msg.c 15.4 KB
Newer Older
P
Per Liden 已提交
1 2
/*
 * net/tipc/msg.c: TIPC message header routines
3
 *
4
 * Copyright (c) 2000-2006, 2014-2015, Ericsson AB
5
 * Copyright (c) 2005, 2010-2011, Wind River Systems
P
Per Liden 已提交
6 7
 * All rights reserved.
 *
P
Per Liden 已提交
8
 * Redistribution and use in source and binary forms, with or without
P
Per Liden 已提交
9 10
 * modification, are permitted provided that the following conditions are met:
 *
P
Per Liden 已提交
11 12 13 14 15 16 17 18
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 * 3. Neither the names of the copyright holders nor the names of its
 *    contributors may be used to endorse or promote products derived from
 *    this software without specific prior written permission.
P
Per Liden 已提交
19
 *
P
Per Liden 已提交
20 21 22 23 24 25 26 27 28 29 30 31 32 33
 * Alternatively, this software may be distributed under the terms of the
 * GNU General Public License ("GPL") version 2 as published by the Free
 * Software Foundation.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
 * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
P
Per Liden 已提交
34 35 36
 * POSSIBILITY OF SUCH DAMAGE.
 */

37
#include <net/sock.h>
P
Per Liden 已提交
38 39
#include "core.h"
#include "msg.h"
40 41
#include "addr.h"
#include "name_table.h"
P
Per Liden 已提交
42

43 44
#define MAX_FORWARD_SIZE 1024

45
static unsigned int align(unsigned int i)
46
{
47
	return (i + 3) & ~3u;
48 49
}

Y
Ying Xue 已提交
50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72
/**
 * tipc_buf_acquire - creates a TIPC message buffer
 * @size: message size (including TIPC header)
 *
 * Returns a new buffer with data pointers set to the specified size.
 *
 * NOTE: Headroom is reserved to allow prepending of a data link header.
 *       There may also be unrequested tailroom present at the buffer's end.
 */
struct sk_buff *tipc_buf_acquire(u32 size)
{
	struct sk_buff *skb;
	unsigned int buf_size = (BUF_HEADROOM + size + 3) & ~3u;

	skb = alloc_skb_fclone(buf_size, GFP_ATOMIC);
	if (skb) {
		skb_reserve(skb, BUF_HEADROOM);
		skb_put(skb, size);
		skb->next = NULL;
	}
	return skb;
}

73 74
void tipc_msg_init(u32 own_node, struct tipc_msg *m, u32 user, u32 type,
		   u32 hsize, u32 dnode)
75 76 77 78 79 80
{
	memset(m, 0, hsize);
	msg_set_version(m);
	msg_set_user(m, user);
	msg_set_hdr_sz(m, hsize);
	msg_set_size(m, hsize);
81
	msg_set_prevnode(m, own_node);
82
	msg_set_type(m, type);
83
	if (hsize > SHORT_H_SIZE) {
84 85
		msg_set_orignode(m, own_node);
		msg_set_destnode(m, dnode);
86 87 88
	}
}

89
struct sk_buff *tipc_msg_create(uint user, uint type,
90 91
				uint hdr_sz, uint data_sz, u32 dnode,
				u32 onode, u32 dport, u32 oport, int errcode)
92 93 94 95 96 97 98 99 100
{
	struct tipc_msg *msg;
	struct sk_buff *buf;

	buf = tipc_buf_acquire(hdr_sz + data_sz);
	if (unlikely(!buf))
		return NULL;

	msg = buf_msg(buf);
101
	tipc_msg_init(onode, msg, user, type, hdr_sz, dnode);
102 103 104 105 106 107 108 109 110
	msg_set_size(msg, hdr_sz + data_sz);
	msg_set_origport(msg, oport);
	msg_set_destport(msg, dport);
	msg_set_errcode(msg, errcode);
	if (hdr_sz > SHORT_H_SIZE) {
		msg_set_orignode(msg, onode);
		msg_set_destnode(msg, dnode);
	}
	return buf;
111 112
}

