summaryrefslogtreecommitdiff
path: root/fs/ubifs/io.c
blob: a0c5fb4b6fa86edde4cbf9a38239fb1d48986882 (plain)
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
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
/*
 * This file is part of UBIFS.
 *
 * Copyright (C) 2006-2008 Nokia Corporation.
 * Copyright (C) 2006, 2007 University of Szeged, Hungary
 *
 * This program is free software; you can redistribute it and/or modify it
 * under the terms of the GNU General Public License version 2 as published by
 * the Free Software Foundation.
 *
 * This program 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
 * this program; if not, write to the Free Software Foundation, Inc., 51
 * Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
 *
 * Authors: Artem Bityutskiy (Битюцкий Артём)
 *          Adrian Hunter
 *          Zoltan Sogor
 */

/*
 * This file implements UBIFS I/O subsystem which provides various I/O-related
 * helper functions (reading/writing/checking/validating nodes) and implements
 * write-buffering support. Write buffers help to save space which otherwise
 * would have been wasted for padding to the nearest minimal I/O unit boundary.
 * Instead, data first goes to the write-buffer and is flushed when the
 * buffer is full or when it is not used for some time (by timer). This is
 * similar to the mechanism is used by JFFS2.
 *
 * UBIFS distinguishes between minimum write size (@c->min_io_size) and maximum
 * write size (@c->max_write_size). The latter is the maximum amount of bytes
 * the underlying flash is able to program at a time, and writing in
 * @c->max_write_size units should presumably be faster. Obviously,
 * @c->min_io_size <= @c->max_write_size. Write-buffers are of
 * @c->max_write_size bytes in size for maximum performance. However, when a
 * write-buffer is flushed, only the portion of it (aligned to @c->min_io_size
 * boundary) which contains data is written, not the whole write-buffer,
 * because this is more space-efficient.
 *
 * This optimization adds few complications to the code. Indeed, on the one
 * hand, we want to write in optimal @c->max_write_size bytes chunks, which
 * also means aligning writes at the @c->max_write_size bytes offsets. On the
 * other hand, we do not want to waste space when synchronizing the write
 * buffer, so during synchronization we writes in smaller chunks. And this makes
 * the next write offset to be not aligned to @c->max_write_size bytes. So the
 * have to make sure that the write-buffer offset (@wbuf->offs) becomes aligned
 * to @c->max_write_size bytes again. We do this by temporarily shrinking
 * write-buffer size (@wbuf->size).
 *
 * Write-buffers are defined by 'struct ubifs_wbuf' objects and protected by
 * mutexes defined inside these objects. Since sometimes upper-level code
 * has to lock the write-buffer (e.g. journal space reservation code), many
 * functions related to write-buffers have "nolock" suffix which means that the
 * caller has to lock the write-buffer before calling this function.
 *
 * UBIFS stores nodes at 64 bit-aligned addresses. If the node length is not
 * aligned, UBIFS starts the next node from the aligned address, and the padded
 * bytes may contain any rubbish. In other words, UBIFS does not put padding
 * bytes in those small gaps. Common headers of nodes store real node lengths,
 * not aligned lengths. Indexing nodes also store real lengths in branches.
 *
 * UBIFS uses padding when it pads to the next min. I/O unit. In this case it
 * uses padding nodes or padding bytes, if the padding node does not fit.
 *
 * All UBIFS nodes are protected by CRC checksums and UBIFS checks CRC when
 * they are read from the flash media.
 */

#include <linux/crc32.h>
#include <linux/slab.h>
#include "ubifs.h"

/**
 * ubifs_ro_mode - switch UBIFS to read read-only mode.
 * @c: UBIFS file-system description object
 * @err: error code which is the reason of switching to R/O mode
 */
void ubifs_ro_mode(struct ubifs_info *c, int err)
{
	if (!c->ro_error) {
		c->ro_error = 1;
		c->no_chk_data_crc = 0;
		c->vfs_sb->s_flags |= MS_RDONLY;
		ubifs_warn("switched to read-only mode, error %d", err);
		dump_stack();
	}
}

