skbuff.c 99.8 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
#include <linux/module.h>
#include <linux/types.h>
#include <linux/kernel.h>
44
#include <linux/kmemcheck.h>
L
Linus Torvalds 已提交
45 46 47 48 49
#include <linux/mm.h>
#include <linux/interrupt.h>
#include <linux/in.h>
#include <linux/inet.h>
#include <linux/slab.h>
50 51
#include <linux/tcp.h>
#include <linux/udp.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 74
#include <net/xfrm.h>

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

78
struct kmem_cache *skbuff_head_cache __read_mostly;
79
static struct kmem_cache *skbuff_fclone_cache __read_mostly;
L
Linus Torvalds 已提交
80 81

/**
82 83 84 85
 *	skb_panic - private function for out-of-line support
 *	@skb:	buffer
 *	@sz:	size
 *	@addr:	address
86
 *	@msg:	skb_over_panic or skb_under_panic
L
Linus Torvalds 已提交
87
 *
88 89 90 91
 *	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
92
 */
93
static void skb_panic(struct sk_buff *skb, unsigned int sz, void *addr,
94
		      const char msg[])
L
Linus Torvalds 已提交
95
{
J
Joe Perches 已提交
96
	pr_emerg("%s: text:%p len:%d put:%d head:%p data:%p tail:%#lx end:%#lx dev:%s\n",
97
		 msg, addr, skb->len, sz, skb->head, skb->data,
J
Joe Perches 已提交
98 99
		 (unsigned long)skb->tail, (unsigned long)skb->end,
		 skb->dev ? skb->dev->name : "<NULL>");
L
Linus Torvalds 已提交
100 101 102
	BUG();
}

103
static void skb_over_panic(struct sk_buff *skb, unsigned int sz, void *addr)
L
Linus Torvalds 已提交
104
{
105
	skb_panic(skb, sz, addr, __func__);
L
Linus Torvalds 已提交
106 107
}

108 109 110 111
static void skb_under_panic(struct sk_buff *skb, unsigned int sz, void *addr)
{
	skb_panic(skb, sz, addr, __func__);
}
112 113 114 115 116 117 118 119 120 121

/*
 * 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)
122 123 124

static void *__kmalloc_reserve(size_t size, gfp_t flags, int node,
			       unsigned long ip, bool *pfmemalloc)
125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149
{
	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
150 151 152 153 154 155
/* 	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.
 *
 */

156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171
struct sk_buff *__alloc_skb_head(gfp_t gfp_mask, int node)
{
	struct sk_buff *skb;

	/* Get the HEAD */
	skb = kmem_cache_alloc_node(skbuff_head_cache,
				    gfp_mask & ~__GFP_DMA, node);
	if (!skb)
		goto out;

	/*
	 * 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!
	 */
	memset(skb, 0, offsetof(struct sk_buff, tail));
172
	skb->head = NULL;
173 174 175
	skb->truesize = sizeof(struct sk_buff);
	atomic_set(&skb->users, 1);

C
Cong Wang 已提交
176
	skb->mac_header = (typeof(skb->mac_header))~0U;
177 178 179 180
out:
	return skb;
}

L
Linus Torvalds 已提交
181
/**
182
 *	__alloc_skb	-	allocate a network buffer
L
Linus Torvalds 已提交
183 184
 *	@size: size to allocate
 *	@gfp_mask: allocation mask
185 186 187 188
 *	@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
189
 *	@node: numa node to allocate memory on
L
Linus Torvalds 已提交
190 191
 *
 *	Allocate a new &sk_buff. The returned buffer has no headroom and a
192 193
 *	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
194 195 196 197
 *
 *	Buffers may only be allocated from interrupts using a @gfp_mask of
 *	%GFP_ATOMIC.
 */
A
Al Viro 已提交
198
struct sk_buff *__alloc_skb(unsigned int size, gfp_t gfp_mask,
199
			    int flags, int node)
L
Linus Torvalds 已提交
200
{
201
	struct kmem_cache *cache;
B
Benjamin LaHaise 已提交
202
	struct skb_shared_info *shinfo;
L
Linus Torvalds 已提交
203 204
	struct sk_buff *skb;
	u8 *data;
205
	bool pfmemalloc;
L
Linus Torvalds 已提交
206

207 208 209 210 211
	cache = (flags & SKB_ALLOC_FCLONE)
		? skbuff_fclone_cache : skbuff_head_cache;

	if (sk_memalloc_socks() && (flags & SKB_ALLOC_RX))
		gfp_mask |= __GFP_MEMALLOC;
212

L
Linus Torvalds 已提交
213
	/* Get the HEAD */
214
	skb = kmem_cache_alloc_node(cache, gfp_mask & ~__GFP_DMA, node);
L
Linus Torvalds 已提交
215 216
	if (!skb)
		goto out;
E
Eric Dumazet 已提交
217
	prefetchw(skb);
L
Linus Torvalds 已提交
218

E
Eric Dumazet 已提交
219 220 221 222 223
	/* 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.
	 */
224
	size = SKB_DATA_ALIGN(size);
E
Eric Dumazet 已提交
225
	size += SKB_DATA_ALIGN(sizeof(struct skb_shared_info));
226
	data = kmalloc_reserve(size, gfp_mask, node, &pfmemalloc);
L
Linus Torvalds 已提交
227 228
	if (!data)
		goto nodata;
E
Eric Dumazet 已提交
229 230 231 232 233
	/* 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
234
	prefetchw(data + size);
L
Linus Torvalds 已提交
235

236
	/*
237 238 239
	 * 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!
240 241
	 */
	memset(skb, 0, offsetof(struct sk_buff, tail));
E
Eric Dumazet 已提交
242 243
	/* Account for allocated memory : skb + skb->head */
	skb->truesize = SKB_TRUESIZE(size);
244
	skb->pfmemalloc = pfmemalloc;
L
Linus Torvalds 已提交
245 246 247
	atomic_set(&skb->users, 1);
	skb->head = data;
	skb->data = data;
248
	skb_reset_tail_pointer(skb);
249
	skb->end = skb->tail + size;
C
Cong Wang 已提交
250 251
	skb->mac_header = (typeof(skb->mac_header))~0U;
	skb->transport_header = (typeof(skb->transport_header))~0U;
252

B
Benjamin LaHaise 已提交
253 254
	/* make sure we initialize shinfo sequentially */
	shinfo = skb_shinfo(skb);
E
Eric Dumazet 已提交
255
	memset(shinfo, 0, offsetof(struct skb_shared_info, dataref));
B
Benjamin LaHaise 已提交
256
	atomic_set(&shinfo->dataref, 1);
257
	kmemcheck_annotate_variable(shinfo->destructor_arg);
B
Benjamin LaHaise 已提交
258

259
	if (flags & SKB_ALLOC_FCLONE) {
260 261
		struct sk_buff *child = skb + 1;
		atomic_t *fclone_ref = (atomic_t *) (child + 1);
L
Linus Torvalds 已提交
262

263 264
		kmemcheck_annotate_bitfield(child, flags1);
		kmemcheck_annotate_bitfield(child, flags2);
265 266 267 268
		skb->fclone = SKB_FCLONE_ORIG;
		atomic_set(fclone_ref, 1);

		child->fclone = SKB_FCLONE_UNAVAILABLE;
269
		child->pfmemalloc = pfmemalloc;
270
	}
L
Linus Torvalds 已提交
271 272 273
out:
	return skb;
nodata:
274
	kmem_cache_free(cache, skb);
L
Linus Torvalds 已提交
275 276 277
	skb = NULL;
	goto out;
}
278
EXPORT_SYMBOL(__alloc_skb);
L
Linus Torvalds 已提交
279

E
Eric Dumazet 已提交
280 281 282
/**
 * build_skb - build a network buffer
 * @data: data buffer provided by caller
283
 * @frag_size: size of fragment, or 0 if head was kmalloced
E
Eric Dumazet 已提交
284 285
 *
 * Allocate a new &sk_buff. Caller provides space holding head and
286 287
 * skb_shared_info. @data must have been allocated by kmalloc() only if
 * @frag_size is 0, otherwise data should come from the page allocator.
E
Eric Dumazet 已提交
288 289 290 291 292 293 294 295 296 297
 * 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.
 */
298
struct sk_buff *build_skb(void *data, unsigned int frag_size)
E
Eric Dumazet 已提交
299 300 301
{
	struct skb_shared_info *shinfo;
	struct sk_buff *skb;
302
	unsigned int size = frag_size ? : ksize(data);
E
Eric Dumazet 已提交
303 304 305 306 307

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

308
	size -= SKB_DATA_ALIGN(sizeof(struct skb_shared_info));
E
Eric Dumazet 已提交
309 310 311

	memset(skb, 0, offsetof(struct sk_buff, tail));
	skb->truesize = SKB_TRUESIZE(size);
312
	skb->head_frag = frag_size != 0;
E
Eric Dumazet 已提交
313 314 315 316 317
	atomic_set(&skb->users, 1);
	skb->head = data;
	skb->data = data;
	skb_reset_tail_pointer(skb);
	skb->end = skb->tail + size;
C
Cong Wang 已提交
318 319
	skb->mac_header = (typeof(skb->mac_header))~0U;
	skb->transport_header = (typeof(skb->transport_header))~0U;
E
Eric Dumazet 已提交
320 321 322 323 324 325 326 327 328 329 330

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

	return skb;
}
EXPORT_SYMBOL(build_skb);

331
struct netdev_alloc_cache {
332 333 334 335 336
	struct page_frag	frag;
	/* we maintain a pagecount bias, so that we dont dirty cache line
	 * containing page->_count every time we allocate a fragment.
	 */
	unsigned int		pagecnt_bias;
337 338 339
};
static DEFINE_PER_CPU(struct netdev_alloc_cache, netdev_alloc_cache);

340
static void *__netdev_alloc_frag(unsigned int fragsz, gfp_t gfp_mask)
341 342 343
{
	struct netdev_alloc_cache *nc;
	void *data = NULL;
344
	int order;
345 346 347 348
	unsigned long flags;

	local_irq_save(flags);
	nc = &__get_cpu_var(netdev_alloc_cache);
349
	if (unlikely(!nc->frag.page)) {
350
refill:
351 352 353 354 355 356 357 358 359 360 361 362
		for (order = NETDEV_FRAG_PAGE_MAX_ORDER; ;) {
			gfp_t gfp = gfp_mask;

			if (order)
				gfp |= __GFP_COMP | __GFP_NOWARN;
			nc->frag.page = alloc_pages(gfp, order);
			if (likely(nc->frag.page))
				break;
			if (--order < 0)
				goto end;
		}
		nc->frag.size = PAGE_SIZE << order;
363
recycle:
364 365 366
		atomic_set(&nc->frag.page->_count, NETDEV_PAGECNT_MAX_BIAS);
		nc->pagecnt_bias = NETDEV_PAGECNT_MAX_BIAS;
		nc->frag.offset = 0;
367
	}
368

369
	if (nc->frag.offset + fragsz > nc->frag.size) {
370
		/* avoid unnecessary locked operations if possible */
371 372
		if ((atomic_read(&nc->frag.page->_count) == nc->pagecnt_bias) ||
		    atomic_sub_and_test(nc->pagecnt_bias, &nc->frag.page->_count))
373 374
			goto recycle;
		goto refill;
375
	}
376

377 378
	data = page_address(nc->frag.page) + nc->frag.offset;
	nc->frag.offset += fragsz;
379 380
	nc->pagecnt_bias--;
end:
381 382 383
	local_irq_restore(flags);
	return data;
}
384 385 386 387 388 389 390 391 392 393 394 395

/**
 * 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)
{
	return __netdev_alloc_frag(fragsz, GFP_ATOMIC | __GFP_COLD);
}
396 397
EXPORT_SYMBOL(netdev_alloc_frag);

398 399 400 401 402 403 404 405 406 407 408 409 410 411
/**
 *	__netdev_alloc_skb - allocate an skbuff for rx on a specific device
 *	@dev: network device to receive on
 *	@length: length to allocate
 *	@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 unspecified headroom built in. Users should allocate
 *	the headroom they think they need without accounting for the
 *	built in space. The built in space is used for optimisations.
 *
 *	%NULL is returned if there is no free memory.
 */
struct sk_buff *__netdev_alloc_skb(struct net_device *dev,
412
				   unsigned int length, gfp_t gfp_mask)
413
{
414
	struct sk_buff *skb = NULL;
415 416 417
	unsigned int fragsz = SKB_DATA_ALIGN(length + NET_SKB_PAD) +
			      SKB_DATA_ALIGN(sizeof(struct skb_shared_info));

418
	if (fragsz <= PAGE_SIZE && !(gfp_mask & (__GFP_WAIT | GFP_DMA))) {
419 420 421 422 423 424
		void *data;

		if (sk_memalloc_socks())
			gfp_mask |= __GFP_MEMALLOC;

		data = __netdev_alloc_frag(fragsz, gfp_mask);
425

426 427 428 429
		if (likely(data)) {
			skb = build_skb(data, fragsz);
			if (unlikely(!skb))
				put_page(virt_to_head_page(data));
430 431
		}
	} else {
432 433
		skb = __alloc_skb(length + NET_SKB_PAD, gfp_mask,
				  SKB_ALLOC_RX, NUMA_NO_NODE);
434
	}
435
	if (likely(skb)) {
436
		skb_reserve(skb, NET_SKB_PAD);
437 438
		skb->dev = dev;
	}
439 440
	return skb;
}
441
EXPORT_SYMBOL(__netdev_alloc_skb);
L
Linus Torvalds 已提交
442

