aboutsummaryrefslogtreecommitdiffstats
path: root/net/sctp/ulpqueue.c
blob: 0fa4d4d4df17f566a198a0685f1bcdf8015b4a82 (plain) (blame)
1
2
3
4
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
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
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
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
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
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
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
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
/* SCTP kernel reference Implementation
 * (C) Copyright IBM Corp. 2001, 2004
 * Copyright (c) 1999-2000 Cisco, Inc.
 * Copyright (c) 1999-2001 Motorola, Inc.
 * Copyright (c) 2001 Intel Corp.
 * Copyright (c) 2001 Nokia, Inc.
 * Copyright (c) 2001 La Monte H.P. Yarroll
 *
 * This abstraction carries sctp events to the ULP (sockets).
 *
 * The SCTP reference implementation 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, or (at your option)
 * any later version.
 *
 * The SCTP reference implementation is distributed in the hope that it
 * will be useful, but WITHOUT ANY WARRANTY; without even the implied
 *                 ************************
 * warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
 * See the GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with GNU CC; see the file COPYING.  If not, write to
 * the Free Software Foundation, 59 Temple Place - Suite 330,
 * Boston, MA 02111-1307, USA.
 *
 * Please send any bug reports or fixes you make to the
 * email address(es):
 *    lksctp developers <lksctp-developers@lists.sourceforge.net>
 *
 * Or submit a bug report through the following website:
 *    http://www.sf.net/projects/lksctp
 *
 * Written or modified by:
 *    Jon Grimm             <jgrimm@us.ibm.com>
 *    La Monte H.P. Yarroll <piggy@acm.org>
 *    Sridhar Samudrala     <sri@us.ibm.com>
 *
 * Any bugs reported given to us we will try to fix... any fixes shared will
 * be incorporated into the next SCTP release.
 */

#include <linux/types.h>
#include <linux/skbuff.h>
#include <net/sock.h>
#include <net/sctp/structs.h>
#include <net/sctp/sctp.h>
#include <net/sctp/sm.h>

/* Forward declarations for internal helpers.  */
static struct sctp_ulpevent * sctp_ulpq_reasm(struct sctp_ulpq *ulpq,
					      struct sctp_ulpevent *);
static struct sctp_ulpevent * sctp_ulpq_order(struct sctp_ulpq *,
					      struct sctp_ulpevent *);

/* 1st Level Abstractions */

/* Initialize a ULP queue from a block of memory.  */
struct sctp_ulpq *sctp_ulpq_init(struct sctp_ulpq *ulpq,
				 struct sctp_association *asoc)
{
	memset(ulpq, 0, sizeof(struct sctp_ulpq));

	ulpq->asoc = asoc;
	skb_queue_head_init(&ulpq->reasm);
	skb_queue_head_init(&ulpq->lobby);
	ulpq->pd_mode  = 0;
	ulpq->malloced = 0;

	return ulpq;
}


/* Flush the reassembly and ordering queues.  */
void sctp_ulpq_flush(struct sctp_ulpq *ulpq)
{
	struct sk_buff *skb;
	struct sctp_ulpevent *event;

	while ((skb = __skb_dequeue(&ulpq->lobby)) != NULL) {
		event = sctp_skb2event(skb);
		sctp_ulpevent_free(event);
	}

	while ((skb = __skb_dequeue(&ulpq->reasm)) != NULL) {
		event = sctp_skb2event(skb);
		sctp_ulpevent_free(event);
	}

}

/* Dispose of a ulpqueue.  */
void sctp_ulpq_free(struct sctp_ulpq *ulpq)
{
	sctp_ulpq_flush(ulpq);
	if (ulpq->malloced)
		kfree(ulpq);
}

/* Process an incoming DATA chunk.  */
int sctp_ulpq_tail_data(struct sctp_ulpq *ulpq, struct sctp_chunk *chunk,
			gfp_t gfp)
{
	struct sk_buff_head temp;
	sctp_data_chunk_t *hdr;
	struct sctp_ulpevent *event;

	hdr = (sctp_data_chunk_t *) chunk->chunk_hdr;