/*
 * Below are simple wrappers over UBI I/O functions which include some
 * additional checks and UBIFS debugging stuff. See corresponding UBI function
 * for more information.
 */

int ubifs_leb_read(const struct ubifs_info *c, int lnum, void *buf, int offs,
		   int len, int even_ebadmsg)
{
	int err;

	err = ubi_read(c->ubi, lnum, buf, offs, len);
	/*
	 * In case of %-EBADMSG print the error message only if the
	 * @even_ebadmsg is true.
	 */
	if (err && (err != -EBADMSG || even_ebadmsg)) {
		ubifs_err("reading %d bytes from LEB %d:%d failed, error %d",
			  len, lnum, offs, err);
		dbg_dump_stack();
	}
	return err;
}

int ubifs_leb_write(struct ubifs_info *c, int lnum, const void *buf, int offs,
		    int len, int dtype)
{
	int err;

	ubifs_assert(!c->ro_media && !c->ro_mount);
	if (c->ro_error)
		return -EROFS;
	if (!dbg_is_tst_rcvry(c))
		err = ubi_leb_write(c->ubi, lnum, buf, offs, len, dtype);
	else
		err = dbg_leb_write(c->ubi, lnum, buf, offs, len, dtype);
	if (err) {
		ubifs_err("writing %d bytes to LEB %d:%d failed, error %d",
			  len, lnum, offs, err);
		ubifs_ro_mode(c, err);
		dbg_dump_stack();
	}
	return err;
}

int ubifs_leb_change(struct ubifs_info *c, int lnum, const void *buf, int len,
		     int dtype)
{
	int err;

	ubifs_assert(!c->ro_media && !c->ro_mount);
	if (c->ro_error)
		return -EROFS;
	if (!dbg_is_tst_rcvry(c))
		err = ubi_leb_change(c->ubi, lnum, buf, len, dtype);
	else
		err = dbg_leb_change(c->ubi, lnum, buf, len, dtype);
	if (err) {
		ubifs_err("changing %d bytes in LEB %d failed, error %d",
			  len, lnum, err);
		ubifs_ro_mode(c, err);
		dbg_dump_stack();
	}
	return err;
}

int ubifs_leb_unmap(struct ubifs_info *c, int lnum)
{
	int err;

	ubifs_assert(!c->ro_media && !c->ro_mount);
	if (c->ro_error)
		return -EROFS;
	if (!dbg_is_tst_rcvry(c))
		err = ubi_leb_unmap(c->ubi, lnum);
	else
		err = dbg_leb_unmap(c->ubi, lnum);
	if (err) {
		ubifs_err("unmap LEB %d failed, error %d", lnum, err);
		ubifs_ro_mode(c, err);
		dbg_dump_stack();
	}
	return err;
}

int ubifs_leb_map(struct ubifs_info *c, int lnum, int dtype)
{
	int err;

	ubifs_assert(!c->ro_media && !c->ro_mount);
	if (c->ro_error)
		return -EROFS;
	if (!dbg_is_tst_rcvry(c))
		err = ubi_leb_map(c->ubi, lnum, dtype);
	else
		err = dbg_leb_map(c->ubi, lnum, dtype);
	if (err) {
		ubifs_err("mapping LEB %d failed, error %d", lnum, err);
		ubifs_ro_mode(c, err);
		dbg_dump_stack();
	}
	return err;
}

int ubifs_is_mapped(const struct ubifs_info *c, int lnum)
{
	int err;

	err = ubi_is_mapped(c->ubi, lnum);
	if (err < 0) {
		ubifs_err("ubi_is_mapped failed for LEB %d, error %d",
			  lnum, err);
		dbg_dump_stack();
	}
	return err;
}