P
Peter Zijlstra 已提交
443
void skb_add_rx_frag(struct sk_buff *skb, int i, struct page *page, int off,
444
		     int size, unsigned int truesize)
P
Peter Zijlstra 已提交
445 446 447 448
{
	skb_fill_page_desc(skb, i, page, off, size);
	skb->len += size;
	skb->data_len += size;
449
	skb->truesize += truesize;
P
Peter Zijlstra 已提交
450 451 452
}
EXPORT_SYMBOL(skb_add_rx_frag);

J
Jason Wang 已提交
453 454 455 456 457 458 459 460 461 462 463 464
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);

465
static void skb_drop_list(struct sk_buff **listp)
L
Linus Torvalds 已提交
466
{
E
Eric Dumazet 已提交
467
	kfree_skb_list(*listp);
468
	*listp = NULL;
L
Linus Torvalds 已提交
469 470
}

471 472 473 474 475
static inline void skb_drop_fraglist(struct sk_buff *skb)
{
	skb_drop_list(&skb_shinfo(skb)->frag_list);
}

L
Linus Torvalds 已提交
476 477 478 479
static void skb_clone_fraglist(struct sk_buff *skb)
{
	struct sk_buff *list;

480
	skb_walk_frags(skb, list)
L
Linus Torvalds 已提交
481 482 483
		skb_get(list);
}

484 485 486 487 488 489 490 491
static void skb_free_head(struct sk_buff *skb)
{
	if (skb->head_frag)
		put_page(virt_to_head_page(skb->head));
	else
		kfree(skb->head);
}

492
static void skb_release_data(struct sk_buff *skb)
L
Linus Torvalds 已提交
493 494 495 496 497 498 499
{
	if (!skb->cloned ||
	    !atomic_sub_return(skb->nohdr ? (1 << SKB_DATAREF_SHIFT) + 1 : 1,
			       &skb_shinfo(skb)->dataref)) {
		if (skb_shinfo(skb)->nr_frags) {
			int i;
			for (i = 0; i < skb_shinfo(skb)->nr_frags; i++)
500
				skb_frag_unref(skb, i);
L
Linus Torvalds 已提交
501 502
		}

503 504 505 506 507 508 509 510 511
		/*
		 * If skb buf is from userspace, we need to notify the caller
		 * the lower device DMA has done;
		 */
		if (skb_shinfo(skb)->tx_flags & SKBTX_DEV_ZEROCOPY) {
			struct ubuf_info *uarg;

			uarg = skb_shinfo(skb)->destructor_arg;
			if (uarg->callback)
512
				uarg->callback(uarg, true);
513 514
		}

515
		if (skb_has_frag_list(skb))
L
Linus Torvalds 已提交
516 517
			skb_drop_fraglist(skb);

518
		skb_free_head(skb);
L
Linus Torvalds 已提交
519 520 521 522 523 524
	}
}

/*
 *	Free an skbuff by memory without cleaning the state.
 */
525
static void kfree_skbmem(struct sk_buff *skb)
L
Linus Torvalds 已提交
526
{
527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552
	struct sk_buff *other;
	atomic_t *fclone_ref;

	switch (skb->fclone) {
	case SKB_FCLONE_UNAVAILABLE:
		kmem_cache_free(skbuff_head_cache, skb);
		break;

	case SKB_FCLONE_ORIG:
		fclone_ref = (atomic_t *) (skb + 2);
		if (atomic_dec_and_test(fclone_ref))
			kmem_cache_free(skbuff_fclone_cache, skb);
		break;

	case SKB_FCLONE_CLONE:
		fclone_ref = (atomic_t *) (skb + 1);
		other = skb - 1;

		/* The clone portion is available for
		 * fast-cloning again.
		 */
		skb->fclone = SKB_FCLONE_UNAVAILABLE;

		if (atomic_dec_and_test(fclone_ref))
			kmem_cache_free(skbuff_fclone_cache, other);
		break;
553
	}
L
Linus Torvalds 已提交
554 555
}

556
static void skb_release_head_state(struct sk_buff *skb)
L
Linus Torvalds 已提交
557
{
E
Eric Dumazet 已提交
558
	skb_dst_drop(skb);
L
Linus Torvalds 已提交
559 560 561
#ifdef CONFIG_XFRM
	secpath_put(skb->sp);
#endif
562 563
	if (skb->destructor) {
		WARN_ON(in_irq());
L
Linus Torvalds 已提交
564 565
		skb->destructor(skb);
	}
I
Igor Maravić 已提交
566
#if IS_ENABLED(CONFIG_NF_CONNTRACK)
567
	nf_conntrack_put(skb->nfct);
568
#endif
L
Linus Torvalds 已提交
569 570 571 572 573 574 575 576 577 578
#ifdef CONFIG_BRIDGE_NETFILTER
	nf_bridge_put(skb->nf_bridge);
#endif
/* XXX: IS this still necessary? - JHS */
#ifdef CONFIG_NET_SCHED
	skb->tc_index = 0;
#ifdef CONFIG_NET_CLS_ACT
	skb->tc_verd = 0;
#endif
#endif
579 580 581 582 583 584
}

/* Free everything but the sk_buff shell. */
static void skb_release_all(struct sk_buff *skb)
{
	skb_release_head_state(skb);
585
	if (likely(skb->head))
586
		skb_release_data(skb);
587 588 589 590 591 592 593 594 595 596
}

/**
 *	__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
597

598 599 600
void __kfree_skb(struct sk_buff *skb)
{
	skb_release_all(skb);
L
Linus Torvalds 已提交
601 602
	kfree_skbmem(skb);
}
603
EXPORT_SYMBOL(__kfree_skb);
L
Linus Torvalds 已提交
604

605 606 607 608 609 610 611 612 613 614 615 616 617 618 619
/**
 *	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)
{
	if (unlikely(!skb))
		return;
	if (likely(atomic_read(&skb->users) == 1))
		smp_rmb();
	else if (likely(!atomic_dec_and_test(&skb->users)))
		return;
620
	trace_kfree_skb(skb, __builtin_return_address(0));
621 622
	__kfree_skb(skb);
}
623
EXPORT_SYMBOL(kfree_skb);
624

E
Eric Dumazet 已提交
625 626 627 628 629 630 631 632 633 634 635
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);

636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655
/**
 *	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)
{
	if (skb_shinfo(skb)->tx_flags & SKBTX_DEV_ZEROCOPY) {
		struct ubuf_info *uarg;

		uarg = skb_shinfo(skb)->destructor_arg;
		if (uarg->callback)
			uarg->callback(uarg, false);
		skb_shinfo(skb)->tx_flags &= ~SKBTX_DEV_ZEROCOPY;
	}
}
EXPORT_SYMBOL(skb_tx_error);

656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671
/**
 *	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)
{
	if (unlikely(!skb))
		return;
	if (likely(atomic_read(&skb->users) == 1))
		smp_rmb();
	else if (likely(!atomic_dec_and_test(&skb->users)))
		return;
672
	trace_consume_skb(skb);
673 674 675 676
	__kfree_skb(skb);
}
EXPORT_SYMBOL(consume_skb);

677 678 679 680 681 682 683
static void __copy_skb_header(struct sk_buff *new, const struct sk_buff *old)
{
	new->tstamp		= old->tstamp;
	new->dev		= old->dev;
	new->transport_header	= old->transport_header;
	new->network_header	= old->network_header;
	new->mac_header		= old->mac_header;
684
	new->inner_protocol	= old->inner_protocol;
685
	new->inner_transport_header = old->inner_transport_header;
686
	new->inner_network_header = old->inner_network_header;
687
	new->inner_mac_header = old->inner_mac_header;
E
Eric Dumazet 已提交
688
	skb_dst_copy(new, old);
689
	skb_copy_hash(new, old);
690
	new->ooo_okay		= old->ooo_okay;
691
	new->no_fcs		= old->no_fcs;
692
	new->encapsulation	= old->encapsulation;
693 694 695
	new->encap_hdr_csum	= old->encap_hdr_csum;
	new->csum_valid		= old->csum_valid;
	new->csum_complete_sw	= old->csum_complete_sw;
696
#ifdef CONFIG_XFRM
697 698 699
	new->sp			= secpath_get(old->sp);
#endif
	memcpy(new->cb, old->cb, sizeof(old->cb));
700
	new->csum		= old->csum;
W
WANG Cong 已提交
701
	new->ignore_df		= old->ignore_df;
702 703 704 705
	new->pkt_type		= old->pkt_type;
	new->ip_summed		= old->ip_summed;
	skb_copy_queue_mapping(new, old);
	new->priority		= old->priority;
I
Igor Maravić 已提交
706
#if IS_ENABLED(CONFIG_IP_VS)
707 708
	new->ipvs_property	= old->ipvs_property;
#endif
709
	new->pfmemalloc		= old->pfmemalloc;
710 711
	new->protocol		= old->protocol;
	new->mark		= old->mark;
712
	new->skb_iif		= old->skb_iif;
713 714 715 716 717 718 719
	__nf_copy(new, old);
#ifdef CONFIG_NET_SCHED
	new->tc_index		= old->tc_index;
#ifdef CONFIG_NET_CLS_ACT
	new->tc_verd		= old->tc_verd;
#endif
#endif
720
	new->vlan_proto		= old->vlan_proto;
721 722
	new->vlan_tci		= old->vlan_tci;

723
	skb_copy_secmark(new, old);
E
Eliezer Tamir 已提交
724

725
#ifdef CONFIG_NET_RX_BUSY_POLL
E
Eliezer Tamir 已提交
726 727
	new->napi_id	= old->napi_id;
#endif
728 729
}

730 731 732 733
/*
 * You should not add any new code to this function.  Add it to
 * __copy_skb_header above instead.
 */
H
Herbert Xu 已提交
734
static struct sk_buff *__skb_clone(struct sk_buff *n, struct sk_buff *skb)
L
Linus Torvalds 已提交
735 736 737 738 739
{
#define C(x) n->x = skb->x

	n->next = n->prev = NULL;
	n->sk = NULL;
740 741
	__copy_skb_header(n, skb);

L
Linus Torvalds 已提交
742 743
	C(len);
	C(data_len);
744
	C(mac_len);
745
	n->hdr_len = skb->nohdr ? skb_headroom(skb) : skb->hdr_len;
746
	n->cloned = 1;
L
Linus Torvalds 已提交
747 748 749 750
	n->nohdr = 0;
	n->destructor = NULL;
	C(tail);
	C(end);
751
	C(head);
752
	C(head_frag);
753 754 755
	C(data);
	C(truesize);
	atomic_set(&n->users, 1);
L
Linus Torvalds 已提交
756 757 758 759 760

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

	return n;
H
Herbert Xu 已提交
761 762 763 764 765 766 767 768 769 770 771 772 773 774 775
#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)
{
776
	skb_release_all(dst);
H
Herbert Xu 已提交
777 778 779 780
	return __skb_clone(dst, src);
}
EXPORT_SYMBOL_GPL(skb_morph);

781 782
/**
 *	skb_copy_ubufs	-	copy userspace skb frags buffers to kernel
783 784 785 786 787 788 789 790 791 792 793 794 795 796
 *	@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)
797 798 799 800 801 802 803 804 805 806
{
	int i;
	int num_frags = skb_shinfo(skb)->nr_frags;
	struct page *page, *head = NULL;
	struct ubuf_info *uarg = skb_shinfo(skb)->destructor_arg;

	for (i = 0; i < num_frags; i++) {
		u8 *vaddr;
		skb_frag_t *f = &skb_shinfo(skb)->frags[i];

807
		page = alloc_page(gfp_mask);
808 809
		if (!page) {
			while (head) {
810
				struct page *next = (struct page *)page_private(head);
811 812 813 814 815
				put_page(head);
				head = next;
			}
			return -ENOMEM;
		}
E
Eric Dumazet 已提交
816
		vaddr = kmap_atomic(skb_frag_page(f));
817
		memcpy(page_address(page),
E
Eric Dumazet 已提交
818
		       vaddr + f->page_offset, skb_frag_size(f));
E
Eric Dumazet 已提交
819
		kunmap_atomic(vaddr);
820
		set_page_private(page, (unsigned long)head);
821 822 823 824
		head = page;
	}

	/* skb frags release userspace buffers */
825
	for (i = 0; i < num_frags; i++)
826
		skb_frag_unref(skb, i);
827

828
	uarg->callback(uarg, false);
829 830

	/* skb frags point to kernel buffers */
831 832 833
	for (i = num_frags - 1; i >= 0; i--) {
		__skb_fill_page_desc(skb, i, head, 0,
				     skb_shinfo(skb)->frags[i].size);
834
		head = (struct page *)page_private(head);
835
	}
836 837

	skb_shinfo(skb)->tx_flags &= ~SKBTX_DEV_ZEROCOPY;
838 839
	return 0;
}
840
EXPORT_SYMBOL_GPL(skb_copy_ubufs);
841

