交
3218 3219
}

3220
/**
3221
 *	__skb_put_padto - increase size and pad an skbuff up to a minimal size
3222 3223
 *	@skb: buffer to pad
 *	@len: minimal length
3224
 *	@free_on_error: free buffer on error
3225 3226 3227 3228
 *
 *	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
3229
 *	success. The skb is freed on error if @free_on_error is true.
3230
 */
3231 3232 3233
static inline int __must_check __skb_put_padto(struct sk_buff *skb,
					       unsigned int len,
					       bool free_on_error)
3234 3235 3236 3237 3238
{
	unsigned int size = skb->len;

	if (unlikely(size < len)) {
		len -= size;
3239
		if (__skb_pad(skb, len, free_on_error))
3240 3241 3242 3243 3244 3245
			return -ENOMEM;
		__skb_put(skb, len);
	}
	return 0;
}

3246 3247 3248 3249 3250 3251 3252 3253 3254 3255
/**
 *	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.
 */
3256
static inline int __must_check skb_put_padto(struct sk_buff *skb, unsigned int len)
3257 3258 3259 3260
{
	return __skb_put_padto(skb, len, true);
}

L
Linus Torvalds 已提交
3261
static inline int skb_add_data(struct sk_buff *skb,
3262
			       struct iov_iter *from, int copy)
L
Linus Torvalds 已提交
3263 3264 3265 3266
{
	const int off = skb->len;

	if (skb->ip_summed == CHECKSUM_NONE) {
3267
		__wsum csum = 0;
3268 3269
		if (csum_and_copy_from_iter_full(skb_put(skb, copy), copy,
					         &csum, from)) {
L
Linus Torvalds 已提交
3270 3271 3272
			skb->csum = csum_block_add(skb->csum, csum, off);
			return 0;
		}
3273
	} else if (copy_from_iter_full(skb_put(skb, copy), copy, from))
L
Linus Torvalds 已提交
3274 3275 3276 3277 3278 3279
		return 0;

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

3280 3281
static inline bool skb_can_coalesce(struct sk_buff *skb, int i,
				    const struct page *page, int off)
L
Linus Torvalds 已提交
3282
{
W
Willem de Bruijn 已提交
3283 3284
	if (skb_zcopy(skb))
		return false;
L
Linus Torvalds 已提交
3285
	if (i) {
3286
		const skb_frag_t *frag = &skb_shinfo(skb)->frags[i - 1];
L
Linus Torvalds 已提交
3287

3288
		return page == skb_frag_page(frag) &&
3289
		       off == skb_frag_off(frag) + skb_frag_size(frag);
L
Linus Torvalds 已提交
3290
	}
3291
	return false;
L
Linus Torvalds 已提交
3292 3293
}

H
Herbert Xu 已提交
3294 3295 3296 3297 3298
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
3299 3300 3301 3302 3303 3304 3305
/**
 *	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
3306 3307 3308 3309 3310
static inline int skb_linearize(struct sk_buff *skb)
{
	return skb_is_nonlinear(skb) ? __skb_linearize(skb) : 0;
}

3311 3312 3313 3314 3315 3316 3317 3318 3319
/**
 * 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)
{
3320 3321
	return skb_is_nonlinear(skb) &&
	       skb_shinfo(skb)->tx_flags & SKBTX_SHARED_FRAG;
3322 3323
}

H
Herbert Xu 已提交
3324 3325 3326 3327 3328 3329 3330 3331
/**
 *	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
3332
{
H
Herbert Xu 已提交
3333 3334
	return skb_is_nonlinear(skb) || skb_cloned(skb) ?
	       __skb_linearize(skb) : 0;
L
Linus Torvalds 已提交
3335 3336
}

3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348
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
3349 3350 3351 3352 3353 3354 3355
/**
 *	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
3356 3357
 *	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
3358 3359
 */
static inline void skb_postpull_rcsum(struct sk_buff *skb,
3360
				      const void *start, unsigned int len)
L
Linus Torvalds 已提交
3361
{
3362
	__skb_postpull_rcsum(skb, start, len, 0);
L
Linus Torvalds 已提交
3363 3364
}

3365 3366 3367 3368 3369 3370 3371 3372
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);
}
3373

3374 3375 3376 3377 3378 3379 3380 3381 3382
/**
 *	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.
 */
3383 3384 3385
static inline void skb_postpush_rcsum(struct sk_buff *skb,
				      const void *start, unsigned int len)
{
3386
	__skb_postpush_rcsum(skb, start, len, 0);
3387 3388
}

3389
void *skb_pull_rcsum(struct sk_buff *skb, unsigned int len);
3390

3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401
/**
 *	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.
 */
3402
static inline void *skb_push_rcsum(struct sk_buff *skb, unsigned int len)
3403 3404 3405 3406 3407 3408
{
	skb_push(skb, len);
	skb_postpush_rcsum(skb, skb->data, len);
	return skb->data;
}

3409
int pskb_trim_rcsum_slow(struct sk_buff *skb, unsigned int len);
3410 3411 3412 3413 3414 3415 3416
/**
 *	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.
3417
 *	It can change skb pointers.
3418 3419 3420 3421 3422 3423
 */

static inline int pskb_trim_rcsum(struct sk_buff *skb, unsigned int len)
{
	if (likely(len >= skb->len))
		return 0;
3424
	return pskb_trim_rcsum_slow(skb, len);
3425 3426
}

3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437 3438 3439 3440 3441
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);
}