113
/* tipc_buf_append(): Append a buffer to the fragment list of another buffer
114 115 116
 * @*headbuf: in:  NULL for first frag, otherwise value returned from prev call
 *            out: set when successful non-complete reassembly, otherwise NULL
 * @*buf:     in:  the buffer to append. Always defined
S
stephen hemminger 已提交
117
 *            out: head buf after successful complete reassembly, otherwise NULL
118
 * Returns 1 when reassembly complete, otherwise 0
119 120 121 122 123 124
 */
int tipc_buf_append(struct sk_buff **headbuf, struct sk_buff **buf)
{
	struct sk_buff *head = *headbuf;
	struct sk_buff *frag = *buf;
	struct sk_buff *tail;
125 126
	struct tipc_msg *msg;
	u32 fragid;
127
	int delta;
128
	bool headstolen;
129

130 131 132 133 134 135
	if (!frag)
		goto err;

	msg = buf_msg(frag);
	fragid = msg_type(msg);
	frag->next = NULL;
136 137 138
	skb_pull(frag, msg_hdr_sz(msg));

	if (fragid == FIRST_FRAGMENT) {
139 140 141 142
		if (unlikely(head))
			goto err;
		if (unlikely(skb_unclone(frag, GFP_ATOMIC)))
			goto err;
143 144
		head = *headbuf = frag;
		skb_frag_list_init(head);
145
		TIPC_SKB_CB(head)->tail = NULL;
146
		*buf = NULL;
147 148
		return 0;
	}
149

150
	if (!head)
151 152
		goto err;

153 154 155
	if (skb_try_coalesce(head, frag, &headstolen, &delta)) {
		kfree_skb_partial(frag, headstolen);
	} else {
156
		tail = TIPC_SKB_CB(head)->tail;
157 158 159 160 161 162 163 164 165
		if (!skb_has_frag_list(head))
			skb_shinfo(head)->frag_list = frag;
		else
			tail->next = frag;
		head->truesize += frag->truesize;
		head->data_len += frag->len;
		head->len += frag->len;
		TIPC_SKB_CB(head)->tail = frag;
	}
166

167 168 169 170 171 172 173 174
	if (fragid == LAST_FRAGMENT) {
		*buf = head;
		TIPC_SKB_CB(head)->tail = NULL;
		*headbuf = NULL;
		return 1;
	}
	*buf = NULL;
	return 0;
175 176

err:
177 178
	pr_warn_ratelimited("Unable to build fragment list\n");
	kfree_skb(*buf);
179 180
	kfree_skb(*headbuf);
	*buf = *headbuf = NULL;
181 182
	return 0;
}
183

184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225
/* tipc_msg_validate - validate basic format of received message
 *
 * This routine ensures a TIPC message has an acceptable header, and at least
 * as much data as the header indicates it should.  The routine also ensures
 * that the entire message header is stored in the main fragment of the message
 * buffer, to simplify future access to message header fields.
 *
 * Note: Having extra info present in the message header or data areas is OK.
 * TIPC will ignore the excess, under the assumption that it is optional info
 * introduced by a later release of the protocol.
 */
bool tipc_msg_validate(struct sk_buff *skb)
{
	struct tipc_msg *msg;
	int msz, hsz;

	if (unlikely(TIPC_SKB_CB(skb)->validated))
		return true;
	if (unlikely(!pskb_may_pull(skb, MIN_H_SIZE)))
		return false;

	hsz = msg_hdr_sz(buf_msg(skb));
	if (unlikely(hsz < MIN_H_SIZE) || (hsz > MAX_H_SIZE))
		return false;
	if (unlikely(!pskb_may_pull(skb, hsz)))
		return false;

	msg = buf_msg(skb);
	if (unlikely(msg_version(msg) != TIPC_VERSION))
		return false;

	msz = msg_size(msg);
	if (unlikely(msz < hsz))
		return false;
	if (unlikely((msz - hsz) > TIPC_MAX_USER_MSG_SIZE))
		return false;
	if (unlikely(skb->len < msz))
		return false;

	TIPC_SKB_CB(skb)->validated = true;
	return true;
}
226 227

/**
228
 * tipc_msg_build - create buffer chain containing specified header and data
229
 * @mhdr: Message header, to be prepended to data
230
 * @m: User message
231 232
 * @dsz: Total length of user data
 * @pktmax: Max packet size that can be used
233 234
 * @list: Buffer or chain of buffers to be returned to caller
 *
235 236
 * Returns message data size or errno: -ENOMEM, -EFAULT
 */
237
int tipc_msg_build(struct tipc_msg *mhdr, struct msghdr *m,
238
		   int offset, int dsz, int pktmax, struct sk_buff_head *list)
239 240 241 242 243 244 245 246
{
	int mhsz = msg_hdr_sz(mhdr);
	int msz = mhsz + dsz;
	int pktno = 1;
	int pktsz;
	int pktrem = pktmax;
	int drem = dsz;
	struct tipc_msg pkthdr;
247
	struct sk_buff *skb;
248 249
	char *pktpos;
	int rc;
250