H
Herbert Xu 已提交
842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859
/**
 *	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)
{
	struct sk_buff *n;

860 861
	if (skb_orphan_frags(skb, gfp_mask))
		return NULL;
862

H
Herbert Xu 已提交
863 864 865 866 867 868 869
	n = skb + 1;
	if (skb->fclone == SKB_FCLONE_ORIG &&
	    n->fclone == SKB_FCLONE_UNAVAILABLE) {
		atomic_t *fclone_ref = (atomic_t *) (n + 1);
		n->fclone = SKB_FCLONE_CLONE;
		atomic_inc(fclone_ref);
	} else {
870 871 872
		if (skb_pfmemalloc(skb))
			gfp_mask |= __GFP_MEMALLOC;

H
Herbert Xu 已提交
873 874 875
		n = kmem_cache_alloc(skbuff_head_cache, gfp_mask);
		if (!n)
			return NULL;
876 877 878

		kmemcheck_annotate_bitfield(n, flags1);
		kmemcheck_annotate_bitfield(n, flags2);
H
Herbert Xu 已提交
879 880 881 882
		n->fclone = SKB_FCLONE_UNAVAILABLE;
	}

	return __skb_clone(n, skb);
L
Linus Torvalds 已提交
883
}
884
EXPORT_SYMBOL(skb_clone);
L
Linus Torvalds 已提交
885

886 887
static void skb_headers_offset_update(struct sk_buff *skb, int off)
{
E
Eric Dumazet 已提交
888 889 890
	/* Only adjust this if it actually is csum_start rather than csum */
	if (skb->ip_summed == CHECKSUM_PARTIAL)
		skb->csum_start += off;
891 892 893 894 895 896 897
	/* {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;
898
	skb->inner_mac_header += off;
899 900
}

L
Linus Torvalds 已提交
901 902
static void copy_skb_header(struct sk_buff *new, const struct sk_buff *old)
{
903 904
	__copy_skb_header(new, old);

905 906 907
	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
908 909
}

910 911 912 913 914 915 916
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
917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933
/**
 *	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
934
struct sk_buff *skb_copy(const struct sk_buff *skb, gfp_t gfp_mask)
L
Linus Torvalds 已提交
935
{
E
Eric Dumazet 已提交
936
	int headerlen = skb_headroom(skb);
937
	unsigned int size = skb_end_offset(skb) + skb->data_len;
938 939
	struct sk_buff *n = __alloc_skb(size, gfp_mask,
					skb_alloc_rx_flag(skb), NUMA_NO_NODE);
E
Eric Dumazet 已提交
940

L
Linus Torvalds 已提交
941 942 943 944 945 946 947 948 949 950 951 952 953 954
	if (!n)
		return NULL;

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

	if (skb_copy_bits(skb, -headerlen, n->head, headerlen + skb->len))
		BUG();

	copy_skb_header(n, skb);
	return n;
}
955
EXPORT_SYMBOL(skb_copy);
L
Linus Torvalds 已提交
956 957

/**
958
 *	__pskb_copy_fclone	-  create copy of an sk_buff with private head.
L
Linus Torvalds 已提交
959
 *	@skb: buffer to copy
E
Eric Dumazet 已提交
960
 *	@headroom: headroom of new skb
L
Linus Torvalds 已提交
961
 *	@gfp_mask: allocation priority
962 963 964
 *	@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
965 966 967 968 969 970 971 972 973
 *
 *	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.
 */

974 975
struct sk_buff *__pskb_copy_fclone(struct sk_buff *skb, int headroom,
				   gfp_t gfp_mask, bool fclone)
L
Linus Torvalds 已提交
976
{
E
Eric Dumazet 已提交
977
	unsigned int size = skb_headlen(skb) + headroom;
978 979
	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
980

L
Linus Torvalds 已提交
981 982 983 984
	if (!n)
		goto out;

	/* Set the data pointer */
E
Eric Dumazet 已提交
985
	skb_reserve(n, headroom);
L
Linus Torvalds 已提交
986 987 988
	/* Set the tail pointer and length */
	skb_put(n, skb_headlen(skb));
	/* Copy the bytes */
989
	skb_copy_from_linear_data(skb, n->data, n->len);
L
Linus Torvalds 已提交
990

H
Herbert Xu 已提交
991
	n->truesize += skb->data_len;
L
Linus Torvalds 已提交
992 993 994 995 996 997
	n->data_len  = skb->data_len;
	n->len	     = skb->len;

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

998 999 1000 1001
		if (skb_orphan_frags(skb, gfp_mask)) {
			kfree_skb(n);
			n = NULL;
			goto out;
1002
		}
L
Linus Torvalds 已提交
1003 1004
		for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
			skb_shinfo(n)->frags[i] = skb_shinfo(skb)->frags[i];
1005
			skb_frag_ref(skb, i);
L
Linus Torvalds 已提交
1006 1007 1008 1009
		}
		skb_shinfo(n)->nr_frags = i;
	}

1010
	if (skb_has_frag_list(skb)) {
L
Linus Torvalds 已提交
1011 1012 1013 1014 1015 1016 1017 1018
		skb_shinfo(n)->frag_list = skb_shinfo(skb)->frag_list;
		skb_clone_fraglist(n);
	}

	copy_skb_header(n, skb);
out:
	return n;
}
1019
EXPORT_SYMBOL(__pskb_copy_fclone);
L
Linus Torvalds 已提交
1020 1021 1022 1023 1024 1025 1026 1027

/**
 *	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
1028 1029
 *	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
1030 1031 1032 1033 1034 1035 1036
 *	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
1037
int pskb_expand_head(struct sk_buff *skb, int nhead, int ntail,
A
Al Viro 已提交
1038
		     gfp_t gfp_mask)
L
Linus Torvalds 已提交
1039 1040 1041
{
	int i;
	u8 *data;
1042
	int size = nhead + skb_end_offset(skb) + ntail;
L
Linus Torvalds 已提交
1043 1044
	long off;

1045 1046
	BUG_ON(nhead < 0);

L
Linus Torvalds 已提交
1047 1048 1049 1050 1051
	if (skb_shared(skb))
		BUG();

	size = SKB_DATA_ALIGN(size);

1052 1053 1054 1055
	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
1056 1057
	if (!data)
		goto nodata;
1058
	size = SKB_WITH_OVERHEAD(ksize(data));
L
Linus Torvalds 已提交
1059 1060

	/* Copy only real data... and, alas, header. This should be
E
Eric Dumazet 已提交
1061 1062 1063 1064 1065 1066
	 * 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),
1067
	       offsetof(struct skb_shared_info, frags[skb_shinfo(skb)->nr_frags]));
L
Linus Torvalds 已提交
1068

1069 1070 1071 1072 1073 1074
	/*
	 * 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)) {
1075
		/* copy this zero copy skb frags */
1076 1077
		if (skb_orphan_frags(skb, gfp_mask))
			goto nofrags;
1078
		for (i = 0; i < skb_shinfo(skb)->nr_frags; i++)
1079
			skb_frag_ref(skb, i);
L
Linus Torvalds 已提交
1080

1081 1082
		if (skb_has_frag_list(skb))
			skb_clone_fraglist(skb);
L
Linus Torvalds 已提交
1083

1084
		skb_release_data(skb);
1085 1086
	} else {
		skb_free_head(skb);
1087
	}
L
Linus Torvalds 已提交
1088 1089 1090
	off = (data + nhead) - skb->head;

	skb->head     = data;
1091
	skb->head_frag = 0;
L
Linus Torvalds 已提交
1092
	skb->data    += off;
1093 1094
#ifdef NET_SKBUFF_DATA_USES_OFFSET
	skb->end      = size;
1095
	off           = nhead;
1096 1097
#else
	skb->end      = skb->head + size;
1098
#endif
1099
	skb->tail	      += off;
1100
	skb_headers_offset_update(skb, nhead);
L
Linus Torvalds 已提交
1101
	skb->cloned   = 0;
1102
	skb->hdr_len  = 0;
L
Linus Torvalds 已提交
1103 1104 1105 1106
	skb->nohdr    = 0;
	atomic_set(&skb_shinfo(skb)->dataref, 1);
	return 0;

1107 1108
nofrags:
	kfree(data);
L
Linus Torvalds 已提交
1109 1110 1111
nodata:
	return -ENOMEM;
}
1112
EXPORT_SYMBOL(pskb_expand_head);
L
Linus Torvalds 已提交
1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132

/* 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;
}
1133
EXPORT_SYMBOL(skb_realloc_headroom);
L
Linus Torvalds 已提交
1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153

/**
 *	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
1154
				int newheadroom, int newtailroom,
A
Al Viro 已提交
1155
				gfp_t gfp_mask)
L
Linus Torvalds 已提交
1156 1157 1158 1159
{
	/*
	 *	Allocate the copy buffer
	 */
1160 1161 1162
	struct sk_buff *n = __alloc_skb(newheadroom + skb->len + newtailroom,
					gfp_mask, skb_alloc_rx_flag(skb),
					NUMA_NO_NODE);
1163
	int oldheadroom = skb_headroom(skb);
L
Linus Torvalds 已提交
1164 1165 1166 1167 1168 1169 1170 1171 1172 1173
	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);

1174
	head_copy_len = oldheadroom;
L
Linus Torvalds 已提交
1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187
	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. */
	if (skb_copy_bits(skb, -head_copy_len, n->head + head_copy_off,
			  skb->len + head_copy_len))
		BUG();

	copy_skb_header(n, skb);

E
Eric Dumazet 已提交
1188
	skb_headers_offset_update(n, newheadroom - oldheadroom);
1189

L
Linus Torvalds 已提交
1190 1191
	return n;
}
1192
EXPORT_SYMBOL(skb_copy_expand);
L
Linus Torvalds 已提交
1193 1194 1195 1196 1197 1198 1199 1200 1201 1202

/**
 *	skb_pad			-	zero pad the tail of an skb
 *	@skb: buffer to pad
 *	@pad: space to pad
 *
 *	Ensure that a buffer is followed by a padding area that is zero
 *	filled. Used by network drivers which may DMA or transfer data
 *	beyond the buffer end onto the wire.
 *
1203
 *	May return error in out of memory cases. The skb is freed on error.
L
Linus Torvalds 已提交
1204
 */
1205

1206
int skb_pad(struct sk_buff *skb, int pad)
L
Linus Torvalds 已提交
1207
{
1208 1209
	int err;
	int ntail;
1210

L
Linus Torvalds 已提交
1211
	/* If the skbuff is non linear tailroom is always zero.. */
1212
	if (!skb_cloned(skb) && skb_tailroom(skb) >= pad) {
L
Linus Torvalds 已提交
1213
		memset(skb->data+skb->len, 0, pad);
1214
		return 0;
L
Linus Torvalds 已提交
1215
	}
1216

1217
	ntail = skb->data_len + pad - (skb->end - skb->tail);
1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234
	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:
L
Linus Torvalds 已提交
1235
	kfree_skb(skb);
1236
	return err;
1237
}
1238
EXPORT_SYMBOL(skb_pad);
1239

M
Mathias Krause 已提交
1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262
/**
 *	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.
 */

unsigned char *pskb_put(struct sk_buff *skb, struct sk_buff *tail, int len)
{
	if (tail != skb) {
		skb->data_len += len;
		skb->len += len;
	}
	return skb_put(tail, len);
}
EXPORT_SYMBOL_GPL(pskb_put);

1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283
/**
 *	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.
 */
unsigned char *skb_put(struct sk_buff *skb, unsigned int len)
{
	unsigned char *tmp = skb_tail_pointer(skb);
	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);

1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302
/**
 *	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.
 */
unsigned char *skb_push(struct sk_buff *skb, unsigned int len)
{
	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);

1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314
/**
 *	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.
 */
unsigned char *skb_pull(struct sk_buff *skb, unsigned int len)
{
1315
	return skb_pull_inline(skb, len);
1316 1317 1318
}
EXPORT_SYMBOL(skb_pull);

1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334
/**
 *	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);

1335
/* Trims skb to length len. It can change skb pointers.
L
Linus Torvalds 已提交
1336 1337
 */

1338
int ___pskb_trim(struct sk_buff *skb, unsigned int len)
L
Linus Torvalds 已提交
1339
{
1340 1341
	struct sk_buff **fragp;
	struct sk_buff *frag;
L
Linus Torvalds 已提交
1342 1343 1344
	int offset = skb_headlen(skb);
	int nfrags = skb_shinfo(skb)->nr_frags;
	int i;
1345 1346 1347 1348 1349
	int err;

	if (skb_cloned(skb) &&
	    unlikely((err = pskb_expand_head(skb, 0, 0, GFP_ATOMIC))))
		return err;
L
Linus Torvalds 已提交
1350

1351 1352 1353 1354 1355
	i = 0;
	if (offset >= len)
		goto drop_pages;

	for (; i < nfrags; i++) {
E
Eric Dumazet 已提交
1356
		int end = offset + skb_frag_size(&skb_shinfo(skb)->frags[i]);
1357 1358 1359 1360 1361 1362

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

E
Eric Dumazet 已提交
1363
		skb_frag_size_set(&skb_shinfo(skb)->frags[i++], len - offset);
1364