3442 3443 3444 3445 3446 3447
#define rb_to_skb(rb) rb_entry_safe(rb, struct sk_buff, rbnode)
#define skb_rb_first(root) rb_to_skb(rb_first(root))
#define skb_rb_last(root)  rb_to_skb(rb_last(root))
#define skb_rb_next(skb)   rb_to_skb(rb_next(&(skb)->rbnode))
#define skb_rb_prev(skb)   rb_to_skb(rb_prev(&(skb)->rbnode))

L
Linus Torvalds 已提交
3448 3449
#define skb_queue_walk(queue, skb) \
		for (skb = (queue)->next;					\
3450
		     skb != (struct sk_buff *)(queue);				\
L
Linus Torvalds 已提交
3451 3452
		     skb = skb->next)

3453 3454 3455 3456 3457
#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)

3458
#define skb_queue_walk_from(queue, skb)						\
3459
		for (; skb != (struct sk_buff *)(queue);			\
3460 3461
		     skb = skb->next)

3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473
#define skb_rbtree_walk(skb, root)						\
		for (skb = skb_rb_first(root); skb != NULL;			\
		     skb = skb_rb_next(skb))

#define skb_rbtree_walk_from(skb)						\
		for (; skb != NULL;						\
		     skb = skb_rb_next(skb))

#define skb_rbtree_walk_from_safe(skb, tmp)					\
		for (; tmp = skb ? skb_rb_next(skb) : NULL, (skb != NULL);	\
		     skb = tmp)

3474 3475 3476 3477 3478
#define skb_queue_walk_from_safe(queue, skb, tmp)				\
		for (tmp = skb->next;						\
		     skb != (struct sk_buff *)(queue);				\
		     skb = tmp, tmp = skb->next)

3479 3480
#define skb_queue_reverse_walk(queue, skb) \
		for (skb = (queue)->prev;					\
3481
		     skb != (struct sk_buff *)(queue);				\
3482 3483
		     skb = skb->prev)

3484 3485 3486 3487 3488 3489 3490 3491 3492
#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
3493

3494
static inline bool skb_has_frag_list(const struct sk_buff *skb)
3495 3496 3497 3498 3499 3500 3501 3502 3503 3504 3505 3506
{
	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)

3507

3508 3509
int __skb_wait_for_more_packets(struct sock *sk, struct sk_buff_head *queue,
				int *err, long *timeo_p,
3510
				const struct sk_buff *skb);
3511 3512 3513
struct sk_buff *__skb_try_recv_from_queue(struct sock *sk,
					  struct sk_buff_head *queue,
					  unsigned int flags,
3514
					  int *off, int *err,
3515
					  struct sk_buff **last);
3516 3517
struct sk_buff *__skb_try_recv_datagram(struct sock *sk,
					struct sk_buff_head *queue,
3518
					unsigned int flags, int *off, int *err,
3519
					struct sk_buff **last);
3520 3521
struct sk_buff *__skb_recv_datagram(struct sock *sk,
				    struct sk_buff_head *sk_queue,
3522
				    unsigned int flags, int *off, int *err);
3523 3524
struct sk_buff *skb_recv_datagram(struct sock *sk, unsigned flags, int noblock,
				  int *err);
3525 3526
__poll_t datagram_poll(struct file *file, struct socket *sock,
			   struct poll_table_struct *wait);
A
Al Viro 已提交
3527 3528
int skb_copy_datagram_iter(const struct sk_buff *from, int offset,
			   struct iov_iter *to, int size);
3529 3530 3531
static inline int skb_copy_datagram_msg(const struct sk_buff *from, int offset,
					struct msghdr *msg, int size)
{
3532
	return skb_copy_datagram_iter(from, offset, &msg->msg_iter, size);
3533
}
3534 3535
int skb_copy_and_csum_datagram_msg(struct sk_buff *skb, int hlen,
				   struct msghdr *msg);
3536 3537 3538
int skb_copy_and_hash_datagram_iter(const struct sk_buff *skb, int offset,
			   struct iov_iter *to, int len,
			   struct ahash_request *hash);
3539 3540 3541
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);
3542
void skb_free_datagram(struct sock *sk, struct sk_buff *skb);
3543 3544 3545 3546 3547 3548
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);
}
3549 3550 3551 3552
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,
3553
			      int len);
3554
int skb_splice_bits(struct sk_buff *skb, struct sock *sk, unsigned int offset,
3555
		    struct pipe_inode_info *pipe, unsigned int len,
A
Al Viro 已提交
3556
		    unsigned int flags);
3557 3558
int skb_send_sock_locked(struct sock *sk, struct sk_buff *skb, int offset,
			 int len);
3559
void skb_copy_and_csum_dev(const struct sk_buff *skb, u8 *to);
3560
unsigned int skb_zerocopy_headlen(const struct sk_buff *from);
3561 3562
int skb_zerocopy(struct sk_buff *to, struct sk_buff *from,
		 int len, int hlen);
3563 3564 3565
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);
3566
bool skb_gso_validate_network_len(const struct sk_buff *skb, unsigned int mtu);
3567
bool skb_gso_validate_mac_len(const struct sk_buff *skb, unsigned int len);
3568
struct sk_buff *skb_segment(struct sk_buff *skb, netdev_features_t features);
3569 3570
struct sk_buff *skb_segment_list(struct sk_buff *skb, netdev_features_t features,
				 unsigned int offset);
3571
struct sk_buff *skb_vlan_untag(struct sk_buff *skb);
3572
int skb_ensure_writable(struct sk_buff *skb, int write_len);
3573
int __skb_vlan_pop(struct sk_buff *skb, u16 *vlan_tci);
3574 3575
int skb_vlan_pop(struct sk_buff *skb);
int skb_vlan_push(struct sk_buff *skb, __be16 vlan_proto, u16 vlan_tci);
3576 3577 3578
int skb_eth_pop(struct sk_buff *skb);
int skb_eth_push(struct sk_buff *skb, const unsigned char *dst,
		 const unsigned char *src);
3579
int skb_mpls_push(struct sk_buff *skb, __be32 mpls_lse, __be16 mpls_proto,
3580
		  int mac_len, bool ethernet);
3581 3582
int skb_mpls_pop(struct sk_buff *skb, __be16 next_proto, int mac_len,
		 bool ethernet);
3583
int skb_mpls_update_lse(struct sk_buff *skb, __be32 mpls_lse);
3584
int skb_mpls_dec_ttl(struct sk_buff *skb);
3585 3586
struct sk_buff *pskb_extract(struct sk_buff *skb, int off, int to_copy,
			     gfp_t gfp);
3587