	/* Create an event from the incoming chunk. */
	event = sctp_ulpevent_make_rcvmsg(chunk->asoc, chunk, gfp);
	if (!event)
		return -ENOMEM;

	/* Do reassembly if needed.  */
	event = sctp_ulpq_reasm(ulpq, event);

	/* Do ordering if needed.  */
	if ((event) && (event->msg_flags & MSG_EOR)){
		/* Create a temporary list to collect chunks on.  */
		skb_queue_head_init(&temp);
		__skb_queue_tail(&temp, sctp_event2skb(event));

		event = sctp_ulpq_order(ulpq, event);
	}

	/* Send event to the ULP.  'event' is the sctp_ulpevent for
	 * very first SKB on the 'temp' list.
	 */
	if (event)
		sctp_ulpq_tail_event(ulpq, event);

	return 0;
}

/* Add a new event for propagation to the ULP.  */
/* Clear the partial delivery mode for this socket.   Note: This
 * assumes that no association is currently in partial delivery mode.
 */
int sctp_clear_pd(struct sock *sk, struct sctp_association *asoc)
{
	struct sctp_sock *sp = sctp_sk(sk);

	if (atomic_dec_and_test(&sp->pd_mode)) {
		/* This means there are no other associations in PD, so
		 * we can go ahead and clear out the lobby in one shot
		 */
		if (!skb_queue_empty(&sp->pd_lobby)) {
			struct list_head *list;
			sctp_skb_list_tail(&sp->pd_lobby, &sk->sk_receive_queue);
			list = (struct list_head *)&sctp_sk(sk)->pd_lobby;
			INIT_LIST_HEAD(list);
			return 1;
		}
	} else {
		/* There are other associations in PD, so we only need to
		 * pull stuff out of the lobby that belongs to the
		 * associations that is exiting PD (all of its notifications
		 * are posted here).
		 */
		if (!skb_queue_empty(&sp->pd_lobby) && asoc) {
			struct sk_buff *skb, *tmp;
			struct sctp_ulpevent *event;

			sctp_skb_for_each(skb, &sp->pd_lobby, tmp) {
				event = sctp_skb2event(skb);
				if (event->asoc == asoc) {
					__skb_unlink(skb, &sp->pd_lobby);
					__skb_queue_tail(&sk->sk_receive_queue,
							 skb);
				}
			}
		}
	}

	return 0;
}

/* Set the pd_mode on the socket and ulpq */
static void sctp_ulpq_set_pd(struct sctp_ulpq *ulpq)
{
	struct sctp_sock *sp = sctp_sk(ulpq->asoc->base.sk);

	atomic_inc(&sp->pd_mode);
	ulpq->pd_mode = 1;
}

/* Clear the pd_mode and restart any pending messages waiting for delivery. */
static int sctp_ulpq_clear_pd(struct sctp_ulpq *ulpq)
{
	ulpq->pd_mode = 0;
	return sctp_clear_pd(ulpq->asoc->base.sk, ulpq->asoc);
}

/* If the SKB of 'event' is on a list, it is the first such member
 * of that list.
 */
int sctp_ulpq_tail_event(struct sctp_ulpq *ulpq, struct sctp_ulpevent *event)
{
	struct sock *sk = ulpq->asoc->base.sk;
	struct sk_buff_head *queue, *skb_list;
	struct sk_buff *skb = sctp_event2skb(event);
	int clear_pd = 0;

	skb_list = (struct sk_buff_head *) skb->prev;

	/* If the socket is just going to throw this away, do not
	 * even try to deliver it.
	 */
	if (sock_flag(sk, SOCK_DEAD) || (sk->sk_shutdown & RCV_SHUTDOWN))
		goto out_free;

	/* Check if the user wishes to receive this event.  */
	if (!sctp_ulpevent_is_enabled(event, &sctp_sk(sk)->subscribe))
		goto out_free;