1365
drop_pages:
1366 1367 1368
		skb_shinfo(skb)->nr_frags = i;

		for (; i < nfrags; i++)
1369
			skb_frag_unref(skb, i);
1370

1371
		if (skb_has_frag_list(skb))
1372
			skb_drop_fraglist(skb);
1373
		goto done;
1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387
	}

	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;
1388
			consume_skb(frag);
1389 1390
			frag = nfrag;
			*fragp = frag;
L
Linus Torvalds 已提交
1391
		}
1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404

		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
1405 1406
	}

1407
done:
1408
	if (len > skb_headlen(skb)) {
L
Linus Torvalds 已提交
1409 1410 1411
		skb->data_len -= skb->len - len;
		skb->len       = len;
	} else {
1412 1413
		skb->len       = len;
		skb->data_len  = 0;
1414
		skb_set_tail_pointer(skb, len);
L
Linus Torvalds 已提交
1415 1416 1417 1418
	}

	return 0;
}
1419
EXPORT_SYMBOL(___pskb_trim);
L
Linus Torvalds 已提交
1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451

/**
 *	__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.
 */
unsigned char *__pskb_pull_tail(struct sk_buff *skb, int delta)
{
	/* 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.
	 */
1452
	int i, k, eat = (skb->tail + delta) - skb->end;
L
Linus Torvalds 已提交
1453 1454 1455 1456 1457 1458 1459

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

1460
	if (skb_copy_bits(skb, skb_headlen(skb), skb_tail_pointer(skb), delta))
L
Linus Torvalds 已提交
1461 1462 1463 1464 1465
		BUG();

	/* Optimization: no fragments, no reasons to preestimate
	 * size of pulled pages. Superb.
	 */
1466
	if (!skb_has_frag_list(skb))
L
Linus Torvalds 已提交
1467 1468 1469 1470 1471
		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
1472 1473 1474
		int size = skb_frag_size(&skb_shinfo(skb)->frags[i]);

		if (size >= eat)
L
Linus Torvalds 已提交
1475
			goto pull_pages;
E
Eric Dumazet 已提交
1476
		eat -= size;
L
Linus Torvalds 已提交
1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491
	}

	/* If we need update frag list, we are in troubles.
	 * Certainly, it possible to add an offset to skb data,
	 * 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 {
1492
			BUG_ON(!list);
L
Linus Torvalds 已提交
1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514

			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)) {
1515
					kfree_skb(clone);
L
Linus Torvalds 已提交
1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538
					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
1539 1540 1541
		int size = skb_frag_size(&skb_shinfo(skb)->frags[i]);

		if (size <= eat) {
1542
			skb_frag_unref(skb, i);
E
Eric Dumazet 已提交
1543
			eat -= size;
L
Linus Torvalds 已提交
1544 1545 1546 1547
		} 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
1548
				skb_frag_size_sub(&skb_shinfo(skb)->frags[k], eat);
L
Linus Torvalds 已提交
1549 1550 1551 1552 1553 1554 1555 1556 1557 1558
				eat = 0;
			}
			k++;
		}
	}
	skb_shinfo(skb)->nr_frags = k;

	skb->tail     += delta;
	skb->data_len -= delta;

1559
	return skb_tail_pointer(skb);
L
Linus Torvalds 已提交
1560
}
1561
EXPORT_SYMBOL(__pskb_pull_tail);
L
Linus Torvalds 已提交
1562

1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577
/**
 *	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
1578 1579
int skb_copy_bits(const struct sk_buff *skb, int offset, void *to, int len)
{
1580
	int start = skb_headlen(skb);
1581 1582
	struct sk_buff *frag_iter;
	int i, copy;
L
Linus Torvalds 已提交
1583 1584 1585 1586 1587

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

	/* Copy header. */
1588
	if ((copy = start - offset) > 0) {
L
Linus Torvalds 已提交
1589 1590
		if (copy > len)
			copy = len;
1591
		skb_copy_from_linear_data_offset(skb, offset, to, copy);
L
Linus Torvalds 已提交
1592 1593 1594 1595 1596 1597 1598
		if ((len -= copy) == 0)
			return 0;
		offset += copy;
		to     += copy;
	}

	for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
1599
		int end;
E
Eric Dumazet 已提交
1600
		skb_frag_t *f = &skb_shinfo(skb)->frags[i];
L
Linus Torvalds 已提交
1601

1602
		WARN_ON(start > offset + len);
1603

E
Eric Dumazet 已提交
1604
		end = start + skb_frag_size(f);
L
Linus Torvalds 已提交
1605 1606 1607 1608 1609 1610
		if ((copy = end - offset) > 0) {
			u8 *vaddr;

			if (copy > len)
				copy = len;

E
Eric Dumazet 已提交
1611
			vaddr = kmap_atomic(skb_frag_page(f));
L
Linus Torvalds 已提交
1612
			memcpy(to,
E
Eric Dumazet 已提交
1613 1614 1615
			       vaddr + f->page_offset + offset - start,
			       copy);
			kunmap_atomic(vaddr);
L
Linus Torvalds 已提交
1616 1617 1618 1619 1620 1621

			if ((len -= copy) == 0)
				return 0;
			offset += copy;
			to     += copy;
		}
1622
		start = end;
L
Linus Torvalds 已提交
1623 1624
	}

1625 1626
	skb_walk_frags(skb, frag_iter) {
		int end;
L
Linus Torvalds 已提交
1627

1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639
		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
1640
		}
1641
		start = end;
L
Linus Torvalds 已提交
1642
	}
1643

L
Linus Torvalds 已提交
1644 1645 1646 1647 1648 1649
	if (!len)
		return 0;

fault:
	return -EFAULT;
}
1650
EXPORT_SYMBOL(skb_copy_bits);
L
Linus Torvalds 已提交
1651

J
Jens Axboe 已提交
1652 1653 1654 1655 1656 1657
/*
 * 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)
{
1658 1659
	put_page(spd->pages[i]);
}
J
Jens Axboe 已提交
1660

1661 1662
static struct page *linear_to_page(struct page *page, unsigned int *len,
				   unsigned int *offset,
1663
				   struct sock *sk)
1664
{
1665
	struct page_frag *pfrag = sk_page_frag(sk);
1666

1667 1668
	if (!sk_page_frag_refill(sk, pfrag))
		return NULL;
1669

1670
	*len = min_t(unsigned int, *len, pfrag->size - pfrag->offset);
1671

1672 1673 1674 1675
	memcpy(page_address(pfrag->page) + pfrag->offset,
	       page_address(page) + *offset, *len);
	*offset = pfrag->offset;
	pfrag->offset += *len;
1676

1677
	return pfrag->page;
J
Jens Axboe 已提交
1678 1679
}

E
Eric Dumazet 已提交
1680 1681 1682 1683 1684 1685 1686 1687 1688 1689
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
1690 1691 1692
/*
 * Fill page/offset/length into spd, if it can hold more pages.
 */
1693 1694 1695
static bool spd_fill_page(struct splice_pipe_desc *spd,
			  struct pipe_inode_info *pipe, struct page *page,
			  unsigned int *len, unsigned int offset,
1696
			  bool linear,
1697
			  struct sock *sk)
J
Jens Axboe 已提交
1698
{
E
Eric Dumazet 已提交
1699
	if (unlikely(spd->nr_pages == MAX_SKB_FRAGS))
1700
		return true;
J
Jens Axboe 已提交
1701

1702
	if (linear) {
1703
		page = linear_to_page(page, len, &offset, sk);
1704
		if (!page)
1705
			return true;
E
Eric Dumazet 已提交
1706 1707 1708
	}
	if (spd_can_coalesce(spd, page, offset)) {
		spd->partial[spd->nr_pages - 1].len += *len;
1709
		return false;
E
Eric Dumazet 已提交
1710 1711
	}
	get_page(page);
J
Jens Axboe 已提交
1712
	spd->pages[spd->nr_pages] = page;
1713
	spd->partial[spd->nr_pages].len = *len;
J
Jens Axboe 已提交
1714 1715
	spd->partial[spd->nr_pages].offset = offset;
	spd->nr_pages++;
1716

1717
	return false;
J
Jens Axboe 已提交
1718 1719
}

1720 1721
static bool __splice_segment(struct page *page, unsigned int poff,
			     unsigned int plen, unsigned int *off,
1722
			     unsigned int *len,
1723
			     struct splice_pipe_desc *spd, bool linear,
1724 1725
			     struct sock *sk,
			     struct pipe_inode_info *pipe)
J
Jens Axboe 已提交
1726
{
1727
	if (!*len)
1728
		return true;
1729 1730 1731 1732

	/* skip this segment if already processed */
	if (*off >= plen) {
		*off -= plen;
1733
		return false;
1734
	}
J
Jens Axboe 已提交
1735

1736
	/* ignore any bits we already processed */
1737 1738 1739
	poff += *off;
	plen -= *off;
	*off = 0;
J
Jens Axboe 已提交
1740

1741 1742
	do {
		unsigned int flen = min(*len, plen);
1743

1744 1745 1746 1747 1748 1749 1750
		if (spd_fill_page(spd, pipe, page, &flen, poff,
				  linear, sk))
			return true;
		poff += flen;
		plen -= flen;
		*len -= flen;
	} while (*len && plen);
1751

1752
	return false;
1753 1754 1755
}

/*
1756
 * Map linear and fragment data from the skb to spd. It reports true if the
1757 1758
 * pipe is full or if we already spliced the requested length.
 */
1759 1760 1761
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)
1762 1763 1764
{
	int seg;

1765
	/* map the linear part :
1766 1767 1768
	 * 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.
1769 1770 1771 1772
	 */
	if (__splice_segment(virt_to_page(skb->data),
			     (unsigned long) skb->data & (PAGE_SIZE - 1),
			     skb_headlen(skb),
1773
			     offset, len, spd,
1774
			     skb_head_is_locked(skb),
1775
			     sk, pipe))
1776
		return true;
J
Jens Axboe 已提交
1777 1778 1779 1780 1781 1782 1783

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

1784
		if (__splice_segment(skb_frag_page(f),
E
Eric Dumazet 已提交
1785
				     f->page_offset, skb_frag_size(f),
1786
				     offset, len, spd, false, sk, pipe))
1787
			return true;
J
Jens Axboe 已提交
1788 1789
	}

1790
	return false;
J
Jens Axboe 已提交
1791 1792 1793 1794 1795 1796 1797 1798
}

/*
 * Map data from the skb to a pipe. Should handle both the linear part,
 * the fragments, and the frag list. It does NOT handle frag lists within
 * the frag list, if such a thing exists. We'd probably need to recurse to
 * handle that cleanly.
 */
1799
int skb_splice_bits(struct sk_buff *skb, unsigned int offset,
J
Jens Axboe 已提交
1800 1801 1802
		    struct pipe_inode_info *pipe, unsigned int tlen,
		    unsigned int flags)
{
E
Eric Dumazet 已提交
1803 1804
	struct partial_page partial[MAX_SKB_FRAGS];
	struct page *pages[MAX_SKB_FRAGS];
J
Jens Axboe 已提交
1805 1806 1807
	struct splice_pipe_desc spd = {
		.pages = pages,
		.partial = partial,
1808
		.nr_pages_max = MAX_SKB_FRAGS,
J
Jens Axboe 已提交
1809
		.flags = flags,
M
Miklos Szeredi 已提交
1810
		.ops = &nosteal_pipe_buf_ops,
J
Jens Axboe 已提交
1811 1812
		.spd_release = sock_spd_release,
	};
1813
	struct sk_buff *frag_iter;
1814
	struct sock *sk = skb->sk;
1815 1816
	int ret = 0;

J
Jens Axboe 已提交
1817 1818 1819 1820
	/*
	 * __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.
	 */
1821
	if (__skb_splice_bits(skb, pipe, &offset, &tlen, &spd, sk))
J
Jens Axboe 已提交
1822 1823 1824 1825 1826 1827 1828
		goto done;
	else if (!tlen)
		goto done;

	/*
	 * now see if we have a frag_list to map
	 */
1829 1830 1831
	skb_walk_frags(skb, frag_iter) {
		if (!tlen)
			break;
1832
		if (__skb_splice_bits(frag_iter, pipe, &offset, &tlen, &spd, sk))
1833
			break;
J
Jens Axboe 已提交
1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846
	}

done:
	if (spd.nr_pages) {
		/*
		 * Drop the socket lock, otherwise we have reverse
		 * locking dependencies between sk_lock and i_mutex
		 * here as compared to sendfile(). We enter here
		 * with the socket lock held, and splice_to_pipe() will
		 * grab the pipe inode lock. For sendfile() emulation,
		 * we call into ->sendpage() with the i_mutex lock held
		 * and networking will grab the socket lock.
		 */
1847
		release_sock(sk);
J
Jens Axboe 已提交
1848
		ret = splice_to_pipe(pipe, &spd);
1849
		lock_sock(sk);
J
Jens Axboe 已提交
1850 1851
	}

1852
	return ret;
J
Jens Axboe 已提交
1853 1854
}

1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866
/**
 *	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.
 */

1867
int skb_store_bits(struct sk_buff *skb, int offset, const void *from, int len)
1868
{
1869
	int start = skb_headlen(skb);
1870 1871
	struct sk_buff *frag_iter;
	int i, copy;
1872 1873 1874 1875