A
Al Viro 已提交
3588 3589
static inline int memcpy_from_msg(void *data, struct msghdr *msg, int len)
{
3590
	return copy_from_iter_full(data, len, &msg->msg_iter) ? 0 : -EFAULT;
A
Al Viro 已提交
3591 3592
}

A
Al Viro 已提交
3593 3594
static inline int memcpy_to_msg(struct msghdr *msg, void *data, int len)
{
3595
	return copy_to_iter(data, len, &msg->msg_iter) == len ? 0 : -EFAULT;
A
Al Viro 已提交
3596 3597
}

3598 3599 3600 3601 3602
struct skb_checksum_ops {
	__wsum (*update)(const void *mem, int len, __wsum wsum);
	__wsum (*combine)(__wsum csum, __wsum csum2, int offset, int len);
};

3603 3604
extern const struct skb_checksum_ops *crc32c_csum_stub __read_mostly;

3605 3606 3607 3608 3609
__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);

3610 3611 3612
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
3613
{
3614
	if (hlen - offset >= len)
3615
		return data + offset;
L
Linus Torvalds 已提交
3616

3617 3618
	if (!skb ||
	    skb_copy_bits(skb, offset, buffer, len) < 0)
L
Linus Torvalds 已提交
3619 3620 3621 3622 3623
		return NULL;

	return buffer;
}

3624 3625
static inline void * __must_check
skb_header_pointer(const struct sk_buff *skb, int offset, int len, void *buffer)
3626 3627 3628 3629 3630
{
	return __skb_header_pointer(skb, offset, len, skb->data,
				    skb_headlen(skb), buffer);
}

3631 3632 3633 3634 3635 3636 3637 3638 3639 3640 3641 3642 3643 3644 3645 3646 3647 3648
/**
 *	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)));
}

3649 3650 3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662
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);
}

3663 3664 3665 3666 3667 3668 3669 3670 3671 3672 3673 3674 3675 3676 3677
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);
}

3678
void skb_init(void);
L
Linus Torvalds 已提交
3679

3680 3681 3682 3683 3684
static inline ktime_t skb_get_ktime(const struct sk_buff *skb)
{
	return skb->tstamp;
}

3685 3686 3687
/**
 *	skb_get_timestamp - get timestamp from a skb
 *	@skb: skb to get stamp from
3688
 *	@stamp: pointer to struct __kernel_old_timeval to store stamp in
3689 3690 3691 3692 3693
 *
 *	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.
 */
3694
static inline void skb_get_timestamp(const struct sk_buff *skb,
3695
				     struct __kernel_old_timeval *stamp)
3696
{
3697
	*stamp = ns_to_kernel_old_timeval(skb->tstamp);
3698 3699
}

3700 3701 3702 3703 3704 3705 3706 3707 3708
static inline void skb_get_new_timestamp(const struct sk_buff *skb,
					 struct __kernel_sock_timeval *stamp)
{
	struct timespec64 ts = ktime_to_timespec64(skb->tstamp);

	stamp->tv_sec = ts.tv_sec;
	stamp->tv_usec = ts.tv_nsec / 1000;
}

3709
static inline void skb_get_timestampns(const struct sk_buff *skb,
3710
				       struct __kernel_old_timespec *stamp)
3711
{
3712 3713 3714 3715
	struct timespec64 ts = ktime_to_timespec64(skb->tstamp);

	stamp->tv_sec = ts.tv_sec;
	stamp->tv_nsec = ts.tv_nsec;
3716 3717
}

3718 3719 3720 3721 3722 3723 3724 3725 3726
static inline void skb_get_new_timestampns(const struct sk_buff *skb,
					   struct __kernel_timespec *stamp)
{
	struct timespec64 ts = ktime_to_timespec64(skb->tstamp);

	stamp->tv_sec = ts.tv_sec;
	stamp->tv_nsec = ts.tv_nsec;
}

3727
static inline void __net_timestamp(struct sk_buff *skb)
3728
{
3729
	skb->tstamp = ktime_get_real();
3730 3731
}

3732 3733 3734 3735 3736
static inline ktime_t net_timedelta(ktime_t t)
{
	return ktime_sub(ktime_get_real(), t);
}

3737 3738
static inline ktime_t net_invalid_timestamp(void)
{
T
Thomas Gleixner 已提交
3739
	return 0;
3740
}
3741