/**
 * ubifs_check_node - check node.
 * @c: UBIFS file-system description object
 * @buf: node to check
 * @lnum: logical eraseblock number
 * @offs: offset within the logical eraseblock
 * @quiet: print no messages
 * @must_chk_crc: indicates whether to always check the CRC
 *
 * This function checks node magic number and CRC checksum. This function also
 * validates node length to prevent UBIFS from becoming crazy when an attacker
 * feeds it a file-system image with incorrect nodes. For example, too large
 * node length in the common header could cause UBIFS to read memory outside of
 * allocated buffer when checking the CRC checksum.
 *
 * This function may skip data nodes CRC checking if @c->no_chk_data_crc is
 * true, which is controlled by corresponding UBIFS mount option. However, if
 * @must_chk_crc is true, then @c->no_chk_data_crc is ignored and CRC is
 * checked. Similarly, if @c->mounting or @c->remounting_rw is true (we are
 * mounting or re-mounting to R/W mode), @c->no_chk_data_crc is ignored and CRC
 * is checked. This is because during mounting or re-mounting from R/O mode to
 * R/W mode we may read journal nodes (when replying the journal or doing the
 * recovery) and the journal nodes may potentially be corrupted, so checking is
 * required.
 *
 * This function returns zero in case of success and %-EUCLEAN in case of bad
 * CRC or magic.
 */
int ubifs_check_node(const struct ubifs_info *c, const void *buf, int lnum,
		     int offs, int quiet, int must_chk_crc)
{
	int err = -EINVAL, type, node_len;
	uint32_t crc, node_crc, magic;
	const struct ubifs_ch *ch = buf;

	ubifs_assert(lnum >= 0 && lnum < c->leb_cnt && offs >= 0);
	ubifs_assert(!(offs & 7) && offs < c->leb_size);

	magic = le32_to_cpu(ch->magic);
	if (magic != UBIFS_NODE_MAGIC) {
		if (!quiet)
			ubifs_err("bad magic %#08x, expected %#08x",
				  magic, UBIFS_NODE_MAGIC);
		err = -EUCLEAN;
		goto out;
	}

	type = ch->node_type;
	if (type < 0 || type >= UBIFS_NODE_TYPES_CNT) {
		if (!quiet)
			ubifs_err("bad node type %d", type);
		goto out;
	}

	node_len = le32_to_cpu(ch->len);
	if (node_len + offs > c->leb_size)
		goto out_len;

	if (c->ranges[type].max_len == 0) {
		if (node_len != c->ranges[type].len)
			goto out_len;
	} else if (node_len < c->ranges[type].min_len ||
		   node_len > c->ranges[type].max_len)
		goto out_len;

	if (!must_chk_crc && type == UBIFS_DATA_NODE && !c->mounting &&
	    !c->remounting_rw && c->no_chk_data_crc)
		return 0;

	crc = crc32(UBIFS_CRC32_INIT, buf + 8, node_len - 8);
	node_crc = le32_to_cpu(ch->crc);
	if (crc != node_crc) {
		if (!quiet)
			ubifs_err("bad CRC: calculated %#08x, read %#08x",
				  crc, node_crc);
		err = -EUCLEAN;
		goto out;
	}

	return 0;

out_len:
	if (!quiet)
		ubifs_err("bad node length %d", node_len);
out:
	if (!quiet) {
		ubifs_err("bad node at LEB %d:%d", lnum, offs);
		dbg_dump_node(c, buf);
		dbg_dump_stack();
	}
	return err;
}

/**
 * ubifs_pad - pad flash space.
 * @c: UBIFS file-system description object
 * @buf: buffer to put padding to
 * @pad: how many bytes to pad
 *
 * The flash media obliges us to write only in chunks of %c->min_io_size and
 * when we have to write less data we add padding node to the write-buffer and
 * pad it to the next minimal I/O unit's boundary. Padding nodes help when the
 * media is being scanned. If the amount of wasted space is not enough to fit a
 * padding node which takes %UBIFS_PAD_NODE_SZ bytes, we write padding bytes
 * pattern (%UBIFS_PADDING_BYTE).
 *
 * Padding nodes are also used to fill gaps when the "commit-in-gaps" method is
 * used.
 */
void ubifs_pad(const struct ubifs_info *c, void *buf, int pad)
{
	uint32_t crc;

	ubifs_assert(pad >= 0 && !(pad & 7));

	if (pad >= UBIFS_PAD_NODE_SZ) {
		struct ubifs_ch *ch = buf;
		struct ubifs_pad_node *pad_node = buf;

		ch->magic = cpu_to_le32(UBIFS_NODE_MAGIC);
		ch->node_type = UBIFS_PAD_NODE;
		ch->group_type = UBIFS_NO_NODE_GROUP;
		ch->padding[0] = ch->padding[1] = 0;
		ch->sqnum = 0;
		ch->len = cpu_to_le32(UBIFS_PAD_NODE_SZ);
		pad -= UBIFS_PAD_NODE_SZ;
		pad_node->pad_len = cpu_to_le32(pad);
		crc = crc32(UBIFS_CRC32_INIT, buf + 8, UBIFS_PAD_NODE_SZ - 8);
		ch->crc = cpu_to_le32(crc);
		memset(buf + UBIFS_PAD_NODE_SZ, 0, pad);
	} else if (pad > 0)
		/* Too little space, padding node won't fit */
		memset(buf, UBIFS_PADDING_BYTE, pad);
}

