skbuff.c 137.0 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3
/*
 *	Routines having to do with the 'struct sk_buff' memory handlers.
 *
4
 *	Authors:	Alan Cox <alan@lxorguk.ukuu.org.uk>
L
Linus Torvalds 已提交
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38
 *			Florian La Roche <rzsfl@rz.uni-sb.de>
 *
 *	Fixes:
 *		Alan Cox	:	Fixed the worst of the load
 *					balancer bugs.
 *		Dave Platt	:	Interrupt stacking fix.
 *	Richard Kooijman	:	Timestamp fixes.
 *		Alan Cox	:	Changed buffer format.
 *		Alan Cox	:	destructor hook for AF_UNIX etc.
 *		Linus Torvalds	:	Better skb_clone.
 *		Alan Cox	:	Added skb_copy.
 *		Alan Cox	:	Added all the changed routines Linus
 *					only put in the headers
 *		Ray VanTassle	:	Fixed --skb->lock in free
 *		Alan Cox	:	skb_copy copy arp field
 *		Andi Kleen	:	slabified it.
 *		Robert Olsson	:	Removed skb_head_pool
 *
 *	NOTE:
 *		The __skb_ routines should be called with interrupts
 *	disabled, or you better be *real* sure that the operation is atomic
 *	with respect to whatever list is being frobbed (e.g. via lock_sock()
 *	or via disabling bottom half handlers, etc).
 *
 *	This program is free software; you can redistribute it and/or
 *	modify it under the terms of the GNU General Public License
 *	as published by the Free Software Foundation; either version
 *	2 of the License, or (at your option) any later version.
 */

/*
 *	The functions in this file will not compile correctly with gcc 2.4.x
 */

J
Joe Perches 已提交
39 40
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

L
Linus Torvalds 已提交
41 42 43 44 45 46 47 48
#include <linux/module.h>
#include <linux/types.h>
#include <linux/kernel.h>
#include <linux/mm.h>
#include <linux/interrupt.h>
#include <linux/in.h>
#include <linux/inet.h>
#include <linux/slab.h>
49 50
#include <linux/tcp.h>
#include <linux/udp.h>
M
Marcelo Ricardo Leitner 已提交
51
#include <linux/sctp.h>
L
Linus Torvalds 已提交
52 53 54 55 56 57
#include <linux/netdevice.h>
#ifdef CONFIG_NET_CLS_ACT
#include <net/pkt_sched.h>
#endif
#include <linux/string.h>
#include <linux/skbuff.h>
J
Jens Axboe 已提交
58
#include <linux/splice.h>
L
Linus Torvalds 已提交
59 60 61
#include <linux/cache.h>
#include <linux/rtnetlink.h>
#include <linux/init.h>
62
#include <linux/scatterlist.h>
63
#include <linux/errqueue.h>
64
#include <linux/prefetch.h>
65
#include <linux/if_vlan.h>
L
Linus Torvalds 已提交
66 67 68 69 70

#include <net/protocol.h>
#include <net/dst.h>
#include <net/sock.h>
#include <net/checksum.h>
P
Paul Durrant 已提交
71
#include <net/ip6_checksum.h>
L
Linus Torvalds 已提交
72 73
#include <net/xfrm.h>

74
#include <linux/uaccess.h>
75
#include <trace/events/skb.h>
E
Eric Dumazet 已提交
76
#include <linux/highmem.h>
77 78
#include <linux/capability.h>
#include <linux/user_namespace.h>
A
Al Viro 已提交
79

80 81
struct kmem_cache *skbuff_head_cache __ro_after_init;
static struct kmem_cache *skbuff_fclone_cache __ro_after_init;
H
Hans Westgaard Ry 已提交
82 83
int sysctl_max_skb_frags __read_mostly = MAX_SKB_FRAGS;
EXPORT_SYMBOL(sysctl_max_skb_frags);
L
Linus Torvalds 已提交
84 85

/**
86 87 88 89
 *	skb_panic - private function for out-of-line support
 *	@skb:	buffer
 *	@sz:	size
 *	@addr:	address
90
 *	@msg:	skb_over_panic or skb_under_panic
L
Linus Torvalds 已提交
91
 *
92 93 94 95
 *	Out-of-line support for skb_put() and skb_push().
 *	Called via the wrapper skb_over_panic() or skb_under_panic().
 *	Keep out of line to prevent kernel bloat.
 *	__builtin_return_address is not used because it is not always reliable.
L
Linus Torvalds 已提交
96
 */
97
static void skb_panic(struct sk_buff *skb, unsigned int sz, void *addr,
98
		      const char msg[])
L
Linus Torvalds 已提交
99
{
J
Joe Perches 已提交
100
	pr_emerg("%s: text:%p len:%d put:%d head:%p data:%p tail:%#lx end:%#lx dev:%s\n",
101
		 msg, addr, skb->len, sz, skb->head, skb->data,
J
Joe Perches 已提交
102 103
		 (unsigned long)skb->tail, (unsigned long)skb->end,
		 skb->dev ? skb->dev->name : "<NULL>");
L
Linus Torvalds 已提交
104 105 106
	BUG();
}

107
static void skb_over_panic(struct sk_buff *skb, unsigned int sz, void *addr)
L
Linus Torvalds 已提交
108
{
109
	skb_panic(skb, sz, addr, __func__);
L
Linus Torvalds 已提交
110 111
}

112 113 114 115
static void skb_under_panic(struct sk_buff *skb, unsigned int sz, void *addr)
{
	skb_panic(skb, sz, addr, __func__);
}
116 117 118 119 120 121 122 123 124 125

/*
 * kmalloc_reserve is a wrapper around kmalloc_node_track_caller that tells
 * the caller if emergency pfmemalloc reserves are being used. If it is and
 * the socket is later found to be SOCK_MEMALLOC then PFMEMALLOC reserves
 * may be used. Otherwise, the packet data may be discarded until enough
 * memory is free
 */
#define kmalloc_reserve(size, gfp, node, pfmemalloc) \
	 __kmalloc_reserve(size, gfp, node, _RET_IP_, pfmemalloc)
126 127 128

static void *__kmalloc_reserve(size_t size, gfp_t flags, int node,
			       unsigned long ip, bool *pfmemalloc)
129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153
{
	void *obj;
	bool ret_pfmemalloc = false;

	/*
	 * Try a regular allocation, when that fails and we're not entitled
	 * to the reserves, fail.
	 */
	obj = kmalloc_node_track_caller(size,
					flags | __GFP_NOMEMALLOC | __GFP_NOWARN,
					node);
	if (obj || !(gfp_pfmemalloc_allowed(flags)))
		goto out;

	/* Try again but now we are using pfmemalloc reserves */
	ret_pfmemalloc = true;
	obj = kmalloc_node_track_caller(size, flags, node);

out:
	if (pfmemalloc)
		*pfmemalloc = ret_pfmemalloc;

	return obj;
}

L
Linus Torvalds 已提交
154 155 156 157 158 159 160
/* 	Allocate a new skbuff. We do this ourselves so we can fill in a few
 *	'private' fields and also do memory statistics to find all the
 *	[BEEP] leaks.
 *
 */

/**
161
 *	__alloc_skb	-	allocate a network buffer
L
Linus Torvalds 已提交
162 163
 *	@size: size to allocate
 *	@gfp_mask: allocation mask
164 165 166 167
 *	@flags: If SKB_ALLOC_FCLONE is set, allocate from fclone cache
 *		instead of head cache and allocate a cloned (child) skb.
 *		If SKB_ALLOC_RX is set, __GFP_MEMALLOC will be used for
 *		allocations in case the data is required for writeback
168
 *	@node: numa node to allocate memory on
L
Linus Torvalds 已提交
169 170
 *
 *	Allocate a new &sk_buff. The returned buffer has no headroom and a
171 172
 *	tail room of at least size bytes. The object has a reference count
 *	of one. The return is the buffer. On a failure the return is %NULL.
L
Linus Torvalds 已提交
173 174 175 176
 *
 *	Buffers may only be allocated from interrupts using a @gfp_mask of
 *	%GFP_ATOMIC.
 */
A
Al Viro 已提交
177
struct sk_buff *__alloc_skb(unsigned int size, gfp_t gfp_mask,
178
			    int flags, int node)
L
Linus Torvalds 已提交
179
{
180
	struct kmem_cache *cache;
B
Benjamin LaHaise 已提交
181
	struct skb_shared_info *shinfo;
L
Linus Torvalds 已提交
182 183
	struct sk_buff *skb;
	u8 *data;
184
	bool pfmemalloc;
L
Linus Torvalds 已提交
185

186 187 188 189 190
	cache = (flags & SKB_ALLOC_FCLONE)
		? skbuff_fclone_cache : skbuff_head_cache;

	if (sk_memalloc_socks() && (flags & SKB_ALLOC_RX))
		gfp_mask |= __GFP_MEMALLOC;
191

L
Linus Torvalds 已提交
192
	/* Get the HEAD */
193
	skb = kmem_cache_alloc_node(cache, gfp_mask & ~__GFP_DMA, node);
L
Linus Torvalds 已提交
194 195
	if (!skb)
		goto out;
E
Eric Dumazet 已提交
196
	prefetchw(skb);
L
Linus Torvalds 已提交
197

E
Eric Dumazet 已提交
198 199 200 201 202
	/* We do our best to align skb_shared_info on a separate cache
	 * line. It usually works because kmalloc(X > SMP_CACHE_BYTES) gives
	 * aligned memory blocks, unless SLUB/SLAB debug is enabled.
	 * Both skb->head and skb_shared_info are cache line aligned.
	 */
203
	size = SKB_DATA_ALIGN(size);
E
Eric Dumazet 已提交
204
	size += SKB_DATA_ALIGN(sizeof(struct skb_shared_info));
205
	data = kmalloc_reserve(size, gfp_mask, node, &pfmemalloc);
L
Linus Torvalds 已提交
206 207
	if (!data)
		goto nodata;
E
Eric Dumazet 已提交
208 209 210 211 212
	/* kmalloc(size) might give us more room than requested.
	 * Put skb_shared_info exactly at the end of allocated zone,
	 * to allow max possible filling before reallocation.
	 */
	size = SKB_WITH_OVERHEAD(ksize(data));
E
Eric Dumazet 已提交
213
	prefetchw(data + size);
L
Linus Torvalds 已提交
214

215
	/*
216 217 218
	 * Only clear those fields we need to clear, not those that we will
	 * actually initialise below. Hence, don't put any more fields after
	 * the tail pointer in struct sk_buff!
219 220
	 */
	memset(skb, 0, offsetof(struct sk_buff, tail));
E
Eric Dumazet 已提交
221 222
	/* Account for allocated memory : skb + skb->head */
	skb->truesize = SKB_TRUESIZE(size);
223
	skb->pfmemalloc = pfmemalloc;
224
	refcount_set(&skb->users, 1);
L
Linus Torvalds 已提交
225 226
	skb->head = data;
	skb->data = data;
227
	skb_reset_tail_pointer(skb);
228
	skb->end = skb->tail + size;
C
Cong Wang 已提交
229 230
	skb->mac_header = (typeof(skb->mac_header))~0U;
	skb->transport_header = (typeof(skb->transport_header))~0U;
231

B
Benjamin LaHaise 已提交
232 233
	/* make sure we initialize shinfo sequentially */
	shinfo = skb_shinfo(skb);
E
Eric Dumazet 已提交
234
	memset(shinfo, 0, offsetof(struct skb_shared_info, dataref));
B
Benjamin LaHaise 已提交
235 236
	atomic_set(&shinfo->dataref, 1);

237
	if (flags & SKB_ALLOC_FCLONE) {
238
		struct sk_buff_fclones *fclones;
L
Linus Torvalds 已提交
239

240 241
		fclones = container_of(skb, struct sk_buff_fclones, skb1);

242
		skb->fclone = SKB_FCLONE_ORIG;
243
		refcount_set(&fclones->fclone_ref, 1);
244

245
		fclones->skb2.fclone = SKB_FCLONE_CLONE;
246
	}
L
Linus Torvalds 已提交
247 248 249
out:
	return skb;
nodata:
250
	kmem_cache_free(cache, skb);
L
Linus Torvalds 已提交
251 252 253
	skb = NULL;
	goto out;
}
254
EXPORT_SYMBOL(__alloc_skb);
L
Linus Torvalds 已提交
255

E
Eric Dumazet 已提交
256
/**
E
Eric Dumazet 已提交
257
 * __build_skb - build a network buffer
E
Eric Dumazet 已提交
258
 * @data: data buffer provided by caller
E
Eric Dumazet 已提交
259
 * @frag_size: size of data, or 0 if head was kmalloced
E
Eric Dumazet 已提交
260 261
 *
 * Allocate a new &sk_buff. Caller provides space holding head and
262
 * skb_shared_info. @data must have been allocated by kmalloc() only if
E
Eric Dumazet 已提交
263 264
 * @frag_size is 0, otherwise data should come from the page allocator
 *  or vmalloc()
E
Eric Dumazet 已提交
265 266 267 268 269 270 271 272 273 274
 * The return is the new skb buffer.
 * On a failure the return is %NULL, and @data is not freed.
 * Notes :
 *  Before IO, driver allocates only data buffer where NIC put incoming frame
 *  Driver should add room at head (NET_SKB_PAD) and
 *  MUST add room at tail (SKB_DATA_ALIGN(skb_shared_info))
 *  After IO, driver calls build_skb(), to allocate sk_buff and populate it
 *  before giving packet to stack.
 *  RX rings only contains data buffers, not full skbs.
 */
E
Eric Dumazet 已提交
275
struct sk_buff *__build_skb(void *data, unsigned int frag_size)
E
Eric Dumazet 已提交
276 277 278
{
	struct skb_shared_info *shinfo;
	struct sk_buff *skb;
279
	unsigned int size = frag_size ? : ksize(data);
E
Eric Dumazet 已提交
280 281 282 283 284

	skb = kmem_cache_alloc(skbuff_head_cache, GFP_ATOMIC);
	if (!skb)
		return NULL;

285
	size -= SKB_DATA_ALIGN(sizeof(struct skb_shared_info));
E
Eric Dumazet 已提交
286 287 288

	memset(skb, 0, offsetof(struct sk_buff, tail));
	skb->truesize = SKB_TRUESIZE(size);
289
	refcount_set(&skb->users, 1);
E
Eric Dumazet 已提交
290 291 292 293
	skb->head = data;
	skb->data = data;
	skb_reset_tail_pointer(skb);
	skb->end = skb->tail + size;
C
Cong Wang 已提交
294 295
	skb->mac_header = (typeof(skb->mac_header))~0U;
	skb->transport_header = (typeof(skb->transport_header))~0U;
E
Eric Dumazet 已提交
296 297 298 299 300 301 302 303

	/* make sure we initialize shinfo sequentially */
	shinfo = skb_shinfo(skb);
	memset(shinfo, 0, offsetof(struct skb_shared_info, dataref));
	atomic_set(&shinfo->dataref, 1);

	return skb;
}
E
Eric Dumazet 已提交
304 305 306 307 308 309 310 311 312 313 314 315

/* build_skb() is wrapper over __build_skb(), that specifically
 * takes care of skb->head and skb->pfmemalloc
 * This means that if @frag_size is not zero, then @data must be backed
 * by a page fragment, not kmalloc() or vmalloc()
 */
struct sk_buff *build_skb(void *data, unsigned int frag_size)
{
	struct sk_buff *skb = __build_skb(data, frag_size);

	if (skb && frag_size) {
		skb->head_frag = 1;
316
		if (page_is_pfmemalloc(virt_to_head_page(data)))
E
Eric Dumazet 已提交
317 318 319 320
			skb->pfmemalloc = 1;
	}
	return skb;
}
E
Eric Dumazet 已提交
321 322
EXPORT_SYMBOL(build_skb);

323 324 325 326
#define NAPI_SKB_CACHE_SIZE	64

struct napi_alloc_cache {
	struct page_frag_cache page;
327
	unsigned int skb_count;
328 329 330
	void *skb_cache[NAPI_SKB_CACHE_SIZE];
};

331
static DEFINE_PER_CPU(struct page_frag_cache, netdev_alloc_cache);
332
static DEFINE_PER_CPU(struct napi_alloc_cache, napi_alloc_cache);
333 334 335

static void *__netdev_alloc_frag(unsigned int fragsz, gfp_t gfp_mask)
{
336
	struct page_frag_cache *nc;
337 338 339 340
	unsigned long flags;
	void *data;

	local_irq_save(flags);
341
	nc = this_cpu_ptr(&netdev_alloc_cache);
342
	data = page_frag_alloc(nc, fragsz, gfp_mask);
343 344 345
	local_irq_restore(flags);
	return data;
}
346 347 348 349 350 351 352 353 354 355

/**
 * netdev_alloc_frag - allocate a page fragment
 * @fragsz: fragment size
 *
 * Allocates a frag from a page for receive buffer.
 * Uses GFP_ATOMIC allocations.
 */
void *netdev_alloc_frag(unsigned int fragsz)
{
M
Mel Gorman 已提交
356
	return __netdev_alloc_frag(fragsz, GFP_ATOMIC);
357
}
358 359
EXPORT_SYMBOL(netdev_alloc_frag);

360 361
static void *__napi_alloc_frag(unsigned int fragsz, gfp_t gfp_mask)
{
362
	struct napi_alloc_cache *nc = this_cpu_ptr(&napi_alloc_cache);
363

364
	return page_frag_alloc(&nc->page, fragsz, gfp_mask);
365 366 367 368
}

void *napi_alloc_frag(unsigned int fragsz)
{
M
Mel Gorman 已提交
369
	return __napi_alloc_frag(fragsz, GFP_ATOMIC);
370 371 372
}
EXPORT_SYMBOL(napi_alloc_frag);

373 374 375
/**
 *	__netdev_alloc_skb - allocate an skbuff for rx on a specific device
 *	@dev: network device to receive on
376
 *	@len: length to allocate
377 378 379 380 381 382 383 384 385
 *	@gfp_mask: get_free_pages mask, passed to alloc_skb
 *
 *	Allocate a new &sk_buff and assign it a usage count of one. The
 *	buffer has NET_SKB_PAD 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.
 */
386 387
struct sk_buff *__netdev_alloc_skb(struct net_device *dev, unsigned int len,
				   gfp_t gfp_mask)
388
{
389
	struct page_frag_cache *nc;
390
	unsigned long flags;
391
	struct sk_buff *skb;
392 393 394 395
	bool pfmemalloc;
	void *data;

	len += NET_SKB_PAD;
396

397
	if ((len > SKB_WITH_OVERHEAD(PAGE_SIZE)) ||
398
	    (gfp_mask & (__GFP_DIRECT_RECLAIM | GFP_DMA))) {
399 400 401 402 403
		skb = __alloc_skb(len, gfp_mask, SKB_ALLOC_RX, NUMA_NO_NODE);
		if (!skb)
			goto skb_fail;
		goto skb_success;
	}
404

405 406 407 408 409 410 411 412 413
	len += SKB_DATA_ALIGN(sizeof(struct skb_shared_info));
	len = SKB_DATA_ALIGN(len);

	if (sk_memalloc_socks())
		gfp_mask |= __GFP_MEMALLOC;

	local_irq_save(flags);

	nc = this_cpu_ptr(&netdev_alloc_cache);
414
	data = page_frag_alloc(nc, len, gfp_mask);
415 416 417 418 419 420 421 422 423
	pfmemalloc = nc->pfmemalloc;

	local_irq_restore(flags);

	if (unlikely(!data))
		return NULL;

	skb = __build_skb(data, len);
	if (unlikely(!skb)) {
424
		skb_free_frag(data);
425
		return NULL;
426
	}
427

428 429 430 431 432
	/* use OR instead of assignment to avoid clearing of bits in mask */
	if (pfmemalloc)
		skb->pfmemalloc = 1;
	skb->head_frag = 1;

433
skb_success:
434 435 436
	skb_reserve(skb, NET_SKB_PAD);
	skb->dev = dev;

437
skb_fail:
438 439
	return skb;
}
440
EXPORT_SYMBOL(__netdev_alloc_skb);
L
Linus Torvalds 已提交
441

442 443 444
/**
 *	__napi_alloc_skb - allocate skbuff for rx in a specific NAPI instance
 *	@napi: napi instance this buffer was allocated for
445
 *	@len: length to allocate
446 447 448 449 450 451 452 453 454
 *	@gfp_mask: get_free_pages mask, passed to alloc_skb and alloc_pages
 *
 *	Allocate a new sk_buff for use in NAPI receive.  This buffer will
 *	attempt to allocate the head from a special reserved region used
 *	only for NAPI Rx allocation.  By doing this we can save several
 *	CPU cycles by avoiding having to disable and re-enable IRQs.
 *
 *	%NULL is returned if there is no free memory.
 */
455 456
struct sk_buff *__napi_alloc_skb(struct napi_struct *napi, unsigned int len,
				 gfp_t gfp_mask)
457
{
458
	struct napi_alloc_cache *nc = this_cpu_ptr(&napi_alloc_cache);
459
	struct sk_buff *skb;
460 461 462
	void *data;

	len += NET_SKB_PAD + NET_IP_ALIGN;
463

464
	if ((len > SKB_WITH_OVERHEAD(PAGE_SIZE)) ||
465
	    (gfp_mask & (__GFP_DIRECT_RECLAIM | GFP_DMA))) {
466 467 468 469 470
		skb = __alloc_skb(len, gfp_mask, SKB_ALLOC_RX, NUMA_NO_NODE);
		if (!skb)
			goto skb_fail;
		goto skb_success;
	}
471 472 473 474 475 476

	len += SKB_DATA_ALIGN(sizeof(struct skb_shared_info));
	len = SKB_DATA_ALIGN(len);

	if (sk_memalloc_socks())
		gfp_mask |= __GFP_MEMALLOC;
477

478
	data = page_frag_alloc(&nc->page, len, gfp_mask);
479 480 481 482 483
	if (unlikely(!data))
		return NULL;

	skb = __build_skb(data, len);
	if (unlikely(!skb)) {
484
		skb_free_frag(data);
485
		return NULL;
486 487
	}

488
	/* use OR instead of assignment to avoid clearing of bits in mask */
489
	if (nc->page.pfmemalloc)
490 491 492
		skb->pfmemalloc = 1;
	skb->head_frag = 1;

493
skb_success:
494 495 496
	skb_reserve(skb, NET_SKB_PAD + NET_IP_ALIGN);
	skb->dev = napi->dev;

497
skb_fail:
498 499 500 501
	return skb;
}
EXPORT_SYMBOL(__napi_alloc_skb);

P
Peter Zijlstra 已提交
502
void skb_add_rx_frag(struct sk_buff *skb, int i, struct page *page, int off,
503
		     int size, unsigned int truesize)
P
Peter Zijlstra 已提交
504 505 506 507
{
	skb_fill_page_desc(skb, i, page, off, size);
	skb->len += size;
	skb->data_len += size;
508
	skb->truesize += truesize;
P
Peter Zijlstra 已提交
509 510 511
}
EXPORT_SYMBOL(skb_add_rx_frag);

J
Jason Wang 已提交
512 513 514 515 516 517 518 519 520 521 522 523
void skb_coalesce_rx_frag(struct sk_buff *skb, int i, int size,
			  unsigned int truesize)
{
	skb_frag_t *frag = &skb_shinfo(skb)->frags[i];

	skb_frag_size_add(frag, size);
	skb->len += size;
	skb->data_len += size;
	skb->truesize += truesize;
}
EXPORT_SYMBOL(skb_coalesce_rx_frag);

524
static void skb_drop_list(struct sk_buff **listp)
L
Linus Torvalds 已提交
525
{
E
Eric Dumazet 已提交
526
	kfree_skb_list(*listp);
527
	*listp = NULL;
L
Linus Torvalds 已提交
528 529
}

530 531 532 533 534
static inline void skb_drop_fraglist(struct sk_buff *skb)
{
	skb_drop_list(&skb_shinfo(skb)->frag_list);
}

L
Linus Torvalds 已提交
535 536 537 538
static void skb_clone_fraglist(struct sk_buff *skb)
{
	struct sk_buff *list;

539
	skb_walk_frags(skb, list)
L
Linus Torvalds 已提交
540 541 542
		skb_get(list);
}

543 544
static void skb_free_head(struct sk_buff *skb)
{
545 546
	unsigned char *head = skb->head;

547
	if (skb->head_frag)
548
		skb_free_frag(head);
549
	else
550
		kfree(head);
551 552
}

553
static void skb_release_data(struct sk_buff *skb)
L
Linus Torvalds 已提交
554
{
E
Eric Dumazet 已提交
555 556
	struct skb_shared_info *shinfo = skb_shinfo(skb);
	int i;
L
Linus Torvalds 已提交
557

E
Eric Dumazet 已提交
558 559 560 561
	if (skb->cloned &&
	    atomic_sub_return(skb->nohdr ? (1 << SKB_DATAREF_SHIFT) + 1 : 1,
			      &shinfo->dataref))
		return;
562

E
Eric Dumazet 已提交
563 564
	for (i = 0; i < shinfo->nr_frags; i++)
		__skb_frag_unref(&shinfo->frags[i]);
565

E
Eric Dumazet 已提交
566 567 568
	if (shinfo->frag_list)
		kfree_skb_list(shinfo->frag_list);

W
Willem de Bruijn 已提交
569
	skb_zcopy_clear(skb, true);
E
Eric Dumazet 已提交
570
	skb_free_head(skb);
L
Linus Torvalds 已提交
571 572 573 574 575
}

/*
 *	Free an skbuff by memory without cleaning the state.
 */
576
static void kfree_skbmem(struct sk_buff *skb)
L
Linus Torvalds 已提交
577
{
578
	struct sk_buff_fclones *fclones;
579 580 581 582

	switch (skb->fclone) {
	case SKB_FCLONE_UNAVAILABLE:
		kmem_cache_free(skbuff_head_cache, skb);
583
		return;
584 585

	case SKB_FCLONE_ORIG:
586
		fclones = container_of(skb, struct sk_buff_fclones, skb1);
587

588 589 590
		/* We usually free the clone (TX completion) before original skb
		 * This test would have no chance to be true for the clone,
		 * while here, branch prediction will be good.
591
		 */
592
		if (refcount_read(&fclones->fclone_ref) == 1)
593 594
			goto fastpath;
		break;
595

596 597
	default: /* SKB_FCLONE_CLONE */
		fclones = container_of(skb, struct sk_buff_fclones, skb2);
598
		break;
599
	}
600
	if (!refcount_dec_and_test(&fclones->fclone_ref))
601 602 603
		return;
fastpath:
	kmem_cache_free(skbuff_fclone_cache, fclones);
L
Linus Torvalds 已提交
604 605
}

P
Paolo Abeni 已提交
606
void skb_release_head_state(struct sk_buff *skb)
L
Linus Torvalds 已提交
607
{
E
Eric Dumazet 已提交
608
	skb_dst_drop(skb);
P
Paolo Abeni 已提交
609
	secpath_reset(skb);
610 611
	if (skb->destructor) {
		WARN_ON(in_irq());
L
Linus Torvalds 已提交
612 613
		skb->destructor(skb);
	}
I
Igor Maravić 已提交
614
#if IS_ENABLED(CONFIG_NF_CONNTRACK)
615
	nf_conntrack_put(skb_nfct(skb));
616
#endif
617
#if IS_ENABLED(CONFIG_BRIDGE_NETFILTER)
L
Linus Torvalds 已提交
618 619
	nf_bridge_put(skb->nf_bridge);
#endif
620 621 622 623 624 625
}

/* Free everything but the sk_buff shell. */
static void skb_release_all(struct sk_buff *skb)
{
	skb_release_head_state(skb);
626 627
	if (likely(skb->head))
		skb_release_data(skb);
628 629 630 631 632 633 634 635 636 637
}