	/* If we are in partial delivery mode, post to the lobby until
	 * partial delivery is cleared, unless, of course _this_ is
	 * the association the cause of the partial delivery.
	 */

	if (atomic_read(&sctp_sk(sk)->pd_mode) == 0) {
		queue = &sk->sk_receive_queue;
	} else {
		if (ulpq->pd_mode) {
			/* If the association is in partial delivery, we
			 * need to finish delivering the partially processed
			 * packet before passing any other data.  This is
			 * because we don't truly support stream interleaving.
			 */
			if ((event->msg_flags & MSG_NOTIFICATION) ||
			    (SCTP_DATA_NOT_FRAG ==
				    (event->msg_flags & SCTP_DATA_FRAG_MASK)))
				queue = &sctp_sk(sk)->pd_lobby;
			else {
				clear_pd = event->msg_flags & MSG_EOR;
				queue = &sk->sk_receive_queue;
			}
		} else {
			/*
			 * If fragment interleave is enabled, we
			 * can queue this to the recieve queue instead
			 * of the lobby.
			 */
			if (sctp_sk(sk)->frag_interleave)
				queue = &sk->sk_receive_queue;
			else
				queue = &sctp_sk(sk)->pd_lobby;
		}
	}

	/* If we are harvesting multiple skbs they will be
	 * collected on a list.
	 */
	if (skb_list)
		sctp_skb_list_tail(skb_list, queue);
	else
		__skb_queue_tail(queue, skb);

	/* Did we just complete partial delivery and need to get
	 * rolling again?  Move pending data to the receive
	 * queue.
	 */
	if (clear_pd)
		sctp_ulpq_clear_pd(ulpq);

	if (queue == &sk->sk_receive_queue)
		sk->sk_data_ready(sk, 0);
	return 1;

out_free:
	if (skb_list)
		sctp_queue_purge_ulpevents(skb_list);
	else
		sctp_ulpevent_free(event);

	return 0;
}

/* 2nd Level Abstractions */

/* Helper function to store chunks that need to be reassembled.  */
static inline void sctp_ulpq_store_reasm(struct sctp_ulpq *ulpq,
					 struct sctp_ulpevent *event)
{
	struct sk_buff *pos;
	struct sctp_ulpevent *cevent;
	__u32 tsn, ctsn;

	tsn = event->tsn;

	/* See if it belongs at the end. */
	pos = skb_peek_tail(&ulpq->reasm);
	if (!pos) {
		__skb_queue_tail(&ulpq->reasm, sctp_event2skb(event));
		return;
	}

	/* Short circuit just dropping it at the end. */
	cevent = sctp_skb2event(pos);
	ctsn = cevent->tsn;
	if (TSN_lt(ctsn, tsn)) {
		__skb_queue_tail(&ulpq->reasm, sctp_event2skb(event));
		return;
	}

	/* Find the right place in this list. We store them by TSN.  */
	skb_queue_walk(&ulpq->reasm, pos) {
		cevent = sctp_skb2event(pos);
		ctsn = cevent->tsn;

		if (TSN_lt(tsn, ctsn))
			break;
	}

	/* Insert before pos. */
	__skb_insert(sctp_event2skb(event), pos->prev, pos, &ulpq->reasm);

}

/* Helper function to return an event corresponding to the reassembled
 * datagram.
 * This routine creates a re-assembled skb given the first and last skb's
 * as stored in the reassembly queue. The skb's may be non-linear if the sctp
 * payload was fragmented on the way and ip had to reassemble them.
 * We add the rest of skb's to the first skb's fraglist.
 */
static struct sctp_ulpevent *sctp_make_reassembled_event(struct sk_buff_head *queue, struct sk_buff *f_frag, struct sk_buff *l_frag)
{
	struct sk_buff *pos;
	struct sk_buff *new = NULL;
	struct sctp_ulpevent *event;
	struct sk_buff *pnext, *last;
	struct sk_buff *list = skb_shinfo(f_frag)->frag_list;

	/* Store the pointer to the 2nd skb */
	if (f_frag == l_frag)
		pos = NULL;
	else
		pos = f_frag->next;

	/* Get the last skb in the f_frag's frag_list if present. */
	for (last = list; list; last = list, list = list->next);