/**
 * next_sqnum - get next sequence number.
 * @c: UBIFS file-system description object
 */
static unsigned long long next_sqnum(struct ubifs_info *c)
{
	unsigned long long sqnum;

	spin_lock(&c->cnt_lock);
	sqnum = ++c->max_sqnum;
	spin_unlock(&c->cnt_lock);

	if (unlikely(sqnum >= SQNUM_WARN_WATERMARK)) {
		if (sqnum >= SQNUM_WATERMARK) {
			ubifs_err("sequence number overflow %llu, end of life",
				  sqnum);
			ubifs_ro_mode(c, -EINVAL);
		}
		ubifs_warn("running out of sequence numbers, end of life soon");
	}

	return sqnum;
}

/**
 * ubifs_prepare_node - prepare node to be written to flash.
 * @c: UBIFS file-system description object
 * @node: the node to pad
 * @len: node length
 * @pad: if the buffer has to be padded
 *
 * This function prepares node at @node to be written to the media - it
 * calculates node CRC, fills the common header, and adds proper padding up to
 * the next minimum I/O unit if @pad is not zero.
 */
void ubifs_prepare_node(struct ubifs_info *c, void *node, int len, int pad)
{
	uint32_t crc;
	struct ubifs_ch *ch = node;
	unsigned long long sqnum = next_sqnum(c);

	ubifs_assert(len >= UBIFS_CH_SZ);

	ch->magic = cpu_to_le32(UBIFS_NODE_MAGIC);
	ch->len = cpu_to_le32(len);
	ch->group_type = UBIFS_NO_NODE_GROUP;
	ch->sqnum = cpu_to_le64(sqnum);
	ch->padding[0] = ch->padding[1] = 0;
	crc = crc32(UBIFS_CRC32_INIT, node + 8, len - 8);
	ch->crc = cpu_to_le32(crc);

	if (pad) {
		len = ALIGN(len, 8);
		pad = ALIGN(len, c->min_io_size) - len;
		ubifs_pad(c, node + len, pad);
	}
}

/**
 * ubifs_prep_grp_node - prepare node of a group to be written to flash.
 * @c: UBIFS file-system description object
 * @node: the node to pad
 * @len: node length
 * @last: indicates the last node of the group
 *
 * This function prepares node at @node to be written to the media - it
 * calculates node CRC and fills the common header.
 */
void ubifs_prep_grp_node(struct ubifs_info *c, void *node, int len, int last)
{
	uint32_t crc;
	struct ubifs_ch *ch = node;
	unsigned long long sqnum = next_sqnum(c);

	ubifs_assert(len >= UBIFS_CH_SZ);

	ch->magic = cpu_to_le32(UBIFS_NODE_MAGIC);
	ch->len = cpu_to_le32(len);
	if (last)
		ch->group_type = UBIFS_LAST_OF_NODE_GROUP;
	else
		ch->group_type = UBIFS_IN_NODE_GROUP;
	ch->sqnum = cpu_to_le64(sqnum);
	ch->padding[0] = ch->padding[1] = 0;
	crc = crc32(UBIFS_CRC32_INIT, node + 8, len - 8);
	ch->crc = cpu_to_le32(crc);
}

/**
 * wbuf_timer_callback - write-buffer timer callback function.
 * @data: timer data (write-buffer descriptor)
 *
 * This function is called when the write-buffer timer expires.
 */
static enum hrtimer_restart wbuf_timer_callback_nolock(struct hrtimer *timer)
{
	struct ubifs_wbuf *wbuf = container_of(timer, struct ubifs_wbuf, timer);

	dbg_io("jhead %s", dbg_jhead(wbuf->jhead));
	wbuf->need_sync = 1;
	wbuf->c->need_wbuf_sync = 1;
	ubifs_wake_up_bgt(wbuf->c);
	return HRTIMER_NORESTART;
}