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

1876
	if ((copy = start - offset) > 0) {
1877 1878
		if (copy > len)
			copy = len;
1879
		skb_copy_to_linear_data_offset(skb, offset, from, copy);
1880 1881 1882 1883 1884 1885 1886 1887
		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];
1888 1889
		int end;

1890
		WARN_ON(start > offset + len);
1891

E
Eric Dumazet 已提交
1892
		end = start + skb_frag_size(frag);
1893 1894 1895 1896 1897 1898
		if ((copy = end - offset) > 0) {
			u8 *vaddr;

			if (copy > len)
				copy = len;

E
Eric Dumazet 已提交
1899
			vaddr = kmap_atomic(skb_frag_page(frag));
1900 1901
			memcpy(vaddr + frag->page_offset + offset - start,
			       from, copy);
E
Eric Dumazet 已提交
1902
			kunmap_atomic(vaddr);
1903 1904 1905 1906 1907 1908

			if ((len -= copy) == 0)
				return 0;
			offset += copy;
			from += copy;
		}
1909
		start = end;
1910 1911
	}

1912 1913
	skb_walk_frags(skb, frag_iter) {
		int end;
1914

1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927
		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;
1928
		}
1929
		start = end;
1930 1931 1932 1933 1934 1935 1936 1937 1938
	}
	if (!len)
		return 0;

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

L
Linus Torvalds 已提交
1939
/* Checksum skb data. */
1940 1941
__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
1942
{
1943 1944
	int start = skb_headlen(skb);
	int i, copy = start - offset;
1945
	struct sk_buff *frag_iter;
L
Linus Torvalds 已提交
1946 1947 1948 1949 1950 1951
	int pos = 0;

	/* Checksum header. */
	if (copy > 0) {
		if (copy > len)
			copy = len;
1952
		csum = ops->update(skb->data + offset, copy, csum);
L
Linus Torvalds 已提交
1953 1954 1955 1956 1957 1958 1959
		if ((len -= copy) == 0)
			return csum;
		offset += copy;
		pos	= copy;
	}

	for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
1960
		int end;
E
Eric Dumazet 已提交
1961
		skb_frag_t *frag = &skb_shinfo(skb)->frags[i];
1962

1963
		WARN_ON(start > offset + len);
L
Linus Torvalds 已提交
1964

E
Eric Dumazet 已提交
1965
		end = start + skb_frag_size(frag);
L
Linus Torvalds 已提交
1966
		if ((copy = end - offset) > 0) {
1967
			__wsum csum2;
L
Linus Torvalds 已提交
1968 1969 1970 1971
			u8 *vaddr;

			if (copy > len)
				copy = len;
E
Eric Dumazet 已提交
1972
			vaddr = kmap_atomic(skb_frag_page(frag));
1973 1974
			csum2 = ops->update(vaddr + frag->page_offset +
					    offset - start, copy, 0);
E
Eric Dumazet 已提交
1975
			kunmap_atomic(vaddr);
1976
			csum = ops->combine(csum, csum2, pos, copy);
L
Linus Torvalds 已提交
1977 1978 1979 1980 1981
			if (!(len -= copy))
				return csum;
			offset += copy;
			pos    += copy;
		}
1982
		start = end;
L
Linus Torvalds 已提交
1983 1984
	}

1985 1986
	skb_walk_frags(skb, frag_iter) {
		int end;
L
Linus Torvalds 已提交
1987

1988 1989 1990 1991 1992 1993 1994
		WARN_ON(start > offset + len);

		end = start + frag_iter->len;
		if ((copy = end - offset) > 0) {
			__wsum csum2;
			if (copy > len)
				copy = len;
1995 1996 1997
			csum2 = __skb_checksum(frag_iter, offset - start,
					       copy, 0, ops);
			csum = ops->combine(csum, csum2, pos, copy);
1998 1999 2000 2001
			if ((len -= copy) == 0)
				return csum;
			offset += copy;
			pos    += copy;
L
Linus Torvalds 已提交
2002
		}
2003
		start = end;
L
Linus Torvalds 已提交
2004
	}
2005
	BUG_ON(len);
L
Linus Torvalds 已提交
2006 2007 2008

	return csum;
}
2009 2010 2011 2012 2013 2014
EXPORT_SYMBOL(__skb_checksum);

__wsum skb_checksum(const struct sk_buff *skb, int offset,
		    int len, __wsum csum)
{
	const struct skb_checksum_ops ops = {
2015
		.update  = csum_partial_ext,
2016 2017 2018 2019 2020
		.combine = csum_block_add_ext,
	};

	return __skb_checksum(skb, offset, len, csum, &ops);
}
2021
EXPORT_SYMBOL(skb_checksum);
L
Linus Torvalds 已提交
2022 2023 2024

/* Both of above in one bottle. */

2025 2026
__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
2027
{
2028 2029
	int start = skb_headlen(skb);
	int i, copy = start - offset;
2030
	struct sk_buff *frag_iter;
L
Linus Torvalds 已提交
2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046
	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++) {
2047 2048
		int end;

2049
		WARN_ON(start > offset + len);
L
Linus Torvalds 已提交
2050

E
Eric Dumazet 已提交
2051
		end = start + skb_frag_size(&skb_shinfo(skb)->frags[i]);
L
Linus Torvalds 已提交
2052
		if ((copy = end - offset) > 0) {
2053
			__wsum csum2;
L
Linus Torvalds 已提交
2054 2055 2056 2057 2058
			u8 *vaddr;
			skb_frag_t *frag = &skb_shinfo(skb)->frags[i];

			if (copy > len)
				copy = len;
E
Eric Dumazet 已提交
2059
			vaddr = kmap_atomic(skb_frag_page(frag));
L
Linus Torvalds 已提交
2060
			csum2 = csum_partial_copy_nocheck(vaddr +
2061 2062 2063
							  frag->page_offset +
							  offset - start, to,
							  copy, 0);
E
Eric Dumazet 已提交
2064
			kunmap_atomic(vaddr);
L
Linus Torvalds 已提交
2065 2066 2067 2068 2069 2070 2071
			csum = csum_block_add(csum, csum2, pos);
			if (!(len -= copy))
				return csum;
			offset += copy;
			to     += copy;
			pos    += copy;
		}
2072
		start = end;
L
Linus Torvalds 已提交
2073 2074
	}

2075 2076 2077
	skb_walk_frags(skb, frag_iter) {
		__wsum csum2;
		int end;
L
Linus Torvalds 已提交
2078

2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093
		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
2094
		}
2095
		start = end;
L
Linus Torvalds 已提交
2096
	}
2097
	BUG_ON(len);
L
Linus Torvalds 已提交
2098 2099
	return csum;
}
2100
EXPORT_SYMBOL(skb_copy_and_csum_bits);
L
Linus Torvalds 已提交
2101

2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128
 /**
 *	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
2129
 *	@from: source buffer
2130 2131 2132 2133 2134 2135 2136 2137
 *	@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.
2138 2139 2140 2141 2142
 *
 *	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
2143
 */
2144 2145
int
skb_zerocopy(struct sk_buff *to, struct sk_buff *from, int len, int hlen)
2146 2147 2148
{
	int i, j = 0;
	int plen = 0; /* length of skb->head fragment */
2149
	int ret;
2150 2151 2152 2153 2154 2155
	struct page *page;
	unsigned int offset;

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

	/* dont bother with small payloads */
2156 2157
	if (len <= skb_tailroom(to))
		return skb_copy_bits(from, 0, skb_put(to, len), len);
2158 2159

	if (hlen) {
2160 2161 2162
		ret = skb_copy_bits(from, 0, skb_put(to, hlen), hlen);
		if (unlikely(ret))
			return ret;
2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179
		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;

2180 2181 2182 2183 2184
	if (unlikely(skb_orphan_frags(from, GFP_ATOMIC))) {
		skb_tx_error(from);
		return -ENOMEM;
	}

2185 2186 2187 2188 2189 2190 2191 2192 2193 2194
	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;
2195 2196

	return 0;
2197 2198 2199
}
EXPORT_SYMBOL_GPL(skb_zerocopy);

L
Linus Torvalds 已提交
2200 2201
void skb_copy_and_csum_dev(const struct sk_buff *skb, u8 *to)
{
2202
	__wsum csum;
L
Linus Torvalds 已提交
2203 2204
	long csstart;

2205
	if (skb->ip_summed == CHECKSUM_PARTIAL)
2206
		csstart = skb_checksum_start_offset(skb);
L
Linus Torvalds 已提交
2207 2208 2209
	else
		csstart = skb_headlen(skb);

2210
	BUG_ON(csstart > skb_headlen(skb));
L
Linus Torvalds 已提交
2211

2212
	skb_copy_from_linear_data(skb, to, csstart);
L
Linus Torvalds 已提交
2213 2214 2215 2216 2217 2218

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

2219
	if (skb->ip_summed == CHECKSUM_PARTIAL) {
A
Al Viro 已提交
2220
		long csstuff = csstart + skb->csum_offset;
L
Linus Torvalds 已提交
2221

2222
		*((__sum16 *)(to + csstuff)) = csum_fold(csum);
L
Linus Torvalds 已提交
2223 2224
	}
}
2225
EXPORT_SYMBOL(skb_copy_and_csum_dev);
L
Linus Torvalds 已提交
2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245

/**
 *	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;
}
2246
EXPORT_SYMBOL(skb_dequeue);
L
Linus Torvalds 已提交
2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265

/**
 *	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;
}
2266
EXPORT_SYMBOL(skb_dequeue_tail);
L
Linus Torvalds 已提交
2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281

/**
 *	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);
}
2282
EXPORT_SYMBOL(skb_queue_purge);
L
Linus Torvalds 已提交
2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302

/**
 *	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);
}
2303
EXPORT_SYMBOL(skb_queue_head);
L
Linus Torvalds 已提交
2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323

/**
 *	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);
}
2324
EXPORT_SYMBOL(skb_queue_tail);
D
David S. Miller 已提交
2325

L
Linus Torvalds 已提交
2326 2327 2328
/**
 *	skb_unlink	-	remove a buffer from a list
 *	@skb: buffer to remove
D
David S. Miller 已提交
2329
 *	@list: list to use
L
Linus Torvalds 已提交
2330
 *
D
David S. Miller 已提交
2331 2332
 *	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
2333
 *
D
David S. Miller 已提交
2334
 *	You must know what list the SKB is on.
L
Linus Torvalds 已提交
2335
 */
D
David S. Miller 已提交
2336
void skb_unlink(struct sk_buff *skb, struct sk_buff_head *list)
L
Linus Torvalds 已提交
2337
{
D
David S. Miller 已提交
2338
	unsigned long flags;
L
Linus Torvalds 已提交
2339

D
David S. Miller 已提交
2340 2341 2342
	spin_lock_irqsave(&list->lock, flags);
	__skb_unlink(skb, list);
	spin_unlock_irqrestore(&list->lock, flags);
L
Linus Torvalds 已提交
2343
}
2344
EXPORT_SYMBOL(skb_unlink);
L
Linus Torvalds 已提交
2345 2346 2347 2348 2349

/**
 *	skb_append	-	append a buffer
 *	@old: buffer to insert after
 *	@newsk: buffer to insert
D
David S. Miller 已提交
2350
 *	@list: list to use
L
Linus Torvalds 已提交
2351 2352 2353 2354 2355
 *
 *	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
2356
void skb_append(struct sk_buff *old, struct sk_buff *newsk, struct sk_buff_head *list)
L
Linus Torvalds 已提交
2357 2358 2359
{
	unsigned long flags;

D
David S. Miller 已提交
2360
	spin_lock_irqsave(&list->lock, flags);
2361
	__skb_queue_after(list, old, newsk);
D
David S. Miller 已提交
2362
	spin_unlock_irqrestore(&list->lock, flags);
L
Linus Torvalds 已提交
2363
}
2364
EXPORT_SYMBOL(skb_append);
L
Linus Torvalds 已提交
2365 2366 2367 2368 2369

/**
 *	skb_insert	-	insert a buffer
 *	@old: buffer to insert before
 *	@newsk: buffer to insert
D
David S. Miller 已提交
2370 2371 2372 2373 2374
 *	@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
2375 2376 2377
 *
 *	A buffer cannot be placed on two lists at the same time.
 */
D
David S. Miller 已提交
2378
void skb_insert(struct sk_buff *old, struct sk_buff *newsk, struct sk_buff_head *list)
L
Linus Torvalds 已提交
2379 2380 2381
{
	unsigned long flags;

D
David S. Miller 已提交
2382 2383 2384
	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
2385
}
2386
EXPORT_SYMBOL(skb_insert);
L
Linus Torvalds 已提交
2387 2388 2389 2390 2391 2392 2393

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

2394 2395
	skb_copy_from_linear_data_offset(skb, len, skb_put(skb1, pos - len),
					 pos - len);
L
Linus Torvalds 已提交
2396 2397 2398 2399 2400 2401 2402 2403 2404 2405
	/* 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;
2406
	skb_set_tail_pointer(skb, len);
L
Linus Torvalds 已提交
2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421
}

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
2422
		int size = skb_frag_size(&skb_shinfo(skb)->frags[i]);
L
Linus Torvalds 已提交
2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435