251 252 253 254
	msg_set_size(mhdr, msz);

	/* No fragmentation needed? */
	if (likely(msz <= pktmax)) {
255 256
		skb = tipc_buf_acquire(msz);
		if (unlikely(!skb))
257
			return -ENOMEM;
258
		skb_orphan(skb);
259 260 261
		__skb_queue_tail(list, skb);
		skb_copy_to_linear_data(skb, mhdr, mhsz);
		pktpos = skb->data + mhsz;
A
Al Viro 已提交
262
		if (copy_from_iter(pktpos, dsz, &m->msg_iter) == dsz)
263 264 265 266 267 268
			return dsz;
		rc = -EFAULT;
		goto error;
	}

	/* Prepare reusable fragment header */
269 270
	tipc_msg_init(msg_prevnode(mhdr), &pkthdr, MSG_FRAGMENTER,
		      FIRST_FRAGMENT, INT_H_SIZE, msg_destnode(mhdr));
271 272 273 274
	msg_set_size(&pkthdr, pktmax);
	msg_set_fragm_no(&pkthdr, pktno);

	/* Prepare first fragment */
275 276
	skb = tipc_buf_acquire(pktmax);
	if (!skb)
277
		return -ENOMEM;
278
	skb_orphan(skb);
279 280 281
	__skb_queue_tail(list, skb);
	pktpos = skb->data;
	skb_copy_to_linear_data(skb, &pkthdr, INT_H_SIZE);
282 283
	pktpos += INT_H_SIZE;
	pktrem -= INT_H_SIZE;
284
	skb_copy_to_linear_data_offset(skb, INT_H_SIZE, mhdr, mhsz);
285 286 287 288 289 290 291
	pktpos += mhsz;
	pktrem -= mhsz;

	do {
		if (drem < pktrem)
			pktrem = drem;

A
Al Viro 已提交
292
		if (copy_from_iter(pktpos, pktrem, &m->msg_iter) != pktrem) {
293 294 295 296 297 298 299 300 301 302 303 304 305
			rc = -EFAULT;
			goto error;
		}
		drem -= pktrem;

		if (!drem)
			break;

		/* Prepare new fragment: */
		if (drem < (pktmax - INT_H_SIZE))
			pktsz = drem + INT_H_SIZE;
		else
			pktsz = pktmax;
306 307
		skb = tipc_buf_acquire(pktsz);
		if (!skb) {
308 309 310
			rc = -ENOMEM;
			goto error;
		}
311
		skb_orphan(skb);
312
		__skb_queue_tail(list, skb);
313 314 315
		msg_set_type(&pkthdr, FRAGMENT);
		msg_set_size(&pkthdr, pktsz);
		msg_set_fragm_no(&pkthdr, ++pktno);
316 317
		skb_copy_to_linear_data(skb, &pkthdr, INT_H_SIZE);
		pktpos = skb->data + INT_H_SIZE;
318 319 320
		pktrem = pktsz - INT_H_SIZE;

	} while (1);
321
	msg_set_type(buf_msg(skb), LAST_FRAGMENT);
322 323
	return dsz;
error:
324 325
	__skb_queue_purge(list);
	__skb_queue_head_init(list);
326 327 328
	return rc;
}

329 330
/**
 * tipc_msg_bundle(): Append contents of a buffer to tail of an existing one
331 332
 * @list: the buffer chain of the existing buffer ("bundle")
 * @skb:  buffer to be appended
333 334 335 336
 * @mtu:  max allowable size for the bundle buffer
 * Consumes buffer if successful
 * Returns true if bundling could be performed, otherwise false
 */
337
bool tipc_msg_bundle(struct sk_buff_head *list, struct sk_buff *skb, u32 mtu)
338
{
339 340 341
	struct sk_buff *bskb = skb_peek_tail(list);
	struct tipc_msg *bmsg = buf_msg(bskb);
	struct tipc_msg *msg = buf_msg(skb);
342 343 344 345 346 347 348 349 350 351 352 353 354 355
	unsigned int bsz = msg_size(bmsg);
	unsigned int msz = msg_size(msg);
	u32 start = align(bsz);
	u32 max = mtu - INT_H_SIZE;
	u32 pad = start - bsz;

	if (likely(msg_user(msg) == MSG_FRAGMENTER))
		return false;
	if (unlikely(msg_user(msg) == CHANGEOVER_PROTOCOL))
		return false;
	if (unlikely(msg_user(msg) == BCAST_PROTOCOL))
		return false;
	if (likely(msg_user(bmsg) != MSG_BUNDLER))
		return false;
356
	if (likely(!TIPC_SKB_CB(bskb)->bundling))
357
		return false;
358
	if (unlikely(skb_tailroom(bskb) < (pad + msz)))
359 360 361 362
		return false;
	if (unlikely(max < (start + msz)))
		return false;