/**
 *	__kfree_skb - private function
 *	@skb: buffer
 *
 *	Free an sk_buff. Release anything attached to the buffer.
 *	Clean the state. This is an internal helper function. Users should
 *	always call kfree_skb
 */
L
Linus Torvalds 已提交
638

639 640 641
void __kfree_skb(struct sk_buff *skb)
{
	skb_release_all(skb);
L
Linus Torvalds 已提交
642 643
	kfree_skbmem(skb);
}
644
EXPORT_SYMBOL(__kfree_skb);
L
Linus Torvalds 已提交
645

646 647 648 649 650 651 652 653 654
/**
 *	kfree_skb - free an sk_buff
 *	@skb: buffer to free
 *
 *	Drop a reference to the buffer and free it if the usage count has
 *	hit zero.
 */
void kfree_skb(struct sk_buff *skb)
{
655
	if (!skb_unref(skb))
656
		return;
657

658
	trace_kfree_skb(skb, __builtin_return_address(0));
659 660
	__kfree_skb(skb);
}
661
EXPORT_SYMBOL(kfree_skb);
662

E
Eric Dumazet 已提交
663 664 665 666 667 668 669 670 671 672 673
void kfree_skb_list(struct sk_buff *segs)
{
	while (segs) {
		struct sk_buff *next = segs->next;

		kfree_skb(segs);
		segs = next;
	}
}
EXPORT_SYMBOL(kfree_skb_list);

674 675 676 677 678 679 680 681 682
/**
 *	skb_tx_error - report an sk_buff xmit error
 *	@skb: buffer that triggered an error
 *
 *	Report xmit error if a device callback is tracking this skb.
 *	skb must be freed afterwards.
 */
void skb_tx_error(struct sk_buff *skb)
{
W
Willem de Bruijn 已提交
683
	skb_zcopy_clear(skb, true);
684 685 686
}
EXPORT_SYMBOL(skb_tx_error);

687 688 689 690 691 692 693 694 695 696
/**
 *	consume_skb - free an skbuff
 *	@skb: buffer to free
 *
 *	Drop a ref to the buffer and free it if the usage count has hit zero
 *	Functions identically to kfree_skb, but kfree_skb assumes that the frame
 *	is being dropped after a failure and notes that
 */
void consume_skb(struct sk_buff *skb)
{
697
	if (!skb_unref(skb))
698
		return;
699

700
	trace_consume_skb(skb);
701 702 703 704
	__kfree_skb(skb);
}
EXPORT_SYMBOL(consume_skb);

P
Paolo Abeni 已提交
705 706 707 708
/**
 *	consume_stateless_skb - free an skbuff, assuming it is stateless
 *	@skb: buffer to free
 *
709 710
 *	Alike consume_skb(), but this variant assumes that this is the last
 *	skb reference and all the head states have been already dropped
P
Paolo Abeni 已提交
711
 */
712
void __consume_stateless_skb(struct sk_buff *skb)
P
Paolo Abeni 已提交
713 714
{
	trace_consume_skb(skb);
715
	skb_release_data(skb);
P
Paolo Abeni 已提交
716 717 718
	kfree_skbmem(skb);
}

719 720 721 722 723 724 725 726 727 728 729 730
void __kfree_skb_flush(void)
{
	struct napi_alloc_cache *nc = this_cpu_ptr(&napi_alloc_cache);

	/* flush skb_cache if containing objects */
	if (nc->skb_count) {
		kmem_cache_free_bulk(skbuff_head_cache, nc->skb_count,
				     nc->skb_cache);
		nc->skb_count = 0;
	}
}

731
static inline void _kfree_skb_defer(struct sk_buff *skb)
732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752
{
	struct napi_alloc_cache *nc = this_cpu_ptr(&napi_alloc_cache);

	/* drop skb->head and call any destructors for packet */
	skb_release_all(skb);

	/* record skb to CPU local list */
	nc->skb_cache[nc->skb_count++] = skb;

#ifdef CONFIG_SLUB
	/* SLUB writes into objects when freeing */
	prefetchw(skb);
#endif

	/* flush skb_cache if it is filled */
	if (unlikely(nc->skb_count == NAPI_SKB_CACHE_SIZE)) {
		kmem_cache_free_bulk(skbuff_head_cache, NAPI_SKB_CACHE_SIZE,
				     nc->skb_cache);
		nc->skb_count = 0;
	}
}
753 754 755 756
void __kfree_skb_defer(struct sk_buff *skb)
{
	_kfree_skb_defer(skb);
}
757 758 759 760 761 762

void napi_consume_skb(struct sk_buff *skb, int budget)
{
	if (unlikely(!skb))
		return;

763
	/* Zero budget indicate non-NAPI context called us, like netpoll */
764
	if (unlikely(!budget)) {
765
		dev_consume_skb_any(skb);
766 767 768
		return;
	}

769
	if (!skb_unref(skb))
770
		return;
771

772 773 774 775
	/* if reaching here SKB is ready to free */
	trace_consume_skb(skb);

	/* if SKB is a clone, don't handle this case */
776
	if (skb->fclone != SKB_FCLONE_UNAVAILABLE) {
777 778 779 780
		__kfree_skb(skb);
		return;
	}

781
	_kfree_skb_defer(skb);
782 783 784
}
EXPORT_SYMBOL(napi_consume_skb);

785 786 787 788 789 790 791
/* Make sure a field is enclosed inside headers_start/headers_end section */
#define CHECK_SKB_FIELD(field) \
	BUILD_BUG_ON(offsetof(struct sk_buff, field) <		\
		     offsetof(struct sk_buff, headers_start));	\
	BUILD_BUG_ON(offsetof(struct sk_buff, field) >		\
		     offsetof(struct sk_buff, headers_end));	\

792 793 794
static void __copy_skb_header(struct sk_buff *new, const struct sk_buff *old)
{
	new->tstamp		= old->tstamp;
795
	/* We do not copy old->sk */
796
	new->dev		= old->dev;
797
	memcpy(new->cb, old->cb, sizeof(old->cb));
E
Eric Dumazet 已提交
798
	skb_dst_copy(new, old);
799
#ifdef CONFIG_XFRM
800 801
	new->sp			= secpath_get(old->sp);
#endif
802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831
	__nf_copy(new, old, false);

	/* Note : this field could be in headers_start/headers_end section
	 * It is not yet because we do not want to have a 16 bit hole
	 */
	new->queue_mapping = old->queue_mapping;

	memcpy(&new->headers_start, &old->headers_start,
	       offsetof(struct sk_buff, headers_end) -
	       offsetof(struct sk_buff, headers_start));
	CHECK_SKB_FIELD(protocol);
	CHECK_SKB_FIELD(csum);
	CHECK_SKB_FIELD(hash);
	CHECK_SKB_FIELD(priority);
	CHECK_SKB_FIELD(skb_iif);
	CHECK_SKB_FIELD(vlan_proto);
	CHECK_SKB_FIELD(vlan_tci);
	CHECK_SKB_FIELD(transport_header);
	CHECK_SKB_FIELD(network_header);
	CHECK_SKB_FIELD(mac_header);
	CHECK_SKB_FIELD(inner_protocol);
	CHECK_SKB_FIELD(inner_transport_header);
	CHECK_SKB_FIELD(inner_network_header);
	CHECK_SKB_FIELD(inner_mac_header);
	CHECK_SKB_FIELD(mark);
#ifdef CONFIG_NETWORK_SECMARK
	CHECK_SKB_FIELD(secmark);
#endif
#ifdef CONFIG_NET_RX_BUSY_POLL
	CHECK_SKB_FIELD(napi_id);
832
#endif
E
Eric Dumazet 已提交
833 834 835
#ifdef CONFIG_XPS
	CHECK_SKB_FIELD(sender_cpu);
#endif
836
#ifdef CONFIG_NET_SCHED
837
	CHECK_SKB_FIELD(tc_index);
838
#endif
E
Eliezer Tamir 已提交
839

840 841
}

842 843 844 845
/*
 * You should not add any new code to this function.  Add it to
 * __copy_skb_header above instead.
 */
H
Herbert Xu 已提交
846
static struct sk_buff *__skb_clone(struct sk_buff *n, struct sk_buff *skb)
L
Linus Torvalds 已提交
847 848 849 850 851
{
#define C(x) n->x = skb->x

	n->next = n->prev = NULL;
	n->sk = NULL;
852 853
	__copy_skb_header(n, skb);

L
Linus Torvalds 已提交
854 855
	C(len);
	C(data_len);
856
	C(mac_len);
857
	n->hdr_len = skb->nohdr ? skb_headroom(skb) : skb->hdr_len;
858
	n->cloned = 1;
L
Linus Torvalds 已提交
859
	n->nohdr = 0;
860
	n->peeked = 0;
861
	C(pfmemalloc);
L
Linus Torvalds 已提交
862 863 864
	n->destructor = NULL;
	C(tail);
	C(end);
865
	C(head);
866
	C(head_frag);
867 868
	C(data);
	C(truesize);
869
	refcount_set(&n->users, 1);
L
Linus Torvalds 已提交
870 871 872 873 874

	atomic_inc(&(skb_shinfo(skb)->dataref));
	skb->cloned = 1;

	return n;
H
Herbert Xu 已提交
875 876 877 878 879 880 881 882 883 884 885 886 887 888 889
#undef C
}

/**
 *	skb_morph	-	morph one skb into another
 *	@dst: the skb to receive the contents
 *	@src: the skb to supply the contents
 *
 *	This is identical to skb_clone except that the target skb is
 *	supplied by the user.
 *
 *	The target skb is returned upon exit.
 */
struct sk_buff *skb_morph(struct sk_buff *dst, struct sk_buff *src)
{
890
	skb_release_all(dst);
H
Herbert Xu 已提交
891 892 893 894
	return __skb_clone(dst, src);
}
EXPORT_SYMBOL_GPL(skb_morph);

895
int mm_account_pinned_pages(struct mmpin *mmp, size_t size)
896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923
{
	unsigned long max_pg, num_pg, new_pg, old_pg;
	struct user_struct *user;

	if (capable(CAP_IPC_LOCK) || !size)
		return 0;

	num_pg = (size >> PAGE_SHIFT) + 2;	/* worst case */
	max_pg = rlimit(RLIMIT_MEMLOCK) >> PAGE_SHIFT;
	user = mmp->user ? : current_user();

	do {
		old_pg = atomic_long_read(&user->locked_vm);
		new_pg = old_pg + num_pg;
		if (new_pg > max_pg)
			return -ENOBUFS;
	} while (atomic_long_cmpxchg(&user->locked_vm, old_pg, new_pg) !=
		 old_pg);

	if (!mmp->user) {
		mmp->user = get_uid(user);
		mmp->num_pg = num_pg;
	} else {
		mmp->num_pg += num_pg;
	}

	return 0;
}
924
EXPORT_SYMBOL_GPL(mm_account_pinned_pages);
925

926
void mm_unaccount_pinned_pages(struct mmpin *mmp)
927 928 929 930 931 932
{
	if (mmp->user) {
		atomic_long_sub(mmp->num_pg, &mmp->user->locked_vm);
		free_uid(mmp->user);
	}
}
933
EXPORT_SYMBOL_GPL(mm_unaccount_pinned_pages);
934

W
Willem de Bruijn 已提交
935 936 937 938 939 940 941
struct ubuf_info *sock_zerocopy_alloc(struct sock *sk, size_t size)
{
	struct ubuf_info *uarg;
	struct sk_buff *skb;

	WARN_ON_ONCE(!in_task());

942 943 944
	if (!sock_flag(sk, SOCK_ZEROCOPY))
		return NULL;

W
Willem de Bruijn 已提交
945 946 947 948 949 950
	skb = sock_omalloc(sk, 0, GFP_KERNEL);
	if (!skb)
		return NULL;

	BUILD_BUG_ON(sizeof(*uarg) > sizeof(skb->cb));
	uarg = (void *)skb->cb;
951 952 953 954 955 956
	uarg->mmp.user = NULL;

	if (mm_account_pinned_pages(&uarg->mmp, size)) {
		kfree_skb(skb);
		return NULL;
	}
W
Willem de Bruijn 已提交
957 958

	uarg->callback = sock_zerocopy_callback;
959 960 961
	uarg->id = ((u32)atomic_inc_return(&sk->sk_zckey)) - 1;
	uarg->len = 1;
	uarg->bytelen = size;
W
Willem de Bruijn 已提交
962
	uarg->zerocopy = 1;
963
	refcount_set(&uarg->refcnt, 1);
W
Willem de Bruijn 已提交
964 965 966 967 968 969 970 971 972 973 974
	sock_hold(sk);

	return uarg;
}
EXPORT_SYMBOL_GPL(sock_zerocopy_alloc);

static inline struct sk_buff *skb_from_uarg(struct ubuf_info *uarg)
{
	return container_of((void *)uarg, struct sk_buff, cb);
}

975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999
struct ubuf_info *sock_zerocopy_realloc(struct sock *sk, size_t size,
					struct ubuf_info *uarg)
{
	if (uarg) {
		const u32 byte_limit = 1 << 19;		/* limit to a few TSO */
		u32 bytelen, next;

		/* realloc only when socket is locked (TCP, UDP cork),
		 * so uarg->len and sk_zckey access is serialized
		 */
		if (!sock_owned_by_user(sk)) {
			WARN_ON_ONCE(1);
			return NULL;
		}

		bytelen = uarg->bytelen + size;
		if (uarg->len == USHRT_MAX - 1 || bytelen > byte_limit) {
			/* TCP can create new skb to attach new uarg */
			if (sk->sk_type == SOCK_STREAM)
				goto new_alloc;
			return NULL;
		}

		next = (u32)atomic_read(&sk->sk_zckey);
		if ((u32)(uarg->id + uarg->len) == next) {
1000 1001
			if (mm_account_pinned_pages(&uarg->mmp, size))
				return NULL;
1002 1003 1004
			uarg->len++;
			uarg->bytelen = bytelen;
			atomic_set(&sk->sk_zckey, ++next);
1005
			sock_zerocopy_get(uarg);
1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034
			return uarg;
		}
	}

new_alloc:
	return sock_zerocopy_alloc(sk, size);
}
EXPORT_SYMBOL_GPL(sock_zerocopy_realloc);

static bool skb_zerocopy_notify_extend(struct sk_buff *skb, u32 lo, u16 len)
{
	struct sock_exterr_skb *serr = SKB_EXT_ERR(skb);
	u32 old_lo, old_hi;
	u64 sum_len;

	old_lo = serr->ee.ee_info;
	old_hi = serr->ee.ee_data;
	sum_len = old_hi - old_lo + 1ULL + len;

	if (sum_len >= (1ULL << 32))
		return false;

	if (lo != old_hi + 1)
		return false;

	serr->ee.ee_data += len;
	return true;
}

W
Willem de Bruijn 已提交
1035 1036
void sock_zerocopy_callback(struct ubuf_info *uarg, bool success)
{
1037
	struct sk_buff *tail, *skb = skb_from_uarg(uarg);
W
Willem de Bruijn 已提交
1038 1039
	struct sock_exterr_skb *serr;
	struct sock *sk = skb->sk;
1040 1041 1042 1043
	struct sk_buff_head *q;
	unsigned long flags;
	u32 lo, hi;
	u16 len;
W
Willem de Bruijn 已提交
1044

1045 1046
	mm_unaccount_pinned_pages(&uarg->mmp);

1047 1048 1049 1050
	/* if !len, there was only 1 call, and it was aborted
	 * so do not queue a completion notification
	 */
	if (!uarg->len || sock_flag(sk, SOCK_DEAD))
W
Willem de Bruijn 已提交
1051 1052
		goto release;

1053 1054 1055 1056
	len = uarg->len;
	lo = uarg->id;
	hi = uarg->id + len - 1;

W
Willem de Bruijn 已提交
1057 1058 1059 1060
	serr = SKB_EXT_ERR(skb);
	memset(serr, 0, sizeof(*serr));
	serr->ee.ee_errno = 0;
	serr->ee.ee_origin = SO_EE_ORIGIN_ZEROCOPY;
1061 1062
	serr->ee.ee_data = hi;
	serr->ee.ee_info = lo;
W
Willem de Bruijn 已提交
1063 1064 1065
	if (!success)
		serr->ee.ee_code |= SO_EE_CODE_ZEROCOPY_COPIED;

1066 1067 1068 1069 1070 1071 1072 1073 1074
	q = &sk->sk_error_queue;
	spin_lock_irqsave(&q->lock, flags);
	tail = skb_peek_tail(q);
	if (!tail || SKB_EXT_ERR(tail)->ee.ee_origin != SO_EE_ORIGIN_ZEROCOPY ||
	    !skb_zerocopy_notify_extend(tail, lo, len)) {
		__skb_queue_tail(q, skb);
		skb = NULL;
	}
	spin_unlock_irqrestore(&q->lock, flags);
W
Willem de Bruijn 已提交
1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085

	sk->sk_error_report(sk);

release:
	consume_skb(skb);
	sock_put(sk);
}
EXPORT_SYMBOL_GPL(sock_zerocopy_callback);

void sock_zerocopy_put(struct ubuf_info *uarg)
{
1086
	if (uarg && refcount_dec_and_test(&uarg->refcnt)) {
W
Willem de Bruijn 已提交
1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100
		if (uarg->callback)
			uarg->callback(uarg, uarg->zerocopy);
		else
			consume_skb(skb_from_uarg(uarg));
	}
}
EXPORT_SYMBOL_GPL(sock_zerocopy_put);

void sock_zerocopy_put_abort(struct ubuf_info *uarg)
{
	if (uarg) {
		struct sock *sk = skb_from_uarg(uarg)->sk;

		atomic_dec(&sk->sk_zckey);
1101
		uarg->len--;
W
Willem de Bruijn 已提交
1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114

		sock_zerocopy_put(uarg);
	}
}
EXPORT_SYMBOL_GPL(sock_zerocopy_put_abort);

extern int __zerocopy_sg_from_iter(struct sock *sk, struct sk_buff *skb,
				   struct iov_iter *from, size_t length);

int skb_zerocopy_iter_stream(struct sock *sk, struct sk_buff *skb,
			     struct msghdr *msg, int len,
			     struct ubuf_info *uarg)
{
1115
	struct ubuf_info *orig_uarg = skb_zcopy(skb);
W
Willem de Bruijn 已提交
1116 1117 1118
	struct iov_iter orig_iter = msg->msg_iter;
	int err, orig_len = skb->len;

1119 1120 1121 1122 1123 1124
	/* An skb can only point to one uarg. This edge case happens when
	 * TCP appends to an skb, but zerocopy_realloc triggered a new alloc.
	 */
	if (orig_uarg && uarg != orig_uarg)
		return -EEXIST;

W
Willem de Bruijn 已提交
1125 1126
	err = __zerocopy_sg_from_iter(sk, skb, &msg->msg_iter, len);
	if (err == -EFAULT || (err == -EMSGSIZE && skb->len == orig_len)) {
1127 1128
		struct sock *save_sk = skb->sk;

W
Willem de Bruijn 已提交
1129 1130
		/* Streams do not free skb on error. Reset to prev state. */
		msg->msg_iter = orig_iter;
1131
		skb->sk = sk;
W
Willem de Bruijn 已提交
1132
		___pskb_trim(skb, orig_len);
1133
		skb->sk = save_sk;
W
Willem de Bruijn 已提交
1134 1135 1136 1137 1138 1139 1140 1141
		return err;
	}

	skb_zcopy_set(skb, uarg);
	return skb->len - orig_len;
}
EXPORT_SYMBOL_GPL(skb_zerocopy_iter_stream);

W
Willem de Bruijn 已提交
1142
static int skb_zerocopy_clone(struct sk_buff *nskb, struct sk_buff *orig,
W
Willem de Bruijn 已提交
1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161
			      gfp_t gfp_mask)
{
	if (skb_zcopy(orig)) {
		if (skb_zcopy(nskb)) {
			/* !gfp_mask callers are verified to !skb_zcopy(nskb) */
			if (!gfp_mask) {
				WARN_ON_ONCE(1);
				return -ENOMEM;
			}
			if (skb_uarg(nskb) == skb_uarg(orig))
				return 0;
			if (skb_copy_ubufs(nskb, GFP_ATOMIC))
				return -EIO;
		}
		skb_zcopy_set(nskb, skb_uarg(orig));
	}
	return 0;
}

1162 1163
/**
 *	skb_copy_ubufs	-	copy userspace skb frags buffers to kernel
1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177
 *	@skb: the skb to modify
 *	@gfp_mask: allocation priority
 *
 *	This must be called on SKBTX_DEV_ZEROCOPY skb.
 *	It will copy all frags into kernel and drop the reference
 *	to userspace pages.
 *
 *	If this function is called from an interrupt gfp_mask() must be
 *	%GFP_ATOMIC.
 *
 *	Returns 0 on success or a negative error code on failure
 *	to allocate kernel memory to copy to.
 */
int skb_copy_ubufs(struct sk_buff *skb, gfp_t gfp_mask)
1178 1179 1180
{
	int num_frags = skb_shinfo(skb)->nr_frags;
	struct page *page, *head = NULL;
1181 1182
	int i, new_frags;
	u32 d_off;
1183

1184 1185
	if (skb_shared(skb) || skb_unclone(skb, gfp_mask))
		return -EINVAL;
1186

1187 1188 1189
	if (!num_frags)
		goto release;

1190 1191
	new_frags = (__skb_pagelen(skb) + PAGE_SIZE - 1) >> PAGE_SHIFT;
	for (i = 0; i < new_frags; i++) {
1192
		page = alloc_page(gfp_mask);
1193 1194
		if (!page) {
			while (head) {
1195
				struct page *next = (struct page *)page_private(head);
1196 1197 1198 1199 1200
				put_page(head);
				head = next;
			}
			return -ENOMEM;
		}
1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211
		set_page_private(page, (unsigned long)head);
		head = page;
	}

	page = head;
	d_off = 0;
	for (i = 0; i < num_frags; i++) {
		skb_frag_t *f = &skb_shinfo(skb)->frags[i];
		u32 p_off, p_len, copied;
		struct page *p;
		u8 *vaddr;
1212 1213 1214

		skb_frag_foreach_page(f, f->page_offset, skb_frag_size(f),
				      p, p_off, p_len, copied) {
1215
			u32 copy, done = 0;
1216
			vaddr = kmap_atomic(p);
1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228

			while (done < p_len) {
				if (d_off == PAGE_SIZE) {
					d_off = 0;
					page = (struct page *)page_private(page);
				}
				copy = min_t(u32, PAGE_SIZE - d_off, p_len - done);
				memcpy(page_address(page) + d_off,
				       vaddr + p_off + done, copy);
				done += copy;
				d_off += copy;
			}
1229 1230
			kunmap_atomic(vaddr);
		}
1231 1232 1233
	}

	/* skb frags release userspace buffers */
1234
	for (i = 0; i < num_frags; i++)
1235
		skb_frag_unref(skb, i);
1236 1237

	/* skb frags point to kernel buffers */
1238 1239
	for (i = 0; i < new_frags - 1; i++) {
		__skb_fill_page_desc(skb, i, head, 0, PAGE_SIZE);
1240
		head = (struct page *)page_private(head);
1241
	}
1242 1243
	__skb_fill_page_desc(skb, new_frags - 1, head, 0, d_off);
	skb_shinfo(skb)->nr_frags = new_frags;
1244

1245
release:
W
Willem de Bruijn 已提交
1246
	skb_zcopy_clear(skb, false);
1247 1248
	return 0;
}
1249
EXPORT_SYMBOL_GPL(skb_copy_ubufs);
1250

H
Herbert Xu 已提交
1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266
/**
 *	skb_clone	-	duplicate an sk_buff
 *	@skb: buffer to clone
 *	@gfp_mask: allocation priority
 *
 *	Duplicate an &sk_buff. The new one is not owned by a socket. Both
 *	copies share the same packet data but not structure. The new
 *	buffer has a reference count of 1. If the allocation fails the
 *	function returns %NULL otherwise the new buffer is returned.
 *
 *	If this function is called from an interrupt gfp_mask() must be
 *	%GFP_ATOMIC.
 */

struct sk_buff *skb_clone(struct sk_buff *skb, gfp_t gfp_mask)
{
1267 1268 1269
	struct sk_buff_fclones *fclones = container_of(skb,
						       struct sk_buff_fclones,
						       skb1);
1270
	struct sk_buff *n;
H
Herbert Xu 已提交
1271

1272 1273
	if (skb_orphan_frags(skb, gfp_mask))
		return NULL;
1274

H
Herbert Xu 已提交
1275
	if (skb->fclone == SKB_FCLONE_ORIG &&
1276
	    refcount_read(&fclones->fclone_ref) == 1) {
1277
		n = &fclones->skb2;
1278
		refcount_set(&fclones->fclone_ref, 2);
H
Herbert Xu 已提交
1279
	} else {
1280 1281 1282
		if (skb_pfmemalloc(skb))
			gfp_mask |= __GFP_MEMALLOC;

H
Herbert Xu 已提交
1283 1284 1285
		n = kmem_cache_alloc(skbuff_head_cache, gfp_mask);
		if (!n)
			return NULL;
1286

H
Herbert Xu 已提交
1287 1288 1289 1290
		n->fclone = SKB_FCLONE_UNAVAILABLE;
	}

	return __skb_clone(n, skb);
L
Linus Torvalds 已提交
1291
}
1292
EXPORT_SYMBOL(skb_clone);
L
Linus Torvalds 已提交
1293

1294 1295
static void skb_headers_offset_update(struct sk_buff *skb, int off)
{
E
Eric Dumazet 已提交
1296 1297 1298
	/* Only adjust this if it actually is csum_start rather than csum */
	if (skb->ip_summed == CHECKSUM_PARTIAL)
		skb->csum_start += off;
1299 1300 1301 1302 1303 1304 1305
	/* {transport,network,mac}_header and tail are relative to skb->head */
	skb->transport_header += off;
	skb->network_header   += off;
	if (skb_mac_header_was_set(skb))
		skb->mac_header += off;
	skb->inner_transport_header += off;
	skb->inner_network_header += off;
1306
	skb->inner_mac_header += off;
1307 1308
}

1309
void skb_copy_header(struct sk_buff *new, const struct sk_buff *old)
L
Linus Torvalds 已提交
1310
{
1311 1312
	__copy_skb_header(new, old);

1313 1314 1315
	skb_shinfo(new)->gso_size = skb_shinfo(old)->gso_size;
	skb_shinfo(new)->gso_segs = skb_shinfo(old)->gso_segs;
	skb_shinfo(new)->gso_type = skb_shinfo(old)->gso_type;
L
Linus Torvalds 已提交
1316
}
1317
EXPORT_SYMBOL(skb_copy_header);
L
Linus Torvalds 已提交
1318

1319 1320 1321 1322 1323 1324 1325
static inline int skb_alloc_rx_flag(const struct sk_buff *skb)
{
	if (skb_pfmemalloc(skb))
		return SKB_ALLOC_RX;
	return 0;
}

L
Linus Torvalds 已提交
1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342
/**
 *	skb_copy	-	create private copy of an sk_buff
 *	@skb: buffer to copy
 *	@gfp_mask: allocation priority
 *
 *	Make a copy of both an &sk_buff and its data. This is used when the
 *	caller wishes to modify the data and needs a private copy of the
 *	data to alter. Returns %NULL on failure or the pointer to the buffer
 *	on success. The returned buffer has a reference count of 1.
 *
 *	As by-product this function converts non-linear &sk_buff to linear
 *	one, so that &sk_buff becomes completely private and caller is allowed
 *	to modify all the data of returned buffer. This means that this
 *	function is not recommended for use in circumstances when only
 *	header is going to be modified. Use pskb_copy() instead.
 */