		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.
				 */
2436
				skb_frag_ref(skb, i);
L
Linus Torvalds 已提交
2437
				skb_shinfo(skb1)->frags[0].page_offset += len - pos;
E
Eric Dumazet 已提交
2438 2439
				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
2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459
				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);

2460
	skb_shinfo(skb1)->tx_flags = skb_shinfo(skb)->tx_flags & SKBTX_SHARED_FRAG;
L
Linus Torvalds 已提交
2461 2462 2463 2464 2465
	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);
}
2466
EXPORT_SYMBOL(skb_split);
L
Linus Torvalds 已提交
2467

2468 2469 2470 2471
/* Shifting from/to a cloned skb is a no-go.
 *
 * Caller cannot keep skb_shinfo related pointers past calling here!
 */
2472 2473
static int skb_prepare_for_shift(struct sk_buff *skb)
{
2474
	return skb_cloned(skb) && pskb_expand_head(skb, 0, 0, GFP_ATOMIC);
2475 2476 2477 2478 2479 2480 2481 2482 2483
}

/**
 * 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
2484
 * the length of the skb, from skb to tgt. Returns number bytes shifted.
2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511
 * 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);
	BUG_ON(skb_headlen(skb));	/* Would corrupt stream */

	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 ||
2512 2513
	    !skb_can_coalesce(tgt, to, skb_frag_page(fragfrom),
			      fragfrom->page_offset)) {
2514 2515 2516 2517
		merge = -1;
	} else {
		merge = to - 1;

E
Eric Dumazet 已提交
2518
		todo -= skb_frag_size(fragfrom);
2519 2520 2521 2522 2523
		if (todo < 0) {
			if (skb_prepare_for_shift(skb) ||
			    skb_prepare_for_shift(tgt))
				return 0;

2524 2525
			/* All previous frag pointers might be stale! */
			fragfrom = &skb_shinfo(skb)->frags[from];
2526 2527
			fragto = &skb_shinfo(tgt)->frags[merge];

E
Eric Dumazet 已提交
2528 2529
			skb_frag_size_add(fragto, shiftlen);
			skb_frag_size_sub(fragfrom, shiftlen);
2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552
			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
2553
		if (todo >= skb_frag_size(fragfrom)) {
2554
			*fragto = *fragfrom;
E
Eric Dumazet 已提交
2555
			todo -= skb_frag_size(fragfrom);
2556 2557 2558 2559
			from++;
			to++;

		} else {
2560
			__skb_frag_ref(fragfrom);
2561 2562
			fragto->page = fragfrom->page;
			fragto->page_offset = fragfrom->page_offset;
E
Eric Dumazet 已提交
2563
			skb_frag_size_set(fragto, todo);
2564 2565

			fragfrom->page_offset += todo;
E
Eric Dumazet 已提交
2566
			skb_frag_size_sub(fragfrom, todo);
2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580
			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
2581
		skb_frag_size_add(fragto, skb_frag_size(fragfrom));
2582
		__skb_frag_unref(fragfrom);
2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610
	}

	/* 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;
}

2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629
/**
 * 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;
}
2630
EXPORT_SYMBOL(skb_prepare_seq_read);
2631 2632 2633 2634 2635 2636 2637

/**
 * 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
2638
 * Reads a block of skb data at @consumed relative to the
2639
 * lower offset specified to skb_prepare_seq_read(). Assigns
M
Mathias Krause 已提交
2640
 * the head of the data block to @data and returns the length
2641 2642 2643 2644
 * 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
2645
 * returned, i.e. @consumed is typically set to the number
2646 2647 2648
 * 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
2649
 * Note 1: The size of each block of data returned can be arbitrary,
2650 2651 2652
 *       this limitation is the cost for zerocopy seqeuental
 *       reads of potentially non linear data.
 *
2653
 * Note 2: Fragment lists within fragments are not implemented
2654 2655 2656 2657 2658 2659 2660 2661 2662
 *       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;

2663 2664 2665 2666 2667
	if (unlikely(abs_offset >= st->upper_offset)) {
		if (st->frag_data) {
			kunmap_atomic(st->frag_data);
			st->frag_data = NULL;
		}
2668
		return 0;
2669
	}
2670 2671

next_skb:
2672
	block_limit = skb_headlen(st->cur_skb) + st->stepped_offset;
2673

2674
	if (abs_offset < block_limit && !st->frag_data) {
2675
		*data = st->cur_skb->data + (abs_offset - st->stepped_offset);
2676 2677 2678 2679 2680 2681 2682 2683
		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
2684
		block_limit = skb_frag_size(frag) + st->stepped_offset;
2685 2686 2687

		if (abs_offset < block_limit) {
			if (!st->frag_data)
E
Eric Dumazet 已提交
2688
				st->frag_data = kmap_atomic(skb_frag_page(frag));
2689 2690 2691 2692 2693 2694 2695 2696

			*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
2697
			kunmap_atomic(st->frag_data);
2698 2699 2700 2701
			st->frag_data = NULL;
		}

		st->frag_idx++;
E
Eric Dumazet 已提交
2702
		st->stepped_offset += skb_frag_size(frag);
2703 2704
	}

2705
	if (st->frag_data) {
E
Eric Dumazet 已提交
2706
		kunmap_atomic(st->frag_data);
2707 2708 2709
		st->frag_data = NULL;
	}

2710
	if (st->root_skb == st->cur_skb && skb_has_frag_list(st->root_skb)) {
S
Shyam Iyer 已提交
2711
		st->cur_skb = skb_shinfo(st->root_skb)->frag_list;
2712 2713
		st->frag_idx = 0;
		goto next_skb;
S
Shyam Iyer 已提交
2714 2715
	} else if (st->cur_skb->next) {
		st->cur_skb = st->cur_skb->next;
2716
		st->frag_idx = 0;
2717 2718 2719 2720 2721
		goto next_skb;
	}

	return 0;
}
2722
EXPORT_SYMBOL(skb_seq_read);
2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733

/**
 * 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
2734
		kunmap_atomic(st->frag_data);
2735
}
2736
EXPORT_SYMBOL(skb_abort_seq_read);
2737

2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768
#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
 * @state: uninitialized textsearch state variable
 *
 * 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,
			   unsigned int to, struct ts_config *config,
			   struct ts_state *state)
{
2769 2770
	unsigned int ret;

2771 2772 2773 2774 2775
	config->get_next_block = skb_ts_get_next_block;
	config->finish = skb_ts_finish;

	skb_prepare_seq_read(skb, from, to, TS_SKB_CB(state));

2776 2777
	ret = textsearch_find(config, state);
	return (ret <= to - from ? ret : UINT_MAX);
2778
}
2779
EXPORT_SYMBOL(skb_find_text);
2780

2781
/**
2782
 * skb_append_datato_frags - append the user data to a skb
2783 2784 2785 2786 2787 2788 2789 2790 2791 2792
 * @sk: sock  structure
 * @skb: skb structure to be appened with user data.
 * @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,
2793
			int (*getfrag)(void *from, char *to, int offset,
2794 2795 2796
					int len, int odd, struct sk_buff *skb),
			void *from, int length)
{
2797 2798
	int frg_cnt = skb_shinfo(skb)->nr_frags;
	int copy;
2799 2800
	int offset = 0;
	int ret;
2801
	struct page_frag *pfrag = &current->task_frag;
2802 2803 2804 2805

	do {
		/* Return error if we don't have space for new frag */
		if (frg_cnt >= MAX_SKB_FRAGS)
2806
			return -EMSGSIZE;
2807

2808
		if (!sk_page_frag_refill(sk, pfrag))
2809 2810 2811
			return -ENOMEM;

		/* copy the user data to page */
2812
		copy = min_t(int, length, pfrag->size - pfrag->offset);
2813

2814 2815
		ret = getfrag(from, page_address(pfrag->page) + pfrag->offset,
			      offset, copy, 0, skb);
2816 2817 2818 2819
		if (ret < 0)
			return -EFAULT;

		/* copy was successful so update the size parameters */
2820 2821 2822 2823 2824 2825 2826 2827
		skb_fill_page_desc(skb, frg_cnt, pfrag->page, pfrag->offset,
				   copy);
		frg_cnt++;
		pfrag->offset += copy;
		get_page(pfrag->page);

		skb->truesize += copy;
		atomic_add(copy, &sk->sk_wmem_alloc);
2828 2829 2830 2831 2832 2833 2834 2835 2836
		skb->len += copy;
		skb->data_len += copy;
		offset += copy;
		length -= copy;

	} while (length > 0);

	return 0;
}
2837
EXPORT_SYMBOL(skb_append_datato_frags);
2838

2839 2840 2841 2842 2843 2844
/**
 *	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
2845
 *	the CHECKSUM_COMPLETE checksum.  It should be used on
2846 2847 2848
 *	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.
2849 2850 2851 2852 2853 2854 2855 2856 2857
 */
unsigned char *skb_pull_rcsum(struct sk_buff *skb, unsigned int len)
{
	BUG_ON(len > skb->len);
	skb->len -= len;
	BUG_ON(skb->len < skb->data_len);
	skb_postpull_rcsum(skb, skb->data, len);
	return skb->data += len;
}
2858 2859
EXPORT_SYMBOL_GPL(skb_pull_rcsum);

H
Herbert Xu 已提交
2860 2861
/**
 *	skb_segment - Perform protocol segmentation on skb.
2862
 *	@head_skb: buffer to segment
2863
 *	@features: features for the output path (see dev->features)
H
Herbert Xu 已提交
2864 2865
 *
 *	This function performs segmentation on the given skb.  It returns
2866 2867
 *	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
2868
 */
2869 2870
struct sk_buff *skb_segment(struct sk_buff *head_skb,
			    netdev_features_t features)
H
Herbert Xu 已提交
2871 2872 2873
{
	struct sk_buff *segs = NULL;
	struct sk_buff *tail = NULL;
2874
	struct sk_buff *list_skb = skb_shinfo(head_skb)->frag_list;
2875 2876 2877
	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);
2878
	struct sk_buff *frag_skb = head_skb;
H
Herbert Xu 已提交
2879
	unsigned int offset = doffset;
2880
	unsigned int tnl_hlen = skb_tnl_header_len(head_skb);
H
Herbert Xu 已提交
2881 2882
	unsigned int headroom;
	unsigned int len;
2883 2884
	__be16 proto;
	bool csum;
2885
	int sg = !!(features & NETIF_F_SG);
2886
	int nfrags = skb_shinfo(head_skb)->nr_frags;
H
Herbert Xu 已提交
2887 2888 2889
	int err = -ENOMEM;
	int i = 0;
	int pos;
2890
	int dummy;
H
Herbert Xu 已提交
2891

2892
	__skb_push(head_skb, doffset);
2893
	proto = skb_network_protocol(head_skb, &dummy);
2894 2895 2896
	if (unlikely(!proto))
		return ERR_PTR(-EINVAL);

2897 2898 2899
	csum = !head_skb->encap_hdr_csum &&
	    !!can_checksum_protocol(features, proto);

2900 2901
	headroom = skb_headroom(head_skb);
	pos = skb_headlen(head_skb);
H
Herbert Xu 已提交
2902 2903 2904

	do {
		struct sk_buff *nskb;
2905
		skb_frag_t *nskb_frag;
2906
		int hsize;
H
Herbert Xu 已提交
2907 2908
		int size;

2909
		len = head_skb->len - offset;
H
Herbert Xu 已提交
2910 2911 2912
		if (len > mss)
			len = mss;

2913
		hsize = skb_headlen(head_skb) - offset;
H
Herbert Xu 已提交
2914 2915
		if (hsize < 0)
			hsize = 0;
2916 2917
		if (hsize > len || !sg)
			hsize = len;
H
Herbert Xu 已提交
2918

2919 2920 2921
		if (!hsize && i >= nfrags && skb_headlen(list_skb) &&
		    (skb_headlen(list_skb) == len || sg)) {
			BUG_ON(skb_headlen(list_skb) > len);
2922 2923

			i = 0;
2924 2925
			nfrags = skb_shinfo(list_skb)->nr_frags;
			frag = skb_shinfo(list_skb)->frags;
2926
			frag_skb = list_skb;
2927
			pos += skb_headlen(list_skb);
2928 2929 2930 2931

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

2932
				size = skb_frag_size(frag);
2933 2934 2935 2936 2937
				if (pos + size > offset + len)
					break;

				i++;
				pos += size;
2938
				frag++;
2939
			}
2940

2941 2942
			nskb = skb_clone(list_skb, GFP_ATOMIC);
			list_skb = list_skb->next;
2943 2944 2945 2946

			if (unlikely(!nskb))
				goto err;

2947 2948 2949 2950 2951
			if (unlikely(pskb_trim(nskb, len))) {
				kfree_skb(nskb);
				goto err;
			}

2952
			hsize = skb_end_offset(nskb);
2953 2954 2955 2956 2957
			if (skb_cow_head(nskb, doffset + headroom)) {
				kfree_skb(nskb);
				goto err;
			}