/**
 * new_wbuf_timer - start new write-buffer timer.
 * @wbuf: write-buffer descriptor
 */
static void new_wbuf_timer_nolock(struct ubifs_wbuf *wbuf)
{
	ubifs_assert(!hrtimer_active(&wbuf->timer));

	if (wbuf->no_timer)
		return;
	dbg_io("set timer for jhead %s, %llu-%llu millisecs",
	       dbg_jhead(wbuf->jhead),
	       div_u64(ktime_to_ns(wbuf->softlimit), USEC_PER_SEC),
	       div_u64(ktime_to_ns(wbuf->softlimit) + wbuf->delta,
		       USEC_PER_SEC));
	hrtimer_start_range_ns(&wbuf->timer, wbuf->softlimit, wbuf->delta,
			       HRTIMER_MODE_REL);
}

/**
 * cancel_wbuf_timer - cancel write-buffer timer.
 * @wbuf: write-buffer descriptor
 */
static void cancel_wbuf_timer_nolock(struct ubifs_wbuf *wbuf)
{
	if (wbuf->no_timer)
		return;
	wbuf->need_sync = 0;
	hrtimer_cancel(&wbuf->timer);
}

/**
 * ubifs_wbuf_sync_nolock - synchronize write-buffer.
 * @wbuf: write-buffer to synchronize
 *
 * This function synchronizes write-buffer @buf and returns zero in case of
 * success or a negative error code in case of failure.
 *
 * Note, although write-buffers are of @c->max_write_size, this function does
 * not necessarily writes all @c->max_write_size bytes to the flash. Instead,
 * if the write-buffer is only partially filled with data, only the used part
 * of the write-buffer (aligned on @c->min_io_size boundary) is synchronized.
 * This way we waste less space.
 */
int ubifs_wbuf_sync_nolock(struct ubifs_wbuf *wbuf)
{
	struct ubifs_info *c = wbuf->c;
	int err, dirt, sync_len;

	cancel_wbuf_timer_nolock(wbuf);
	if (!wbuf->used || wbuf->lnum == -1)
		/* Write-buffer is empty or not seeked */
		return 0;

	dbg_io("LEB %d:%d, %d bytes, jhead %s",
	       wbuf->lnum, wbuf->offs, wbuf->used, dbg_jhead(wbuf->jhead));
	ubifs_assert(!(wbuf->avail & 7));
	ubifs_assert(wbuf->offs + wbuf->size <= c->leb_size);
	ubifs_assert(wbuf->size >= c->min_io_size);
	ubifs_assert(wbuf->size <= c->max_write_size);
	ubifs_assert(wbuf->size % c->min_io_size == 0);
	ubifs_assert(!c->ro_media && !c->ro_mount);
	if (c->leb_size - wbuf->offs >= c->max_write_size)
		ubifs_assert(!((wbuf->offs + wbuf->size) % c->max_write_size));

	if (c->ro_error)
		return -EROFS;

	/*
	 * Do not write whole write buffer but write only the minimum necessary
	 * amount of min. I/O units.
	 */
	sync_len = ALIGN(wbuf->used, c->min_io_size);
	dirt = sync_len - wbuf->used;
	if (dirt)
		ubifs_pad(c, wbuf->buf + wbuf->used, dirt);
	err = ubifs_leb_write(c, wbuf->lnum, wbuf->buf, wbuf->offs, sync_len,
			      wbuf->dtype);
	if (err)
		return err;

	spin_lock(&wbuf->lock);
	wbuf->offs += sync_len;
	/*
	 * Now @wbuf->offs is not necessarily aligned to @c->max_write_size.
	 * But our goal is to optimize writes and make sure we write in
	 * @c->max_write_size chunks and to @c->max_write_size-aligned offset.
	 * Thus, if @wbuf->offs is not aligned to @c->max_write_size now, make
	 * sure that @wbuf->offs + @wbuf->size is aligned to
	 * @c->max_write_size. This way we make sure that after next
	 * write-buffer flush we are again at the optimal offset (aligned to
	 * @c->max_write_size).
	 */
	if (c->leb_size - wbuf->offs < c->max_write_size)
		wbuf->size = c->leb_size - wbuf->offs;
	else if (wbuf->offs & (c->max_write_size - 1))
		wbuf->size = ALIGN(wbuf->offs, c->max_write_size) - wbuf->offs;
	else
		wbuf->size = c->max_write_size;
	wbuf->avail = wbuf->size;
	wbuf->used = 0;
	wbuf->next_ino = 0;
	spin_unlock(&wbuf->lock);

	if (wbuf->sync_callback)
		err = wbuf->sync_callback(c, wbuf->lnum,
					  c->leb_size - wbuf->offs, dirt);
	return err;
}