	/* Add the list of remaining fragments to the first fragments
	 * frag_list.
	 */
	if (last)
		last->next = pos;
	else {
		if (skb_cloned(f_frag)) {
			/* This is a cloned skb, we can't just modify
			 * the frag_list.  We need a new skb to do that.
			 * Instead of calling skb_unshare(), we'll do it
			 * ourselves since we need to delay the free.
			 */
			new = skb_copy(f_frag, GFP_ATOMIC);
			if (!new)
				return NULL;	/* try again later */

			sctp_skb_set_owner_r(new, f_frag->sk);

			skb_shinfo(new)->frag_list = pos;
		} else
			skb_shinfo(f_frag)->frag_list = pos;
	}

	/* Remove the first fragment from the reassembly queue.  */
	__skb_unlink(f_frag, queue);

	/* if we did unshare, then free the old skb and re-assign */
	if (new) {
		kfree_skb(f_frag);
		f_frag = new;
	}

	while (pos) {

		pnext = pos->next;

		/* Update the len and data_len fields of the first fragment. */
		f_frag->len += pos->len;
		f_frag->data_len += pos->len;

		/* Remove the fragment from the reassembly queue.  */
		__skb_unlink(pos, queue);

		/* Break if we have reached the last fragment.  */
		if (pos == l_frag)
			break;
		pos->next = pnext;
		pos = pnext;
	};

	event = sctp_skb2event(f_frag);
	SCTP_INC_STATS(SCTP_MIB_REASMUSRMSGS);

	return event;
}


/* Helper function to check if an incoming chunk has filled up the last
 * missing fragment in a SCTP datagram and return the corresponding event.
 */
static inline struct sctp_ulpevent *sctp_ulpq_retrieve_reassembled(struct sctp_ulpq *ulpq)
{
	struct sk_buff *pos;
	struct sctp_ulpevent *cevent;
	struct sk_buff *first_frag = NULL;
	__u32 ctsn, next_tsn;
	struct sctp_ulpevent *retval = NULL;
	struct sk_buff *pd_first = NULL;
	struct sk_buff *pd_last = NULL;
	size_t pd_len = 0;
	struct sctp_association *asoc;
	u32 pd_point;

	/* Initialized to 0 just to avoid compiler warning message.  Will
	 * never be used with this value. It is referenced only after it
	 * is set when we find the first fragment of a message.
	 */
	next_tsn = 0;

	/* The chunks are held in the reasm queue sorted by TSN.
	 * Walk through the queue sequentially and look for a sequence of
	 * fragmented chunks that complete a datagram.
	 * 'first_frag' and next_tsn are reset when we find a chunk which
	 * is the first fragment of a datagram. Once these 2 fields are set
	 * we expect to find the remaining middle fragments and the last
	 * fragment in order. If not, first_frag is reset to NULL and we
	 * start the next pass when we find another first fragment.
	 *
	 * There is a potential to do partial delivery if user sets
	 * SCTP_PARTIAL_DELIVERY_POINT option. Lets count some things here
	 * to see if can do PD.
	 */
	skb_queue_walk(&ulpq->reasm, pos) {
		cevent = sctp_skb2event(pos);
		ctsn = cevent->tsn;

		switch (cevent->msg_flags & SCTP_DATA_FRAG_MASK) {
		case SCTP_DATA_FIRST_FRAG:
			/* If this "FIRST_FRAG" is the first
			 * element in the queue, then count it towards
			 * possible PD.
			 */
			if (pos == ulpq->reasm.next) {
			    pd_first = pos;
			    pd_last = pos;
			    pd_len = pos->len;
			} else {
			    pd_first = NULL;
			    pd_last = NULL;
			    pd_len = 0;
			}

			first_frag = pos;
			next_tsn = ctsn + 1;
			break;

		case SCTP_DATA_MIDDLE_FRAG:
			if ((first_frag) && (ctsn == next_tsn)) {
				next_tsn++;
				if (pd_first) {
				    pd_last = pos;
				    pd_len += pos->len;
				}
			} else
				first_frag = NULL;
			break;

		case SCTP_DATA_LAST_FRAG:
			if (first_frag && (ctsn == next_tsn))
				goto found;
			else
				first_frag = NULL;
			break;
		};
	}

	asoc = ulpq->asoc;
	if (pd_first) {
		/* Make sure we can enter partial deliver.
		 * We can trigger partial delivery only if framgent
		 * interleave is set, or the socket is not already
		 * in  partial delivery.
		 */
		if (!sctp_sk(asoc->base.sk)->frag_interleave &&
		    atomic_read(&sctp_sk(asoc->base.sk)->pd_mode))
			goto done;