363 364
	skb_put(bskb, pad + msz);
	skb_copy_to_linear_data_offset(bskb, start, skb->data, msz);
365 366
	msg_set_size(bmsg, start + msz);
	msg_set_msgcnt(bmsg, msg_msgcnt(bmsg) + 1);
367
	kfree_skb(skb);
368 369 370
	return true;
}

371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404
/**
 *  tipc_msg_extract(): extract bundled inner packet from buffer
 *  @skb: linear outer buffer, to be extracted from.
 *  @iskb: extracted inner buffer, to be returned
 *  @pos: position of msg to be extracted. Returns with pointer of next msg
 *  Consumes outer buffer when last packet extracted
 *  Returns true when when there is an extracted buffer, otherwise false
 */
bool tipc_msg_extract(struct sk_buff *skb, struct sk_buff **iskb, int *pos)
{
	struct tipc_msg *msg = buf_msg(skb);
	int imsz;
	struct tipc_msg *imsg = (struct tipc_msg *)(msg_data(msg) + *pos);

	/* Is there space left for shortest possible message? */
	if (*pos > (msg_data_sz(msg) - SHORT_H_SIZE))
		goto none;
	imsz = msg_size(imsg);

	/* Is there space left for current message ? */
	if ((*pos + imsz) > msg_data_sz(msg))
		goto none;
	*iskb = tipc_buf_acquire(imsz);
	if (!*iskb)
		goto none;
	skb_copy_to_linear_data(*iskb, imsg, imsz);
	*pos += align(imsz);
	return true;
none:
	kfree_skb(skb);
	*iskb = NULL;
	return false;
}

405 406
/**
 * tipc_msg_make_bundle(): Create bundle buf and append message to its tail
407 408 409
 * @list: the buffer chain
 * @skb: buffer to be appended and replaced
 * @mtu: max allowable size for the bundle buffer, inclusive header
410 411
 * @dnode: destination node for message. (Not always present in header)
 * Replaces buffer if successful
S
stephen hemminger 已提交
412
 * Returns true if success, otherwise false
413
 */
414
bool tipc_msg_make_bundle(struct sk_buff_head *list,
415
			  struct sk_buff *skb, u32 mtu, u32 dnode)
416
{
417
	struct sk_buff *bskb;
418
	struct tipc_msg *bmsg;
419
	struct tipc_msg *msg = buf_msg(skb);
420 421 422 423 424 425 426 427 428 429 430 431
	u32 msz = msg_size(msg);
	u32 max = mtu - INT_H_SIZE;

	if (msg_user(msg) == MSG_FRAGMENTER)
		return false;
	if (msg_user(msg) == CHANGEOVER_PROTOCOL)
		return false;
	if (msg_user(msg) == BCAST_PROTOCOL)
		return false;
	if (msz > (max / 2))
		return false;

432 433
	bskb = tipc_buf_acquire(max);
	if (!bskb)
434 435
		return false;

436 437
	skb_trim(bskb, INT_H_SIZE);
	bmsg = buf_msg(bskb);
438 439
	tipc_msg_init(msg_prevnode(msg), bmsg, MSG_BUNDLER, 0,
		      INT_H_SIZE, dnode);
440 441 442
	msg_set_seqno(bmsg, msg_seqno(msg));
	msg_set_ack(bmsg, msg_ack(msg));
	msg_set_bcast_ack(bmsg, msg_bcast_ack(msg));
443 444 445
	TIPC_SKB_CB(bskb)->bundling = true;
	__skb_queue_tail(list, bskb);
	return tipc_msg_bundle(list, skb, mtu);
446
}
447 448 449 450 451 452 453 454 455