3742 3743 3744 3745 3746 3747 3748 3749 3750 3751 3752 3753 3754 3755 3756 3757 3758 3759 3760 3761 3762 3763 3764 3765
static inline u8 skb_metadata_len(const struct sk_buff *skb)
{
	return skb_shinfo(skb)->meta_len;
}

static inline void *skb_metadata_end(const struct sk_buff *skb)
{
	return skb_mac_header(skb);
}

static inline bool __skb_metadata_differs(const struct sk_buff *skb_a,
					  const struct sk_buff *skb_b,
					  u8 meta_len)
{
	const void *a = skb_metadata_end(skb_a);
	const void *b = skb_metadata_end(skb_b);
	/* Using more efficient varaiant than plain call to memcmp(). */
#if defined(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS) && BITS_PER_LONG == 64
	u64 diffs = 0;

	switch (meta_len) {
#define __it(x, op) (x -= sizeof(u##op))
#define __it_diff(a, b, op) (*(u##op *)__it(a, op)) ^ (*(u##op *)__it(b, op))
	case 32: diffs |= __it_diff(a, b, 64);
3766
		fallthrough;
3767
	case 24: diffs |= __it_diff(a, b, 64);
3768
		fallthrough;
3769
	case 16: diffs |= __it_diff(a, b, 64);
3770
		fallthrough;
3771 3772 3773
	case  8: diffs |= __it_diff(a, b, 64);
		break;
	case 28: diffs |= __it_diff(a, b, 64);
3774
		fallthrough;
3775
	case 20: diffs |= __it_diff(a, b, 64);
3776
		fallthrough;
3777
	case 12: diffs |= __it_diff(a, b, 64);
3778
		fallthrough;
3779 3780 3781 3782 3783 3784 3785 3786 3787 3788 3789 3790 3791 3792 3793 3794 3795 3796 3797 3798 3799 3800 3801 3802 3803 3804 3805 3806 3807 3808 3809 3810
	case  4: diffs |= __it_diff(a, b, 32);
		break;
	}
	return diffs;
#else
	return memcmp(a - meta_len, b - meta_len, meta_len);
#endif
}

static inline bool skb_metadata_differs(const struct sk_buff *skb_a,
					const struct sk_buff *skb_b)
{
	u8 len_a = skb_metadata_len(skb_a);
	u8 len_b = skb_metadata_len(skb_b);

	if (!(len_a | len_b))
		return false;

	return len_a != len_b ?
	       true : __skb_metadata_differs(skb_a, skb_b, len_a);
}

static inline void skb_metadata_set(struct sk_buff *skb, u8 meta_len)
{
	skb_shinfo(skb)->meta_len = meta_len;
}

static inline void skb_metadata_clear(struct sk_buff *skb)
{
	skb_metadata_set(skb, 0);
}

3811 3812
struct sk_buff *skb_clone_sk(struct sk_buff *skb);

3813 3814
#ifdef CONFIG_NETWORK_PHY_TIMESTAMPING

3815 3816
void skb_clone_tx_timestamp(struct sk_buff *skb);
bool skb_defer_rx_timestamp(struct sk_buff *skb);
3817 3818 3819 3820 3821 3822 3823 3824 3825 3826 3827 3828 3829 3830 3831 3832 3833

#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
 *
3834 3835
 * PHY drivers may accept clones of transmitted packets for
 * timestamping via their phy_driver.txtstamp method. These drivers
3836 3837
 * must call this function to return the skb back to the stack with a
 * timestamp.
3838
 *
3839
 * @skb: clone of the original outgoing packet
3840
 * @hwtstamps: hardware time stamps
3841 3842 3843 3844 3845
 *
 */
void skb_complete_tx_timestamp(struct sk_buff *skb,
			       struct skb_shared_hwtstamps *hwtstamps);

3846 3847 3848 3849
void __skb_tstamp_tx(struct sk_buff *orig_skb,
		     struct skb_shared_hwtstamps *hwtstamps,
		     struct sock *sk, int tstype);

3850 3851 3852 3853 3854 3855 3856 3857 3858 3859 3860
/**
 * 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.
 */
3861 3862
void skb_tstamp_tx(struct sk_buff *orig_skb,
		   struct skb_shared_hwtstamps *hwtstamps);
3863

3864 3865 3866 3867
/**
 * skb_tx_timestamp() - Driver hook for transmit timestamping
 *
 * Ethernet MAC Drivers should call this function in their hard_xmit()
3868
 * function immediately before giving the sk_buff to the MAC hardware.
3869
 *
3870 3871 3872 3873
 * 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.
 *
3874 3875 3876 3877
 * @skb: A socket buffer.
 */
static inline void skb_tx_timestamp(struct sk_buff *skb)
{
3878
	skb_clone_tx_timestamp(skb);
3879 3880
	if (skb_shinfo(skb)->tx_flags & SKBTX_SW_TSTAMP)
		skb_tstamp_tx(skb, NULL);
3881 3882
}

3883 3884 3885 3886 3887 3888 3889 3890 3891
/**
 * 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);

3892 3893
__sum16 __skb_checksum_complete_head(struct sk_buff *skb, int len);
__sum16 __skb_checksum_complete(struct sk_buff *skb);
3894