		cevent = sctp_skb2event(pd_first);
		pd_point = sctp_sk(asoc->base.sk)->pd_point;
		if (pd_point && pd_point <= pd_len) {
			retval = sctp_make_reassembled_event(&ulpq->reasm,
							     pd_first,
							     pd_last);
			if (retval)
				sctp_ulpq_set_pd(ulpq);
		}
	}
done:
	return retval;
found:
	retval = sctp_make_reassembled_event(&ulpq->reasm, first_frag, pos);
	if (retval)
		retval->msg_flags |= MSG_EOR;
	goto done;
}

/* Retrieve the next set of fragments of a partial message. */
static inline struct sctp_ulpevent *sctp_ulpq_retrieve_partial(struct sctp_ulpq *ulpq)
{
	struct sk_buff *pos, *last_frag, *first_frag;
	struct sctp_ulpevent *cevent;
	__u32 ctsn, next_tsn;
	int is_last;
	struct sctp_ulpevent *retval;

	/* The chunks are held in the reasm queue sorted by TSN.
	 * Walk through the queue sequentially and look for the first
	 * sequence of fragmented chunks.
	 */

	if (skb_queue_empty(&ulpq->reasm))
		return NULL;

	last_frag = first_frag = NULL;
	retval = NULL;
	next_tsn = 0;
	is_last = 0;

	skb_queue_walk(&ulpq->reasm, pos) {
		cevent = sctp_skb2event(pos);
		ctsn = cevent->tsn;

		switch (cevent->msg_flags & SCTP_DATA_FRAG_MASK) {
		case SCTP_DATA_MIDDLE_FRAG:
			if (!first_frag) {
				first_frag = pos;
				next_tsn = ctsn + 1;
				last_frag = pos;
			} else if (next_tsn == ctsn)
				next_tsn++;
			else
				goto done;
			break;
		case SCTP_DATA_LAST_FRAG:
			if (!first_frag)
				first_frag = pos;
			else if (ctsn != next_tsn)
				goto done;
			last_frag = pos;
			is_last = 1;
			goto done;
		default:
			return NULL;
		};
	}

	/* We have the reassembled event. There is no need to look
	 * further.
	 */
done:
	retval = sctp_make_reassembled_event(&ulpq->reasm, first_frag, last_frag);
	if (retval && is_last)
		retval->msg_flags |= MSG_EOR;

	return retval;
}


/* Helper function to reassemble chunks.  Hold chunks on the reasm queue that
 * need reassembling.
 */
static struct sctp_ulpevent *sctp_ulpq_reasm(struct sctp_ulpq *ulpq,
						struct sctp_ulpevent *event)
{
	struct sctp_ulpevent *retval = NULL;

	/* Check if this is part of a fragmented message.  */
	if (SCTP_DATA_NOT_FRAG == (event->msg_flags & SCTP_DATA_FRAG_MASK)) {
		event->msg_flags |= MSG_EOR;
		return event;
	}

	sctp_ulpq_store_reasm(ulpq, event);
	if (!ulpq->pd_mode)
		retval = sctp_ulpq_retrieve_reassembled(ulpq);
	else {
		__u32 ctsn, ctsnap;

		/* Do not even bother unless this is the next tsn to
		 * be delivered.
		 */
		ctsn = event->tsn;
		ctsnap = sctp_tsnmap_get_ctsn(&ulpq->asoc->peer.tsn_map);
		if (TSN_lte(ctsn, ctsnap))
			retval = sctp_ulpq_retrieve_partial(ulpq);
	}

	return retval;
}

/* Retrieve the first part (sequential fragments) for partial delivery.  */
static inline struct sctp_ulpevent *sctp_ulpq_retrieve_first(struct sctp_ulpq *ulpq)
{
	struct sk_buff *pos, *last_frag, *first_frag;
	struct sctp_ulpevent *cevent;
	__u32 ctsn, next_tsn;
	struct sctp_ulpevent *retval;