2958
			nskb->truesize += skb_end_offset(nskb) - hsize;
2959 2960 2961
			skb_release_head_state(nskb);
			__skb_push(nskb, doffset);
		} else {
2962
			nskb = __alloc_skb(hsize + doffset + headroom,
2963
					   GFP_ATOMIC, skb_alloc_rx_flag(head_skb),
2964
					   NUMA_NO_NODE);
2965 2966 2967 2968 2969 2970 2971

			if (unlikely(!nskb))
				goto err;

			skb_reserve(nskb, headroom);
			__skb_put(nskb, doffset);
		}
H
Herbert Xu 已提交
2972 2973 2974 2975 2976 2977 2978

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

2979
		__copy_skb_header(nskb, head_skb);
H
Herbert Xu 已提交
2980

E
Eric Dumazet 已提交
2981
		skb_headers_offset_update(nskb, skb_headroom(nskb) - headroom);
2982
		skb_reset_mac_len(nskb);
2983

2984
		skb_copy_from_linear_data_offset(head_skb, -tnl_hlen,
2985 2986
						 nskb->data - tnl_hlen,
						 doffset + tnl_hlen);
2987

2988
		if (nskb->len == len + doffset)
2989
			goto perform_csum_check;
2990

H
Herbert Xu 已提交
2991
		if (!sg) {
2992
			nskb->ip_summed = CHECKSUM_NONE;
2993
			nskb->csum = skb_copy_and_csum_bits(head_skb, offset,
H
Herbert Xu 已提交
2994 2995
							    skb_put(nskb, len),
							    len, 0);
2996
			SKB_GSO_CB(nskb)->csum_start =
2997
			    skb_headroom(nskb) + doffset;
H
Herbert Xu 已提交
2998 2999 3000
			continue;
		}

3001
		nskb_frag = skb_shinfo(nskb)->frags;
H
Herbert Xu 已提交
3002

3003
		skb_copy_from_linear_data_offset(head_skb, offset,
3004
						 skb_put(nskb, hsize), hsize);
H
Herbert Xu 已提交
3005

3006 3007
		skb_shinfo(nskb)->tx_flags = skb_shinfo(head_skb)->tx_flags &
			SKBTX_SHARED_FRAG;
3008

3009 3010
		while (pos < offset + len) {
			if (i >= nfrags) {
3011
				BUG_ON(skb_headlen(list_skb));
3012 3013

				i = 0;
3014 3015
				nfrags = skb_shinfo(list_skb)->nr_frags;
				frag = skb_shinfo(list_skb)->frags;
3016
				frag_skb = list_skb;
3017 3018 3019

				BUG_ON(!nfrags);

3020
				list_skb = list_skb->next;
3021 3022 3023 3024 3025 3026 3027 3028 3029 3030
			}

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

3031 3032 3033
			if (unlikely(skb_orphan_frags(frag_skb, GFP_ATOMIC)))
				goto err;

3034
			*nskb_frag = *frag;
3035 3036
			__skb_frag_ref(nskb_frag);
			size = skb_frag_size(nskb_frag);
H
Herbert Xu 已提交
3037 3038

			if (pos < offset) {
3039 3040
				nskb_frag->page_offset += offset - pos;
				skb_frag_size_sub(nskb_frag, offset - pos);
H
Herbert Xu 已提交
3041 3042
			}

3043
			skb_shinfo(nskb)->nr_frags++;
H
Herbert Xu 已提交
3044 3045 3046

			if (pos + size <= offset + len) {
				i++;
3047
				frag++;
H
Herbert Xu 已提交
3048 3049
				pos += size;
			} else {
3050
				skb_frag_size_sub(nskb_frag, pos + size - (offset + len));
3051
				goto skip_fraglist;
H
Herbert Xu 已提交
3052 3053
			}

3054
			nskb_frag++;
H
Herbert Xu 已提交
3055 3056
		}

3057
skip_fraglist:
H
Herbert Xu 已提交
3058 3059 3060
		nskb->data_len = len - hsize;
		nskb->len += nskb->data_len;
		nskb->truesize += nskb->data_len;
3061

3062
perform_csum_check:
3063 3064 3065 3066
		if (!csum) {
			nskb->csum = skb_checksum(nskb, doffset,
						  nskb->len - doffset, 0);
			nskb->ip_summed = CHECKSUM_NONE;
3067 3068
			SKB_GSO_CB(nskb)->csum_start =
			    skb_headroom(nskb) + doffset;
3069
		}
3070
	} while ((offset += len) < head_skb->len);
H
Herbert Xu 已提交
3071 3072 3073 3074

	return segs;

err:
E
Eric Dumazet 已提交
3075
	kfree_skb_list(segs);
H
Herbert Xu 已提交
3076 3077 3078 3079
	return ERR_PTR(err);
}
EXPORT_SYMBOL_GPL(skb_segment);

H
Herbert Xu 已提交
3080 3081
int skb_gro_receive(struct sk_buff **head, struct sk_buff *skb)
{
3082
	struct skb_shared_info *pinfo, *skbinfo = skb_shinfo(skb);
3083 3084
	unsigned int offset = skb_gro_offset(skb);
	unsigned int headlen = skb_headlen(skb);
3085 3086
	struct sk_buff *nskb, *lp, *p = *head;
	unsigned int len = skb_gro_len(skb);
3087
	unsigned int delta_truesize;
3088
	unsigned int headroom;
H
Herbert Xu 已提交
3089

3090
	if (unlikely(p->len + len >= 65536))
H
Herbert Xu 已提交
3091 3092
		return -E2BIG;

3093
	lp = NAPI_GRO_CB(p)->last;
3094 3095 3096
	pinfo = skb_shinfo(lp);

	if (headlen <= offset) {
3097
		skb_frag_t *frag;
3098
		skb_frag_t *frag2;
3099 3100
		int i = skbinfo->nr_frags;
		int nr_frags = pinfo->nr_frags + i;
3101 3102

		if (nr_frags > MAX_SKB_FRAGS)
3103
			goto merge;
3104

3105
		offset -= headlen;
3106 3107
		pinfo->nr_frags = nr_frags;
		skbinfo->nr_frags = 0;
3108

3109 3110
		frag = pinfo->frags + nr_frags;
		frag2 = skbinfo->frags + i;
3111 3112 3113
		do {
			*--frag = *--frag2;
		} while (--i);
H
Herbert Xu 已提交
3114

3115
		frag->page_offset += offset;
E
Eric Dumazet 已提交
3116
		skb_frag_size_sub(frag, offset);
3117

3118
		/* all fragments truesize : remove (head size + sk_buff) */
3119 3120
		delta_truesize = skb->truesize -
				 SKB_TRUESIZE(skb_end_offset(skb));
3121

3122 3123 3124 3125
		skb->truesize -= skb->data_len;
		skb->len -= skb->data_len;
		skb->data_len = 0;

3126
		NAPI_GRO_CB(skb)->free = NAPI_GRO_FREE;
H
Herbert Xu 已提交
3127
		goto done;
3128 3129 3130 3131 3132 3133 3134 3135
	} 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)
3136
			goto merge;
3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150

		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 */

3151
		delta_truesize = skb->truesize - SKB_DATA_ALIGN(sizeof(struct sk_buff));
3152 3153
		NAPI_GRO_CB(skb)->free = NAPI_GRO_FREE_STOLEN_HEAD;
		goto done;
3154 3155 3156 3157
	}
	if (pinfo->frag_list)
		goto merge;
	if (skb_gro_len(p) != pinfo->gso_size)
3158
		return -E2BIG;
H
Herbert Xu 已提交
3159 3160

	headroom = skb_headroom(p);
E
Eric Dumazet 已提交
3161
	nskb = alloc_skb(headroom + skb_gro_offset(p), GFP_ATOMIC);
H
Herbert Xu 已提交
3162 3163 3164 3165 3166 3167 3168
	if (unlikely(!nskb))
		return -ENOMEM;

	__copy_skb_header(nskb, p);
	nskb->mac_len = p->mac_len;

	skb_reserve(nskb, headroom);
3169
	__skb_put(nskb, skb_gro_offset(p));
H
Herbert Xu 已提交
3170

3171
	skb_set_mac_header(nskb, skb_mac_header(p) - p->data);
H
Herbert Xu 已提交
3172 3173 3174
	skb_set_network_header(nskb, skb_network_offset(p));
	skb_set_transport_header(nskb, skb_transport_offset(p));

3175 3176 3177
	__skb_pull(p, skb_gro_offset(p));
	memcpy(skb_mac_header(nskb), skb_mac_header(p),
	       p->data - skb_mac_header(p));
H
Herbert Xu 已提交
3178 3179

	skb_shinfo(nskb)->frag_list = p;
3180
	skb_shinfo(nskb)->gso_size = pinfo->gso_size;
3181
	pinfo->gso_size = 0;
H
Herbert Xu 已提交
3182
	skb_header_release(p);
3183
	NAPI_GRO_CB(nskb)->last = p;
H
Herbert Xu 已提交
3184 3185

	nskb->data_len += p->len;
E
Eric Dumazet 已提交
3186
	nskb->truesize += p->truesize;
H
Herbert Xu 已提交
3187 3188 3189 3190 3191 3192 3193 3194 3195
	nskb->len += p->len;

	*head = nskb;
	nskb->next = p->next;
	p->next = NULL;

	p = nskb;

merge:
3196
	delta_truesize = skb->truesize;
3197
	if (offset > headlen) {
3198 3199 3200
		unsigned int eat = offset - headlen;

		skbinfo->frags[0].page_offset += eat;
E
Eric Dumazet 已提交
3201
		skb_frag_size_sub(&skbinfo->frags[0], eat);
3202 3203
		skb->data_len -= eat;
		skb->len -= eat;
3204
		offset = headlen;
3205 3206
	}

3207
	__skb_pull(skb, offset);
3208

3209
	if (NAPI_GRO_CB(p)->last == p)
3210 3211 3212
		skb_shinfo(p)->frag_list = skb;
	else
		NAPI_GRO_CB(p)->last->next = skb;
3213
	NAPI_GRO_CB(p)->last = skb;
H
Herbert Xu 已提交
3214
	skb_header_release(skb);
3215
	lp = p;
H
Herbert Xu 已提交
3216

H
Herbert Xu 已提交
3217 3218
done:
	NAPI_GRO_CB(p)->count++;
H
Herbert Xu 已提交
3219
	p->data_len += len;
3220
	p->truesize += delta_truesize;
H
Herbert Xu 已提交
3221
	p->len += len;
3222 3223 3224 3225 3226
	if (lp != p) {
		lp->data_len += len;
		lp->truesize += delta_truesize;
		lp->len += len;
	}
H
Herbert Xu 已提交
3227 3228 3229 3230 3231
	NAPI_GRO_CB(skb)->same_flow = 1;
	return 0;
}
EXPORT_SYMBOL_GPL(skb_gro_receive);

L
Linus Torvalds 已提交
3232 3233 3234 3235 3236
void __init skb_init(void)
{
	skbuff_head_cache = kmem_cache_create("skbuff_head_cache",
					      sizeof(struct sk_buff),
					      0,
A
Alexey Dobriyan 已提交
3237
					      SLAB_HWCACHE_ALIGN|SLAB_PANIC,
3238
					      NULL);
3239 3240 3241 3242
	skbuff_fclone_cache = kmem_cache_create("skbuff_fclone_cache",
						(2*sizeof(struct sk_buff)) +
						sizeof(atomic_t),
						0,
A
Alexey Dobriyan 已提交
3243
						SLAB_HWCACHE_ALIGN|SLAB_PANIC,
3244
						NULL);
L
Linus Torvalds 已提交
3245 3246
}

3247 3248 3249 3250 3251 3252 3253 3254 3255 3256
/**
 *	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.
 */
3257 3258
static int
__skb_to_sgvec(struct sk_buff *skb, struct scatterlist *sg, int offset, int len)
3259
{
3260 3261
	int start = skb_headlen(skb);
	int i, copy = start - offset;
3262
	struct sk_buff *frag_iter;
3263 3264 3265 3266 3267
	int elt = 0;

	if (copy > 0) {
		if (copy > len)
			copy = len;
3268
		sg_set_buf(sg, skb->data + offset, copy);
3269 3270 3271 3272 3273 3274 3275
		elt++;
		if ((len -= copy) == 0)
			return elt;
		offset += copy;
	}

	for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
3276
		int end;
3277

3278
		WARN_ON(start > offset + len);
3279

E
Eric Dumazet 已提交
3280
		end = start + skb_frag_size(&skb_shinfo(skb)->frags[i]);
3281 3282 3283 3284 3285
		if ((copy = end - offset) > 0) {
			skb_frag_t *frag = &skb_shinfo(skb)->frags[i];

			if (copy > len)
				copy = len;
3286
			sg_set_page(&sg[elt], skb_frag_page(frag), copy,
3287
					frag->page_offset+offset-start);
3288 3289 3290 3291 3292
			elt++;
			if (!(len -= copy))
				return elt;
			offset += copy;
		}
3293
		start = end;
3294 3295
	}

3296 3297
	skb_walk_frags(skb, frag_iter) {
		int end;
3298

3299
		WARN_ON(start > offset + len);
3300

3301 3302 3303 3304 3305 3306 3307 3308 3309
		end = start + frag_iter->len;
		if ((copy = end - offset) > 0) {
			if (copy > len)
				copy = len;
			elt += __skb_to_sgvec(frag_iter, sg+elt, offset - start,
					      copy);
			if ((len -= copy) == 0)
				return elt;
			offset += copy;
3310
		}
3311
		start = end;
3312 3313 3314 3315 3316
	}
	BUG_ON(len);
	return elt;
}