3895 3896
static inline int skb_csum_unnecessary(const struct sk_buff *skb)
{
3897 3898 3899 3900
	return ((skb->ip_summed == CHECKSUM_UNNECESSARY) ||
		skb->csum_valid ||
		(skb->ip_summed == CHECKSUM_PARTIAL &&
		 skb_checksum_start_offset(skb) >= 0));
3901 3902
}

3903 3904 3905 3906 3907 3908 3909 3910 3911 3912 3913 3914 3915 3916 3917 3918
/**
 *	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.
 */
3919
static inline __sum16 skb_checksum_complete(struct sk_buff *skb)
3920
{
3921 3922
	return skb_csum_unnecessary(skb) ?
	       0 : __skb_checksum_complete(skb);
3923 3924
}

3925 3926 3927 3928 3929 3930 3931 3932 3933 3934 3935 3936 3937 3938 3939 3940 3941 3942 3943 3944 3945
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;
	}
}

3946 3947 3948 3949 3950 3951 3952 3953
static inline void __skb_reset_checksum_unnecessary(struct sk_buff *skb)
{
	if (skb->ip_summed == CHECKSUM_UNNECESSARY) {
		skb->ip_summed = CHECKSUM_NONE;
		skb->csum_level = 0;
	}
}

3954 3955 3956 3957 3958 3959 3960 3961 3962
/* 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)
{
3963 3964
	if (skb_csum_unnecessary(skb) || (zero_okay && !check)) {
		skb->csum_valid = 1;
3965
		__skb_decr_checksum_unnecessary(skb);
3966 3967 3968 3969 3970 3971
		return false;
	}

	return true;
}

3972
/* For small packets <= CHECKSUM_BREAK perform checksum complete directly
3973 3974 3975 3976
 * in checksum_init.
 */
#define CHECKSUM_BREAK 76

3977 3978 3979 3980 3981 3982 3983 3984 3985 3986 3987 3988
/* 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;
}

3989 3990 3991 3992 3993 3994 3995 3996 3997 3998 3999 4000 4001 4002 4003
/* 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))) {
4004
			skb->csum_valid = 1;
4005 4006 4007 4008 4009 4010
			return 0;
		}
	}

	skb->csum = psum;

4011 4012 4013 4014 4015 4016 4017
	if (complete || skb->len <= CHECKSUM_BREAK) {
		__sum16 csum;

		csum = __skb_checksum_complete(skb);
		skb->csum_valid = !csum;
		return csum;
	}
4018 4019 4020 4021 4022 4023 4024 4025 4026 4027 4028 4029 4030 4031 4032 4033 4034 4035 4036 4037 4038 4039 4040

	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;						\
4041
	skb->csum_valid = 0;						\
4042 4043 4044 4045 4046 4047 4048 4049 4050 4051 4052 4053 4054 4055 4056 4057 4058
	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)		\
4059
	__skb_checksum_validate(skb, proto, true, true, check, compute_pseudo)
4060 4061 4062 4063

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

4064 4065
static inline bool __skb_checksum_convert_check(struct sk_buff *skb)
{
4066
	return (skb->ip_summed == CHECKSUM_NONE && skb->csum_valid);
4067 4068
}

4069
static inline void __skb_checksum_convert(struct sk_buff *skb, __wsum pseudo)
4070 4071 4072 4073 4074
{
	skb->csum = ~pseudo;
	skb->ip_summed = CHECKSUM_COMPLETE;
}

4075
#define skb_checksum_try_convert(skb, proto, compute_pseudo)	\
4076 4077
do {									\
	if (__skb_checksum_convert_check(skb))				\
4078
		__skb_checksum_convert(skb, compute_pseudo(skb, proto)); \
4079 4080
} while (0)

4081 4082 4083 4084 4085 4086 4087 4088
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;
}

4089 4090 4091 4092 4093 4094
/* 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,
4095
				       int start, int offset, bool nopartial)
4096 4097 4098
{
	__wsum delta;

4099 4100 4101 4102 4103
	if (!nopartial) {
		skb_remcsum_adjust_partial(skb, ptr, start, offset);
		return;
	}

4104 4105 4106 4107 4108 4109 4110 4111 4112 4113 4114
	 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);
}

4115 4116 4117
static inline struct nf_conntrack *skb_nfct(const struct sk_buff *skb)
{
#if IS_ENABLED(CONFIG_NF_CONNTRACK)
4118
	return (void *)(skb->_nfct & NFCT_PTRMASK);
4119 4120 4121 4122 4123
#else
	return NULL;
#endif
}

4124
static inline unsigned long skb_get_nfct(const struct sk_buff *skb)
L
Linus Torvalds 已提交
4125
{
4126 4127 4128 4129 4130
#if IS_ENABLED(CONFIG_NF_CONNTRACK)
	return skb->_nfct;
#else
	return 0UL;
#endif
L
Linus Torvalds 已提交
4131
}
4132 4133

static inline void skb_set_nfct(struct sk_buff *skb, unsigned long nfct)
L
Linus Torvalds 已提交
4134
{
4135 4136
#if IS_ENABLED(CONFIG_NF_CONNTRACK)
	skb->_nfct = nfct;
4137
#endif
4138
}
4139 4140 4141 4142 4143