A
Al Viro 已提交
1343
struct sk_buff *skb_copy(const struct sk_buff *skb, gfp_t gfp_mask)
L
Linus Torvalds 已提交
1344
{
E
Eric Dumazet 已提交
1345
	int headerlen = skb_headroom(skb);
1346
	unsigned int size = skb_end_offset(skb) + skb->data_len;
1347 1348
	struct sk_buff *n = __alloc_skb(size, gfp_mask,
					skb_alloc_rx_flag(skb), NUMA_NO_NODE);
E
Eric Dumazet 已提交
1349

L
Linus Torvalds 已提交
1350 1351 1352 1353 1354 1355 1356 1357
	if (!n)
		return NULL;

	/* Set the data pointer */
	skb_reserve(n, headerlen);
	/* Set the tail pointer and length */
	skb_put(n, skb->len);

1358
	BUG_ON(skb_copy_bits(skb, -headerlen, n->head, headerlen + skb->len));
L
Linus Torvalds 已提交
1359

1360
	skb_copy_header(n, skb);
L
Linus Torvalds 已提交
1361 1362
	return n;
}
1363
EXPORT_SYMBOL(skb_copy);
L
Linus Torvalds 已提交
1364 1365

/**
1366
 *	__pskb_copy_fclone	-  create copy of an sk_buff with private head.
L
Linus Torvalds 已提交
1367
 *	@skb: buffer to copy
E
Eric Dumazet 已提交
1368
 *	@headroom: headroom of new skb
L
Linus Torvalds 已提交
1369
 *	@gfp_mask: allocation priority
1370 1371 1372
 *	@fclone: if true allocate the copy of the skb from the fclone
 *	cache instead of the head cache; it is recommended to set this
 *	to true for the cases where the copy will likely be cloned
L
Linus Torvalds 已提交
1373 1374 1375 1376 1377 1378 1379 1380 1381
 *
 *	Make a copy of both an &sk_buff and part of its data, located
 *	in header. Fragmented data remain shared. This is used when
 *	the caller wishes to modify only header of &sk_buff and needs
 *	private copy of the header to alter. Returns %NULL on failure
 *	or the pointer to the buffer on success.
 *	The returned buffer has a reference count of 1.
 */

1382 1383
struct sk_buff *__pskb_copy_fclone(struct sk_buff *skb, int headroom,
				   gfp_t gfp_mask, bool fclone)
L
Linus Torvalds 已提交
1384
{
E
Eric Dumazet 已提交
1385
	unsigned int size = skb_headlen(skb) + headroom;
1386 1387
	int flags = skb_alloc_rx_flag(skb) | (fclone ? SKB_ALLOC_FCLONE : 0);
	struct sk_buff *n = __alloc_skb(size, gfp_mask, flags, NUMA_NO_NODE);
E
Eric Dumazet 已提交
1388

L
Linus Torvalds 已提交
1389 1390 1391 1392
	if (!n)
		goto out;

	/* Set the data pointer */
E
Eric Dumazet 已提交
1393
	skb_reserve(n, headroom);
L
Linus Torvalds 已提交
1394 1395 1396
	/* Set the tail pointer and length */
	skb_put(n, skb_headlen(skb));
	/* Copy the bytes */
1397
	skb_copy_from_linear_data(skb, n->data, n->len);
L
Linus Torvalds 已提交
1398

H
Herbert Xu 已提交
1399
	n->truesize += skb->data_len;
L
Linus Torvalds 已提交
1400 1401 1402 1403 1404 1405
	n->data_len  = skb->data_len;
	n->len	     = skb->len;

	if (skb_shinfo(skb)->nr_frags) {
		int i;

W
Willem de Bruijn 已提交
1406 1407
		if (skb_orphan_frags(skb, gfp_mask) ||
		    skb_zerocopy_clone(n, skb, gfp_mask)) {
1408 1409 1410
			kfree_skb(n);
			n = NULL;
			goto out;
1411
		}
L
Linus Torvalds 已提交
1412 1413
		for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
			skb_shinfo(n)->frags[i] = skb_shinfo(skb)->frags[i];
1414
			skb_frag_ref(skb, i);
L
Linus Torvalds 已提交
1415 1416 1417 1418
		}
		skb_shinfo(n)->nr_frags = i;
	}

1419
	if (skb_has_frag_list(skb)) {
L
Linus Torvalds 已提交
1420 1421 1422 1423
		skb_shinfo(n)->frag_list = skb_shinfo(skb)->frag_list;
		skb_clone_fraglist(n);
	}

1424
	skb_copy_header(n, skb);
L
Linus Torvalds 已提交
1425 1426 1427
out:
	return n;
}
1428
EXPORT_SYMBOL(__pskb_copy_fclone);
L
Linus Torvalds 已提交
1429 1430 1431 1432 1433 1434 1435 1436

/**
 *	pskb_expand_head - reallocate header of &sk_buff
 *	@skb: buffer to reallocate
 *	@nhead: room to add at head
 *	@ntail: room to add at tail
 *	@gfp_mask: allocation priority
 *
M
Mathias Krause 已提交
1437 1438
 *	Expands (or creates identical copy, if @nhead and @ntail are zero)
 *	header of @skb. &sk_buff itself is not changed. &sk_buff MUST have
L
Linus Torvalds 已提交
1439 1440 1441 1442 1443 1444 1445
 *	reference count of 1. Returns zero in the case of success or error,
 *	if expansion failed. In the last case, &sk_buff is not changed.
 *
 *	All the pointers pointing into skb header may change and must be
 *	reloaded after call to this function.
 */

V
Victor Fusco 已提交
1446
int pskb_expand_head(struct sk_buff *skb, int nhead, int ntail,
A
Al Viro 已提交
1447
		     gfp_t gfp_mask)
L
Linus Torvalds 已提交
1448
{
1449 1450
	int i, osize = skb_end_offset(skb);
	int size = osize + nhead + ntail;
L
Linus Torvalds 已提交
1451
	long off;
1452
	u8 *data;
L
Linus Torvalds 已提交
1453

1454 1455
	BUG_ON(nhead < 0);

1456
	BUG_ON(skb_shared(skb));
L
Linus Torvalds 已提交
1457 1458 1459

	size = SKB_DATA_ALIGN(size);

1460 1461 1462 1463
	if (skb_pfmemalloc(skb))
		gfp_mask |= __GFP_MEMALLOC;
	data = kmalloc_reserve(size + SKB_DATA_ALIGN(sizeof(struct skb_shared_info)),
			       gfp_mask, NUMA_NO_NODE, NULL);
L
Linus Torvalds 已提交
1464 1465
	if (!data)
		goto nodata;
1466
	size = SKB_WITH_OVERHEAD(ksize(data));
L
Linus Torvalds 已提交
1467 1468

	/* Copy only real data... and, alas, header. This should be
E
Eric Dumazet 已提交
1469 1470 1471 1472 1473 1474
	 * optimized for the cases when header is void.
	 */
	memcpy(data + nhead, skb->head, skb_tail_pointer(skb) - skb->head);

	memcpy((struct skb_shared_info *)(data + size),
	       skb_shinfo(skb),
1475
	       offsetof(struct skb_shared_info, frags[skb_shinfo(skb)->nr_frags]));
L
Linus Torvalds 已提交
1476

1477 1478 1479 1480 1481 1482
	/*
	 * if shinfo is shared we must drop the old head gracefully, but if it
	 * is not we can just drop the old head and let the existing refcount
	 * be since all we did is relocate the values
	 */
	if (skb_cloned(skb)) {
1483 1484
		if (skb_orphan_frags(skb, gfp_mask))
			goto nofrags;
W
Willem de Bruijn 已提交
1485
		if (skb_zcopy(skb))
1486
			refcount_inc(&skb_uarg(skb)->refcnt);
1487
		for (i = 0; i < skb_shinfo(skb)->nr_frags; i++)
1488
			skb_frag_ref(skb, i);
L
Linus Torvalds 已提交
1489

1490 1491
		if (skb_has_frag_list(skb))
			skb_clone_fraglist(skb);
L
Linus Torvalds 已提交
1492

1493
		skb_release_data(skb);
1494 1495
	} else {
		skb_free_head(skb);
1496
	}
L
Linus Torvalds 已提交
1497 1498 1499
	off = (data + nhead) - skb->head;

	skb->head     = data;
1500
	skb->head_frag = 0;
L
Linus Torvalds 已提交
1501
	skb->data    += off;
1502 1503
#ifdef NET_SKBUFF_DATA_USES_OFFSET
	skb->end      = size;
1504
	off           = nhead;
1505 1506
#else
	skb->end      = skb->head + size;
1507
#endif
1508
	skb->tail	      += off;
1509
	skb_headers_offset_update(skb, nhead);
L
Linus Torvalds 已提交
1510
	skb->cloned   = 0;
1511
	skb->hdr_len  = 0;
L
Linus Torvalds 已提交
1512 1513
	skb->nohdr    = 0;
	atomic_set(&skb_shinfo(skb)->dataref, 1);
1514

1515 1516
	skb_metadata_clear(skb);

1517 1518 1519 1520 1521 1522 1523
	/* It is not generally safe to change skb->truesize.
	 * For the moment, we really care of rx path, or
	 * when skb is orphaned (not attached to a socket).
	 */
	if (!skb->sk || skb->destructor == sock_edemux)
		skb->truesize += size - osize;

L
Linus Torvalds 已提交
1524 1525
	return 0;

1526 1527
nofrags:
	kfree(data);
L
Linus Torvalds 已提交
1528 1529 1530
nodata:
	return -ENOMEM;
}
1531
EXPORT_SYMBOL(pskb_expand_head);
L
Linus Torvalds 已提交
1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551

/* Make private copy of skb with writable head and some headroom */

struct sk_buff *skb_realloc_headroom(struct sk_buff *skb, unsigned int headroom)
{
	struct sk_buff *skb2;
	int delta = headroom - skb_headroom(skb);

	if (delta <= 0)
		skb2 = pskb_copy(skb, GFP_ATOMIC);
	else {
		skb2 = skb_clone(skb, GFP_ATOMIC);
		if (skb2 && pskb_expand_head(skb2, SKB_DATA_ALIGN(delta), 0,
					     GFP_ATOMIC)) {
			kfree_skb(skb2);
			skb2 = NULL;
		}
	}
	return skb2;
}
1552
EXPORT_SYMBOL(skb_realloc_headroom);
L
Linus Torvalds 已提交
1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572

/**
 *	skb_copy_expand	-	copy and expand sk_buff
 *	@skb: buffer to copy
 *	@newheadroom: new free bytes at head
 *	@newtailroom: new free bytes at tail
 *	@gfp_mask: allocation priority
 *
 *	Make a copy of both an &sk_buff and its data and while doing so
 *	allocate additional space.
 *
 *	This is used when the caller wishes to modify the data and needs a
 *	private copy of the data to alter as well as more space for new fields.
 *	Returns %NULL on failure or the pointer to the buffer
 *	on success. The returned buffer has a reference count of 1.
 *
 *	You must pass %GFP_ATOMIC as the allocation priority if this function
 *	is called from an interrupt.
 */
struct sk_buff *skb_copy_expand(const struct sk_buff *skb,
V
Victor Fusco 已提交
1573
				int newheadroom, int newtailroom,
A
Al Viro 已提交
1574
				gfp_t gfp_mask)
L
Linus Torvalds 已提交
1575 1576 1577 1578
{
	/*
	 *	Allocate the copy buffer
	 */
1579 1580 1581
	struct sk_buff *n = __alloc_skb(newheadroom + skb->len + newtailroom,
					gfp_mask, skb_alloc_rx_flag(skb),
					NUMA_NO_NODE);
1582
	int oldheadroom = skb_headroom(skb);
L
Linus Torvalds 已提交
1583 1584 1585 1586 1587 1588 1589 1590 1591 1592
	int head_copy_len, head_copy_off;

	if (!n)
		return NULL;

	skb_reserve(n, newheadroom);

	/* Set the tail pointer and length */
	skb_put(n, skb->len);

1593
	head_copy_len = oldheadroom;
L
Linus Torvalds 已提交
1594 1595 1596 1597 1598 1599 1600
	head_copy_off = 0;
	if (newheadroom <= head_copy_len)
		head_copy_len = newheadroom;
	else
		head_copy_off = newheadroom - head_copy_len;

	/* Copy the linear header and data. */
1601 1602
	BUG_ON(skb_copy_bits(skb, -head_copy_len, n->head + head_copy_off,
			     skb->len + head_copy_len));
L
Linus Torvalds 已提交
1603

1604
	skb_copy_header(n, skb);
L
Linus Torvalds 已提交
1605

E
Eric Dumazet 已提交
1606
	skb_headers_offset_update(n, newheadroom - oldheadroom);
1607

L
Linus Torvalds 已提交
1608 1609
	return n;
}
1610
EXPORT_SYMBOL(skb_copy_expand);
L
Linus Torvalds 已提交
1611 1612

/**
1613
 *	__skb_pad		-	zero pad the tail of an skb
L
Linus Torvalds 已提交
1614 1615
 *	@skb: buffer to pad
 *	@pad: space to pad
1616
 *	@free_on_error: free buffer on error
L
Linus Torvalds 已提交
1617 1618 1619 1620 1621
 *
 *	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.
 *
1622 1623
 *	May return error in out of memory cases. The skb is freed on error
 *	if @free_on_error is true.
L
Linus Torvalds 已提交
1624
 */
1625

1626
int __skb_pad(struct sk_buff *skb, int pad, bool free_on_error)
L
Linus Torvalds 已提交
1627
{
1628 1629
	int err;
	int ntail;
1630

L
Linus Torvalds 已提交
1631
	/* If the skbuff is non linear tailroom is always zero.. */
1632
	if (!skb_cloned(skb) && skb_tailroom(skb) >= pad) {
L
Linus Torvalds 已提交
1633
		memset(skb->data+skb->len, 0, pad);
1634
		return 0;
L
Linus Torvalds 已提交
1635
	}
1636

1637
	ntail = skb->data_len + pad - (skb->end - skb->tail);
1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654
	if (likely(skb_cloned(skb) || ntail > 0)) {
		err = pskb_expand_head(skb, 0, ntail, GFP_ATOMIC);
		if (unlikely(err))
			goto free_skb;
	}

	/* FIXME: The use of this function with non-linear skb's really needs
	 * to be audited.
	 */
	err = skb_linearize(skb);
	if (unlikely(err))
		goto free_skb;

	memset(skb->data + skb->len, 0, pad);
	return 0;

free_skb:
1655 1656
	if (free_on_error)
		kfree_skb(skb);
1657
	return err;
1658
}
1659
EXPORT_SYMBOL(__skb_pad);
1660

M
Mathias Krause 已提交
1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673
/**
 *	pskb_put - add data to the tail of a potentially fragmented buffer
 *	@skb: start of the buffer to use
 *	@tail: tail fragment of the buffer to use
 *	@len: amount of data to add
 *
 *	This function extends the used data area of the potentially
 *	fragmented buffer. @tail must be the last fragment of @skb -- or
 *	@skb itself. If this would exceed the total buffer size the kernel
 *	will panic. A pointer to the first byte of the extra data is
 *	returned.
 */

1674
void *pskb_put(struct sk_buff *skb, struct sk_buff *tail, int len)
M
Mathias Krause 已提交
1675 1676 1677 1678 1679 1680 1681 1682 1683
{
	if (tail != skb) {
		skb->data_len += len;
		skb->len += len;
	}
	return skb_put(tail, len);
}
EXPORT_SYMBOL_GPL(pskb_put);

1684 1685 1686 1687 1688 1689 1690 1691 1692
/**
 *	skb_put - add data to a buffer
 *	@skb: buffer to use
 *	@len: amount of data to add
 *
 *	This function extends the used data area of the buffer. If this would
 *	exceed the total buffer size the kernel will panic. A pointer to the
 *	first byte of the extra data is returned.
 */
1693
void *skb_put(struct sk_buff *skb, unsigned int len)
1694
{
1695
	void *tmp = skb_tail_pointer(skb);
1696 1697 1698 1699 1700 1701 1702 1703 1704
	SKB_LINEAR_ASSERT(skb);
	skb->tail += len;
	skb->len  += len;
	if (unlikely(skb->tail > skb->end))
		skb_over_panic(skb, len, __builtin_return_address(0));
	return tmp;
}
EXPORT_SYMBOL(skb_put);

1705 1706 1707 1708 1709 1710 1711 1712 1713
/**
 *	skb_push - add data to the start of a buffer
 *	@skb: buffer to use
 *	@len: amount of data to add
 *
 *	This function extends the used data area of the buffer at the buffer
 *	start. If this would exceed the total buffer headroom the kernel will
 *	panic. A pointer to the first byte of the extra data is returned.
 */
1714
void *skb_push(struct sk_buff *skb, unsigned int len)
1715 1716 1717 1718 1719 1720 1721 1722 1723
{
	skb->data -= len;
	skb->len  += len;
	if (unlikely(skb->data<skb->head))
		skb_under_panic(skb, len, __builtin_return_address(0));
	return skb->data;
}
EXPORT_SYMBOL(skb_push);

1724 1725 1726 1727 1728 1729 1730 1731 1732 1733
/**
 *	skb_pull - remove data from the start of a buffer
 *	@skb: buffer to use
 *	@len: amount of data to remove
 *
 *	This function removes data from the start of a buffer, returning
 *	the memory to the headroom. A pointer to the next data in the buffer
 *	is returned. Once the data has been pulled future pushes will overwrite
 *	the old data.
 */
1734
void *skb_pull(struct sk_buff *skb, unsigned int len)
1735
{
1736
	return skb_pull_inline(skb, len);
1737 1738 1739
}
EXPORT_SYMBOL(skb_pull);

1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755
/**
 *	skb_trim - remove end from a buffer
 *	@skb: buffer to alter
 *	@len: new length
 *
 *	Cut the length of a buffer down by removing data from the tail. If
 *	the buffer is already under the length specified it is not modified.
 *	The skb must be linear.
 */
void skb_trim(struct sk_buff *skb, unsigned int len)
{
	if (skb->len > len)
		__skb_trim(skb, len);
}
EXPORT_SYMBOL(skb_trim);

1756
/* Trims skb to length len. It can change skb pointers.
L
Linus Torvalds 已提交
1757 1758
 */

1759
int ___pskb_trim(struct sk_buff *skb, unsigned int len)
L
Linus Torvalds 已提交
1760
{
1761 1762
	struct sk_buff **fragp;
	struct sk_buff *frag;
L
Linus Torvalds 已提交
1763 1764 1765
	int offset = skb_headlen(skb);
	int nfrags = skb_shinfo(skb)->nr_frags;
	int i;
1766 1767 1768 1769 1770
	int err;

	if (skb_cloned(skb) &&
	    unlikely((err = pskb_expand_head(skb, 0, 0, GFP_ATOMIC))))
		return err;
L
Linus Torvalds 已提交
1771

1772 1773 1774 1775 1776
	i = 0;
	if (offset >= len)
		goto drop_pages;

	for (; i < nfrags; i++) {
E
Eric Dumazet 已提交
1777
		int end = offset + skb_frag_size(&skb_shinfo(skb)->frags[i]);
1778 1779 1780 1781 1782 1783

		if (end < len) {
			offset = end;
			continue;
		}

E
Eric Dumazet 已提交
1784
		skb_frag_size_set(&skb_shinfo(skb)->frags[i++], len - offset);
1785

1786
drop_pages:
1787 1788 1789
		skb_shinfo(skb)->nr_frags = i;

		for (; i < nfrags; i++)
1790
			skb_frag_unref(skb, i);
1791

1792
		if (skb_has_frag_list(skb))
1793
			skb_drop_fraglist(skb);
1794
		goto done;
1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808
	}

	for (fragp = &skb_shinfo(skb)->frag_list; (frag = *fragp);
	     fragp = &frag->next) {
		int end = offset + frag->len;

		if (skb_shared(frag)) {
			struct sk_buff *nfrag;

			nfrag = skb_clone(frag, GFP_ATOMIC);
			if (unlikely(!nfrag))
				return -ENOMEM;

			nfrag->next = frag->next;
1809
			consume_skb(frag);
1810 1811
			frag = nfrag;
			*fragp = frag;
L
Linus Torvalds 已提交
1812
		}
1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825

		if (end < len) {
			offset = end;
			continue;
		}

		if (end > len &&
		    unlikely((err = pskb_trim(frag, len - offset))))
			return err;

		if (frag->next)
			skb_drop_list(&frag->next);
		break;
L
Linus Torvalds 已提交
1826 1827
	}

1828
done:
1829
	if (len > skb_headlen(skb)) {
L
Linus Torvalds 已提交
1830 1831 1832
		skb->data_len -= skb->len - len;
		skb->len       = len;
	} else {
1833 1834
		skb->len       = len;
		skb->data_len  = 0;
1835
		skb_set_tail_pointer(skb, len);
L
Linus Torvalds 已提交
1836 1837
	}

1838 1839
	if (!skb->sk || skb->destructor == sock_edemux)
		skb_condense(skb);
L
Linus Torvalds 已提交
1840 1841
	return 0;
}
1842
EXPORT_SYMBOL(___pskb_trim);
L
Linus Torvalds 已提交
1843

1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857
/* Note : use pskb_trim_rcsum() instead of calling this directly
 */
int pskb_trim_rcsum_slow(struct sk_buff *skb, unsigned int len)
{
	if (skb->ip_summed == CHECKSUM_COMPLETE) {
		int delta = skb->len - len;

		skb->csum = csum_sub(skb->csum,
				     skb_checksum(skb, len, delta, 0));
	}
	return __pskb_trim(skb, len);
}
EXPORT_SYMBOL(pskb_trim_rcsum_slow);

L
Linus Torvalds 已提交
1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882
/**
 *	__pskb_pull_tail - advance tail of skb header
 *	@skb: buffer to reallocate
 *	@delta: number of bytes to advance tail
 *
 *	The function makes a sense only on a fragmented &sk_buff,
 *	it expands header moving its tail forward and copying necessary
 *	data from fragmented part.
 *
 *	&sk_buff MUST have reference count of 1.
 *
 *	Returns %NULL (and &sk_buff does not change) if pull failed
 *	or value of new tail of skb in the case of success.
 *
 *	All the pointers pointing into skb header may change and must be
 *	reloaded after call to this function.
 */

/* Moves tail of skb head forward, copying data from fragmented part,
 * when it is necessary.
 * 1. It may fail due to malloc failure.
 * 2. It may change skb pointers.
 *
 * It is pretty complicated. Luckily, it is called only in exceptional cases.
 */
1883
void *__pskb_pull_tail(struct sk_buff *skb, int delta)
L
Linus Torvalds 已提交
1884 1885 1886 1887 1888
{
	/* If skb has not enough free space at tail, get new one
	 * plus 128 bytes for future expansions. If we have enough
	 * room at tail, reallocate without expansion only if skb is cloned.
	 */
1889
	int i, k, eat = (skb->tail + delta) - skb->end;
L
Linus Torvalds 已提交
1890 1891 1892 1893 1894 1895 1896

	if (eat > 0 || skb_cloned(skb)) {
		if (pskb_expand_head(skb, 0, eat > 0 ? eat + 128 : 0,
				     GFP_ATOMIC))
			return NULL;
	}

1897 1898
	BUG_ON(skb_copy_bits(skb, skb_headlen(skb),
			     skb_tail_pointer(skb), delta));
L
Linus Torvalds 已提交
1899 1900 1901 1902

	/* Optimization: no fragments, no reasons to preestimate
	 * size of pulled pages. Superb.
	 */
1903
	if (!skb_has_frag_list(skb))
L
Linus Torvalds 已提交
1904 1905 1906 1907 1908
		goto pull_pages;

	/* Estimate size of pulled pages. */
	eat = delta;
	for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
E
Eric Dumazet 已提交
1909 1910 1911
		int size = skb_frag_size(&skb_shinfo(skb)->frags[i]);

		if (size >= eat)
L
Linus Torvalds 已提交
1912
			goto pull_pages;
E
Eric Dumazet 已提交
1913
		eat -= size;
L
Linus Torvalds 已提交
1914 1915 1916
	}

	/* If we need update frag list, we are in troubles.
W
Wenhua Shi 已提交
1917
	 * Certainly, it is possible to add an offset to skb data,
L
Linus Torvalds 已提交
1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928
	 * but taking into account that pulling is expected to
	 * be very rare operation, it is worth to fight against
	 * further bloating skb head and crucify ourselves here instead.
	 * Pure masohism, indeed. 8)8)
	 */
	if (eat) {
		struct sk_buff *list = skb_shinfo(skb)->frag_list;
		struct sk_buff *clone = NULL;
		struct sk_buff *insp = NULL;

		do {
1929
			BUG_ON(!list);
L
Linus Torvalds 已提交
1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951

			if (list->len <= eat) {
				/* Eaten as whole. */
				eat -= list->len;
				list = list->next;
				insp = list;
			} else {
				/* Eaten partially. */

				if (skb_shared(list)) {
					/* Sucks! We need to fork list. :-( */
					clone = skb_clone(list, GFP_ATOMIC);
					if (!clone)
						return NULL;
					insp = list->next;
					list = clone;
				} else {
					/* This may be pulled without
					 * problems. */
					insp = list;
				}
				if (!pskb_pull(list, eat)) {
1952
					kfree_skb(clone);
L
Linus Torvalds 已提交
1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975
					return NULL;
				}
				break;
			}
		} while (eat);

		/* Free pulled out fragments. */
		while ((list = skb_shinfo(skb)->frag_list) != insp) {
			skb_shinfo(skb)->frag_list = list->next;
			kfree_skb(list);
		}
		/* And insert new clone at head. */
		if (clone) {
			clone->next = list;
			skb_shinfo(skb)->frag_list = clone;
		}
	}
	/* Success! Now we may commit changes to skb data. */

pull_pages:
	eat = delta;
	k = 0;
	for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
E
Eric Dumazet 已提交
1976 1977 1978
		int size = skb_frag_size(&skb_shinfo(skb)->frags[i]);

		if (size <= eat) {
1979
			skb_frag_unref(skb, i);
E
Eric Dumazet 已提交
1980
			eat -= size;
L
Linus Torvalds 已提交
1981 1982 1983 1984
		} else {
			skb_shinfo(skb)->frags[k] = skb_shinfo(skb)->frags[i];
			if (eat) {
				skb_shinfo(skb)->frags[k].page_offset += eat;
E
Eric Dumazet 已提交
1985
				skb_frag_size_sub(&skb_shinfo(skb)->frags[k], eat);
1986 1987
				if (!i)
					goto end;
L
Linus Torvalds 已提交
1988 1989 1990 1991 1992 1993 1994
				eat = 0;
			}
			k++;
		}
	}
	skb_shinfo(skb)->nr_frags = k;

1995
end:
L
Linus Torvalds 已提交
1996 1997 1998
	skb->tail     += delta;
	skb->data_len -= delta;

W
Willem de Bruijn 已提交
1999 2000 2001
	if (!skb->data_len)
		skb_zcopy_clear(skb, false);

2002
	return skb_tail_pointer(skb);
L
Linus Torvalds 已提交
2003
}
2004
EXPORT_SYMBOL(__pskb_pull_tail);
L
Linus Torvalds 已提交
2005

2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
/**
 *	skb_copy_bits - copy bits from skb to kernel buffer
 *	@skb: source skb
 *	@offset: offset in source
 *	@to: destination buffer
 *	@len: number of bytes to copy
 *
 *	Copy the specified number of bytes from the source skb to the
 *	destination buffer.
 *
 *	CAUTION ! :
 *		If its prototype is ever changed,
 *		check arch/{*}/net/{*}.S files,
 *		since it is called from BPF assembly code.
 */
L
Linus Torvalds 已提交
2021 2022
int skb_copy_bits(const struct sk_buff *skb, int offset, void *to, int len)
{
2023
	int start = skb_headlen(skb);
2024 2025
	struct sk_buff *frag_iter;
	int i, copy;
L
Linus Torvalds 已提交
2026 2027 2028 2029 2030

	if (offset > (int)skb->len - len)
		goto fault;