	/* The chunks are held in the reasm queue sorted by TSN.
	 * Walk through the queue sequentially and look for a sequence of
	 * fragmented chunks that start a datagram.
	 */

	if (skb_queue_empty(&ulpq->reasm))
		return NULL;

	last_frag = first_frag = NULL;
	retval = NULL;
	next_tsn = 0;

	skb_queue_walk(&ulpq->reasm, pos) {
		cevent = sctp_skb2event(pos);
		ctsn = cevent->tsn;

		switch (cevent->msg_flags & SCTP_DATA_FRAG_MASK) {
		case SCTP_DATA_FIRST_FRAG:
			if (!first_frag) {
				first_frag = pos;
				next_tsn = ctsn + 1;
				last_frag = pos;
			} else
				goto done;
			break;

		case SCTP_DATA_MIDDLE_FRAG:
			if (!first_frag)
				return NULL;
			if (ctsn == next_tsn) {
				next_tsn++;
				last_frag = pos;
			} else
				goto done;
			break;
		default:
			return NULL;
		};
	}

	/* We have the reassembled event. There is no need to look
	 * further.
	 */
done:
	retval = sctp_make_reassembled_event(&ulpq->reasm, first_frag, last_frag);
	return retval;
}

/* Helper function to gather skbs that have possibly become
 * ordered by an an incoming chunk.
 */
static inline void sctp_ulpq_retrieve_ordered(struct sctp_ulpq *ulpq,
					      struct sctp_ulpevent *event)
{
	struct sk_buff_head *event_list;
	struct sk_buff *pos, *tmp;
	struct sctp_ulpevent *cevent;
	struct sctp_stream *in;
	__u16 sid, csid;
	__u16 ssn, cssn;

	sid = event->stream;
	ssn = event->ssn;
	in  = &ulpq->asoc->ssnmap->in;

	event_list = (struct sk_buff_head *) sctp_event2skb(event)->prev;

	/* We are holding the chunks by stream, by SSN.  */
	sctp_skb_for_each(pos, &ulpq->lobby, tmp) {
		cevent = (struct sctp_ulpevent *) pos->cb;
		csid = cevent->stream;
		cssn = cevent->ssn;

		/* Have we gone too far?  */
		if (csid > sid)
			break;

		/* Have we not gone far enough?  */
		if (csid < sid)
			continue;

		if (cssn != sctp_ssn_peek(in, sid))
			break;

		/* Found it, so mark in the ssnmap. */
		sctp_ssn_next(in, sid);

		__skb_unlink(pos, &ulpq->lobby);

		/* Attach all gathered skbs to the event.  */
		__skb_queue_tail(event_list, pos);
	}
}

/* Helper function to store chunks needing ordering.  */
static inline void sctp_ulpq_store_ordered(struct sctp_ulpq *ulpq,
					   struct sctp_ulpevent *event)
{
	struct sk_buff *pos;
	struct sctp_ulpevent *cevent;
	__u16 sid, csid;
	__u16 ssn, cssn;

	pos = skb_peek_tail(&ulpq->lobby);
	if (!pos) {
		__skb_queue_tail(&ulpq->lobby, sctp_event2skb(event));
		return;
	}

	sid = event->stream;
	ssn = event->ssn;

	cevent = (struct sctp_ulpevent *) pos->cb;
	csid = cevent->stream;
	cssn = cevent->ssn;
	if (sid > csid) {
		__skb_queue_tail(&ulpq->lobby, sctp_event2skb(event));
		return;
	}

	if ((sid == csid) && SSN_lt(cssn, ssn)) {
		__skb_queue_tail(&ulpq->lobby, sctp_event2skb(event));
		return;
	}

	/* Find the right place in this list.  We store them by
	 * stream ID and then by SSN.
	 */
	skb_queue_walk(&ulpq->lobby, pos) {
		cevent = (struct sctp_ulpevent *) pos->cb;
		csid = cevent->stream;
		cssn = cevent->ssn;

		if (csid > sid)
			break;
		if (csid == sid && SSN_lt(ssn, cssn))
			break;
	}


	/* Insert before pos. */
	__skb_insert(sctp_event2skb(event), pos->prev, pos, &ulpq->lobby);