/**
 * tipc_msg_reverse(): swap source and destination addresses and add error code
 * @buf:  buffer containing message to be reversed
 * @dnode: return value: node where to send message after reversal
 * @err:  error code to be set in message
 * Consumes buffer if failure
 * Returns true if success, otherwise false
 */
456
bool tipc_msg_reverse(u32 own_addr,  struct sk_buff *buf, u32 *dnode,
457
		      int err)
458 459 460 461 462 463
{
	struct tipc_msg *msg = buf_msg(buf);
	uint imp = msg_importance(msg);
	struct tipc_msg ohdr;
	uint rdsz = min_t(uint, msg_data_sz(msg), MAX_FORWARD_SIZE);

464
	if (skb_linearize(buf))
465
		goto exit;
466 467 468
	if (msg_dest_droppable(msg))
		goto exit;
	if (msg_errcode(msg))
469 470 471
		goto exit;

	memcpy(&ohdr, msg, msg_hdr_sz(msg));
472 473 474
	imp = min_t(uint, imp + 1, TIPC_CRITICAL_IMPORTANCE);
	if (msg_isdata(msg))
		msg_set_importance(msg, imp);
475 476 477
	msg_set_errcode(msg, err);
	msg_set_origport(msg, msg_destport(&ohdr));
	msg_set_destport(msg, msg_origport(&ohdr));
478
	msg_set_prevnode(msg, own_addr);
479 480 481 482 483 484 485 486 487 488 489
	if (!msg_short(msg)) {
		msg_set_orignode(msg, msg_destnode(&ohdr));
		msg_set_destnode(msg, msg_orignode(&ohdr));
	}
	msg_set_size(msg, msg_hdr_sz(msg) + rdsz);
	skb_trim(buf, msg_size(msg));
	skb_orphan(buf);
	*dnode = msg_orignode(&ohdr);
	return true;
exit:
	kfree_skb(buf);
490
	*dnode = 0;
491 492
	return false;
}
493 494

/**
495 496 497 498
 * tipc_msg_lookup_dest(): try to find new destination for named message
 * @skb: the buffer containing the message.
 * @dnode: return value: next-hop node, if destination found
 * @err: return value: error code to use, if message to be rejected
499
 * Does not consume buffer
500
 * Returns true if a destination is found, false otherwise
501
 */
502 503
bool tipc_msg_lookup_dest(struct net *net, struct sk_buff *skb,
			  u32 *dnode, int *err)
504
{
505
	struct tipc_msg *msg = buf_msg(skb);
506 507
	u32 dport;

508 509 510 511 512 513 514
	if (!msg_isdata(msg))
		return false;
	if (!msg_named(msg))
		return false;
	*err = -TIPC_ERR_NO_NAME;
	if (skb_linearize(skb))
		return false;
515
	if (msg_reroute_cnt(msg) > 0)
516
		return false;
517
	*dnode = addr_domain(net, msg_lookup_scope(msg));
518
	dport = tipc_nametbl_translate(net, msg_nametype(msg),
519
				       msg_nameinst(msg), dnode);
520
	if (!dport)
521
		return false;
522 523 524
	msg_incr_reroute_cnt(msg);
	msg_set_destnode(msg, *dnode);
	msg_set_destport(msg, dport);
525 526
	*err = TIPC_OK;
	return true;
527
}
528 529 530 531

/* tipc_msg_reassemble() - clone a buffer chain of fragments and
 *                         reassemble the clones into one message
 */
532
struct sk_buff *tipc_msg_reassemble(struct sk_buff_head *list)
533
{
534 535
	struct sk_buff *skb;
	struct sk_buff *frag = NULL;
536 537 538 539
	struct sk_buff *head = NULL;
	int hdr_sz;

	/* Copy header if single buffer */
540 541 542 543
	if (skb_queue_len(list) == 1) {
		skb = skb_peek(list);
		hdr_sz = skb_headroom(skb) + msg_hdr_sz(buf_msg(skb));
		return __pskb_copy(skb, hdr_sz, GFP_ATOMIC);
544 545 546
	}

	/* Clone all fragments and reassemble */
547 548
	skb_queue_walk(list, skb) {
		frag = skb_clone(skb, GFP_ATOMIC);
549 550 551 552 553 554 555 556 557 558 559 560 561 562
		if (!frag)
			goto error;
		frag->next = NULL;
		if (tipc_buf_append(&head, &frag))
			break;
		if (!head)
			goto error;
	}
	return frag;
error:
	pr_warn("Failed do clone local mcast rcv buffer\n");
	kfree_skb(head);
	return NULL;
}