#ifdef CONFIG_SKB_EXTENSIONS
enum skb_ext_id {
#if IS_ENABLED(CONFIG_BRIDGE_NETFILTER)
	SKB_EXT_BRIDGE_NF,
4144 4145 4146
#endif
#ifdef CONFIG_XFRM
	SKB_EXT_SEC_PATH,
4147 4148 4149
#endif
#if IS_ENABLED(CONFIG_NET_TC_SKB_EXT)
	TC_SKB_EXT,
4150 4151 4152
#endif
#if IS_ENABLED(CONFIG_MPTCP)
	SKB_EXT_MPTCP,
4153 4154 4155
#endif
#if IS_ENABLED(CONFIG_KCOV)
	SKB_EXT_KCOV_HANDLE,
4156 4157 4158 4159 4160 4161 4162 4163 4164 4165 4166 4167 4168 4169 4170 4171 4172 4173
#endif
	SKB_EXT_NUM, /* must be last */
};

/**
 *	struct skb_ext - sk_buff extensions
 *	@refcnt: 1 on allocation, deallocated on 0
 *	@offset: offset to add to @data to obtain extension address
 *	@chunks: size currently allocated, stored in SKB_EXT_ALIGN_SHIFT units
 *	@data: start of extension data, variable sized
 *
 *	Note: offsets/lengths are stored in chunks of 8 bytes, this allows
 *	to use 'u8' types while allowing up to 2kb worth of extension data.
 */
struct skb_ext {
	refcount_t refcnt;
	u8 offset[SKB_EXT_NUM]; /* in chunks of 8 bytes */
	u8 chunks;		/* same */
4174
	char data[] __aligned(8);
4175 4176
};

4177
struct skb_ext *__skb_ext_alloc(gfp_t flags);
4178 4179
void *__skb_ext_set(struct sk_buff *skb, enum skb_ext_id id,
		    struct skb_ext *ext);
4180 4181 4182 4183 4184 4185 4186 4187 4188 4189 4190 4191 4192 4193 4194 4195 4196 4197 4198 4199 4200 4201 4202 4203 4204 4205 4206 4207 4208 4209 4210 4211 4212 4213 4214 4215 4216 4217 4218 4219 4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230 4231 4232 4233 4234
void *skb_ext_add(struct sk_buff *skb, enum skb_ext_id id);
void __skb_ext_del(struct sk_buff *skb, enum skb_ext_id id);
void __skb_ext_put(struct skb_ext *ext);

static inline void skb_ext_put(struct sk_buff *skb)
{
	if (skb->active_extensions)
		__skb_ext_put(skb->extensions);
}

static inline void __skb_ext_copy(struct sk_buff *dst,
				  const struct sk_buff *src)
{
	dst->active_extensions = src->active_extensions;

	if (src->active_extensions) {
		struct skb_ext *ext = src->extensions;

		refcount_inc(&ext->refcnt);
		dst->extensions = ext;
	}
}

static inline void skb_ext_copy(struct sk_buff *dst, const struct sk_buff *src)
{
	skb_ext_put(dst);
	__skb_ext_copy(dst, src);
}

static inline bool __skb_ext_exist(const struct skb_ext *ext, enum skb_ext_id i)
{
	return !!ext->offset[i];
}

static inline bool skb_ext_exist(const struct sk_buff *skb, enum skb_ext_id id)
{
	return skb->active_extensions & (1 << id);
}

static inline void skb_ext_del(struct sk_buff *skb, enum skb_ext_id id)
{
	if (skb_ext_exist(skb, id))
		__skb_ext_del(skb, id);
}

static inline void *skb_ext_find(const struct sk_buff *skb, enum skb_ext_id id)
{
	if (skb_ext_exist(skb, id)) {
		struct skb_ext *ext = skb->extensions;

		return (void *)ext + (ext->offset[id] << 3);
	}

	return NULL;
}
4235 4236 4237 4238 4239 4240 4241 4242

static inline void skb_ext_reset(struct sk_buff *skb)
{
	if (unlikely(skb->active_extensions)) {
		__skb_ext_put(skb->extensions);
		skb->active_extensions = 0;
	}
}
4243 4244 4245 4246 4247

static inline bool skb_has_extensions(struct sk_buff *skb)
{
	return unlikely(skb->active_extensions);
}
4248 4249
#else
static inline void skb_ext_put(struct sk_buff *skb) {}
4250
static inline void skb_ext_reset(struct sk_buff *skb) {}
4251 4252 4253
static inline void skb_ext_del(struct sk_buff *skb, int unused) {}
static inline void __skb_ext_copy(struct sk_buff *d, const struct sk_buff *s) {}
static inline void skb_ext_copy(struct sk_buff *dst, const struct sk_buff *s) {}
4254
static inline bool skb_has_extensions(struct sk_buff *skb) { return false; }
4255 4256
#endif /* CONFIG_SKB_EXTENSIONS */

4257
static inline void nf_reset_ct(struct sk_buff *skb)
4258
{
4259
#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
4260 4261
	nf_conntrack_put(skb_nfct(skb));
	skb->_nfct = 0;
4262
#endif
4263 4264
}

4265 4266
static inline void nf_reset_trace(struct sk_buff *skb)
{
4267
#if IS_ENABLED(CONFIG_NETFILTER_XT_TARGET_TRACE) || defined(CONFIG_NF_TABLES)
G
Gao feng 已提交
4268 4269
	skb->nf_trace = 0;
#endif
4270 4271
}

4272 4273 4274 4275 4276 4277 4278
static inline void ipvs_reset(struct sk_buff *skb)
{
#if IS_ENABLED(CONFIG_IP_VS)
	skb->ipvs_property = 0;
#endif
}