3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 3339 3340 3341 3342
/* 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)
{
	return __skb_to_sgvec(skb, sg, offset, len);
}
EXPORT_SYMBOL_GPL(skb_to_sgvec_nomark);

3343 3344 3345 3346
int skb_to_sgvec(struct sk_buff *skb, struct scatterlist *sg, int offset, int len)
{
	int nsg = __skb_to_sgvec(skb, sg, offset, len);

J
Jens Axboe 已提交
3347
	sg_mark_end(&sg[nsg - 1]);
3348 3349 3350

	return nsg;
}
3351
EXPORT_SYMBOL_GPL(skb_to_sgvec);
3352

3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384
/**
 *	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. */
3385
	if (!skb_has_frag_list(skb)) {
3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419
		/* 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 ||
3420
			    skb_has_frag_list(skb1) ||
3421 3422 3423 3424 3425 3426 3427 3428
			    skb_tailroom(skb1) < tailbits)
				ntail = tailbits + 128;
		}

		if (copyflag ||
		    skb_cloned(skb1) ||
		    ntail ||
		    skb_shinfo(skb1)->nr_frags ||
3429
		    skb_has_frag_list(skb1)) {
3430 3431 3432 3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460
			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;
}
3461
EXPORT_SYMBOL_GPL(skb_cow_data);
3462

3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475
static void sock_rmem_free(struct sk_buff *skb)
{
	struct sock *sk = skb->sk;

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

/*
 * 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 >=
3476
	    (unsigned int)sk->sk_rcvbuf)
3477 3478 3479 3480 3481 3482 3483
		return -ENOMEM;

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

3484 3485 3486
	/* before exiting rcu section, make sure dst is refcounted */
	skb_dst_force(skb);

3487 3488
	skb_queue_tail(&sk->sk_error_queue, skb);
	if (!sock_flag(sk, SOCK_DEAD))
3489
		sk->sk_data_ready(sk);
3490 3491 3492 3493
	return 0;
}
EXPORT_SYMBOL(sock_queue_err_skb);

3494 3495 3496
void __skb_tstamp_tx(struct sk_buff *orig_skb,
		     struct skb_shared_hwtstamps *hwtstamps,
		     struct sock *sk, int tstype)
3497 3498 3499 3500 3501 3502 3503 3504 3505
{
	struct sock_exterr_skb *serr;
	struct sk_buff *skb;
	int err;

	if (!sk)
		return;

	if (hwtstamps) {
3506
		*skb_hwtstamps(orig_skb) =
3507 3508 3509 3510
			*hwtstamps;
	} else {
		/*
		 * no hardware time stamps available,
3511
		 * so keep the shared tx_flags and only
3512 3513
		 * store software time stamp
		 */
3514
		orig_skb->tstamp = ktime_get_real();
3515 3516
	}

3517 3518 3519 3520
	skb = skb_clone(orig_skb, GFP_ATOMIC);
	if (!skb)
		return;

3521 3522 3523 3524
	serr = SKB_EXT_ERR(skb);
	memset(serr, 0, sizeof(*serr));
	serr->ee.ee_errno = ENOMSG;
	serr->ee.ee_origin = SO_EE_ORIGIN_TIMESTAMPING;
3525
	serr->ee.ee_info = tstype;
3526
	if (sk->sk_tsflags & SOF_TIMESTAMPING_OPT_ID) {
3527
		serr->ee.ee_data = skb_shinfo(skb)->tskey;
3528 3529 3530
		if (sk->sk_protocol == IPPROTO_TCP)
			serr->ee.ee_data -= sk->sk_tskey;
	}
3531

3532
	err = sock_queue_err_skb(sk, skb);
3533

3534 3535 3536
	if (err)
		kfree_skb(skb);
}
3537 3538 3539 3540 3541 3542 3543 3544
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);
}
3545 3546
EXPORT_SYMBOL_GPL(skb_tstamp_tx);

3547 3548 3549 3550 3551 3552 3553 3554 3555 3556 3557 3558 3559 3560 3561 3562 3563 3564 3565 3566
void skb_complete_wifi_ack(struct sk_buff *skb, bool acked)
{
	struct sock *sk = skb->sk;
	struct sock_exterr_skb *serr;
	int err;

	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;

	err = sock_queue_err_skb(sk, skb);
	if (err)
		kfree_skb(skb);
}
EXPORT_SYMBOL_GPL(skb_complete_wifi_ack);

3567

3568 3569 3570 3571 3572 3573 3574 3575 3576 3577 3578 3579 3580 3581
/**
 * 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)
{
3582 3583
	if (unlikely(start > skb_headlen(skb)) ||
	    unlikely((int)start + off > skb_headlen(skb) - 2)) {
3584 3585
		net_warn_ratelimited("bad partial csum: csum=%u/%u len=%u\n",
				     start, off, skb_headlen(skb));
3586 3587 3588 3589 3590
		return false;
	}
	skb->ip_summed = CHECKSUM_PARTIAL;
	skb->csum_start = skb_headroom(skb) + start;
	skb->csum_offset = off;
3591
	skb_set_transport_header(skb, start);
3592 3593
	return true;
}
3594
EXPORT_SYMBOL_GPL(skb_partial_csum_set);
3595

P
Paul Durrant 已提交
3596 3597 3598 3599 3600 3601 3602 3603 3604 3605 3606 3607 3608 3609 3610 3611 3612 3613 3614 3615 3616
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;
}

3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637 3638 3639 3640 3641 3642 3643 3644 3645 3646 3647
#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
3648 3649 3650 3651 3652
/* 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

3653
static int skb_checksum_setup_ipv4(struct sk_buff *skb, bool recalculate)
P
Paul Durrant 已提交
3654 3655 3656
{
	unsigned int off;
	bool fragment;
3657
	__sum16 *csum;
P
Paul Durrant 已提交
3658 3659 3660 3661 3662 3663 3664 3665 3666 3667 3668 3669 3670 3671 3672 3673 3674 3675 3676 3677
	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;

3678 3679 3680
	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
3681

3682 3683 3684 3685 3686
	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
3687 3688 3689 3690 3691 3692 3693 3694 3695 3696 3697 3698 3699 3700 3701 3702 3703 3704 3705 3706 3707 3708
	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;
3709
	__sum16 *csum;
P
Paul Durrant 已提交
3710 3711 3712 3713 3714 3715 3716 3717 3718 3719 3720 3721 3722 3723 3724 3725 3726 3727 3728 3729 3730 3731 3732 3733 3734 3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750 3751 3752 3753 3754 3755 3756 3757 3758 3759 3760 3761 3762 3763 3764 3765 3766 3767 3768 3769 3770 3771 3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785 3786

	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;

3787 3788 3789
	csum = skb_checksum_setup_ip(skb, nexthdr, off);
	if (IS_ERR(csum))
		return PTR_ERR(csum);
P
Paul Durrant 已提交
3790

3791 3792 3793 3794
	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
3795 3796 3797 3798 3799 3800 3801 3802 3803 3804 3805 3806 3807 3808 3809 3810 3811
	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):
3812
		err = skb_checksum_setup_ipv4(skb, recalculate);
P
Paul Durrant 已提交
3813 3814 3815 3816 3817 3818 3819 3820 3821 3822 3823 3824 3825 3826 3827
		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);

3828 3829
void __skb_warn_lro_forwarding(const struct sk_buff *skb)
{
3830 3831
	net_warn_ratelimited("%s: received packets cannot be forwarded while LRO is enabled\n",
			     skb->dev->name);
3832 3833
}
EXPORT_SYMBOL(__skb_warn_lro_forwarding);
E
Eric Dumazet 已提交
3834 3835 3836

void kfree_skb_partial(struct sk_buff *skb, bool head_stolen)
{
E
Eric Dumazet 已提交
3837 3838
	if (head_stolen) {
		skb_release_head_state(skb);
E
Eric Dumazet 已提交
3839
		kmem_cache_free(skbuff_head_cache, skb);
E
Eric Dumazet 已提交
3840
	} else {
E
Eric Dumazet 已提交
3841
		__kfree_skb(skb);
E
Eric Dumazet 已提交
3842
	}
E
Eric Dumazet 已提交
3843 3844 3845 3846 3847 3848 3849 3850
}
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
3851
 * @delta_truesize: how much more was allocated than was requested
E
Eric Dumazet 已提交
3852 3853 3854 3855 3856 3857 3858 3859 3860 3861 3862 3863 3864 3865 3866 3867 3868 3869 3870 3871 3872 3873 3874 3875 3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891 3892 3893 3894 3895
 */
bool skb_try_coalesce(struct sk_buff *to, struct sk_buff *from,
		      bool *fragstolen, int *delta_truesize)
{
	int i, delta, len = from->len;

	*fragstolen = false;

	if (skb_cloned(to))
		return false;

	if (len <= skb_tailroom(to)) {
		BUG_ON(skb_copy_bits(from, 0, skb_put(to, len), len));
		*delta_truesize = 0;
		return true;
	}

	if (skb_has_frag_list(to) || skb_has_frag_list(from))
		return false;

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

		if (skb_shinfo(to)->nr_frags +
		    skb_shinfo(from)->nr_frags >= MAX_SKB_FRAGS)
			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);

		skb_fill_page_desc(to, skb_shinfo(to)->nr_frags,
				   page, offset, skb_headlen(from));
		*fragstolen = true;
	} else {
		if (skb_shinfo(to)->nr_frags +
		    skb_shinfo(from)->nr_frags > MAX_SKB_FRAGS)
			return false;

3896
		delta = from->truesize - SKB_TRUESIZE(skb_end_offset(from));
E
Eric Dumazet 已提交
3897 3898 3899 3900 3901 3902 3903 3904 3905 3906 3907 3908
	}

	WARN_ON_ONCE(delta < len);

	memcpy(skb_shinfo(to)->frags + skb_shinfo(to)->nr_frags,
	       skb_shinfo(from)->frags,
	       skb_shinfo(from)->nr_frags * sizeof(skb_frag_t));
	skb_shinfo(to)->nr_frags += skb_shinfo(from)->nr_frags;

	if (!skb_cloned(from))
		skb_shinfo(from)->nr_frags = 0;

3909 3910 3911
	/* if the skb is not cloned this does nothing
	 * since we set nr_frags to 0.
	 */
E
Eric Dumazet 已提交
3912 3913 3914 3915 3916 3917 3918 3919 3920 3921 3922
	for (i = 0; i < skb_shinfo(from)->nr_frags; i++)
		skb_frag_ref(from, i);

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

	*delta_truesize = delta;
	return true;
}
EXPORT_SYMBOL(skb_try_coalesce);
3923 3924

/**
3925
 * skb_scrub_packet - scrub an skb
3926 3927
 *
 * @skb: buffer to clean
3928 3929 3930 3931 3932 3933 3934 3935
 * @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.
3936
 */
3937
void skb_scrub_packet(struct sk_buff *skb, bool xnet)
3938
{
3939 3940
	if (xnet)
		skb_orphan(skb);
3941 3942 3943
	skb->tstamp.tv64 = 0;
	skb->pkt_type = PACKET_HOST;
	skb->skb_iif = 0;
W
WANG Cong 已提交
3944
	skb->ignore_df = 0;
3945 3946 3947 3948 3949 3950 3951
	skb_dst_drop(skb);
	skb->mark = 0;
	secpath_reset(skb);
	nf_reset(skb);
	nf_reset_trace(skb);
}
EXPORT_SYMBOL_GPL(skb_scrub_packet);
3952 3953 3954 3955 3956 3957 3958 3959 3960 3961 3962 3963 3964 3965 3966 3967

/**
 * 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.
 */
unsigned int skb_gso_transport_seglen(const struct sk_buff *skb)
{
	const struct skb_shared_info *shinfo = skb_shinfo(skb);

	if (likely(shinfo->gso_type & (SKB_GSO_TCPV4 | SKB_GSO_TCPV6)))
3968 3969 3970 3971 3972 3973 3974
		return tcp_hdrlen(skb) + shinfo->gso_size;

	/* UFO sets gso_size to the size of the fragmentation
	 * payload, i.e. the size of the L4 (UDP) header is already
	 * accounted for.
	 */
	return shinfo->gso_size;
3975 3976
}
EXPORT_SYMBOL_GPL(skb_gso_transport_seglen);
3977 3978 3979 3980 3981 3982 3983 3984 3985 3986 3987 3988 3989 3990 3991 3992 3993 3994 3995 3996 3997 3998 3999 4000 4001 4002 4003 4004 4005 4006 4007 4008 4009 4010 4011 4012 4013 4014 4015 4016 4017 4018 4019 4020 4021 4022 4023 4024 4025 4026 4027 4028

static struct sk_buff *skb_reorder_vlan_header(struct sk_buff *skb)
{
	if (skb_cow(skb, skb_headroom(skb)) < 0) {
		kfree_skb(skb);
		return NULL;
	}

	memmove(skb->data - ETH_HLEN, skb->data - VLAN_ETH_HLEN, 2 * ETH_ALEN);
	skb->mac_header += VLAN_HLEN;
	return skb;
}

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

	if (unlikely(vlan_tx_tag_present(skb))) {
		/* 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);