	/* Copy header. */
2031
	if ((copy = start - offset) > 0) {
L
Linus Torvalds 已提交
2032 2033
		if (copy > len)
			copy = len;
2034
		skb_copy_from_linear_data_offset(skb, offset, to, copy);
L
Linus Torvalds 已提交
2035 2036 2037 2038 2039 2040 2041
		if ((len -= copy) == 0)
			return 0;
		offset += copy;
		to     += copy;
	}

	for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
2042
		int end;
E
Eric Dumazet 已提交
2043
		skb_frag_t *f = &skb_shinfo(skb)->frags[i];
L
Linus Torvalds 已提交
2044

2045
		WARN_ON(start > offset + len);
2046

E
Eric Dumazet 已提交
2047
		end = start + skb_frag_size(f);
L
Linus Torvalds 已提交
2048
		if ((copy = end - offset) > 0) {
2049 2050
			u32 p_off, p_len, copied;
			struct page *p;
L
Linus Torvalds 已提交
2051 2052 2053 2054 2055
			u8 *vaddr;

			if (copy > len)
				copy = len;

2056 2057 2058 2059 2060 2061 2062
			skb_frag_foreach_page(f,
					      f->page_offset + offset - start,
					      copy, p, p_off, p_len, copied) {
				vaddr = kmap_atomic(p);
				memcpy(to + copied, vaddr + p_off, p_len);
				kunmap_atomic(vaddr);
			}
L
Linus Torvalds 已提交
2063 2064 2065 2066 2067 2068

			if ((len -= copy) == 0)
				return 0;
			offset += copy;
			to     += copy;
		}
2069
		start = end;
L
Linus Torvalds 已提交
2070 2071
	}

2072 2073
	skb_walk_frags(skb, frag_iter) {
		int end;
L
Linus Torvalds 已提交
2074

2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086
		WARN_ON(start > offset + len);

		end = start + frag_iter->len;
		if ((copy = end - offset) > 0) {
			if (copy > len)
				copy = len;
			if (skb_copy_bits(frag_iter, offset - start, to, copy))
				goto fault;
			if ((len -= copy) == 0)
				return 0;
			offset += copy;
			to     += copy;
L
Linus Torvalds 已提交
2087
		}
2088
		start = end;
L
Linus Torvalds 已提交
2089
	}
2090

L
Linus Torvalds 已提交
2091 2092 2093 2094 2095 2096
	if (!len)
		return 0;

fault:
	return -EFAULT;
}
2097
EXPORT_SYMBOL(skb_copy_bits);
L
Linus Torvalds 已提交
2098

J
Jens Axboe 已提交
2099 2100 2101 2102 2103 2104
/*
 * Callback from splice_to_pipe(), if we need to release some pages
 * at the end of the spd in case we error'ed out in filling the pipe.
 */
static void sock_spd_release(struct splice_pipe_desc *spd, unsigned int i)
{
2105 2106
	put_page(spd->pages[i]);
}
J
Jens Axboe 已提交
2107

2108 2109
static struct page *linear_to_page(struct page *page, unsigned int *len,
				   unsigned int *offset,
2110
				   struct sock *sk)
2111
{
2112
	struct page_frag *pfrag = sk_page_frag(sk);
2113

2114 2115
	if (!sk_page_frag_refill(sk, pfrag))
		return NULL;
2116

2117
	*len = min_t(unsigned int, *len, pfrag->size - pfrag->offset);
2118

2119 2120 2121 2122
	memcpy(page_address(pfrag->page) + pfrag->offset,
	       page_address(page) + *offset, *len);
	*offset = pfrag->offset;
	pfrag->offset += *len;
2123

2124
	return pfrag->page;
J
Jens Axboe 已提交
2125 2126
}

E
Eric Dumazet 已提交
2127 2128 2129 2130 2131 2132 2133 2134 2135 2136
static bool spd_can_coalesce(const struct splice_pipe_desc *spd,
			     struct page *page,
			     unsigned int offset)
{
	return	spd->nr_pages &&
		spd->pages[spd->nr_pages - 1] == page &&
		(spd->partial[spd->nr_pages - 1].offset +
		 spd->partial[spd->nr_pages - 1].len == offset);
}

J
Jens Axboe 已提交
2137 2138 2139
/*
 * Fill page/offset/length into spd, if it can hold more pages.
 */
2140 2141 2142
static bool spd_fill_page(struct splice_pipe_desc *spd,
			  struct pipe_inode_info *pipe, struct page *page,
			  unsigned int *len, unsigned int offset,
2143
			  bool linear,
2144
			  struct sock *sk)
J
Jens Axboe 已提交
2145
{
E
Eric Dumazet 已提交
2146
	if (unlikely(spd->nr_pages == MAX_SKB_FRAGS))
2147
		return true;
J
Jens Axboe 已提交
2148

2149
	if (linear) {
2150
		page = linear_to_page(page, len, &offset, sk);
2151
		if (!page)
2152
			return true;
E
Eric Dumazet 已提交
2153 2154 2155
	}
	if (spd_can_coalesce(spd, page, offset)) {
		spd->partial[spd->nr_pages - 1].len += *len;
2156
		return false;
E
Eric Dumazet 已提交
2157 2158
	}
	get_page(page);
J
Jens Axboe 已提交
2159
	spd->pages[spd->nr_pages] = page;
2160
	spd->partial[spd->nr_pages].len = *len;
J
Jens Axboe 已提交
2161 2162
	spd->partial[spd->nr_pages].offset = offset;
	spd->nr_pages++;
2163

2164
	return false;
J
Jens Axboe 已提交
2165 2166
}

2167 2168
static bool __splice_segment(struct page *page, unsigned int poff,
			     unsigned int plen, unsigned int *off,
2169
			     unsigned int *len,
2170
			     struct splice_pipe_desc *spd, bool linear,
2171 2172
			     struct sock *sk,
			     struct pipe_inode_info *pipe)
J
Jens Axboe 已提交
2173
{
2174
	if (!*len)
2175
		return true;
2176 2177 2178 2179

	/* skip this segment if already processed */
	if (*off >= plen) {
		*off -= plen;
2180
		return false;
2181
	}
J
Jens Axboe 已提交
2182

2183
	/* ignore any bits we already processed */
2184 2185 2186
	poff += *off;
	plen -= *off;
	*off = 0;
J
Jens Axboe 已提交
2187

2188 2189
	do {
		unsigned int flen = min(*len, plen);
2190

2191 2192 2193 2194 2195 2196 2197
		if (spd_fill_page(spd, pipe, page, &flen, poff,
				  linear, sk))
			return true;
		poff += flen;
		plen -= flen;
		*len -= flen;
	} while (*len && plen);
2198

2199
	return false;
2200 2201 2202
}

/*
2203
 * Map linear and fragment data from the skb to spd. It reports true if the
2204 2205
 * pipe is full or if we already spliced the requested length.
 */
2206 2207 2208
static bool __skb_splice_bits(struct sk_buff *skb, struct pipe_inode_info *pipe,
			      unsigned int *offset, unsigned int *len,
			      struct splice_pipe_desc *spd, struct sock *sk)
2209 2210
{
	int seg;
2211
	struct sk_buff *iter;
2212

2213
	/* map the linear part :
2214 2215 2216
	 * If skb->head_frag is set, this 'linear' part is backed by a
	 * fragment, and if the head is not shared with any clones then
	 * we can avoid a copy since we own the head portion of this page.
2217 2218 2219 2220
	 */
	if (__splice_segment(virt_to_page(skb->data),
			     (unsigned long) skb->data & (PAGE_SIZE - 1),
			     skb_headlen(skb),
2221
			     offset, len, spd,
2222
			     skb_head_is_locked(skb),
2223
			     sk, pipe))
2224
		return true;
J
Jens Axboe 已提交
2225 2226 2227 2228 2229 2230 2231

	/*
	 * then map the fragments
	 */
	for (seg = 0; seg < skb_shinfo(skb)->nr_frags; seg++) {
		const skb_frag_t *f = &skb_shinfo(skb)->frags[seg];

2232
		if (__splice_segment(skb_frag_page(f),
E
Eric Dumazet 已提交
2233
				     f->page_offset, skb_frag_size(f),
2234
				     offset, len, spd, false, sk, pipe))
2235
			return true;
J
Jens Axboe 已提交
2236 2237
	}

2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250
	skb_walk_frags(skb, iter) {
		if (*offset >= iter->len) {
			*offset -= iter->len;
			continue;
		}
		/* __skb_splice_bits() only fails if the output has no room
		 * left, so no point in going over the frag_list for the error
		 * case.
		 */
		if (__skb_splice_bits(iter, pipe, offset, len, spd, sk))
			return true;
	}

2251
	return false;
J
Jens Axboe 已提交
2252 2253 2254 2255
}

/*
 * Map data from the skb to a pipe. Should handle both the linear part,
2256
 * the fragments, and the frag list.
J
Jens Axboe 已提交
2257
 */
2258
int skb_splice_bits(struct sk_buff *skb, struct sock *sk, unsigned int offset,
J
Jens Axboe 已提交
2259
		    struct pipe_inode_info *pipe, unsigned int tlen,
A
Al Viro 已提交
2260
		    unsigned int flags)
J
Jens Axboe 已提交
2261
{
E
Eric Dumazet 已提交
2262 2263
	struct partial_page partial[MAX_SKB_FRAGS];
	struct page *pages[MAX_SKB_FRAGS];
J
Jens Axboe 已提交
2264 2265 2266
	struct splice_pipe_desc spd = {
		.pages = pages,
		.partial = partial,
2267
		.nr_pages_max = MAX_SKB_FRAGS,
M
Miklos Szeredi 已提交
2268
		.ops = &nosteal_pipe_buf_ops,
J
Jens Axboe 已提交
2269 2270
		.spd_release = sock_spd_release,
	};
2271 2272
	int ret = 0;

2273
	__skb_splice_bits(skb, pipe, &offset, &tlen, &spd, sk);
J
Jens Axboe 已提交
2274

2275
	if (spd.nr_pages)
A
Al Viro 已提交
2276
		ret = splice_to_pipe(pipe, &spd);
J
Jens Axboe 已提交
2277

2278
	return ret;
J
Jens Axboe 已提交
2279
}
2280
EXPORT_SYMBOL_GPL(skb_splice_bits);
J
Jens Axboe 已提交
2281

2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299
/* Send skb data on a socket. Socket must be locked. */
int skb_send_sock_locked(struct sock *sk, struct sk_buff *skb, int offset,
			 int len)
{
	unsigned int orig_len = len;
	struct sk_buff *head = skb;
	unsigned short fragidx;
	int slen, ret;

do_frag_list:

	/* Deal with head data */
	while (offset < skb_headlen(skb) && len) {
		struct kvec kv;
		struct msghdr msg;

		slen = min_t(int, len, skb_headlen(skb) - offset);
		kv.iov_base = skb->data + offset;
J
John Fastabend 已提交
2300
		kv.iov_len = slen;
2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382
		memset(&msg, 0, sizeof(msg));

		ret = kernel_sendmsg_locked(sk, &msg, &kv, 1, slen);
		if (ret <= 0)
			goto error;

		offset += ret;
		len -= ret;
	}

	/* All the data was skb head? */
	if (!len)
		goto out;

	/* Make offset relative to start of frags */
	offset -= skb_headlen(skb);

	/* Find where we are in frag list */
	for (fragidx = 0; fragidx < skb_shinfo(skb)->nr_frags; fragidx++) {
		skb_frag_t *frag  = &skb_shinfo(skb)->frags[fragidx];

		if (offset < frag->size)
			break;

		offset -= frag->size;
	}

	for (; len && fragidx < skb_shinfo(skb)->nr_frags; fragidx++) {
		skb_frag_t *frag  = &skb_shinfo(skb)->frags[fragidx];

		slen = min_t(size_t, len, frag->size - offset);

		while (slen) {
			ret = kernel_sendpage_locked(sk, frag->page.p,
						     frag->page_offset + offset,
						     slen, MSG_DONTWAIT);
			if (ret <= 0)
				goto error;

			len -= ret;
			offset += ret;
			slen -= ret;
		}

		offset = 0;
	}

	if (len) {
		/* Process any frag lists */

		if (skb == head) {
			if (skb_has_frag_list(skb)) {
				skb = skb_shinfo(skb)->frag_list;
				goto do_frag_list;
			}
		} else if (skb->next) {
			skb = skb->next;
			goto do_frag_list;
		}
	}

out:
	return orig_len - len;

error:
	return orig_len == len ? ret : orig_len - len;
}
EXPORT_SYMBOL_GPL(skb_send_sock_locked);

/* Send skb data on a socket. */
int skb_send_sock(struct sock *sk, struct sk_buff *skb, int offset, int len)
{
	int ret = 0;

	lock_sock(sk);
	ret = skb_send_sock_locked(sk, skb, offset, len);
	release_sock(sk);

	return ret;
}
EXPORT_SYMBOL_GPL(skb_send_sock);

2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394
/**
 *	skb_store_bits - store bits from kernel buffer to skb
 *	@skb: destination buffer
 *	@offset: offset in destination
 *	@from: source buffer
 *	@len: number of bytes to copy
 *
 *	Copy the specified number of bytes from the source buffer to the
 *	destination skb.  This function handles all the messy bits of
 *	traversing fragment lists and such.
 */

2395
int skb_store_bits(struct sk_buff *skb, int offset, const void *from, int len)
2396
{
2397
	int start = skb_headlen(skb);
2398 2399
	struct sk_buff *frag_iter;
	int i, copy;
2400 2401 2402 2403

	if (offset > (int)skb->len - len)
		goto fault;

2404
	if ((copy = start - offset) > 0) {
2405 2406
		if (copy > len)
			copy = len;
2407
		skb_copy_to_linear_data_offset(skb, offset, from, copy);
2408 2409 2410 2411 2412 2413 2414 2415
		if ((len -= copy) == 0)
			return 0;
		offset += copy;
		from += copy;
	}

	for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
		skb_frag_t *frag = &skb_shinfo(skb)->frags[i];
2416 2417
		int end;

2418
		WARN_ON(start > offset + len);
2419

E
Eric Dumazet 已提交
2420
		end = start + skb_frag_size(frag);
2421
		if ((copy = end - offset) > 0) {
2422 2423
			u32 p_off, p_len, copied;
			struct page *p;
2424 2425 2426 2427 2428
			u8 *vaddr;

			if (copy > len)
				copy = len;

2429 2430 2431 2432 2433 2434 2435
			skb_frag_foreach_page(frag,
					      frag->page_offset + offset - start,
					      copy, p, p_off, p_len, copied) {
				vaddr = kmap_atomic(p);
				memcpy(vaddr + p_off, from + copied, p_len);
				kunmap_atomic(vaddr);
			}
2436 2437 2438 2439 2440 2441

			if ((len -= copy) == 0)
				return 0;
			offset += copy;
			from += copy;
		}
2442
		start = end;
2443 2444
	}

2445 2446
	skb_walk_frags(skb, frag_iter) {
		int end;
2447

2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460
		WARN_ON(start > offset + len);

		end = start + frag_iter->len;
		if ((copy = end - offset) > 0) {
			if (copy > len)
				copy = len;
			if (skb_store_bits(frag_iter, offset - start,
					   from, copy))
				goto fault;
			if ((len -= copy) == 0)
				return 0;
			offset += copy;
			from += copy;
2461
		}
2462
		start = end;
2463 2464 2465 2466 2467 2468 2469 2470 2471
	}
	if (!len)
		return 0;

fault:
	return -EFAULT;
}
EXPORT_SYMBOL(skb_store_bits);

L
Linus Torvalds 已提交
2472
/* Checksum skb data. */
2473 2474
__wsum __skb_checksum(const struct sk_buff *skb, int offset, int len,
		      __wsum csum, const struct skb_checksum_ops *ops)
L
Linus Torvalds 已提交
2475
{
2476 2477
	int start = skb_headlen(skb);
	int i, copy = start - offset;
2478
	struct sk_buff *frag_iter;
L
Linus Torvalds 已提交
2479 2480 2481 2482 2483 2484
	int pos = 0;

	/* Checksum header. */
	if (copy > 0) {
		if (copy > len)
			copy = len;
2485
		csum = ops->update(skb->data + offset, copy, csum);
L
Linus Torvalds 已提交
2486 2487 2488 2489 2490 2491 2492
		if ((len -= copy) == 0)
			return csum;
		offset += copy;
		pos	= copy;
	}

	for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
2493
		int end;
E
Eric Dumazet 已提交
2494
		skb_frag_t *frag = &skb_shinfo(skb)->frags[i];
2495

2496
		WARN_ON(start > offset + len);
L
Linus Torvalds 已提交
2497

E
Eric Dumazet 已提交
2498
		end = start + skb_frag_size(frag);
L
Linus Torvalds 已提交
2499
		if ((copy = end - offset) > 0) {
2500 2501
			u32 p_off, p_len, copied;
			struct page *p;
2502
			__wsum csum2;
L
Linus Torvalds 已提交
2503 2504 2505 2506
			u8 *vaddr;

			if (copy > len)
				copy = len;
2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517

			skb_frag_foreach_page(frag,
					      frag->page_offset + offset - start,
					      copy, p, p_off, p_len, copied) {
				vaddr = kmap_atomic(p);
				csum2 = ops->update(vaddr + p_off, p_len, 0);
				kunmap_atomic(vaddr);
				csum = ops->combine(csum, csum2, pos, p_len);
				pos += p_len;
			}

L
Linus Torvalds 已提交
2518 2519 2520 2521
			if (!(len -= copy))
				return csum;
			offset += copy;
		}
2522
		start = end;
L
Linus Torvalds 已提交
2523 2524
	}

2525 2526
	skb_walk_frags(skb, frag_iter) {
		int end;
L
Linus Torvalds 已提交
2527

2528 2529 2530 2531 2532 2533 2534
		WARN_ON(start > offset + len);

		end = start + frag_iter->len;
		if ((copy = end - offset) > 0) {
			__wsum csum2;
			if (copy > len)
				copy = len;
2535 2536 2537
			csum2 = __skb_checksum(frag_iter, offset - start,
					       copy, 0, ops);
			csum = ops->combine(csum, csum2, pos, copy);
2538 2539 2540 2541
			if ((len -= copy) == 0)
				return csum;
			offset += copy;
			pos    += copy;
L
Linus Torvalds 已提交
2542
		}
2543
		start = end;
L
Linus Torvalds 已提交
2544
	}
2545
	BUG_ON(len);
L
Linus Torvalds 已提交
2546 2547 2548

	return csum;
}
2549 2550 2551 2552 2553 2554
EXPORT_SYMBOL(__skb_checksum);

__wsum skb_checksum(const struct sk_buff *skb, int offset,
		    int len, __wsum csum)
{
	const struct skb_checksum_ops ops = {
2555
		.update  = csum_partial_ext,
2556 2557 2558 2559 2560
		.combine = csum_block_add_ext,
	};

	return __skb_checksum(skb, offset, len, csum, &ops);
}
2561
EXPORT_SYMBOL(skb_checksum);
L
Linus Torvalds 已提交
2562 2563 2564

/* Both of above in one bottle. */

2565 2566
__wsum skb_copy_and_csum_bits(const struct sk_buff *skb, int offset,
				    u8 *to, int len, __wsum csum)
L
Linus Torvalds 已提交
2567
{
2568 2569
	int start = skb_headlen(skb);
	int i, copy = start - offset;
2570
	struct sk_buff *frag_iter;
L
Linus Torvalds 已提交
2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586
	int pos = 0;

	/* Copy header. */
	if (copy > 0) {
		if (copy > len)
			copy = len;
		csum = csum_partial_copy_nocheck(skb->data + offset, to,
						 copy, csum);
		if ((len -= copy) == 0)
			return csum;
		offset += copy;
		to     += copy;
		pos	= copy;
	}

	for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
2587 2588
		int end;

2589
		WARN_ON(start > offset + len);
L
Linus Torvalds 已提交
2590

E
Eric Dumazet 已提交
2591
		end = start + skb_frag_size(&skb_shinfo(skb)->frags[i]);
L
Linus Torvalds 已提交
2592
		if ((copy = end - offset) > 0) {
2593 2594 2595
			skb_frag_t *frag = &skb_shinfo(skb)->frags[i];
			u32 p_off, p_len, copied;
			struct page *p;
2596
			__wsum csum2;
L
Linus Torvalds 已提交
2597 2598 2599 2600
			u8 *vaddr;

			if (copy > len)
				copy = len;
2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613

			skb_frag_foreach_page(frag,
					      frag->page_offset + offset - start,
					      copy, p, p_off, p_len, copied) {
				vaddr = kmap_atomic(p);
				csum2 = csum_partial_copy_nocheck(vaddr + p_off,
								  to + copied,
								  p_len, 0);
				kunmap_atomic(vaddr);
				csum = csum_block_add(csum, csum2, pos);
				pos += p_len;
			}

L
Linus Torvalds 已提交
2614 2615 2616 2617 2618
			if (!(len -= copy))
				return csum;
			offset += copy;
			to     += copy;
		}
2619
		start = end;
L
Linus Torvalds 已提交
2620 2621
	}

2622 2623 2624
	skb_walk_frags(skb, frag_iter) {
		__wsum csum2;
		int end;
L
Linus Torvalds 已提交
2625

2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640
		WARN_ON(start > offset + len);

		end = start + frag_iter->len;
		if ((copy = end - offset) > 0) {
			if (copy > len)
				copy = len;
			csum2 = skb_copy_and_csum_bits(frag_iter,
						       offset - start,
						       to, copy, 0);
			csum = csum_block_add(csum, csum2, pos);
			if ((len -= copy) == 0)
				return csum;
			offset += copy;
			to     += copy;
			pos    += copy;
L
Linus Torvalds 已提交
2641
		}
2642
		start = end;
L
Linus Torvalds 已提交
2643
	}
2644
	BUG_ON(len);
L
Linus Torvalds 已提交
2645 2646
	return csum;
}
2647
EXPORT_SYMBOL(skb_copy_and_csum_bits);
L
Linus Torvalds 已提交
2648

2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674
static __wsum warn_crc32c_csum_update(const void *buff, int len, __wsum sum)
{
	net_warn_ratelimited(
		"%s: attempt to compute crc32c without libcrc32c.ko\n",
		__func__);
	return 0;
}

static __wsum warn_crc32c_csum_combine(__wsum csum, __wsum csum2,
				       int offset, int len)
{
	net_warn_ratelimited(
		"%s: attempt to compute crc32c without libcrc32c.ko\n",
		__func__);
	return 0;
}

static const struct skb_checksum_ops default_crc32c_ops = {
	.update  = warn_crc32c_csum_update,
	.combine = warn_crc32c_csum_combine,
};

const struct skb_checksum_ops *crc32c_csum_stub __read_mostly =
	&default_crc32c_ops;
EXPORT_SYMBOL(crc32c_csum_stub);

2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701
 /**
 *	skb_zerocopy_headlen - Calculate headroom needed for skb_zerocopy()
 *	@from: source buffer
 *
 *	Calculates the amount of linear headroom needed in the 'to' skb passed
 *	into skb_zerocopy().
 */
unsigned int
skb_zerocopy_headlen(const struct sk_buff *from)
{
	unsigned int hlen = 0;

	if (!from->head_frag ||
	    skb_headlen(from) < L1_CACHE_BYTES ||
	    skb_shinfo(from)->nr_frags >= MAX_SKB_FRAGS)
		hlen = skb_headlen(from);

	if (skb_has_frag_list(from))
		hlen = from->len;

	return hlen;
}
EXPORT_SYMBOL_GPL(skb_zerocopy_headlen);

/**
 *	skb_zerocopy - Zero copy skb to skb
 *	@to: destination buffer
2702
 *	@from: source buffer
2703 2704 2705 2706 2707 2708 2709 2710
 *	@len: number of bytes to copy from source buffer
 *	@hlen: size of linear headroom in destination buffer
 *
 *	Copies up to `len` bytes from `from` to `to` by creating references
 *	to the frags in the source buffer.
 *
 *	The `hlen` as calculated by skb_zerocopy_headlen() specifies the
 *	headroom in the `to` buffer.
2711 2712 2713 2714 2715
 *
 *	Return value:
 *	0: everything is OK
 *	-ENOMEM: couldn't orphan frags of @from due to lack of memory
 *	-EFAULT: skb_copy_bits() found some problem with skb geometry
2716
 */
2717 2718
int
skb_zerocopy(struct sk_buff *to, struct sk_buff *from, int len, int hlen)
2719 2720 2721
{
	int i, j = 0;
	int plen = 0; /* length of skb->head fragment */
2722
	int ret;
2723 2724 2725 2726 2727 2728
	struct page *page;
	unsigned int offset;

	BUG_ON(!from->head_frag && !hlen);

	/* dont bother with small payloads */
2729 2730
	if (len <= skb_tailroom(to))
		return skb_copy_bits(from, 0, skb_put(to, len), len);
2731 2732

	if (hlen) {
2733 2734 2735
		ret = skb_copy_bits(from, 0, skb_put(to, hlen), hlen);
		if (unlikely(ret))
			return ret;
2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752
		len -= hlen;
	} else {
		plen = min_t(int, skb_headlen(from), len);
		if (plen) {
			page = virt_to_head_page(from->head);
			offset = from->data - (unsigned char *)page_address(page);
			__skb_fill_page_desc(to, 0, page, offset, plen);
			get_page(page);
			j = 1;
			len -= plen;
		}
	}

	to->truesize += len + plen;
	to->len += len + plen;
	to->data_len += len + plen;

2753 2754 2755 2756
	if (unlikely(skb_orphan_frags(from, GFP_ATOMIC))) {
		skb_tx_error(from);
		return -ENOMEM;
	}
W
Willem de Bruijn 已提交
2757
	skb_zerocopy_clone(to, from, GFP_ATOMIC);
2758

2759 2760 2761 2762 2763 2764 2765 2766 2767 2768
	for (i = 0; i < skb_shinfo(from)->nr_frags; i++) {
		if (!len)
			break;
		skb_shinfo(to)->frags[j] = skb_shinfo(from)->frags[i];
		skb_shinfo(to)->frags[j].size = min_t(int, skb_shinfo(to)->frags[j].size, len);
		len -= skb_shinfo(to)->frags[j].size;
		skb_frag_ref(to, j);
		j++;
	}
	skb_shinfo(to)->nr_frags = j;
2769 2770

	return 0;
2771 2772 2773
}
EXPORT_SYMBOL_GPL(skb_zerocopy);

L
Linus Torvalds 已提交
2774 2775
void skb_copy_and_csum_dev(const struct sk_buff *skb, u8 *to)
{
2776
	__wsum csum;
L
Linus Torvalds 已提交
2777 2778
	long csstart;

2779
	if (skb->ip_summed == CHECKSUM_PARTIAL)
2780
		csstart = skb_checksum_start_offset(skb);
L
Linus Torvalds 已提交
2781 2782 2783
	else
		csstart = skb_headlen(skb);

2784
	BUG_ON(csstart > skb_headlen(skb));
L
Linus Torvalds 已提交
2785

2786
	skb_copy_from_linear_data(skb, to, csstart);
L
Linus Torvalds 已提交
2787 2788 2789 2790 2791 2792

	csum = 0;
	if (csstart != skb->len)
		csum = skb_copy_and_csum_bits(skb, csstart, to + csstart,
					      skb->len - csstart, 0);

2793
	if (skb->ip_summed == CHECKSUM_PARTIAL) {
A
Al Viro 已提交
2794
		long csstuff = csstart + skb->csum_offset;
L
Linus Torvalds 已提交
2795