}

static struct sctp_ulpevent *sctp_ulpq_order(struct sctp_ulpq *ulpq,
					     struct sctp_ulpevent *event)
{
	__u16 sid, ssn;
	struct sctp_stream *in;

	/* Check if this message needs ordering.  */
	if (SCTP_DATA_UNORDERED & event->msg_flags)
		return event;

	/* Note: The stream ID must be verified before this routine.  */
	sid = event->stream;
	ssn = event->ssn;
	in  = &ulpq->asoc->ssnmap->in;

	/* Is this the expected SSN for this stream ID?  */
	if (ssn != sctp_ssn_peek(in, sid)) {
		/* We've received something out of order, so find where it
		 * needs to be placed.  We order by stream and then by SSN.
		 */
		sctp_ulpq_store_ordered(ulpq, event);
		return NULL;
	}

	/* Mark that the next chunk has been found.  */
	sctp_ssn_next(in, sid);

	/* Go find any other chunks that were waiting for
	 * ordering.
	 */
	sctp_ulpq_retrieve_ordered(ulpq, event);

	return event;
}

/* Helper function to gather skbs that have possibly become
 * ordered by forward tsn skipping their dependencies.
 */
static inline void sctp_ulpq_reap_ordered(struct sctp_ulpq *ulpq)
{
	struct sk_buff *pos, *tmp;
	struct sctp_ulpevent *cevent;
	struct sctp_ulpevent *event;
	struct sctp_stream *in;
	struct sk_buff_head temp;
	__u16 csid, cssn;

	in  = &ulpq->asoc->ssnmap->in;

	/* We are holding the chunks by stream, by SSN.  */
	skb_queue_head_init(&temp);
	event = NULL;
	sctp_skb_for_each(pos, &ulpq->lobby, tmp) {
		cevent = (struct sctp_ulpevent *) pos->cb;
		csid = cevent->stream;
		cssn = cevent->ssn;

		if (cssn != sctp_ssn_peek(in, csid))
			break;

		/* Found it, so mark in the ssnmap. */
		sctp_ssn_next(in, csid);

		__skb_unlink(pos, &ulpq->lobby);
		if (!event) {
			/* Create a temporary list to collect chunks on.  */
			event = sctp_skb2event(pos);
			__skb_queue_tail(&temp, sctp_event2skb(event));
		} else {
			/* Attach all gathered skbs to the event.  */
			__skb_queue_tail(&temp, pos);
		}
	}

	/* Send event to the ULP.  'event' is the sctp_ulpevent for
	 * very first SKB on the 'temp' list.
	 */
	if (event)
		sctp_ulpq_tail_event(ulpq, event);
}

/* Skip over an SSN. */
void sctp_ulpq_skip(struct sctp_ulpq *ulpq, __u16 sid, __u16 ssn)
{
	struct sctp_stream *in;

	/* Note: The stream ID must be verified before this routine.  */
	in  = &ulpq->asoc->ssnmap->in;

	/* Is this an old SSN?  If so ignore. */
	if (SSN_lt(ssn, sctp_ssn_peek(in, sid)))
		return;

	/* Mark that we are no longer expecting this SSN or lower. */
	sctp_ssn_skip(in, sid, ssn);

	/* Go find any other chunks that were waiting for
	 * ordering and deliver them if needed.
	 */
	sctp_ulpq_reap_ordered(ulpq);
	return;
}

/* Renege 'needed' bytes from the ordering queue. */
static __u16 sctp_ulpq_renege_order(struct sctp_ulpq *ulpq, __u16 needed)
{
	__u16 freed = 0;
	__u32 tsn;
	struct sk_buff *skb;
	struct sctp_ulpevent *event;
	struct sctp_tsnmap *tsnmap;

	tsnmap = &ulpq->asoc->peer.tsn_map;

	while ((skb = __skb_dequeue_tail(&ulpq->lobby)) != NULL) {
		freed += skb_headlen(skb);
		event = sctp_skb2event(skb);
		tsn = event->tsn;

		sctp_ulpevent_free(event);
		sctp_tsnmap_renege(tsnmap, tsn);
		if (freed >= needed)
			return freed;
	}

	return freed;
}

/* Renege 'needed' bytes from the reassembly queue. */
static __u16 sctp_ulpq_renege_frags(struct sctp_ulpq *ulpq, __u16 needed)
{
	__u16 freed = 0;
	__u32 tsn;
	struct sk_buff *skb;
	struct sctp_ulpevent *event;
	struct sctp_tsnmap *tsnmap;

	tsnmap = &ulpq->asoc->peer.tsn_map;