4279
/* Note: This doesn't put any conntrack info in dst. */
4280 4281
static inline void __nf_copy(struct sk_buff *dst, const struct sk_buff *src,
			     bool copy)
4282
{
4283
#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
4284 4285
	dst->_nfct = src->_nfct;
	nf_conntrack_get(skb_nfct(src));
4286
#endif
4287
#if IS_ENABLED(CONFIG_NETFILTER_XT_TARGET_TRACE) || defined(CONFIG_NF_TABLES)
4288 4289
	if (copy)
		dst->nf_trace = src->nf_trace;
4290
#endif
4291 4292
}

4293 4294 4295
static inline void nf_copy(struct sk_buff *dst, const struct sk_buff *src)
{
#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
4296
	nf_conntrack_put(skb_nfct(dst));
4297
#endif
4298
	__nf_copy(dst, src, true);
4299 4300
}

4301 4302 4303 4304 4305 4306 4307 4308 4309 4310 4311 4312 4313 4314 4315 4316 4317 4318
#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

4319 4320 4321
static inline int secpath_exists(const struct sk_buff *skb)
{
#ifdef CONFIG_XFRM
4322
	return skb_ext_exist(skb, SKB_EXT_SEC_PATH);
4323 4324 4325 4326 4327
#else
	return 0;
#endif
}

4328 4329 4330
static inline bool skb_irq_freeable(const struct sk_buff *skb)
{
	return !skb->destructor &&
4331
		!secpath_exists(skb) &&
4332
		!skb_nfct(skb) &&
4333 4334 4335 4336
		!skb->_skb_refdst &&
		!skb_has_frag_list(skb);
}

4337 4338 4339 4340 4341
static inline void skb_set_queue_mapping(struct sk_buff *skb, u16 queue_mapping)
{
	skb->queue_mapping = queue_mapping;
}

4342
static inline u16 skb_get_queue_mapping(const struct sk_buff *skb)
4343 4344 4345 4346
{
	return skb->queue_mapping;
}

4347 4348 4349 4350 4351
static inline void skb_copy_queue_mapping(struct sk_buff *to, const struct sk_buff *from)
{
	to->queue_mapping = from->queue_mapping;
}

4352 4353 4354 4355 4356
static inline void skb_record_rx_queue(struct sk_buff *skb, u16 rx_queue)
{
	skb->queue_mapping = rx_queue + 1;
}

4357
static inline u16 skb_get_rx_queue(const struct sk_buff *skb)
4358 4359 4360 4361
{
	return skb->queue_mapping - 1;
}

4362
static inline bool skb_rx_queue_recorded(const struct sk_buff *skb)
4363
{
E
Eric Dumazet 已提交
4364
	return skb->queue_mapping != 0;
4365 4366
}

4367 4368 4369 4370 4371 4372 4373 4374 4375 4376
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;
}

4377
static inline struct sec_path *skb_sec_path(const struct sk_buff *skb)
4378
{
4379
#ifdef CONFIG_XFRM
4380
	return skb_ext_find(skb, SKB_EXT_SEC_PATH);
4381 4382 4383
#else
	return NULL;
#endif
4384
}
4385

4386 4387 4388
/* 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
4389 4390 4391
 * tunnel skb it points to outer mac header.
 * Keeps track of level of encapsulation of network headers.
 */
4392
struct skb_gso_cb {
4393 4394 4395 4396
	union {
		int	mac_offset;
		int	data_offset;
	};
4397
	int	encap_level;
4398
	__wsum	csum;
4399
	__u16	csum_start;
4400
};
C
Cambda Zhu 已提交
4401 4402
#define SKB_GSO_CB_OFFSET	32
#define SKB_GSO_CB(skb) ((struct skb_gso_cb *)((skb)->cb + SKB_GSO_CB_OFFSET))
4403 4404 4405 4406 4407 4408 4409

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;
}

4410 4411 4412 4413 4414 4415 4416 4417 4418 4419 4420 4421 4422 4423 4424
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;
}

4425 4426 4427 4428 4429 4430 4431 4432 4433 4434
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;
}

4435 4436 4437 4438 4439 4440 4441 4442 4443 4444
/* 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)
{
4445 4446 4447
	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;
4448

4449 4450
	SKB_GSO_CB(skb)->csum = res;
	SKB_GSO_CB(skb)->csum_start = csum_start - skb->head;
4451

4452
	return csum_fold(csum_partial(csum_start, plen, partial));
4453 4454
}

4455
static inline bool skb_is_gso(const struct sk_buff *skb)
H
Herbert Xu 已提交
4456 4457 4458 4459
{
	return skb_shinfo(skb)->gso_size;
}

4460
/* Note: Should be called only if skb_is_gso(skb) is true */
4461
static inline bool skb_is_gso_v6(const struct sk_buff *skb)
B
Brice Goglin 已提交
4462 4463 4464 4465
{
	return skb_shinfo(skb)->gso_type & SKB_GSO_TCPV6;
}

4466 4467 4468 4469 4470 4471
/* Note: Should be called only if skb_is_gso(skb) is true */
static inline bool skb_is_gso_sctp(const struct sk_buff *skb)
{
	return skb_shinfo(skb)->gso_type & SKB_GSO_SCTP;
}

4472
/* Note: Should be called only if skb_is_gso(skb) is true */
4473 4474
static inline bool skb_is_gso_tcp(const struct sk_buff *skb)
{
4475
	return skb_shinfo(skb)->gso_type & (SKB_GSO_TCPV4 | SKB_GSO_TCPV6);
4476 4477
}