2796
		*((__sum16 *)(to + csstuff)) = csum_fold(csum);
L
Linus Torvalds 已提交
2797 2798
	}
}
2799
EXPORT_SYMBOL(skb_copy_and_csum_dev);
L
Linus Torvalds 已提交
2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819

/**
 *	skb_dequeue - remove from the head of the queue
 *	@list: list to dequeue from
 *
 *	Remove the head of the list. The list lock is taken so the function
 *	may be used safely with other locking list functions. The head item is
 *	returned or %NULL if the list is empty.
 */

struct sk_buff *skb_dequeue(struct sk_buff_head *list)
{
	unsigned long flags;
	struct sk_buff *result;

	spin_lock_irqsave(&list->lock, flags);
	result = __skb_dequeue(list);
	spin_unlock_irqrestore(&list->lock, flags);
	return result;
}
2820
EXPORT_SYMBOL(skb_dequeue);
L
Linus Torvalds 已提交
2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839

/**
 *	skb_dequeue_tail - remove from the tail of the queue
 *	@list: list to dequeue from
 *
 *	Remove the tail of the list. The list lock is taken so the function
 *	may be used safely with other locking list functions. The tail item is
 *	returned or %NULL if the list is empty.
 */
struct sk_buff *skb_dequeue_tail(struct sk_buff_head *list)
{
	unsigned long flags;
	struct sk_buff *result;

	spin_lock_irqsave(&list->lock, flags);
	result = __skb_dequeue_tail(list);
	spin_unlock_irqrestore(&list->lock, flags);
	return result;
}
2840
EXPORT_SYMBOL(skb_dequeue_tail);
L
Linus Torvalds 已提交
2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855

/**
 *	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 takes the list
 *	lock and is atomic with respect to other list locking functions.
 */
void skb_queue_purge(struct sk_buff_head *list)
{
	struct sk_buff *skb;
	while ((skb = skb_dequeue(list)) != NULL)
		kfree_skb(skb);
}
2856
EXPORT_SYMBOL(skb_queue_purge);
L
Linus Torvalds 已提交
2857

2858 2859 2860 2861 2862 2863 2864 2865 2866 2867 2868
/**
 *	skb_rbtree_purge - empty a skb rbtree
 *	@root: root of the rbtree to empty
 *
 *	Delete all buffers on an &sk_buff rbtree. Each buffer is removed from
 *	the list and one reference dropped. This function does not take
 *	any lock. Synchronization should be handled by the caller (e.g., TCP
 *	out-of-order queue is protected by the socket lock).
 */
void skb_rbtree_purge(struct rb_root *root)
{
E
Eric Dumazet 已提交
2869
	struct rb_node *p = rb_first(root);
2870

E
Eric Dumazet 已提交
2871 2872
	while (p) {
		struct sk_buff *skb = rb_entry(p, struct sk_buff, rbnode);
2873

E
Eric Dumazet 已提交
2874 2875 2876 2877
		p = rb_next(p);
		rb_erase(&skb->rbnode, root);
		kfree_skb(skb);
	}
2878 2879
}

L
Linus Torvalds 已提交
2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898
/**
 *	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 the list. This function takes the
 *	list lock and can be used safely with other locking &sk_buff functions
 *	safely.
 *
 *	A buffer cannot be placed on two lists at the same time.
 */
void skb_queue_head(struct sk_buff_head *list, struct sk_buff *newsk)
{
	unsigned long flags;

	spin_lock_irqsave(&list->lock, flags);
	__skb_queue_head(list, newsk);
	spin_unlock_irqrestore(&list->lock, flags);
}
2899
EXPORT_SYMBOL(skb_queue_head);
L
Linus Torvalds 已提交
2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919

/**
 *	skb_queue_tail - queue a buffer at the list tail
 *	@list: list to use
 *	@newsk: buffer to queue
 *
 *	Queue a buffer at the tail of the list. This function takes the
 *	list lock and can be used safely with other locking &sk_buff functions
 *	safely.
 *
 *	A buffer cannot be placed on two lists at the same time.
 */
void skb_queue_tail(struct sk_buff_head *list, struct sk_buff *newsk)
{
	unsigned long flags;

	spin_lock_irqsave(&list->lock, flags);
	__skb_queue_tail(list, newsk);
	spin_unlock_irqrestore(&list->lock, flags);
}
2920
EXPORT_SYMBOL(skb_queue_tail);
D
David S. Miller 已提交
2921

L
Linus Torvalds 已提交
2922 2923 2924
/**
 *	skb_unlink	-	remove a buffer from a list
 *	@skb: buffer to remove
D
David S. Miller 已提交
2925
 *	@list: list to use
L
Linus Torvalds 已提交
2926
 *
D
David S. Miller 已提交
2927 2928
 *	Remove a packet from a list. The list locks are taken and this
 *	function is atomic with respect to other list locked calls
L
Linus Torvalds 已提交
2929
 *
D
David S. Miller 已提交
2930
 *	You must know what list the SKB is on.
L
Linus Torvalds 已提交
2931
 */
D
David S. Miller 已提交
2932
void skb_unlink(struct sk_buff *skb, struct sk_buff_head *list)
L
Linus Torvalds 已提交
2933
{
D
David S. Miller 已提交
2934
	unsigned long flags;
L
Linus Torvalds 已提交
2935

D
David S. Miller 已提交
2936 2937 2938
	spin_lock_irqsave(&list->lock, flags);
	__skb_unlink(skb, list);
	spin_unlock_irqrestore(&list->lock, flags);
L
Linus Torvalds 已提交
2939
}
2940
EXPORT_SYMBOL(skb_unlink);
L
Linus Torvalds 已提交
2941 2942 2943 2944 2945

/**
 *	skb_append	-	append a buffer
 *	@old: buffer to insert after
 *	@newsk: buffer to insert
D
David S. Miller 已提交
2946
 *	@list: list to use
L
Linus Torvalds 已提交
2947 2948 2949 2950 2951
 *
 *	Place a packet after a given packet in a list. The list locks are taken
 *	and this function is atomic with respect to other list locked calls.
 *	A buffer cannot be placed on two lists at the same time.
 */
D
David S. Miller 已提交
2952
void skb_append(struct sk_buff *old, struct sk_buff *newsk, struct sk_buff_head *list)
L
Linus Torvalds 已提交
2953 2954 2955
{
	unsigned long flags;

D
David S. Miller 已提交
2956
	spin_lock_irqsave(&list->lock, flags);
2957
	__skb_queue_after(list, old, newsk);
D
David S. Miller 已提交
2958
	spin_unlock_irqrestore(&list->lock, flags);
L
Linus Torvalds 已提交
2959
}
2960
EXPORT_SYMBOL(skb_append);
L
Linus Torvalds 已提交
2961 2962 2963 2964 2965

/**
 *	skb_insert	-	insert a buffer
 *	@old: buffer to insert before
 *	@newsk: buffer to insert
D
David S. Miller 已提交
2966 2967 2968 2969 2970
 *	@list: list to use
 *
 *	Place a packet before a given packet in a list. The list locks are
 * 	taken and this function is atomic with respect to other list locked
 *	calls.
L
Linus Torvalds 已提交
2971 2972 2973
 *
 *	A buffer cannot be placed on two lists at the same time.
 */
D
David S. Miller 已提交
2974
void skb_insert(struct sk_buff *old, struct sk_buff *newsk, struct sk_buff_head *list)
L
Linus Torvalds 已提交
2975 2976 2977
{
	unsigned long flags;

D
David S. Miller 已提交
2978 2979 2980
	spin_lock_irqsave(&list->lock, flags);
	__skb_insert(newsk, old->prev, old, list);
	spin_unlock_irqrestore(&list->lock, flags);
L
Linus Torvalds 已提交
2981
}
2982
EXPORT_SYMBOL(skb_insert);
L
Linus Torvalds 已提交
2983 2984 2985 2986 2987 2988 2989

static inline void skb_split_inside_header(struct sk_buff *skb,
					   struct sk_buff* skb1,
					   const u32 len, const int pos)
{
	int i;

2990 2991
	skb_copy_from_linear_data_offset(skb, len, skb_put(skb1, pos - len),
					 pos - len);
L
Linus Torvalds 已提交
2992 2993 2994 2995 2996 2997 2998 2999 3000 3001
	/* And move data appendix as is. */
	for (i = 0; i < skb_shinfo(skb)->nr_frags; i++)
		skb_shinfo(skb1)->frags[i] = skb_shinfo(skb)->frags[i];

	skb_shinfo(skb1)->nr_frags = skb_shinfo(skb)->nr_frags;
	skb_shinfo(skb)->nr_frags  = 0;
	skb1->data_len		   = skb->data_len;
	skb1->len		   += skb1->data_len;
	skb->data_len		   = 0;
	skb->len		   = len;
3002
	skb_set_tail_pointer(skb, len);
L
Linus Torvalds 已提交
3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017
}

static inline void skb_split_no_header(struct sk_buff *skb,
				       struct sk_buff* skb1,
				       const u32 len, int pos)
{
	int i, k = 0;
	const int nfrags = skb_shinfo(skb)->nr_frags;

	skb_shinfo(skb)->nr_frags = 0;
	skb1->len		  = skb1->data_len = skb->len - len;
	skb->len		  = len;
	skb->data_len		  = len - pos;

	for (i = 0; i < nfrags; i++) {
E
Eric Dumazet 已提交
3018
		int size = skb_frag_size(&skb_shinfo(skb)->frags[i]);
L
Linus Torvalds 已提交
3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031

		if (pos + size > len) {
			skb_shinfo(skb1)->frags[k] = skb_shinfo(skb)->frags[i];

			if (pos < len) {
				/* Split frag.
				 * We have two variants in this case:
				 * 1. Move all the frag to the second
				 *    part, if it is possible. F.e.
				 *    this approach is mandatory for TUX,
				 *    where splitting is expensive.
				 * 2. Split is accurately. We make this.
				 */
3032
				skb_frag_ref(skb, i);
L
Linus Torvalds 已提交
3033
				skb_shinfo(skb1)->frags[0].page_offset += len - pos;
E
Eric Dumazet 已提交
3034 3035
				skb_frag_size_sub(&skb_shinfo(skb1)->frags[0], len - pos);
				skb_frag_size_set(&skb_shinfo(skb)->frags[i], len - pos);
L
Linus Torvalds 已提交
3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055
				skb_shinfo(skb)->nr_frags++;
			}
			k++;
		} else
			skb_shinfo(skb)->nr_frags++;
		pos += size;
	}
	skb_shinfo(skb1)->nr_frags = k;
}

/**
 * skb_split - Split fragmented skb to two parts at length len.
 * @skb: the buffer to split
 * @skb1: the buffer to receive the second part
 * @len: new length for skb
 */
void skb_split(struct sk_buff *skb, struct sk_buff *skb1, const u32 len)
{
	int pos = skb_headlen(skb);

3056 3057
	skb_shinfo(skb1)->tx_flags |= skb_shinfo(skb)->tx_flags &
				      SKBTX_SHARED_FRAG;
W
Willem de Bruijn 已提交
3058
	skb_zerocopy_clone(skb1, skb, 0);
L
Linus Torvalds 已提交
3059 3060 3061 3062 3063
	if (len < pos)	/* Split line is inside header. */
		skb_split_inside_header(skb, skb1, len, pos);
	else		/* Second chunk has no header, nothing to copy. */
		skb_split_no_header(skb, skb1, len, pos);
}
3064
EXPORT_SYMBOL(skb_split);
L
Linus Torvalds 已提交
3065

3066 3067 3068 3069
/* Shifting from/to a cloned skb is a no-go.
 *
 * Caller cannot keep skb_shinfo related pointers past calling here!
 */
3070 3071
static int skb_prepare_for_shift(struct sk_buff *skb)
{
3072
	return skb_cloned(skb) && pskb_expand_head(skb, 0, 0, GFP_ATOMIC);
3073 3074 3075 3076 3077 3078 3079 3080 3081
}

/**
 * skb_shift - Shifts paged data partially from skb to another
 * @tgt: buffer into which tail data gets added
 * @skb: buffer from which the paged data comes from
 * @shiftlen: shift up to this many bytes
 *
 * Attempts to shift up to shiftlen worth of bytes, which may be less than
F
Feng King 已提交
3082
 * the length of the skb, from skb to tgt. Returns number bytes shifted.
3083 3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098
 * It's up to caller to free skb if everything was shifted.
 *
 * If @tgt runs out of frags, the whole operation is aborted.
 *
 * Skb cannot include anything else but paged data while tgt is allowed
 * to have non-paged data as well.
 *
 * TODO: full sized shift could be optimized but that would need
 * specialized skb free'er to handle frags without up-to-date nr_frags.
 */
int skb_shift(struct sk_buff *tgt, struct sk_buff *skb, int shiftlen)
{
	int from, to, merge, todo;
	struct skb_frag_struct *fragfrom, *fragto;

	BUG_ON(shiftlen > skb->len);
3099 3100 3101

	if (skb_headlen(skb))
		return 0;
W
Willem de Bruijn 已提交
3102 3103
	if (skb_zcopy(tgt) || skb_zcopy(skb))
		return 0;
3104 3105 3106 3107 3108 3109 3110 3111 3112 3113

	todo = shiftlen;
	from = 0;
	to = skb_shinfo(tgt)->nr_frags;
	fragfrom = &skb_shinfo(skb)->frags[from];

	/* Actual merge is delayed until the point when we know we can
	 * commit all, so that we don't have to undo partial changes
	 */
	if (!to ||
3114 3115
	    !skb_can_coalesce(tgt, to, skb_frag_page(fragfrom),
			      fragfrom->page_offset)) {
3116 3117 3118 3119
		merge = -1;
	} else {
		merge = to - 1;

E
Eric Dumazet 已提交
3120
		todo -= skb_frag_size(fragfrom);
3121 3122 3123 3124 3125
		if (todo < 0) {
			if (skb_prepare_for_shift(skb) ||
			    skb_prepare_for_shift(tgt))
				return 0;

3126 3127
			/* All previous frag pointers might be stale! */
			fragfrom = &skb_shinfo(skb)->frags[from];
3128 3129
			fragto = &skb_shinfo(tgt)->frags[merge];

E
Eric Dumazet 已提交
3130 3131
			skb_frag_size_add(fragto, shiftlen);
			skb_frag_size_sub(fragfrom, shiftlen);
3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154
			fragfrom->page_offset += shiftlen;

			goto onlymerged;
		}

		from++;
	}

	/* Skip full, not-fitting skb to avoid expensive operations */
	if ((shiftlen == skb->len) &&
	    (skb_shinfo(skb)->nr_frags - from) > (MAX_SKB_FRAGS - to))
		return 0;

	if (skb_prepare_for_shift(skb) || skb_prepare_for_shift(tgt))
		return 0;

	while ((todo > 0) && (from < skb_shinfo(skb)->nr_frags)) {
		if (to == MAX_SKB_FRAGS)
			return 0;

		fragfrom = &skb_shinfo(skb)->frags[from];
		fragto = &skb_shinfo(tgt)->frags[to];

E
Eric Dumazet 已提交
3155
		if (todo >= skb_frag_size(fragfrom)) {
3156
			*fragto = *fragfrom;
E
Eric Dumazet 已提交
3157
			todo -= skb_frag_size(fragfrom);
3158 3159 3160 3161
			from++;
			to++;

		} else {
3162
			__skb_frag_ref(fragfrom);
3163 3164
			fragto->page = fragfrom->page;
			fragto->page_offset = fragfrom->page_offset;
E
Eric Dumazet 已提交
3165
			skb_frag_size_set(fragto, todo);
3166 3167

			fragfrom->page_offset += todo;
E
Eric Dumazet 已提交
3168
			skb_frag_size_sub(fragfrom, todo);
3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182
			todo = 0;

			to++;
			break;
		}
	}

	/* Ready to "commit" this state change to tgt */
	skb_shinfo(tgt)->nr_frags = to;

	if (merge >= 0) {
		fragfrom = &skb_shinfo(skb)->frags[0];
		fragto = &skb_shinfo(tgt)->frags[merge];

E
Eric Dumazet 已提交
3183
		skb_frag_size_add(fragto, skb_frag_size(fragfrom));
3184
		__skb_frag_unref(fragfrom);
3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212
	}

	/* Reposition in the original skb */
	to = 0;
	while (from < skb_shinfo(skb)->nr_frags)
		skb_shinfo(skb)->frags[to++] = skb_shinfo(skb)->frags[from++];
	skb_shinfo(skb)->nr_frags = to;

	BUG_ON(todo > 0 && !skb_shinfo(skb)->nr_frags);

onlymerged:
	/* Most likely the tgt won't ever need its checksum anymore, skb on
	 * the other hand might need it if it needs to be resent
	 */
	tgt->ip_summed = CHECKSUM_PARTIAL;
	skb->ip_summed = CHECKSUM_PARTIAL;

	/* Yak, is it really working this way? Some helper please? */
	skb->len -= shiftlen;
	skb->data_len -= shiftlen;
	skb->truesize -= shiftlen;
	tgt->len += shiftlen;
	tgt->data_len += shiftlen;
	tgt->truesize += shiftlen;

	return shiftlen;
}

3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231
/**
 * skb_prepare_seq_read - Prepare a sequential read of skb data
 * @skb: the buffer to read
 * @from: lower offset of data to be read
 * @to: upper offset of data to be read
 * @st: state variable
 *
 * Initializes the specified state variable. Must be called before
 * invoking skb_seq_read() for the first time.
 */
void skb_prepare_seq_read(struct sk_buff *skb, unsigned int from,
			  unsigned int to, struct skb_seq_state *st)
{
	st->lower_offset = from;
	st->upper_offset = to;
	st->root_skb = st->cur_skb = skb;
	st->frag_idx = st->stepped_offset = 0;
	st->frag_data = NULL;
}
3232
EXPORT_SYMBOL(skb_prepare_seq_read);
3233 3234 3235 3236 3237 3238 3239

/**
 * skb_seq_read - Sequentially read skb data
 * @consumed: number of bytes consumed by the caller so far
 * @data: destination pointer for data to be returned
 * @st: state variable
 *
M
Mathias Krause 已提交
3240
 * Reads a block of skb data at @consumed relative to the
3241
 * lower offset specified to skb_prepare_seq_read(). Assigns
M
Mathias Krause 已提交
3242
 * the head of the data block to @data and returns the length
3243 3244 3245 3246
 * of the block or 0 if the end of the skb data or the upper
 * offset has been reached.
 *
 * The caller is not required to consume all of the data
M
Mathias Krause 已提交
3247
 * returned, i.e. @consumed is typically set to the number
3248 3249 3250
 * of bytes already consumed and the next call to
 * skb_seq_read() will return the remaining part of the block.
 *
L
Lucas De Marchi 已提交
3251
 * Note 1: The size of each block of data returned can be arbitrary,
3252
 *       this limitation is the cost for zerocopy sequential
3253 3254
 *       reads of potentially non linear data.
 *
3255
 * Note 2: Fragment lists within fragments are not implemented
3256 3257 3258 3259 3260 3261 3262 3263 3264
 *       at the moment, state->root_skb could be replaced with
 *       a stack for this purpose.
 */
unsigned int skb_seq_read(unsigned int consumed, const u8 **data,
			  struct skb_seq_state *st)
{
	unsigned int block_limit, abs_offset = consumed + st->lower_offset;
	skb_frag_t *frag;

3265 3266 3267 3268 3269
	if (unlikely(abs_offset >= st->upper_offset)) {
		if (st->frag_data) {
			kunmap_atomic(st->frag_data);
			st->frag_data = NULL;
		}
3270
		return 0;
3271
	}
3272 3273

next_skb:
3274
	block_limit = skb_headlen(st->cur_skb) + st->stepped_offset;
3275

3276
	if (abs_offset < block_limit && !st->frag_data) {
3277
		*data = st->cur_skb->data + (abs_offset - st->stepped_offset);
3278 3279 3280 3281 3282 3283 3284 3285
		return block_limit - abs_offset;
	}

	if (st->frag_idx == 0 && !st->frag_data)
		st->stepped_offset += skb_headlen(st->cur_skb);

	while (st->frag_idx < skb_shinfo(st->cur_skb)->nr_frags) {
		frag = &skb_shinfo(st->cur_skb)->frags[st->frag_idx];
E
Eric Dumazet 已提交
3286
		block_limit = skb_frag_size(frag) + st->stepped_offset;
3287 3288 3289

		if (abs_offset < block_limit) {
			if (!st->frag_data)
E
Eric Dumazet 已提交
3290
				st->frag_data = kmap_atomic(skb_frag_page(frag));
3291 3292 3293 3294 3295 3296 3297 3298

			*data = (u8 *) st->frag_data + frag->page_offset +
				(abs_offset - st->stepped_offset);

			return block_limit - abs_offset;
		}

		if (st->frag_data) {
E
Eric Dumazet 已提交
3299
			kunmap_atomic(st->frag_data);
3300 3301 3302 3303
			st->frag_data = NULL;
		}

		st->frag_idx++;
E
Eric Dumazet 已提交
3304
		st->stepped_offset += skb_frag_size(frag);
3305 3306
	}

3307
	if (st->frag_data) {
E
Eric Dumazet 已提交
3308
		kunmap_atomic(st->frag_data);
3309 3310 3311
		st->frag_data = NULL;
	}

3312
	if (st->root_skb == st->cur_skb && skb_has_frag_list(st->root_skb)) {
S
Shyam Iyer 已提交
3313
		st->cur_skb = skb_shinfo(st->root_skb)->frag_list;
3314 3315
		st->frag_idx = 0;
		goto next_skb;
S
Shyam Iyer 已提交
3316 3317
	} else if (st->cur_skb->next) {
		st->cur_skb = st->cur_skb->next;
3318
		st->frag_idx = 0;
3319 3320 3321 3322 3323
		goto next_skb;
	}

	return 0;
}
3324
EXPORT_SYMBOL(skb_seq_read);
3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335

/**
 * skb_abort_seq_read - Abort a sequential read of skb data
 * @st: state variable
 *
 * Must be called if skb_seq_read() was not called until it
 * returned 0.
 */
void skb_abort_seq_read(struct skb_seq_state *st)
{
	if (st->frag_data)
E
Eric Dumazet 已提交
3336
		kunmap_atomic(st->frag_data);
3337
}
3338
EXPORT_SYMBOL(skb_abort_seq_read);
3339

3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366
#define TS_SKB_CB(state)	((struct skb_seq_state *) &((state)->cb))

static unsigned int skb_ts_get_next_block(unsigned int offset, const u8 **text,
					  struct ts_config *conf,
					  struct ts_state *state)
{
	return skb_seq_read(offset, text, TS_SKB_CB(state));
}

static void skb_ts_finish(struct ts_config *conf, struct ts_state *state)
{
	skb_abort_seq_read(TS_SKB_CB(state));
}

/**
 * skb_find_text - Find a text pattern in skb data
 * @skb: the buffer to look in
 * @from: search offset
 * @to: search limit
 * @config: textsearch configuration
 *
 * Finds a pattern in the skb data according to the specified
 * textsearch configuration. Use textsearch_next() to retrieve
 * subsequent occurrences of the pattern. Returns the offset
 * to the first occurrence or UINT_MAX if no match was found.
 */
unsigned int skb_find_text(struct sk_buff *skb, unsigned int from,
3367
			   unsigned int to, struct ts_config *config)
3368
{
3369
	struct ts_state state;
3370 3371
	unsigned int ret;

3372 3373 3374
	config->get_next_block = skb_ts_get_next_block;
	config->finish = skb_ts_finish;

3375
	skb_prepare_seq_read(skb, from, to, TS_SKB_CB(&state));
3376

3377
	ret = textsearch_find(config, &state);
3378
	return (ret <= to - from ? ret : UINT_MAX);
3379
}
3380
EXPORT_SYMBOL(skb_find_text);
3381

3382
/**
3383
 * skb_append_datato_frags - append the user data to a skb
3384
 * @sk: sock  structure
3385
 * @skb: skb structure to be appended with user data.
3386 3387 3388 3389 3390 3391 3392 3393
 * @getfrag: call back function to be used for getting the user data
 * @from: pointer to user message iov
 * @length: length of the iov message
 *
 * Description: This procedure append the user data in the fragment part
 * of the skb if any page alloc fails user this procedure returns  -ENOMEM
 */
int skb_append_datato_frags(struct sock *sk, struct sk_buff *skb,
3394
			int (*getfrag)(void *from, char *to, int offset,
3395 3396 3397
					int len, int odd, struct sk_buff *skb),
			void *from, int length)
{
3398 3399
	int frg_cnt = skb_shinfo(skb)->nr_frags;
	int copy;
3400 3401
	int offset = 0;
	int ret;
3402
	struct page_frag *pfrag = &current->task_frag;
3403 3404 3405 3406

	do {
		/* Return error if we don't have space for new frag */
		if (frg_cnt >= MAX_SKB_FRAGS)
3407
			return -EMSGSIZE;
3408

3409
		if (!sk_page_frag_refill(sk, pfrag))
3410 3411 3412
			return -ENOMEM;

		/* copy the user data to page */
3413
		copy = min_t(int, length, pfrag->size - pfrag->offset);
3414

3415 3416
		ret = getfrag(from, page_address(pfrag->page) + pfrag->offset,
			      offset, copy, 0, skb);
3417 3418 3419 3420
		if (ret < 0)
			return -EFAULT;

		/* copy was successful so update the size parameters */
3421 3422 3423 3424 3425 3426 3427
		skb_fill_page_desc(skb, frg_cnt, pfrag->page, pfrag->offset,
				   copy);
		frg_cnt++;
		pfrag->offset += copy;
		get_page(pfrag->page);

		skb->truesize += copy;
3428
		refcount_add(copy, &sk->sk_wmem_alloc);
3429 3430 3431 3432 3433 3434 3435 3436 3437
		skb->len += copy;
		skb->data_len += copy;
		offset += copy;
		length -= copy;

	} while (length > 0);

	return 0;
}
3438
EXPORT_SYMBOL(skb_append_datato_frags);
3439

3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457
int skb_append_pagefrags(struct sk_buff *skb, struct page *page,
			 int offset, size_t size)
{
	int i = skb_shinfo(skb)->nr_frags;

	if (skb_can_coalesce(skb, i, page, offset)) {
		skb_frag_size_add(&skb_shinfo(skb)->frags[i - 1], size);
	} else if (i < MAX_SKB_FRAGS) {
		get_page(page);
		skb_fill_page_desc(skb, i, page, offset, size);
	} else {
		return -EMSGSIZE;
	}

	return 0;
}
EXPORT_SYMBOL_GPL(skb_append_pagefrags);

3458 3459 3460 3461 3462 3463
/**
 *	skb_pull_rcsum - pull skb and update receive checksum
 *	@skb: buffer to update
 *	@len: length of data pulled
 *
 *	This function performs an skb_pull on the packet and updates
3464
 *	the CHECKSUM_COMPLETE checksum.  It should be used on
3465 3466 3467
 *	receive path processing instead of skb_pull unless you know
 *	that the checksum difference is zero (e.g., a valid IP header)
 *	or you are setting ip_summed to CHECKSUM_NONE.
3468
 */
3469
void *skb_pull_rcsum(struct sk_buff *skb, unsigned int len)
3470
{
3471 3472
	unsigned char *data = skb->data;

3473
	BUG_ON(len > skb->len);
3474 3475 3476
	__skb_pull(skb, len);
	skb_postpull_rcsum(skb, data, len);
	return skb->data;
3477
}
3478 3479
EXPORT_SYMBOL_GPL(skb_pull_rcsum);