	/* Walk backwards through the list, reneges the newest tsns. */
	while ((skb = __skb_dequeue_tail(&ulpq->reasm)) != NULL) {
		freed += skb_headlen(skb);
		event = sctp_skb2event(skb);
		tsn = event->tsn;

		sctp_ulpevent_free(event);
		sctp_tsnmap_renege(tsnmap, tsn);
		if (freed >= needed)
			return freed;
	}

	return freed;
}

/* Partial deliver the first message as there is pressure on rwnd. */
void sctp_ulpq_partial_delivery(struct sctp_ulpq *ulpq,
				struct sctp_chunk *chunk,
				gfp_t gfp)
{
	struct sctp_ulpevent *event;
	struct sctp_association *asoc;
	struct sctp_sock *sp;

	asoc = ulpq->asoc;
	sp = sctp_sk(asoc->base.sk);

	/* If the association is already in Partial Delivery mode
	 * we have noting to do.
	 */
	if (ulpq->pd_mode)
		return;

	/* If the user enabled fragment interleave socket option,
	 * multiple associations can enter partial delivery.
	 * Otherwise, we can only enter partial delivery if the
	 * socket is not in partial deliver mode.
	 */
	if (sp->frag_interleave || atomic_read(&sp->pd_mode) == 0) {
		/* Is partial delivery possible?  */
		event = sctp_ulpq_retrieve_first(ulpq);
		/* Send event to the ULP.   */
		if (event) {
			sctp_ulpq_tail_event(ulpq, event);
			sctp_ulpq_set_pd(ulpq);
			return;
		}
	}
}

/* Renege some packets to make room for an incoming chunk.  */
void sctp_ulpq_renege(struct sctp_ulpq *ulpq, struct sctp_chunk *chunk,
		      gfp_t gfp)
{
	struct sctp_association *asoc;
	__u16 needed, freed;

	asoc = ulpq->asoc;

	if (chunk) {
		needed = ntohs(chunk->chunk_hdr->length);
		needed -= sizeof(sctp_data_chunk_t);
	} else
		needed = SCTP_DEFAULT_MAXWINDOW;

	freed = 0;

	if (skb_queue_empty(&asoc->base.sk->sk_receive_queue)) {
		freed = sctp_ulpq_renege_order(ulpq, needed);
		if (freed < needed) {
			freed += sctp_ulpq_renege_frags(ulpq, needed - freed);
		}
	}
	/* If able to free enough room, accept this chunk. */
	if (chunk && (freed >= needed)) {
		__u32 tsn;
		tsn = ntohl(chunk->subh.data_hdr->tsn);
		sctp_tsnmap_mark(&asoc->peer.tsn_map, tsn);
		sctp_ulpq_tail_data(ulpq, chunk, gfp);

		sctp_ulpq_partial_delivery(ulpq, chunk, gfp);
	}

	return;
}



/* Notify the application if an association is aborted and in
 * partial delivery mode.  Send up any pending received messages.
 */
void sctp_ulpq_abort_pd(struct sctp_ulpq *ulpq, gfp_t gfp)
{
	struct sctp_ulpevent *ev = NULL;
	struct sock *sk;

	if (!ulpq->pd_mode)
		return;

	sk = ulpq->asoc->base.sk;
	if (sctp_ulpevent_type_enabled(SCTP_PARTIAL_DELIVERY_EVENT,
				       &sctp_sk(sk)->subscribe))
		ev = sctp_ulpevent_make_pdapi(ulpq->asoc,
					      SCTP_PARTIAL_DELIVERY_ABORTED,
					      gfp);
	if (ev)
		__skb_queue_tail(&sk->sk_receive_queue, sctp_event2skb(ev));

	/* If there is data waiting, send it up the socket now. */
	if (sctp_ulpq_clear_pd(ulpq) || ev)
		sk->sk_data_ready(sk, 0);
}