4478 4479 4480 4481 4482 4483 4484
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;
}

4485 4486 4487 4488 4489 4490 4491 4492 4493 4494 4495 4496 4497 4498 4499 4500
static inline void skb_increase_gso_size(struct skb_shared_info *shinfo,
					 u16 increment)
{
	if (WARN_ON_ONCE(shinfo->gso_size == GSO_BY_FRAGS))
		return;
	shinfo->gso_size += increment;
}

static inline void skb_decrease_gso_size(struct skb_shared_info *shinfo,
					 u16 decrement)
{
	if (WARN_ON_ONCE(shinfo->gso_size == GSO_BY_FRAGS))
		return;
	shinfo->gso_size -= decrement;
}

4501
void __skb_warn_lro_forwarding(const struct sk_buff *skb);
4502 4503 4504 4505 4506

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. */
4507 4508
	const struct skb_shared_info *shinfo = skb_shinfo(skb);

4509 4510
	if (skb_is_nonlinear(skb) && shinfo->gso_size != 0 &&
	    unlikely(shinfo->gso_type == 0)) {
4511 4512 4513 4514 4515 4516
		__skb_warn_lro_forwarding(skb);
		return true;
	}
	return false;
}

4517 4518 4519 4520 4521 4522 4523
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;
}

4524 4525 4526 4527 4528 4529 4530 4531
/**
 * 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.
 */
4532
static inline void skb_checksum_none_assert(const struct sk_buff *skb)
4533 4534 4535 4536 4537 4538
{
#ifdef DEBUG
	BUG_ON(skb->ip_summed != CHECKSUM_NONE);
#endif
}

4539
bool skb_partial_csum_set(struct sk_buff *skb, u16 start, u16 off);
4540

P
Paul Durrant 已提交
4541
int skb_checksum_setup(struct sk_buff *skb, bool recalculate);
4542 4543 4544
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
4545

4546 4547 4548 4549 4550 4551 4552 4553 4554 4555 4556 4557 4558
/**
 * 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);
}
4559

4560 4561 4562
/* Local Checksum Offload.
 * Compute outer checksum based on the assumption that the
 * inner checksum will be offloaded later.
4563
 * See Documentation/networking/checksum-offloads.rst for
4564
 * explanation of how this works.
4565 4566 4567 4568 4569 4570
 * 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)
{
4571 4572 4573
	unsigned char *csum_start = skb_checksum_start(skb);
	unsigned char *l4_hdr = skb_transport_header(skb);
	__wsum partial;
4574 4575

	/* Start with complement of inner checksum adjustment */
4576 4577 4578
	partial = ~csum_unfold(*(__force __sum16 *)(csum_start +
						    skb->csum_offset));

4579
	/* Add in checksum of our headers (incl. outer checksum
4580
	 * adjustment filled in by caller) and return result.
4581
	 */
4582
	return csum_partial(l4_hdr, csum_start - l4_hdr, partial);
4583 4584
}

4585 4586 4587 4588 4589 4590 4591 4592 4593 4594 4595 4596 4597 4598 4599 4600 4601 4602 4603 4604 4605 4606 4607 4608 4609 4610
static inline bool skb_is_redirected(const struct sk_buff *skb)
{
#ifdef CONFIG_NET_REDIRECT
	return skb->redirected;
#else
	return false;
#endif
}

static inline void skb_set_redirected(struct sk_buff *skb, bool from_ingress)
{
#ifdef CONFIG_NET_REDIRECT
	skb->redirected = 1;
	skb->from_ingress = from_ingress;
	if (skb->from_ingress)
		skb->tstamp = 0;
#endif
}

static inline void skb_reset_redirect(struct sk_buff *skb)
{
#ifdef CONFIG_NET_REDIRECT
	skb->redirected = 0;
#endif
}

4611
#if IS_ENABLED(CONFIG_KCOV) && IS_ENABLED(CONFIG_SKB_EXTENSIONS)
4612 4613 4614 4615 4616 4617 4618 4619 4620 4621 4622 4623 4624 4625 4626 4627 4628 4629 4630 4631 4632 4633 4634 4635 4636 4637 4638
static inline void skb_set_kcov_handle(struct sk_buff *skb,
				       const u64 kcov_handle)
{
	/* Do not allocate skb extensions only to set kcov_handle to zero
	 * (as it is zero by default). However, if the extensions are
	 * already allocated, update kcov_handle anyway since
	 * skb_set_kcov_handle can be called to zero a previously set
	 * value.
	 */
	if (skb_has_extensions(skb) || kcov_handle) {
		u64 *kcov_handle_ptr = skb_ext_add(skb, SKB_EXT_KCOV_HANDLE);

		if (kcov_handle_ptr)
			*kcov_handle_ptr = kcov_handle;
	}
}

static inline u64 skb_get_kcov_handle(struct sk_buff *skb)
{
	u64 *kcov_handle = skb_ext_find(skb, SKB_EXT_KCOV_HANDLE);

	return kcov_handle ? *kcov_handle : 0;
}
#else
static inline void skb_set_kcov_handle(struct sk_buff *skb,
				       const u64 kcov_handle) { }
static inline u64 skb_get_kcov_handle(struct sk_buff *skb) { return 0; }
4639
#endif /* CONFIG_KCOV && CONFIG_SKB_EXTENSIONS */
4640

L
Linus Torvalds 已提交
4641 4642
#endif	/* __KERNEL__ */
#endif	/* _LINUX_SKBUFF_H */