3480 3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492
static inline skb_frag_t skb_head_frag_to_page_desc(struct sk_buff *frag_skb)
{
	skb_frag_t head_frag;
	struct page *page;

	page = virt_to_head_page(frag_skb->head);
	head_frag.page.p = page;
	head_frag.page_offset = frag_skb->data -
		(unsigned char *)page_address(page);
	head_frag.size = skb_headlen(frag_skb);
	return head_frag;
}

H
Herbert Xu 已提交
3493 3494
/**
 *	skb_segment - Perform protocol segmentation on skb.
3495
 *	@head_skb: buffer to segment
3496
 *	@features: features for the output path (see dev->features)
H
Herbert Xu 已提交
3497 3498
 *
 *	This function performs segmentation on the given skb.  It returns
3499 3500
 *	a pointer to the first in a list of new skbs for the segments.
 *	In case of error it returns ERR_PTR(err).
H
Herbert Xu 已提交
3501
 */
3502 3503
struct sk_buff *skb_segment(struct sk_buff *head_skb,
			    netdev_features_t features)
H
Herbert Xu 已提交
3504 3505 3506
{
	struct sk_buff *segs = NULL;
	struct sk_buff *tail = NULL;
3507
	struct sk_buff *list_skb = skb_shinfo(head_skb)->frag_list;
3508 3509 3510
	skb_frag_t *frag = skb_shinfo(head_skb)->frags;
	unsigned int mss = skb_shinfo(head_skb)->gso_size;
	unsigned int doffset = head_skb->data - skb_mac_header(head_skb);
3511
	struct sk_buff *frag_skb = head_skb;
H
Herbert Xu 已提交
3512
	unsigned int offset = doffset;
3513
	unsigned int tnl_hlen = skb_tnl_header_len(head_skb);
3514
	unsigned int partial_segs = 0;
H
Herbert Xu 已提交
3515
	unsigned int headroom;
3516
	unsigned int len = head_skb->len;
3517
	__be16 proto;
3518
	bool csum, sg;
3519
	int nfrags = skb_shinfo(head_skb)->nr_frags;
H
Herbert Xu 已提交
3520 3521 3522
	int err = -ENOMEM;
	int i = 0;
	int pos;
3523
	int dummy;
H
Herbert Xu 已提交
3524

3525
	__skb_push(head_skb, doffset);
3526
	proto = skb_network_protocol(head_skb, &dummy);
3527 3528 3529
	if (unlikely(!proto))
		return ERR_PTR(-EINVAL);

3530
	sg = !!(features & NETIF_F_SG);
3531
	csum = !!can_checksum_protocol(features, proto);
3532

3533 3534 3535
	if (sg && csum && (mss != GSO_BY_FRAGS))  {
		if (!(features & NETIF_F_GSO_PARTIAL)) {
			struct sk_buff *iter;
3536
			unsigned int frag_len;
3537 3538 3539 3540 3541

			if (!list_skb ||
			    !net_gso_ok(features, skb_shinfo(head_skb)->gso_type))
				goto normal;

3542 3543 3544 3545 3546 3547 3548
			/* If we get here then all the required
			 * GSO features except frag_list are supported.
			 * Try to split the SKB to multiple GSO SKBs
			 * with no frag_list.
			 * Currently we can do that only when the buffers don't
			 * have a linear part and all the buffers except
			 * the last are of the same length.
3549
			 */
3550
			frag_len = list_skb->len;
3551
			skb_walk_frags(head_skb, iter) {
3552 3553
				if (frag_len != iter->len && iter->next)
					goto normal;
3554
				if (skb_headlen(iter) && !iter->head_frag)
3555 3556 3557 3558
					goto normal;

				len -= iter->len;
			}
3559 3560 3561

			if (len != frag_len)
				goto normal;
3562 3563 3564 3565 3566 3567
		}

		/* GSO partial only requires that we trim off any excess that
		 * doesn't fit into an MSS sized block, so take care of that
		 * now.
		 */
3568
		partial_segs = len / mss;
3569 3570 3571 3572
		if (partial_segs > 1)
			mss *= partial_segs;
		else
			partial_segs = 0;
3573 3574
	}

3575
normal:
3576 3577
	headroom = skb_headroom(head_skb);
	pos = skb_headlen(head_skb);
H
Herbert Xu 已提交
3578 3579 3580

	do {
		struct sk_buff *nskb;
3581
		skb_frag_t *nskb_frag;
3582
		int hsize;
H
Herbert Xu 已提交
3583 3584
		int size;

3585 3586 3587 3588 3589 3590 3591
		if (unlikely(mss == GSO_BY_FRAGS)) {
			len = list_skb->len;
		} else {
			len = head_skb->len - offset;
			if (len > mss)
				len = mss;
		}
H
Herbert Xu 已提交
3592

3593
		hsize = skb_headlen(head_skb) - offset;
H
Herbert Xu 已提交
3594 3595
		if (hsize < 0)
			hsize = 0;
3596 3597
		if (hsize > len || !sg)
			hsize = len;
H
Herbert Xu 已提交
3598

3599 3600 3601
		if (!hsize && i >= nfrags && skb_headlen(list_skb) &&
		    (skb_headlen(list_skb) == len || sg)) {
			BUG_ON(skb_headlen(list_skb) > len);
3602 3603

			i = 0;
3604 3605
			nfrags = skb_shinfo(list_skb)->nr_frags;
			frag = skb_shinfo(list_skb)->frags;
3606
			frag_skb = list_skb;
3607
			pos += skb_headlen(list_skb);
3608 3609 3610 3611

			while (pos < offset + len) {
				BUG_ON(i >= nfrags);

3612
				size = skb_frag_size(frag);
3613 3614 3615 3616 3617
				if (pos + size > offset + len)
					break;

				i++;
				pos += size;
3618
				frag++;
3619
			}
3620

3621 3622
			nskb = skb_clone(list_skb, GFP_ATOMIC);
			list_skb = list_skb->next;
3623 3624 3625 3626

			if (unlikely(!nskb))
				goto err;

3627 3628 3629 3630 3631
			if (unlikely(pskb_trim(nskb, len))) {
				kfree_skb(nskb);
				goto err;
			}

3632
			hsize = skb_end_offset(nskb);
3633 3634 3635 3636 3637
			if (skb_cow_head(nskb, doffset + headroom)) {
				kfree_skb(nskb);
				goto err;
			}

3638
			nskb->truesize += skb_end_offset(nskb) - hsize;
3639 3640 3641
			skb_release_head_state(nskb);
			__skb_push(nskb, doffset);
		} else {
3642
			nskb = __alloc_skb(hsize + doffset + headroom,
3643
					   GFP_ATOMIC, skb_alloc_rx_flag(head_skb),
3644
					   NUMA_NO_NODE);
3645 3646 3647 3648 3649 3650 3651

			if (unlikely(!nskb))
				goto err;

			skb_reserve(nskb, headroom);
			__skb_put(nskb, doffset);
		}
H
Herbert Xu 已提交
3652 3653 3654 3655 3656 3657 3658

		if (segs)
			tail->next = nskb;
		else
			segs = nskb;
		tail = nskb;

3659
		__copy_skb_header(nskb, head_skb);
H
Herbert Xu 已提交
3660

E
Eric Dumazet 已提交
3661
		skb_headers_offset_update(nskb, skb_headroom(nskb) - headroom);
3662
		skb_reset_mac_len(nskb);
3663

3664
		skb_copy_from_linear_data_offset(head_skb, -tnl_hlen,
3665 3666
						 nskb->data - tnl_hlen,
						 doffset + tnl_hlen);
3667

3668
		if (nskb->len == len + doffset)
3669
			goto perform_csum_check;
3670

3671 3672 3673
		if (!sg) {
			if (!nskb->remcsum_offload)
				nskb->ip_summed = CHECKSUM_NONE;
3674 3675 3676 3677
			SKB_GSO_CB(nskb)->csum =
				skb_copy_and_csum_bits(head_skb, offset,
						       skb_put(nskb, len),
						       len, 0);
3678
			SKB_GSO_CB(nskb)->csum_start =
3679
				skb_headroom(nskb) + doffset;
H
Herbert Xu 已提交
3680 3681 3682
			continue;
		}

3683
		nskb_frag = skb_shinfo(nskb)->frags;
H
Herbert Xu 已提交
3684

3685
		skb_copy_from_linear_data_offset(head_skb, offset,
3686
						 skb_put(nskb, hsize), hsize);
H
Herbert Xu 已提交
3687

3688 3689
		skb_shinfo(nskb)->tx_flags |= skb_shinfo(head_skb)->tx_flags &
					      SKBTX_SHARED_FRAG;
3690

3691 3692 3693 3694
		if (skb_orphan_frags(frag_skb, GFP_ATOMIC) ||
		    skb_zerocopy_clone(nskb, frag_skb, GFP_ATOMIC))
			goto err;

3695 3696 3697
		while (pos < offset + len) {
			if (i >= nfrags) {
				i = 0;
3698 3699
				nfrags = skb_shinfo(list_skb)->nr_frags;
				frag = skb_shinfo(list_skb)->frags;
3700
				frag_skb = list_skb;
3701 3702 3703 3704
				if (!skb_headlen(list_skb)) {
					BUG_ON(!nfrags);
				} else {
					BUG_ON(!list_skb->head_frag);
3705

3706 3707 3708 3709
					/* to make room for head_frag. */
					i--;
					frag--;
				}
3710 3711 3712 3713 3714
				if (skb_orphan_frags(frag_skb, GFP_ATOMIC) ||
				    skb_zerocopy_clone(nskb, frag_skb,
						       GFP_ATOMIC))
					goto err;

3715
				list_skb = list_skb->next;
3716 3717 3718 3719 3720 3721 3722
			}

			if (unlikely(skb_shinfo(nskb)->nr_frags >=
				     MAX_SKB_FRAGS)) {
				net_warn_ratelimited(
					"skb_segment: too many frags: %u %u\n",
					pos, mss);
3723
				err = -EINVAL;
3724 3725 3726
				goto err;
			}

3727
			*nskb_frag = (i < 0) ? skb_head_frag_to_page_desc(frag_skb) : *frag;
3728 3729
			__skb_frag_ref(nskb_frag);
			size = skb_frag_size(nskb_frag);
H
Herbert Xu 已提交
3730 3731

			if (pos < offset) {
3732 3733
				nskb_frag->page_offset += offset - pos;
				skb_frag_size_sub(nskb_frag, offset - pos);
H
Herbert Xu 已提交
3734 3735
			}

3736
			skb_shinfo(nskb)->nr_frags++;
H
Herbert Xu 已提交
3737 3738 3739

			if (pos + size <= offset + len) {
				i++;
3740
				frag++;
H
Herbert Xu 已提交
3741 3742
				pos += size;
			} else {
3743
				skb_frag_size_sub(nskb_frag, pos + size - (offset + len));
3744
				goto skip_fraglist;
H
Herbert Xu 已提交
3745 3746
			}

3747
			nskb_frag++;
H
Herbert Xu 已提交
3748 3749
		}

3750
skip_fraglist:
H
Herbert Xu 已提交
3751 3752 3753
		nskb->data_len = len - hsize;
		nskb->len += nskb->data_len;
		nskb->truesize += nskb->data_len;
3754

3755
perform_csum_check:
3756
		if (!csum) {
3757 3758 3759 3760
			if (skb_has_shared_frag(nskb) &&
			    __skb_linearize(nskb))
				goto err;

3761 3762
			if (!nskb->remcsum_offload)
				nskb->ip_summed = CHECKSUM_NONE;
3763 3764 3765
			SKB_GSO_CB(nskb)->csum =
				skb_checksum(nskb, doffset,
					     nskb->len - doffset, 0);
3766
			SKB_GSO_CB(nskb)->csum_start =
3767
				skb_headroom(nskb) + doffset;
3768
		}
3769
	} while ((offset += len) < head_skb->len);
H
Herbert Xu 已提交
3770

3771 3772 3773 3774 3775
	/* Some callers want to get the end of the list.
	 * Put it in segs->prev to avoid walking the list.
	 * (see validate_xmit_skb_list() for example)
	 */
	segs->prev = tail;
3776

3777
	if (partial_segs) {
3778
		struct sk_buff *iter;
3779
		int type = skb_shinfo(head_skb)->gso_type;
3780
		unsigned short gso_size = skb_shinfo(head_skb)->gso_size;
3781 3782

		/* Update type to add partial and then remove dodgy if set */
3783
		type |= (features & NETIF_F_GSO_PARTIAL) / NETIF_F_GSO_PARTIAL * SKB_GSO_PARTIAL;
3784 3785 3786 3787 3788
		type &= ~SKB_GSO_DODGY;

		/* Update GSO info and prepare to start updating headers on
		 * our way back down the stack of protocols.
		 */
3789 3790 3791 3792 3793 3794 3795 3796 3797 3798 3799
		for (iter = segs; iter; iter = iter->next) {
			skb_shinfo(iter)->gso_size = gso_size;
			skb_shinfo(iter)->gso_segs = partial_segs;
			skb_shinfo(iter)->gso_type = type;
			SKB_GSO_CB(iter)->data_offset = skb_headroom(iter) + doffset;
		}

		if (tail->len - doffset <= gso_size)
			skb_shinfo(tail)->gso_size = 0;
		else if (tail != segs)
			skb_shinfo(tail)->gso_segs = DIV_ROUND_UP(tail->len - doffset, gso_size);
3800 3801
	}

3802 3803 3804 3805 3806 3807 3808 3809 3810
	/* Following permits correct backpressure, for protocols
	 * using skb_set_owner_w().
	 * Idea is to tranfert ownership from head_skb to last segment.
	 */
	if (head_skb->destructor == sock_wfree) {
		swap(tail->truesize, head_skb->truesize);
		swap(tail->destructor, head_skb->destructor);
		swap(tail->sk, head_skb->sk);
	}
H
Herbert Xu 已提交
3811 3812 3813
	return segs;

err:
E
Eric Dumazet 已提交
3814
	kfree_skb_list(segs);
H
Herbert Xu 已提交
3815 3816 3817 3818
	return ERR_PTR(err);
}
EXPORT_SYMBOL_GPL(skb_segment);

3819
int skb_gro_receive(struct sk_buff *p, struct sk_buff *skb)
H
Herbert Xu 已提交
3820
{
3821
	struct skb_shared_info *pinfo, *skbinfo = skb_shinfo(skb);
3822 3823
	unsigned int offset = skb_gro_offset(skb);
	unsigned int headlen = skb_headlen(skb);
3824
	unsigned int len = skb_gro_len(skb);
3825
	unsigned int delta_truesize;
3826
	struct sk_buff *lp;
H
Herbert Xu 已提交
3827

3828
	if (unlikely(p->len + len >= 65536))
H
Herbert Xu 已提交
3829 3830
		return -E2BIG;

3831
	lp = NAPI_GRO_CB(p)->last;
3832 3833 3834
	pinfo = skb_shinfo(lp);

	if (headlen <= offset) {
3835
		skb_frag_t *frag;
3836
		skb_frag_t *frag2;
3837 3838
		int i = skbinfo->nr_frags;
		int nr_frags = pinfo->nr_frags + i;
3839 3840

		if (nr_frags > MAX_SKB_FRAGS)
3841
			goto merge;
3842

3843
		offset -= headlen;
3844 3845
		pinfo->nr_frags = nr_frags;
		skbinfo->nr_frags = 0;
3846

3847 3848
		frag = pinfo->frags + nr_frags;
		frag2 = skbinfo->frags + i;
3849 3850 3851
		do {
			*--frag = *--frag2;
		} while (--i);
H
Herbert Xu 已提交
3852

3853
		frag->page_offset += offset;
E
Eric Dumazet 已提交
3854
		skb_frag_size_sub(frag, offset);
3855

3856
		/* all fragments truesize : remove (head size + sk_buff) */
3857 3858
		delta_truesize = skb->truesize -
				 SKB_TRUESIZE(skb_end_offset(skb));
3859

3860 3861 3862 3863
		skb->truesize -= skb->data_len;
		skb->len -= skb->data_len;
		skb->data_len = 0;

3864
		NAPI_GRO_CB(skb)->free = NAPI_GRO_FREE;
H
Herbert Xu 已提交
3865
		goto done;
3866 3867 3868 3869 3870 3871 3872 3873
	} else if (skb->head_frag) {
		int nr_frags = pinfo->nr_frags;
		skb_frag_t *frag = pinfo->frags + nr_frags;
		struct page *page = virt_to_head_page(skb->head);
		unsigned int first_size = headlen - offset;
		unsigned int first_offset;

		if (nr_frags + 1 + skbinfo->nr_frags > MAX_SKB_FRAGS)
3874
			goto merge;
3875 3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888

		first_offset = skb->data -
			       (unsigned char *)page_address(page) +
			       offset;

		pinfo->nr_frags = nr_frags + 1 + skbinfo->nr_frags;

		frag->page.p	  = page;
		frag->page_offset = first_offset;
		skb_frag_size_set(frag, first_size);

		memcpy(frag + 1, skbinfo->frags, sizeof(*frag) * skbinfo->nr_frags);
		/* We dont need to clear skbinfo->nr_frags here */

3889
		delta_truesize = skb->truesize - SKB_DATA_ALIGN(sizeof(struct sk_buff));
3890 3891
		NAPI_GRO_CB(skb)->free = NAPI_GRO_FREE_STOLEN_HEAD;
		goto done;
3892
	}
H
Herbert Xu 已提交
3893 3894

merge:
3895
	delta_truesize = skb->truesize;
3896
	if (offset > headlen) {
3897 3898 3899
		unsigned int eat = offset - headlen;

		skbinfo->frags[0].page_offset += eat;
E
Eric Dumazet 已提交
3900
		skb_frag_size_sub(&skbinfo->frags[0], eat);
3901 3902
		skb->data_len -= eat;
		skb->len -= eat;
3903
		offset = headlen;
3904 3905
	}

3906
	__skb_pull(skb, offset);
3907

3908
	if (NAPI_GRO_CB(p)->last == p)
3909 3910 3911
		skb_shinfo(p)->frag_list = skb;
	else
		NAPI_GRO_CB(p)->last->next = skb;
3912
	NAPI_GRO_CB(p)->last = skb;
3913
	__skb_header_release(skb);
3914
	lp = p;
H
Herbert Xu 已提交
3915

H
Herbert Xu 已提交
3916 3917
done:
	NAPI_GRO_CB(p)->count++;
H
Herbert Xu 已提交
3918
	p->data_len += len;
3919
	p->truesize += delta_truesize;
H
Herbert Xu 已提交
3920
	p->len += len;
3921 3922 3923 3924 3925
	if (lp != p) {
		lp->data_len += len;
		lp->truesize += delta_truesize;
		lp->len += len;
	}
H
Herbert Xu 已提交
3926 3927 3928
	NAPI_GRO_CB(skb)->same_flow = 1;
	return 0;
}
3929
EXPORT_SYMBOL_GPL(skb_gro_receive);
H
Herbert Xu 已提交
3930

L
Linus Torvalds 已提交
3931 3932
void __init skb_init(void)
{
3933
	skbuff_head_cache = kmem_cache_create_usercopy("skbuff_head_cache",
L
Linus Torvalds 已提交
3934 3935
					      sizeof(struct sk_buff),
					      0,
A
Alexey Dobriyan 已提交
3936
					      SLAB_HWCACHE_ALIGN|SLAB_PANIC,
3937 3938
					      offsetof(struct sk_buff, cb),
					      sizeof_field(struct sk_buff, cb),
3939
					      NULL);
3940
	skbuff_fclone_cache = kmem_cache_create("skbuff_fclone_cache",
3941
						sizeof(struct sk_buff_fclones),
3942
						0,
A
Alexey Dobriyan 已提交
3943
						SLAB_HWCACHE_ALIGN|SLAB_PANIC,
3944
						NULL);
L
Linus Torvalds 已提交
3945 3946
}

3947
static int
3948 3949
__skb_to_sgvec(struct sk_buff *skb, struct scatterlist *sg, int offset, int len,
	       unsigned int recursion_level)
3950
{
3951 3952
	int start = skb_headlen(skb);
	int i, copy = start - offset;
3953
	struct sk_buff *frag_iter;
3954 3955
	int elt = 0;

3956 3957 3958
	if (unlikely(recursion_level >= 24))
		return -EMSGSIZE;

3959 3960 3961
	if (copy > 0) {
		if (copy > len)
			copy = len;
3962
		sg_set_buf(sg, skb->data + offset, copy);
3963 3964 3965 3966 3967 3968 3969
		elt++;
		if ((len -= copy) == 0)
			return elt;
		offset += copy;
	}

	for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
3970
		int end;
3971

3972
		WARN_ON(start > offset + len);
3973

E
Eric Dumazet 已提交
3974
		end = start + skb_frag_size(&skb_shinfo(skb)->frags[i]);
3975 3976
		if ((copy = end - offset) > 0) {
			skb_frag_t *frag = &skb_shinfo(skb)->frags[i];
3977 3978
			if (unlikely(elt && sg_is_last(&sg[elt - 1])))
				return -EMSGSIZE;
3979 3980 3981

			if (copy > len)
				copy = len;
3982
			sg_set_page(&sg[elt], skb_frag_page(frag), copy,
3983
					frag->page_offset+offset-start);
3984 3985 3986 3987 3988
			elt++;
			if (!(len -= copy))
				return elt;
			offset += copy;
		}
3989
		start = end;
3990 3991
	}

3992
	skb_walk_frags(skb, frag_iter) {
3993
		int end, ret;
3994

3995
		WARN_ON(start > offset + len);
3996

3997 3998
		end = start + frag_iter->len;
		if ((copy = end - offset) > 0) {
3999 4000 4001
			if (unlikely(elt && sg_is_last(&sg[elt - 1])))
				return -EMSGSIZE;

4002 4003
			if (copy > len)
				copy = len;
4004 4005 4006 4007 4008
			ret = __skb_to_sgvec(frag_iter, sg+elt, offset - start,
					      copy, recursion_level + 1);
			if (unlikely(ret < 0))
				return ret;
			elt += ret;
4009 4010 4011
			if ((len -= copy) == 0)
				return elt;
			offset += copy;
4012
		}
4013
		start = end;
4014 4015 4016 4017 4018
	}
	BUG_ON(len);
	return elt;
}

4019 4020 4021 4022 4023 4024 4025 4026 4027 4028 4029 4030 4031 4032 4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 4043
/**
 *	skb_to_sgvec - Fill a scatter-gather list from a socket buffer
 *	@skb: Socket buffer containing the buffers to be mapped
 *	@sg: The scatter-gather list to map into
 *	@offset: The offset into the buffer's contents to start mapping
 *	@len: Length of buffer space to be mapped
 *
 *	Fill the specified scatter-gather list with mappings/pointers into a
 *	region of the buffer space attached to a socket buffer. Returns either
 *	the number of scatterlist items used, or -EMSGSIZE if the contents
 *	could not fit.
 */
int skb_to_sgvec(struct sk_buff *skb, struct scatterlist *sg, int offset, int len)
{
	int nsg = __skb_to_sgvec(skb, sg, offset, len, 0);

	if (nsg <= 0)
		return nsg;

	sg_mark_end(&sg[nsg - 1]);

	return nsg;
}
EXPORT_SYMBOL_GPL(skb_to_sgvec);

4044 4045 4046 4047 4048 4049 4050 4051 4052 4053 4054 4055 4056 4057 4058 4059 4060 4061 4062 4063 4064 4065
/* As compared with skb_to_sgvec, skb_to_sgvec_nomark only map skb to given
 * sglist without mark the sg which contain last skb data as the end.
 * So the caller can mannipulate sg list as will when padding new data after
 * the first call without calling sg_unmark_end to expend sg list.
 *
 * Scenario to use skb_to_sgvec_nomark:
 * 1. sg_init_table
 * 2. skb_to_sgvec_nomark(payload1)
 * 3. skb_to_sgvec_nomark(payload2)
 *
 * This is equivalent to:
 * 1. sg_init_table
 * 2. skb_to_sgvec(payload1)
 * 3. sg_unmark_end
 * 4. skb_to_sgvec(payload2)
 *
 * When mapping mutilple payload conditionally, skb_to_sgvec_nomark
 * is more preferable.
 */
int skb_to_sgvec_nomark(struct sk_buff *skb, struct scatterlist *sg,
			int offset, int len)
{
4066
	return __skb_to_sgvec(skb, sg, offset, len, 0);
4067 4068 4069
}
EXPORT_SYMBOL_GPL(skb_to_sgvec_nomark);

4070 4071


4072 4073 4074 4075 4076 4077 4078 4079 4080 4081 4082 4083 4084 4085 4086 4087 4088 4089 4090 4091 4092 4093 4094 4095 4096 4097 4098 4099 4100 4101 4102 4103
/**
 *	skb_cow_data - Check that a socket buffer's data buffers are writable
 *	@skb: The socket buffer to check.
 *	@tailbits: Amount of trailing space to be added
 *	@trailer: Returned pointer to the skb where the @tailbits space begins
 *
 *	Make sure that the data buffers attached to a socket buffer are
 *	writable. If they are not, private copies are made of the data buffers
 *	and the socket buffer is set to use these instead.
 *
 *	If @tailbits is given, make sure that there is space to write @tailbits
 *	bytes of data beyond current end of socket buffer.  @trailer will be
 *	set to point to the skb in which this space begins.
 *
 *	The number of scatterlist elements required to completely map the
 *	COW'd and extended socket buffer will be returned.
 */
int skb_cow_data(struct sk_buff *skb, int tailbits, struct sk_buff **trailer)
{
	int copyflag;
	int elt;
	struct sk_buff *skb1, **skb_p;

	/* If skb is cloned or its head is paged, reallocate
	 * head pulling out all the pages (pages are considered not writable
	 * at the moment even if they are anonymous).
	 */
	if ((skb_cloned(skb) || skb_shinfo(skb)->nr_frags) &&
	    __pskb_pull_tail(skb, skb_pagelen(skb)-skb_headlen(skb)) == NULL)
		return -ENOMEM;

	/* Easy case. Most of packets will go this way. */
4104
	if (!skb_has_frag_list(skb)) {
4105 4106 4107 4108 4109 4110 4111 4112 4113 4114 4115 4116 4117 4118 4119 4120 4121 4122 4123 4124 4125 4126 4127 4128 4129 4130 4131 4132 4133 4134 4135 4136 4137 4138
		/* A little of trouble, not enough of space for trailer.
		 * This should not happen, when stack is tuned to generate
		 * good frames. OK, on miss we reallocate and reserve even more
		 * space, 128 bytes is fair. */

		if (skb_tailroom(skb) < tailbits &&
		    pskb_expand_head(skb, 0, tailbits-skb_tailroom(skb)+128, GFP_ATOMIC))
			return -ENOMEM;

		/* Voila! */
		*trailer = skb;
		return 1;
	}

	/* Misery. We are in troubles, going to mincer fragments... */

	elt = 1;
	skb_p = &skb_shinfo(skb)->frag_list;
	copyflag = 0;

	while ((skb1 = *skb_p) != NULL) {
		int ntail = 0;

		/* The fragment is partially pulled by someone,
		 * this can happen on input. Copy it and everything
		 * after it. */

		if (skb_shared(skb1))
			copyflag = 1;

		/* If the skb is the last, worry about trailer. */

		if (skb1->next == NULL && tailbits) {
			if (skb_shinfo(skb1)->nr_frags ||
4139
			    skb_has_frag_list(skb1) ||
4140 4141 4142 4143 4144 4145 4146 4147
			    skb_tailroom(skb1) < tailbits)
				ntail = tailbits + 128;
		}

		if (copyflag ||
		    skb_cloned(skb1) ||
		    ntail ||
		    skb_shinfo(skb1)->nr_frags ||
4148
		    skb_has_frag_list(skb1)) {
4149 4150 4151 4152 4153 4154 4155 4156 4157 4158 4159 4160 4161 4162 4163 4164 4165 4166 4167 4168 4169 4170 4171 4172 4173 4174 4175 4176 4177 4178 4179
			struct sk_buff *skb2;

			/* Fuck, we are miserable poor guys... */
			if (ntail == 0)
				skb2 = skb_copy(skb1, GFP_ATOMIC);
			else
				skb2 = skb_copy_expand(skb1,
						       skb_headroom(skb1),
						       ntail,
						       GFP_ATOMIC);
			if (unlikely(skb2 == NULL))
				return -ENOMEM;

			if (skb1->sk)
				skb_set_owner_w(skb2, skb1->sk);

			/* Looking around. Are we still alive?
			 * OK, link new skb, drop old one */

			skb2->next = skb1->next;
			*skb_p = skb2;
			kfree_skb(skb1);
			skb1 = skb2;
		}
		elt++;
		*trailer = skb1;
		skb_p = &skb1->next;
	}

	return elt;
}
4180
EXPORT_SYMBOL_GPL(skb_cow_data);
4181

4182 4183 4184 4185 4186 4187 4188
static void sock_rmem_free(struct sk_buff *skb)
{
	struct sock *sk = skb->sk;

	atomic_sub(skb->truesize, &sk->sk_rmem_alloc);
}

4189 4190 4191 4192 4193 4194 4195 4196 4197
static void skb_set_err_queue(struct sk_buff *skb)
{
	/* pkt_type of skbs received on local sockets is never PACKET_OUTGOING.
	 * So, it is safe to (mis)use it to mark skbs on the error queue.
	 */
	skb->pkt_type = PACKET_OUTGOING;
	BUILD_BUG_ON(PACKET_OUTGOING == 0);
}

4198 4199 4200 4201 4202 4203
/*
 * Note: We dont mem charge error packets (no sk_forward_alloc changes)
 */
int sock_queue_err_skb(struct sock *sk, struct sk_buff *skb)
{
	if (atomic_read(&sk->sk_rmem_alloc) + skb->truesize >=
4204
	    (unsigned int)sk->sk_rcvbuf)
4205 4206 4207 4208 4209 4210
		return -ENOMEM;

	skb_orphan(skb);
	skb->sk = sk;
	skb->destructor = sock_rmem_free;
	atomic_add(skb->truesize, &sk->sk_rmem_alloc);
4211
	skb_set_err_queue(skb);
4212

4213 4214 4215
	/* before exiting rcu section, make sure dst is refcounted */
	skb_dst_force(skb);

4216 4217
	skb_queue_tail(&sk->sk_error_queue, skb);
	if (!sock_flag(sk, SOCK_DEAD))
4218
		sk->sk_error_report(sk);
4219 4220 4221 4222
	return 0;
}
EXPORT_SYMBOL(sock_queue_err_skb);

4223 4224 4225 4226 4227 4228
static bool is_icmp_err_skb(const struct sk_buff *skb)
{
	return skb && (SKB_EXT_ERR(skb)->ee.ee_origin == SO_EE_ORIGIN_ICMP ||
		       SKB_EXT_ERR(skb)->ee.ee_origin == SO_EE_ORIGIN_ICMP6);
}

4229 4230 4231
struct sk_buff *sock_dequeue_err_skb(struct sock *sk)
{
	struct sk_buff_head *q = &sk->sk_error_queue;
4232 4233
	struct sk_buff *skb, *skb_next = NULL;
	bool icmp_next = false;
4234
	unsigned long flags;
4235

4236
	spin_lock_irqsave(&q->lock, flags);
4237
	skb = __skb_dequeue(q);
4238
	if (skb && (skb_next = skb_peek(q))) {
4239
		icmp_next = is_icmp_err_skb(skb_next);
4240 4241 4242
		if (icmp_next)
			sk->sk_err = SKB_EXT_ERR(skb_next)->ee.ee_origin;
	}
4243
	spin_unlock_irqrestore(&q->lock, flags);
4244

4245 4246 4247 4248
	if (is_icmp_err_skb(skb) && !icmp_next)
		sk->sk_err = 0;

	if (skb_next)
4249 4250 4251 4252 4253 4254
		sk->sk_error_report(sk);

	return skb;
}
EXPORT_SYMBOL(sock_dequeue_err_skb);

4255 4256 4257 4258 4259 4260 4261 4262 4263 4264 4265 4266 4267
/**
 * skb_clone_sk - create clone of skb, and take reference to socket
 * @skb: the skb to clone
 *
 * This function creates a clone of a buffer that holds a reference on
 * sk_refcnt.  Buffers created via this function are meant to be
 * returned using sock_queue_err_skb, or free via kfree_skb.
 *
 * When passing buffers allocated with this function to sock_queue_err_skb
 * it is necessary to wrap the call with sock_hold/sock_put in order to
 * prevent the socket from being released prior to being enqueued on
 * the sk_error_queue.
 */
4268 4269 4270 4271 4272
struct sk_buff *skb_clone_sk(struct sk_buff *skb)
{
	struct sock *sk = skb->sk;
	struct sk_buff *clone;

4273
	if (!sk || !refcount_inc_not_zero(&sk->sk_refcnt))
4274 4275 4276 4277 4278 4279 4280 4281 4282 4283 4284 4285 4286 4287 4288
		return NULL;

	clone = skb_clone(skb, GFP_ATOMIC);
	if (!clone) {
		sock_put(sk);
		return NULL;
	}

	clone->sk = sk;
	clone->destructor = sock_efree;

	return clone;
}
EXPORT_SYMBOL(skb_clone_sk);

4289 4290
static void __skb_complete_tx_timestamp(struct sk_buff *skb,
					struct sock *sk,
4291 4292
					int tstype,
					bool opt_stats)
4293 4294 4295 4296
{
	struct sock_exterr_skb *serr;
	int err;

4297 4298
	BUILD_BUG_ON(sizeof(struct sock_exterr_skb) > sizeof(skb->cb));

4299 4300 4301 4302
	serr = SKB_EXT_ERR(skb);
	memset(serr, 0, sizeof(*serr));
	serr->ee.ee_errno = ENOMSG;
	serr->ee.ee_origin = SO_EE_ORIGIN_TIMESTAMPING;
4303
	serr->ee.ee_info = tstype;
4304
	serr->opt_stats = opt_stats;
4305
	serr->header.h4.iif = skb->dev ? skb->dev->ifindex : 0;
4306
	if (sk->sk_tsflags & SOF_TIMESTAMPING_OPT_ID) {
4307
		serr->ee.ee_data = skb_shinfo(skb)->tskey;
4308 4309
		if (sk->sk_protocol == IPPROTO_TCP &&
		    sk->sk_type == SOCK_STREAM)
4310 4311
			serr->ee.ee_data -= sk->sk_tskey;
	}
4312

4313
	err = sock_queue_err_skb(sk, skb);
4314

4315 4316 4317
	if (err)
		kfree_skb(skb);
}
4318

4319 4320 4321 4322 4323 4324 4325 4326 4327 4328 4329 4330 4331 4332
static bool skb_may_tx_timestamp(struct sock *sk, bool tsonly)
{
	bool ret;

	if (likely(sysctl_tstamp_allow_data || tsonly))
		return true;

	read_lock_bh(&sk->sk_callback_lock);
	ret = sk->sk_socket && sk->sk_socket->file &&
	      file_ns_capable(sk->sk_socket->file, &init_user_ns, CAP_NET_RAW);
	read_unlock_bh(&sk->sk_callback_lock);
	return ret;
}

4333 4334 4335 4336 4337
void skb_complete_tx_timestamp(struct sk_buff *skb,
			       struct skb_shared_hwtstamps *hwtstamps)
{
	struct sock *sk = skb->sk;

4338
	if (!skb_may_tx_timestamp(sk, false))
4339
		goto err;
4340

4341 4342 4343
	/* Take a reference to prevent skb_orphan() from freeing the socket,
	 * but only if the socket refcount is not zero.
	 */
4344
	if (likely(refcount_inc_not_zero(&sk->sk_refcnt))) {
4345
		*skb_hwtstamps(skb) = *hwtstamps;
4346
		__skb_complete_tx_timestamp(skb, sk, SCM_TSTAMP_SND, false);
4347
		sock_put(sk);
4348
		return;
4349
	}
4350 4351 4352

err:
	kfree_skb(skb);
4353 4354 4355 4356 4357 4358 4359 4360
}
EXPORT_SYMBOL_GPL(skb_complete_tx_timestamp);

void __skb_tstamp_tx(struct sk_buff *orig_skb,
		     struct skb_shared_hwtstamps *hwtstamps,
		     struct sock *sk, int tstype)
{
	struct sk_buff *skb;
4361
	bool tsonly, opt_stats = false;
4362

4363 4364 4365
	if (!sk)
		return;

4366 4367 4368 4369
	if (!hwtstamps && !(sk->sk_tsflags & SOF_TIMESTAMPING_OPT_TX_SWHW) &&
	    skb_shinfo(orig_skb)->tx_flags & SKBTX_IN_PROGRESS)
		return;

4370 4371
	tsonly = sk->sk_tsflags & SOF_TIMESTAMPING_OPT_TSONLY;
	if (!skb_may_tx_timestamp(sk, tsonly))
4372 4373
		return;

4374 4375 4376 4377
	if (tsonly) {
#ifdef CONFIG_INET
		if ((sk->sk_tsflags & SOF_TIMESTAMPING_OPT_STATS) &&
		    sk->sk_protocol == IPPROTO_TCP &&
4378
		    sk->sk_type == SOCK_STREAM) {
4379
			skb = tcp_get_timestamping_opt_stats(sk);
4380 4381
			opt_stats = true;
		} else
4382 4383 4384
#endif
			skb = alloc_skb(0, GFP_ATOMIC);
	} else {
4385
		skb = skb_clone(orig_skb, GFP_ATOMIC);
4386
	}
4387 4388 4389
	if (!skb)
		return;

4390
	if (tsonly) {
4391 4392
		skb_shinfo(skb)->tx_flags |= skb_shinfo(orig_skb)->tx_flags &
					     SKBTX_ANY_TSTAMP;
4393 4394 4395 4396 4397 4398 4399 4400
		skb_shinfo(skb)->tskey = skb_shinfo(orig_skb)->tskey;
	}

	if (hwtstamps)
		*skb_hwtstamps(skb) = *hwtstamps;
	else
		skb->tstamp = ktime_get_real();

4401
	__skb_complete_tx_timestamp(skb, sk, tstype, opt_stats);
4402
}
4403 4404 4405 4406 4407 4408 4409 4410
EXPORT_SYMBOL_GPL(__skb_tstamp_tx);

void skb_tstamp_tx(struct sk_buff *orig_skb,
		   struct skb_shared_hwtstamps *hwtstamps)
{
	return __skb_tstamp_tx(orig_skb, hwtstamps, orig_skb->sk,
			       SCM_TSTAMP_SND);
}
4411 4412
EXPORT_SYMBOL_GPL(skb_tstamp_tx);

4413 4414 4415 4416
void skb_complete_wifi_ack(struct sk_buff *skb, bool acked)
{
	struct sock *sk = skb->sk;
	struct sock_exterr_skb *serr;
4417
	int err = 1;
4418 4419 4420 4421 4422 4423 4424 4425 4426

	skb->wifi_acked_valid = 1;
	skb->wifi_acked = acked;

	serr = SKB_EXT_ERR(skb);
	memset(serr, 0, sizeof(*serr));
	serr->ee.ee_errno = ENOMSG;
	serr->ee.ee_origin = SO_EE_ORIGIN_TXSTATUS;

4427 4428 4429
	/* Take a reference to prevent skb_orphan() from freeing the socket,
	 * but only if the socket refcount is not zero.
	 */
4430
	if (likely(refcount_inc_not_zero(&sk->sk_refcnt))) {
4431 4432 4433
		err = sock_queue_err_skb(sk, skb);
		sock_put(sk);
	}
4434 4435 4436 4437 4438
	if (err)
		kfree_skb(skb);
}
EXPORT_SYMBOL_GPL(skb_complete_wifi_ack);

4439 4440 4441 4442 4443 4444 4445 4446 4447 4448 4449 4450 4451 4452
/**
 * skb_partial_csum_set - set up and verify partial csum values for packet
 * @skb: the skb to set
 * @start: the number of bytes after skb->data to start checksumming.
 * @off: the offset from start to place the checksum.
 *
 * For untrusted partially-checksummed packets, we need to make sure the values
 * for skb->csum_start and skb->csum_offset are valid so we don't oops.
 *
 * This function checks and sets those values and skb->ip_summed: if this
 * returns false you should drop the packet.
 */
bool skb_partial_csum_set(struct sk_buff *skb, u16 start, u16 off)
{
4453 4454
	if (unlikely(start > skb_headlen(skb)) ||
	    unlikely((int)start + off > skb_headlen(skb) - 2)) {
4455 4456
		net_warn_ratelimited("bad partial csum: csum=%u/%u len=%u\n",
				     start, off, skb_headlen(skb));
4457 4458 4459 4460 4461
		return false;
	}
	skb->ip_summed = CHECKSUM_PARTIAL;
	skb->csum_start = skb_headroom(skb) + start;
	skb->csum_offset = off;
4462
	skb_set_transport_header(skb, start);
4463 4464
	return true;
}
4465
EXPORT_SYMBOL_GPL(skb_partial_csum_set);
4466

P
Paul Durrant 已提交
4467 4468 4469 4470 4471 4472 4473 4474 4475 4476 4477 4478 4479 4480 4481 4482 4483 4484 4485 4486 4487
static int skb_maybe_pull_tail(struct sk_buff *skb, unsigned int len,
			       unsigned int max)
{
	if (skb_headlen(skb) >= len)
		return 0;

	/* If we need to pullup then pullup to the max, so we
	 * won't need to do it again.
	 */
	if (max > skb->len)
		max = skb->len;

	if (__pskb_pull_tail(skb, max - skb_headlen(skb)) == NULL)
		return -ENOMEM;

	if (skb_headlen(skb) < len)
		return -EPROTO;

	return 0;
}

4488 4489 4490 4491 4492 4493 4494 4495 4496 4497 4498 4499 4500 4501 4502 4503 4504 4505 4506 4507 4508 4509 4510 4511 4512 4513 4514 4515 4516 4517 4518
#define MAX_TCP_HDR_LEN (15 * 4)

static __sum16 *skb_checksum_setup_ip(struct sk_buff *skb,
				      typeof(IPPROTO_IP) proto,
				      unsigned int off)
{
	switch (proto) {
		int err;

	case IPPROTO_TCP:
		err = skb_maybe_pull_tail(skb, off + sizeof(struct tcphdr),
					  off + MAX_TCP_HDR_LEN);
		if (!err && !skb_partial_csum_set(skb, off,
						  offsetof(struct tcphdr,
							   check)))
			err = -EPROTO;
		return err ? ERR_PTR(err) : &tcp_hdr(skb)->check;

	case IPPROTO_UDP:
		err = skb_maybe_pull_tail(skb, off + sizeof(struct udphdr),
					  off + sizeof(struct udphdr));
		if (!err && !skb_partial_csum_set(skb, off,
						  offsetof(struct udphdr,
							   check)))
			err = -EPROTO;
		return err ? ERR_PTR(err) : &udp_hdr(skb)->check;
	}

	return ERR_PTR(-EPROTO);
}

P
Paul Durrant 已提交
4519 4520 4521 4522 4523
/* This value should be large enough to cover a tagged ethernet header plus
 * maximally sized IP and TCP or UDP headers.
 */
#define MAX_IP_HDR_LEN 128

4524
static int skb_checksum_setup_ipv4(struct sk_buff *skb, bool recalculate)
P
Paul Durrant 已提交
4525 4526 4527
{
	unsigned int off;
	bool fragment;
4528
	__sum16 *csum;
P
Paul Durrant 已提交
4529 4530 4531 4532 4533 4534 4535 4536 4537 4538 4539 4540 4541 4542 4543 4544 4545 4546 4547 4548
	int err;

	fragment = false;

	err = skb_maybe_pull_tail(skb,
				  sizeof(struct iphdr),
				  MAX_IP_HDR_LEN);
	if (err < 0)
		goto out;

	if (ip_hdr(skb)->frag_off & htons(IP_OFFSET | IP_MF))
		fragment = true;

	off = ip_hdrlen(skb);

	err = -EPROTO;

	if (fragment)
		goto out;

4549 4550 4551
	csum = skb_checksum_setup_ip(skb, ip_hdr(skb)->protocol, off);
	if (IS_ERR(csum))
		return PTR_ERR(csum);
P
Paul Durrant 已提交
4552

4553 4554 4555 4556 4557
	if (recalculate)
		*csum = ~csum_tcpudp_magic(ip_hdr(skb)->saddr,
					   ip_hdr(skb)->daddr,
					   skb->len - off,
					   ip_hdr(skb)->protocol, 0);
P
Paul Durrant 已提交
4558 4559 4560 4561 4562 4563 4564 4565 4566 4567 4568 4569 4570 4571 4572 4573 4574 4575 4576 4577 4578 4579
	err = 0;

out:
	return err;
}

/* This value should be large enough to cover a tagged ethernet header plus
 * an IPv6 header, all options, and a maximal TCP or UDP header.
 */
#define MAX_IPV6_HDR_LEN 256

#define OPT_HDR(type, skb, off) \
	(type *)(skb_network_header(skb) + (off))

static int skb_checksum_setup_ipv6(struct sk_buff *skb, bool recalculate)
{
	int err;
	u8 nexthdr;
	unsigned int off;
	unsigned int len;
	bool fragment;
	bool done;
4580
	__sum16 *csum;
P
Paul Durrant 已提交
4581 4582 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 4611 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 4639 4640 4641 4642 4643 4644 4645 4646 4647 4648 4649 4650 4651 4652 4653 4654 4655 4656 4657

	fragment = false;
	done = false;

	off = sizeof(struct ipv6hdr);

	err = skb_maybe_pull_tail(skb, off, MAX_IPV6_HDR_LEN);
	if (err < 0)
		goto out;

	nexthdr = ipv6_hdr(skb)->nexthdr;

	len = sizeof(struct ipv6hdr) + ntohs(ipv6_hdr(skb)->payload_len);
	while (off <= len && !done) {
		switch (nexthdr) {
		case IPPROTO_DSTOPTS:
		case IPPROTO_HOPOPTS:
		case IPPROTO_ROUTING: {
			struct ipv6_opt_hdr *hp;

			err = skb_maybe_pull_tail(skb,
						  off +
						  sizeof(struct ipv6_opt_hdr),
						  MAX_IPV6_HDR_LEN);
			if (err < 0)
				goto out;

			hp = OPT_HDR(struct ipv6_opt_hdr, skb, off);
			nexthdr = hp->nexthdr;
			off += ipv6_optlen(hp);
			break;
		}
		case IPPROTO_AH: {
			struct ip_auth_hdr *hp;

			err = skb_maybe_pull_tail(skb,
						  off +
						  sizeof(struct ip_auth_hdr),
						  MAX_IPV6_HDR_LEN);
			if (err < 0)
				goto out;

			hp = OPT_HDR(struct ip_auth_hdr, skb, off);
			nexthdr = hp->nexthdr;
			off += ipv6_authlen(hp);
			break;
		}
		case IPPROTO_FRAGMENT: {
			struct frag_hdr *hp;

			err = skb_maybe_pull_tail(skb,
						  off +
						  sizeof(struct frag_hdr),
						  MAX_IPV6_HDR_LEN);
			if (err < 0)
				goto out;

			hp = OPT_HDR(struct frag_hdr, skb, off);

			if (hp->frag_off & htons(IP6_OFFSET | IP6_MF))
				fragment = true;

			nexthdr = hp->nexthdr;
			off += sizeof(struct frag_hdr);
			break;
		}
		default:
			done = true;
			break;
		}
	}

	err = -EPROTO;

	if (!done || fragment)
		goto out;

4658 4659 4660
	csum = skb_checksum_setup_ip(skb, nexthdr, off);
	if (IS_ERR(csum))
		return PTR_ERR(csum);
P
Paul Durrant 已提交
4661

4662 4663 4664 4665
	if (recalculate)
		*csum = ~csum_ipv6_magic(&ipv6_hdr(skb)->saddr,
					 &ipv6_hdr(skb)->daddr,
					 skb->len - off, nexthdr, 0);
P
Paul Durrant 已提交
4666 4667 4668 4669 4670 4671 4672 4673 4674 4675 4676 4677 4678 4679 4680 4681 4682
	err = 0;

out:
	return err;
}

/**
 * skb_checksum_setup - set up partial checksum offset
 * @skb: the skb to set up
 * @recalculate: if true the pseudo-header checksum will be recalculated
 */
int skb_checksum_setup(struct sk_buff *skb, bool recalculate)
{
	int err;

	switch (skb->protocol) {
	case htons(ETH_P_IP):
4683
		err = skb_checksum_setup_ipv4(skb, recalculate);
P
Paul Durrant 已提交
4684 4685 4686 4687 4688 4689 4690 4691 4692 4693 4694 4695 4696 4697 4698
		break;

	case htons(ETH_P_IPV6):
		err = skb_checksum_setup_ipv6(skb, recalculate);
		break;

	default:
		err = -EPROTO;
		break;
	}

	return err;
}
EXPORT_SYMBOL(skb_checksum_setup);

4699 4700 4701 4702 4703 4704 4705 4706 4707 4708
/**
 * skb_checksum_maybe_trim - maybe trims the given skb
 * @skb: the skb to check
 * @transport_len: the data length beyond the network header
 *
 * Checks whether the given skb has data beyond the given transport length.
 * If so, returns a cloned skb trimmed to this transport length.
 * Otherwise returns the provided skb. Returns NULL in error cases
 * (e.g. transport_len exceeds skb length or out-of-memory).
 *
4709 4710
 * Caller needs to set the skb transport header and free any returned skb if it
 * differs from the provided skb.
4711 4712 4713 4714 4715 4716 4717 4718
 */
static struct sk_buff *skb_checksum_maybe_trim(struct sk_buff *skb,
					       unsigned int transport_len)
{
	struct sk_buff *skb_chk;
	unsigned int len = skb_transport_offset(skb) + transport_len;
	int ret;

4719
	if (skb->len < len)
4720
		return NULL;
4721
	else if (skb->len == len)
4722 4723 4724 4725 4726 4727 4728 4729 4730 4731 4732 4733 4734 4735 4736 4737 4738 4739 4740 4741 4742 4743 4744 4745 4746 4747 4748
		return skb;

	skb_chk = skb_clone(skb, GFP_ATOMIC);
	if (!skb_chk)
		return NULL;

	ret = pskb_trim_rcsum(skb_chk, len);
	if (ret) {
		kfree_skb(skb_chk);
		return NULL;
	}

	return skb_chk;
}

/**
 * skb_checksum_trimmed - validate checksum of an skb
 * @skb: the skb to check
 * @transport_len: the data length beyond the network header
 * @skb_chkf: checksum function to use
 *
 * Applies the given checksum function skb_chkf to the provided skb.
 * Returns a checked and maybe trimmed skb. Returns NULL on error.
 *
 * If the skb has data beyond the given transport length, then a
 * trimmed & cloned skb is checked and returned.
 *
4749 4750
 * Caller needs to set the skb transport header and free any returned skb if it
 * differs from the provided skb.
4751 4752 4753 4754 4755 4756 4757
 */
struct sk_buff *skb_checksum_trimmed(struct sk_buff *skb,
				     unsigned int transport_len,
				     __sum16(*skb_chkf)(struct sk_buff *skb))
{
	struct sk_buff *skb_chk;
	unsigned int offset = skb_transport_offset(skb);
4758
	__sum16 ret;
4759 4760 4761

	skb_chk = skb_checksum_maybe_trim(skb, transport_len);
	if (!skb_chk)
4762
		goto err;
4763

4764 4765
	if (!pskb_may_pull(skb_chk, offset))
		goto err;
4766

4767
	skb_pull_rcsum(skb_chk, offset);
4768
	ret = skb_chkf(skb_chk);
4769
	skb_push_rcsum(skb_chk, offset);
4770

4771 4772
	if (ret)
		goto err;
4773 4774

	return skb_chk;
4775 4776 4777 4778 4779 4780 4781

err:
	if (skb_chk && skb_chk != skb)
		kfree_skb(skb_chk);

	return NULL;

4782 4783 4784
}
EXPORT_SYMBOL(skb_checksum_trimmed);

4785 4786
void __skb_warn_lro_forwarding(const struct sk_buff *skb)
{
4787 4788
	net_warn_ratelimited("%s: received packets cannot be forwarded while LRO is enabled\n",
			     skb->dev->name);
4789 4790
}
EXPORT_SYMBOL(__skb_warn_lro_forwarding);
E
Eric Dumazet 已提交
4791 4792 4793

void kfree_skb_partial(struct sk_buff *skb, bool head_stolen)
{
E
Eric Dumazet 已提交
4794 4795
	if (head_stolen) {
		skb_release_head_state(skb);
E
Eric Dumazet 已提交
4796
		kmem_cache_free(skbuff_head_cache, skb);
E
Eric Dumazet 已提交
4797
	} else {
E
Eric Dumazet 已提交
4798
		__kfree_skb(skb);
E
Eric Dumazet 已提交
4799
	}
E
Eric Dumazet 已提交
4800 4801 4802 4803 4804 4805 4806 4807
}
EXPORT_SYMBOL(kfree_skb_partial);

/**
 * skb_try_coalesce - try to merge skb to prior one
 * @to: prior buffer
 * @from: buffer to add
 * @fragstolen: pointer to boolean
R
Randy Dunlap 已提交
4808
 * @delta_truesize: how much more was allocated than was requested
E
Eric Dumazet 已提交
4809 4810 4811 4812
 */
bool skb_try_coalesce(struct sk_buff *to, struct sk_buff *from,
		      bool *fragstolen, int *delta_truesize)
{
4813
	struct skb_shared_info *to_shinfo, *from_shinfo;
E
Eric Dumazet 已提交
4814 4815 4816 4817 4818 4819 4820 4821
	int i, delta, len = from->len;

	*fragstolen = false;

	if (skb_cloned(to))
		return false;

	if (len <= skb_tailroom(to)) {
4822 4823
		if (len)
			BUG_ON(skb_copy_bits(from, 0, skb_put(to, len), len));
E
Eric Dumazet 已提交
4824 4825 4826 4827
		*delta_truesize = 0;
		return true;
	}

4828 4829 4830
	to_shinfo = skb_shinfo(to);
	from_shinfo = skb_shinfo(from);
	if (to_shinfo->frag_list || from_shinfo->frag_list)
E
Eric Dumazet 已提交
4831
		return false;
W
Willem de Bruijn 已提交
4832 4833
	if (skb_zcopy(to) || skb_zcopy(from))
		return false;
E
Eric Dumazet 已提交
4834 4835 4836 4837 4838

	if (skb_headlen(from) != 0) {
		struct page *page;
		unsigned int offset;

4839 4840
		if (to_shinfo->nr_frags +
		    from_shinfo->nr_frags >= MAX_SKB_FRAGS)
E
Eric Dumazet 已提交
4841 4842 4843 4844 4845 4846 4847 4848 4849 4850
			return false;

		if (skb_head_is_locked(from))
			return false;

		delta = from->truesize - SKB_DATA_ALIGN(sizeof(struct sk_buff));

		page = virt_to_head_page(from->head);
		offset = from->data - (unsigned char *)page_address(page);

4851
		skb_fill_page_desc(to, to_shinfo->nr_frags,
E
Eric Dumazet 已提交
4852 4853 4854
				   page, offset, skb_headlen(from));
		*fragstolen = true;
	} else {
4855 4856
		if (to_shinfo->nr_frags +
		    from_shinfo->nr_frags > MAX_SKB_FRAGS)
E
Eric Dumazet 已提交
4857 4858
			return false;

4859
		delta = from->truesize - SKB_TRUESIZE(skb_end_offset(from));
E
Eric Dumazet 已提交
4860 4861 4862 4863
	}

	WARN_ON_ONCE(delta < len);

4864 4865 4866 4867
	memcpy(to_shinfo->frags + to_shinfo->nr_frags,
	       from_shinfo->frags,
	       from_shinfo->nr_frags * sizeof(skb_frag_t));
	to_shinfo->nr_frags += from_shinfo->nr_frags;
E
Eric Dumazet 已提交
4868 4869

	if (!skb_cloned(from))
4870
		from_shinfo->nr_frags = 0;
E
Eric Dumazet 已提交
4871

4872 4873 4874
	/* if the skb is not cloned this does nothing
	 * since we set nr_frags to 0.
	 */
4875 4876
	for (i = 0; i < from_shinfo->nr_frags; i++)
		__skb_frag_ref(&from_shinfo->frags[i]);
E
Eric Dumazet 已提交
4877 4878 4879 4880 4881 4882 4883 4884 4885

	to->truesize += delta;
	to->len += len;
	to->data_len += len;

	*delta_truesize = delta;
	return true;
}
EXPORT_SYMBOL(skb_try_coalesce);
4886 4887

/**
4888
 * skb_scrub_packet - scrub an skb
4889 4890
 *
 * @skb: buffer to clean
4891 4892 4893 4894 4895 4896 4897 4898
 * @xnet: packet is crossing netns
 *
 * skb_scrub_packet can be used after encapsulating or decapsulting a packet
 * into/from a tunnel. Some information have to be cleared during these
 * operations.
 * skb_scrub_packet can also be used to clean a skb before injecting it in
 * another namespace (@xnet == true). We have to clear all information in the
 * skb that could impact namespace isolation.
4899
 */
4900
void skb_scrub_packet(struct sk_buff *skb, bool xnet)
4901 4902 4903
{
	skb->pkt_type = PACKET_HOST;
	skb->skb_iif = 0;
W
WANG Cong 已提交
4904
	skb->ignore_df = 0;
4905 4906 4907 4908
	skb_dst_drop(skb);
	secpath_reset(skb);
	nf_reset(skb);
	nf_reset_trace(skb);
4909 4910 4911 4912

	if (!xnet)
		return;

4913
	ipvs_reset(skb);
4914
	skb->mark = 0;
4915
	skb->tstamp = 0;
4916 4917
}
EXPORT_SYMBOL_GPL(skb_scrub_packet);
4918 4919 4920 4921 4922 4923 4924 4925 4926 4927 4928

/**
 * skb_gso_transport_seglen - Return length of individual segments of a gso packet
 *
 * @skb: GSO skb
 *
 * skb_gso_transport_seglen is used to determine the real size of the
 * individual segments, including Layer4 headers (TCP/UDP).
 *
 * The MAC/L2 or network (IP, IPv6) headers are not accounted for.
 */
4929
static unsigned int skb_gso_transport_seglen(const struct sk_buff *skb)
4930 4931
{
	const struct skb_shared_info *shinfo = skb_shinfo(skb);
4932
	unsigned int thlen = 0;
4933

4934 4935 4936
	if (skb->encapsulation) {
		thlen = skb_inner_transport_header(skb) -
			skb_transport_header(skb);
4937

4938 4939 4940 4941
		if (likely(shinfo->gso_type & (SKB_GSO_TCPV4 | SKB_GSO_TCPV6)))
			thlen += inner_tcp_hdrlen(skb);
	} else if (likely(shinfo->gso_type & (SKB_GSO_TCPV4 | SKB_GSO_TCPV6))) {
		thlen = tcp_hdrlen(skb);
4942
	} else if (unlikely(skb_is_gso_sctp(skb))) {
M
Marcelo Ricardo Leitner 已提交
4943
		thlen = sizeof(struct sctphdr);
W
Willem de Bruijn 已提交
4944 4945
	} else if (shinfo->gso_type & SKB_GSO_UDP_L4) {
		thlen = sizeof(struct udphdr);
4946
	}
4947 4948 4949 4950
	/* UFO sets gso_size to the size of the fragmentation
	 * payload, i.e. the size of the L4 (UDP) header is already
	 * accounted for.
	 */
4951
	return thlen + shinfo->gso_size;
4952
}
4953 4954 4955 4956 4957 4958 4959 4960 4961 4962 4963 4964 4965 4966 4967 4968 4969 4970 4971 4972 4973 4974 4975 4976 4977 4978 4979 4980 4981 4982 4983 4984 4985 4986

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

	return hdr_len + skb_gso_transport_seglen(skb);
}

/**
 * skb_gso_mac_seglen - Return length of individual segments of a gso packet
 *
 * @skb: GSO skb
 *
 * skb_gso_mac_seglen is used to determine the real size of the
 * individual segments, including MAC/L2, Layer3 (IP, IPv6) and L4
 * headers (TCP/UDP).
 */
static unsigned int skb_gso_mac_seglen(const struct sk_buff *skb)
{
	unsigned int hdr_len = skb_transport_header(skb) - skb_mac_header(skb);

	return hdr_len + skb_gso_transport_seglen(skb);
}
4987

4988
/**
4989
 * skb_gso_size_check - check the skb size, considering GSO_BY_FRAGS
4990
 *
4991 4992
 * There are a couple of instances where we have a GSO skb, and we
 * want to determine what size it would be after it is segmented.
4993
 *
4994 4995 4996 4997 4998 4999 5000 5001 5002 5003 5004 5005
 * We might want to check:
 * -    L3+L4+payload size (e.g. IP forwarding)
 * - L2+L3+L4+payload size (e.g. sanity check before passing to driver)
 *
 * This is a helper to do that correctly considering GSO_BY_FRAGS.
 *
 * @seg_len: The segmented length (from skb_gso_*_seglen). In the
 *           GSO_BY_FRAGS case this will be [header sizes + GSO_BY_FRAGS].
 *
 * @max_len: The maximum permissible length.
 *
 * Returns true if the segmented length <= max length.
5006
 */
5007 5008 5009
static inline bool skb_gso_size_check(const struct sk_buff *skb,
				      unsigned int seg_len,
				      unsigned int max_len) {
5010 5011 5012 5013
	const struct skb_shared_info *shinfo = skb_shinfo(skb);
	const struct sk_buff *iter;

	if (shinfo->gso_size != GSO_BY_FRAGS)
5014
		return seg_len <= max_len;
5015 5016

	/* Undo this so we can re-use header sizes */
5017
	seg_len -= GSO_BY_FRAGS;
5018 5019

	skb_walk_frags(skb, iter) {
5020
		if (seg_len + skb_headlen(iter) > max_len)
5021 5022 5023 5024 5025
			return false;
	}

	return true;
}
5026 5027

/**
5028
 * skb_gso_validate_network_len - Will a split GSO skb fit into a given MTU?
5029 5030 5031 5032
 *
 * @skb: GSO skb
 * @mtu: MTU to validate against
 *
5033 5034 5035
 * skb_gso_validate_network_len validates if a given skb will fit a
 * wanted MTU once split. It considers L3 headers, L4 headers, and the
 * payload.
5036
 */
5037
bool skb_gso_validate_network_len(const struct sk_buff *skb, unsigned int mtu)
5038 5039 5040
{
	return skb_gso_size_check(skb, skb_gso_network_seglen(skb), mtu);
}
5041
EXPORT_SYMBOL_GPL(skb_gso_validate_network_len);
5042

5043 5044 5045 5046 5047 5048 5049 5050 5051 5052 5053 5054 5055 5056 5057
/**
 * skb_gso_validate_mac_len - Will a split GSO skb fit in a given length?
 *
 * @skb: GSO skb
 * @len: length to validate against
 *
 * skb_gso_validate_mac_len validates if a given skb will fit a wanted
 * length once split, including L2, L3 and L4 headers and the payload.
 */
bool skb_gso_validate_mac_len(const struct sk_buff *skb, unsigned int len)
{
	return skb_gso_size_check(skb, skb_gso_mac_seglen(skb), len);
}
EXPORT_SYMBOL_GPL(skb_gso_validate_mac_len);

5058 5059
static struct sk_buff *skb_reorder_vlan_header(struct sk_buff *skb)
{
5060 5061
	int mac_len;

5062 5063 5064 5065 5066
	if (skb_cow(skb, skb_headroom(skb)) < 0) {
		kfree_skb(skb);
		return NULL;
	}

5067
	mac_len = skb->data - skb_mac_header(skb);
5068 5069 5070 5071
	if (likely(mac_len > VLAN_HLEN + ETH_TLEN)) {
		memmove(skb_mac_header(skb) + VLAN_HLEN, skb_mac_header(skb),
			mac_len - VLAN_HLEN - ETH_TLEN);
	}
5072 5073 5074 5075 5076 5077 5078 5079 5080
	skb->mac_header += VLAN_HLEN;
	return skb;
}

struct sk_buff *skb_vlan_untag(struct sk_buff *skb)
{
	struct vlan_hdr *vhdr;
	u16 vlan_tci;

5081
	if (unlikely(skb_vlan_tag_present(skb))) {
5082 5083 5084 5085 5086 5087 5088 5089 5090 5091 5092 5093 5094 5095 5096 5097 5098 5099 5100 5101 5102 5103 5104 5105 5106 5107 5108 5109 5110 5111 5112 5113 5114
		/* vlan_tci is already set-up so leave this for another time */
		return skb;
	}

	skb = skb_share_check(skb, GFP_ATOMIC);
	if (unlikely(!skb))
		goto err_free;

	if (unlikely(!pskb_may_pull(skb, VLAN_HLEN)))
		goto err_free;

	vhdr = (struct vlan_hdr *)skb->data;
	vlan_tci = ntohs(vhdr->h_vlan_TCI);
	__vlan_hwaccel_put_tag(skb, skb->protocol, vlan_tci);

	skb_pull_rcsum(skb, VLAN_HLEN);
	vlan_set_encap_proto(skb, vhdr);

	skb = skb_reorder_vlan_header(skb);
	if (unlikely(!skb))
		goto err_free;

	skb_reset_network_header(skb);
	skb_reset_transport_header(skb);
	skb_reset_mac_len(skb);

	return skb;

err_free:
	kfree_skb(skb);
	return NULL;
}
EXPORT_SYMBOL(skb_vlan_untag);
5115

5116 5117 5118 5119 5120 5121 5122 5123 5124 5125 5126 5127
int skb_ensure_writable(struct sk_buff *skb, int write_len)
{
	if (!pskb_may_pull(skb, write_len))
		return -ENOMEM;

	if (!skb_cloned(skb) || skb_clone_writable(skb, write_len))
		return 0;

	return pskb_expand_head(skb, 0, 0, GFP_ATOMIC);
}
EXPORT_SYMBOL(skb_ensure_writable);

5128 5129 5130 5131
/* remove VLAN header from packet and update csum accordingly.
 * expects a non skb_vlan_tag_present skb with a vlan tag payload
 */
int __skb_vlan_pop(struct sk_buff *skb, u16 *vlan_tci)
5132 5133
{
	struct vlan_hdr *vhdr;
5134
	int offset = skb->data - skb_mac_header(skb);
5135 5136
	int err;

5137 5138 5139 5140 5141 5142
	if (WARN_ONCE(offset,
		      "__skb_vlan_pop got skb with skb->data not at mac header (offset %d)\n",
		      offset)) {
		return -EINVAL;
	}

5143 5144
	err = skb_ensure_writable(skb, VLAN_ETH_HLEN);
	if (unlikely(err))
5145
		return err;
5146 5147 5148 5149 5150 5151 5152 5153 5154 5155 5156 5157 5158 5159 5160 5161 5162 5163 5164

	skb_postpull_rcsum(skb, skb->data + (2 * ETH_ALEN), VLAN_HLEN);

	vhdr = (struct vlan_hdr *)(skb->data + ETH_HLEN);
	*vlan_tci = ntohs(vhdr->h_vlan_TCI);

	memmove(skb->data + VLAN_HLEN, skb->data, 2 * ETH_ALEN);
	__skb_pull(skb, VLAN_HLEN);

	vlan_set_encap_proto(skb, vhdr);
	skb->mac_header += VLAN_HLEN;

	if (skb_network_offset(skb) < ETH_HLEN)
		skb_set_network_header(skb, ETH_HLEN);

	skb_reset_mac_len(skb);

	return err;
}
5165
EXPORT_SYMBOL(__skb_vlan_pop);
5166

5167 5168 5169
/* Pop a vlan tag either from hwaccel or from payload.
 * Expects skb->data at mac header.
 */
5170 5171 5172 5173 5174 5175
int skb_vlan_pop(struct sk_buff *skb)
{
	u16 vlan_tci;
	__be16 vlan_proto;
	int err;

5176
	if (likely(skb_vlan_tag_present(skb))) {
5177 5178
		skb->vlan_tci = 0;
	} else {
5179
		if (unlikely(!eth_type_vlan(skb->protocol)))
5180 5181 5182 5183 5184 5185 5186
			return 0;

		err = __skb_vlan_pop(skb, &vlan_tci);
		if (err)
			return err;
	}
	/* move next vlan tag to hw accel tag */
5187
	if (likely(!eth_type_vlan(skb->protocol)))
5188 5189 5190 5191 5192 5193 5194 5195 5196 5197 5198 5199
		return 0;

	vlan_proto = skb->protocol;
	err = __skb_vlan_pop(skb, &vlan_tci);
	if (unlikely(err))
		return err;

	__vlan_hwaccel_put_tag(skb, vlan_proto, vlan_tci);
	return 0;
}
EXPORT_SYMBOL(skb_vlan_pop);

5200 5201 5202
/* Push a vlan tag either into hwaccel or into payload (if hwaccel tag present).
 * Expects skb->data at mac header.
 */
5203 5204
int skb_vlan_push(struct sk_buff *skb, __be16 vlan_proto, u16 vlan_tci)
{
5205
	if (skb_vlan_tag_present(skb)) {
5206
		int offset = skb->data - skb_mac_header(skb);
5207 5208
		int err;

5209 5210 5211 5212 5213 5214
		if (WARN_ONCE(offset,
			      "skb_vlan_push got skb with skb->data not at mac header (offset %d)\n",
			      offset)) {
			return -EINVAL;
		}

5215
		err = __vlan_insert_tag(skb, skb->vlan_proto,
5216
					skb_vlan_tag_get(skb));
5217
		if (err)
5218
			return err;
5219

5220 5221 5222
		skb->protocol = skb->vlan_proto;
		skb->mac_len += VLAN_HLEN;

5223
		skb_postpush_rcsum(skb, skb->data + (2 * ETH_ALEN), VLAN_HLEN);
5224 5225 5226 5227 5228 5229
	}
	__vlan_hwaccel_put_tag(skb, vlan_proto, vlan_tci);
	return 0;
}
EXPORT_SYMBOL(skb_vlan_push);

5230 5231 5232
/**
 * alloc_skb_with_frags - allocate skb with page frags
 *
5233 5234 5235 5236 5237
 * @header_len: size of linear part
 * @data_len: needed length in frags
 * @max_page_order: max page order desired.
 * @errcode: pointer to error code if any
 * @gfp_mask: allocation mask
5238 5239 5240 5241 5242 5243 5244 5245 5246 5247 5248 5249 5250 5251 5252 5253 5254 5255 5256 5257 5258 5259 5260 5261
 *
 * This can be used to allocate a paged skb, given a maximal order for frags.
 */
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)
{
	int npages = (data_len + (PAGE_SIZE - 1)) >> PAGE_SHIFT;
	unsigned long chunk;
	struct sk_buff *skb;
	struct page *page;
	gfp_t gfp_head;
	int i;

	*errcode = -EMSGSIZE;
	/* Note this test could be relaxed, if we succeed to allocate
	 * high order pages...
	 */
	if (npages > MAX_SKB_FRAGS)
		return NULL;

	gfp_head = gfp_mask;
5262
	if (gfp_head & __GFP_DIRECT_RECLAIM)
5263
		gfp_head |= __GFP_RETRY_MAYFAIL;
5264 5265 5266 5267 5268 5269 5270 5271 5272 5273 5274 5275 5276

	*errcode = -ENOBUFS;
	skb = alloc_skb(header_len, gfp_head);
	if (!skb)
		return NULL;

	skb->truesize += npages << PAGE_SHIFT;

	for (i = 0; npages > 0; i++) {
		int order = max_page_order;

		while (order) {
			if (npages >= 1 << order) {
5277
				page = alloc_pages((gfp_mask & ~__GFP_DIRECT_RECLAIM) |
5278
						   __GFP_COMP |
5279
						   __GFP_NOWARN,
5280 5281 5282 5283 5284 5285 5286 5287 5288 5289 5290 5291 5292 5293 5294 5295 5296 5297 5298 5299 5300 5301 5302 5303 5304 5305
						   order);
				if (page)
					goto fill_page;
				/* Do not retry other high order allocations */
				order = 1;
				max_page_order = 0;
			}
			order--;
		}
		page = alloc_page(gfp_mask);
		if (!page)
			goto failure;
fill_page:
		chunk = min_t(unsigned long, data_len,
			      PAGE_SIZE << order);
		skb_fill_page_desc(skb, i, page, 0, chunk);
		data_len -= chunk;
		npages -= 1 << order;
	}
	return skb;

failure:
	kfree_skb(skb);
	return NULL;
}
EXPORT_SYMBOL(alloc_skb_with_frags);
5306 5307 5308 5309 5310 5311 5312 5313 5314 5315 5316 5317 5318 5319 5320 5321 5322 5323 5324 5325 5326 5327 5328 5329 5330 5331 5332 5333 5334 5335 5336 5337 5338 5339 5340 5341 5342 5343 5344 5345 5346 5347 5348 5349 5350 5351 5352 5353 5354 5355 5356 5357 5358 5359 5360 5361 5362 5363 5364 5365 5366 5367 5368 5369 5370 5371 5372 5373 5374 5375 5376 5377 5378 5379 5380 5381 5382 5383 5384 5385 5386 5387 5388 5389 5390 5391 5392 5393 5394 5395 5396 5397 5398 5399 5400 5401 5402 5403 5404 5405 5406 5407 5408 5409 5410 5411 5412 5413 5414 5415 5416 5417 5418 5419 5420 5421 5422 5423 5424 5425 5426 5427 5428 5429 5430 5431 5432 5433 5434 5435 5436 5437 5438 5439 5440 5441 5442 5443 5444 5445 5446 5447 5448 5449 5450 5451 5452 5453 5454 5455 5456 5457 5458 5459 5460 5461 5462 5463 5464 5465 5466 5467 5468 5469 5470 5471 5472 5473 5474 5475 5476 5477 5478 5479 5480 5481 5482 5483 5484 5485 5486 5487 5488 5489 5490 5491 5492 5493 5494 5495 5496 5497 5498 5499 5500 5501 5502 5503 5504 5505 5506 5507 5508 5509 5510 5511 5512 5513 5514 5515 5516 5517 5518 5519 5520 5521 5522 5523 5524 5525 5526 5527 5528 5529 5530 5531 5532 5533 5534 5535 5536 5537 5538 5539 5540 5541

/* carve out the first off bytes from skb when off < headlen */
static int pskb_carve_inside_header(struct sk_buff *skb, const u32 off,
				    const int headlen, gfp_t gfp_mask)
{
	int i;
	int size = skb_end_offset(skb);
	int new_hlen = headlen - off;
	u8 *data;

	size = SKB_DATA_ALIGN(size);

	if (skb_pfmemalloc(skb))
		gfp_mask |= __GFP_MEMALLOC;
	data = kmalloc_reserve(size +
			       SKB_DATA_ALIGN(sizeof(struct skb_shared_info)),
			       gfp_mask, NUMA_NO_NODE, NULL);
	if (!data)
		return -ENOMEM;

	size = SKB_WITH_OVERHEAD(ksize(data));

	/* Copy real data, and all frags */
	skb_copy_from_linear_data_offset(skb, off, data, new_hlen);
	skb->len -= off;

	memcpy((struct skb_shared_info *)(data + size),
	       skb_shinfo(skb),
	       offsetof(struct skb_shared_info,
			frags[skb_shinfo(skb)->nr_frags]));
	if (skb_cloned(skb)) {
		/* drop the old head gracefully */
		if (skb_orphan_frags(skb, gfp_mask)) {
			kfree(data);
			return -ENOMEM;
		}
		for (i = 0; i < skb_shinfo(skb)->nr_frags; i++)
			skb_frag_ref(skb, i);
		if (skb_has_frag_list(skb))
			skb_clone_fraglist(skb);
		skb_release_data(skb);
	} else {
		/* we can reuse existing recount- all we did was
		 * relocate values
		 */
		skb_free_head(skb);
	}

	skb->head = data;
	skb->data = data;
	skb->head_frag = 0;
#ifdef NET_SKBUFF_DATA_USES_OFFSET
	skb->end = size;
#else
	skb->end = skb->head + size;
#endif
	skb_set_tail_pointer(skb, skb_headlen(skb));
	skb_headers_offset_update(skb, 0);
	skb->cloned = 0;
	skb->hdr_len = 0;
	skb->nohdr = 0;
	atomic_set(&skb_shinfo(skb)->dataref, 1);

	return 0;
}

static int pskb_carve(struct sk_buff *skb, const u32 off, gfp_t gfp);

/* carve out the first eat bytes from skb's frag_list. May recurse into
 * pskb_carve()
 */
static int pskb_carve_frag_list(struct sk_buff *skb,
				struct skb_shared_info *shinfo, int eat,
				gfp_t gfp_mask)
{
	struct sk_buff *list = shinfo->frag_list;
	struct sk_buff *clone = NULL;
	struct sk_buff *insp = NULL;

	do {
		if (!list) {
			pr_err("Not enough bytes to eat. Want %d\n", eat);
			return -EFAULT;
		}
		if (list->len <= eat) {
			/* Eaten as whole. */
			eat -= list->len;
			list = list->next;
			insp = list;
		} else {
			/* Eaten partially. */
			if (skb_shared(list)) {
				clone = skb_clone(list, gfp_mask);
				if (!clone)
					return -ENOMEM;
				insp = list->next;
				list = clone;
			} else {
				/* This may be pulled without problems. */
				insp = list;
			}
			if (pskb_carve(list, eat, gfp_mask) < 0) {
				kfree_skb(clone);
				return -ENOMEM;
			}
			break;
		}
	} while (eat);

	/* Free pulled out fragments. */
	while ((list = shinfo->frag_list) != insp) {
		shinfo->frag_list = list->next;
		kfree_skb(list);
	}
	/* And insert new clone at head. */
	if (clone) {
		clone->next = list;
		shinfo->frag_list = clone;
	}
	return 0;
}

/* carve off first len bytes from skb. Split line (off) is in the
 * non-linear part of skb
 */
static int pskb_carve_inside_nonlinear(struct sk_buff *skb, const u32 off,
				       int pos, gfp_t gfp_mask)
{
	int i, k = 0;
	int size = skb_end_offset(skb);
	u8 *data;
	const int nfrags = skb_shinfo(skb)->nr_frags;
	struct skb_shared_info *shinfo;

	size = SKB_DATA_ALIGN(size);

	if (skb_pfmemalloc(skb))
		gfp_mask |= __GFP_MEMALLOC;
	data = kmalloc_reserve(size +
			       SKB_DATA_ALIGN(sizeof(struct skb_shared_info)),
			       gfp_mask, NUMA_NO_NODE, NULL);
	if (!data)
		return -ENOMEM;

	size = SKB_WITH_OVERHEAD(ksize(data));

	memcpy((struct skb_shared_info *)(data + size),
	       skb_shinfo(skb), offsetof(struct skb_shared_info,
					 frags[skb_shinfo(skb)->nr_frags]));
	if (skb_orphan_frags(skb, gfp_mask)) {
		kfree(data);
		return -ENOMEM;
	}
	shinfo = (struct skb_shared_info *)(data + size);
	for (i = 0; i < nfrags; i++) {
		int fsize = skb_frag_size(&skb_shinfo(skb)->frags[i]);

		if (pos + fsize > off) {
			shinfo->frags[k] = skb_shinfo(skb)->frags[i];

			if (pos < off) {
				/* Split frag.
				 * We have two variants in this case:
				 * 1. Move all the frag to the second
				 *    part, if it is possible. F.e.
				 *    this approach is mandatory for TUX,
				 *    where splitting is expensive.
				 * 2. Split is accurately. We make this.
				 */
				shinfo->frags[0].page_offset += off - pos;
				skb_frag_size_sub(&shinfo->frags[0], off - pos);
			}
			skb_frag_ref(skb, i);
			k++;
		}
		pos += fsize;
	}
	shinfo->nr_frags = k;
	if (skb_has_frag_list(skb))
		skb_clone_fraglist(skb);

	if (k == 0) {
		/* split line is in frag list */
		pskb_carve_frag_list(skb, shinfo, off - pos, gfp_mask);
	}
	skb_release_data(skb);

	skb->head = data;
	skb->head_frag = 0;
	skb->data = data;
#ifdef NET_SKBUFF_DATA_USES_OFFSET
	skb->end = size;
#else
	skb->end = skb->head + size;
#endif
	skb_reset_tail_pointer(skb);
	skb_headers_offset_update(skb, 0);
	skb->cloned   = 0;
	skb->hdr_len  = 0;
	skb->nohdr    = 0;
	skb->len -= off;
	skb->data_len = skb->len;
	atomic_set(&skb_shinfo(skb)->dataref, 1);
	return 0;
}

/* remove len bytes from the beginning of the skb */
static int pskb_carve(struct sk_buff *skb, const u32 len, gfp_t gfp)
{
	int headlen = skb_headlen(skb);

	if (len < headlen)
		return pskb_carve_inside_header(skb, len, headlen, gfp);
	else
		return pskb_carve_inside_nonlinear(skb, len, headlen, gfp);
}

/* Extract to_copy bytes starting at off from skb, and return this in
 * a new skb
 */
struct sk_buff *pskb_extract(struct sk_buff *skb, int off,
			     int to_copy, gfp_t gfp)
{
	struct sk_buff  *clone = skb_clone(skb, gfp);

	if (!clone)
		return NULL;

	if (pskb_carve(clone, off, gfp) < 0 ||
	    pskb_trim(clone, to_copy)) {
		kfree_skb(clone);
		return NULL;
	}
	return clone;
}
EXPORT_SYMBOL(pskb_extract);
5542 5543 5544 5545 5546 5547 5548 5549 5550 5551 5552 5553 5554 5555 5556

/**
 * skb_condense - try to get rid of fragments/frag_list if possible
 * @skb: buffer
 *
 * Can be used to save memory before skb is added to a busy queue.
 * If packet has bytes in frags and enough tail room in skb->head,
 * pull all of them, so that we can free the frags right now and adjust
 * truesize.
 * Notes:
 *	We do not reallocate skb->head thus can not fail.
 *	Caller must re-evaluate skb->truesize if needed.
 */
void skb_condense(struct sk_buff *skb)
{
5557 5558 5559 5560
	if (skb->data_len) {
		if (skb->data_len > skb->end - skb->tail ||
		    skb_cloned(skb))
			return;
5561

5562 5563 5564 5565 5566 5567 5568 5569 5570
		/* Nice, we can free page frag(s) right now */
		__pskb_pull_tail(skb, skb->data_len);
	}
	/* At this point, skb->truesize might be over estimated,
	 * because skb had a fragment, and fragments do not tell
	 * their truesize.
	 * When we pulled its content into skb->head, fragment
	 * was freed, but __pskb_pull_tail() could not possibly
	 * adjust skb->truesize, not knowing the frag truesize.
5571 5572 5573
	 */
	skb->truesize = SKB_TRUESIZE(skb_end_offset(skb));
}