/**
 * ubifs_wbuf_seek_nolock - seek write-buffer.
 * @wbuf: write-buffer
 * @lnum: logical eraseblock number to seek to
 * @offs: logical eraseblock offset to seek to
 * @dtype: data type
 *
 * This function targets the write-buffer to logical eraseblock @lnum:@offs.
 * The write-buffer has to be empty. Returns zero in case of success and a
 * negative error code in case of failure.
 */
int ubifs_wbuf_seek_nolock(struct ubifs_wbuf *wbuf, int lnum, int offs,
			   int dtype)
{
	const struct ubifs_info *c = wbuf->c;

	dbg_io("LEB %d:%d, jhead %s", lnum, offs, dbg_jhead(wbuf->jhead));
	ubifs_assert(lnum >= 0 && lnum < c->leb_cnt);
	ubifs_assert(offs >= 0 && offs <= c->leb_size);
	ubifs_assert(offs % c->min_io_size == 0 && !(offs & 7));
	ubifs_assert(lnum != wbuf->lnum);
	ubifs_assert(wbuf->used == 0);

	spin_lock(&wbuf->lock);
	wbuf->lnum = lnum;
	wbuf->offs = offs;
	if (c->leb_size - wbuf->offs < c->max_write_size)
		wbuf->size = c->leb_size - wbuf->offs;
	else if (wbuf->offs & (c->max_write_size - 1))
		wbuf->size = ALIGN(wbuf->offs, c->max_write_size) - wbuf->offs;
	else
		wbuf->size = c->max_write_size;
	wbuf->avail = wbuf->size;
	wbuf->used = 0;
	spin_unlock(&wbuf->lock);
	wbuf->dtype = dtype;

	return 0;
}

/**
 * ubifs_bg_wbufs_sync - synchronize write-buffers.
 * @c: UBIFS file-system description object
 *
 * This function is called by background thread to synchronize write-buffers.
 * Returns zero in case of success and a negative error code in case of
 * failure.
 */
int ubifs_bg_wbufs_sync(struct ubifs_info *c)
{
	int err, i;

	ubifs_assert(!c->ro_media && !c->ro_mount);
	if (!c->need_wbuf_sync)
		return 0;
	c->need_wbuf_sync = 0;

	if (c->ro_error) {
		err = -EROFS;
		goto out_timers;
	}

	dbg_io("synchronize");
	for (i = 0; i < c->jhead_cnt; i++) {
		struct ubifs_wbuf *wbuf = &c->jheads[i].wbuf;

		cond_resched();

		/*
		 * If the mutex is locked then wbuf is being changed, so
		 * synchronization is not necessary.
		 */
		if (mutex_is_locked(&wbuf->io_mutex))
			continue;

		mutex_lock_nested(&wbuf->io_mutex, wbuf->jhead);
		if (!wbuf->need_sync) {
			mutex_unlock(&wbuf->io_mutex);
			continue;
		}

		err = ubifs_wbuf_sync_nolock(wbuf);
		mutex_unlock(&wbuf->io_mutex);
		if (err) {
			ubifs_err("cannot sync write-buffer, error %d", err);
			ubifs_ro_mode(c, err);
			goto out_timers;
		}
	}

	return 0;

out_timers:
	/* Cancel all timers to prevent repeated errors */
	for (i = 0; i < c->jhead_cnt; i++) {
		struct ubifs_wbuf *wbuf = &c->jheads[i].wbuf;

		mutex_lock_nested(&wbuf->io_mutex, wbuf->jhead);
		cancel_wbuf_timer_nolock(wbuf);
		mutex_unlock(&wbuf->io_mutex);
	}
	return err;
}

/**
 * ubifs_wbuf_write_nolock - write data to flash via write-buffer.
 * @wbuf: write-buffer
 * @buf: node to write
 * @len: node length
 *
 * This function writes data to flash via write-buffer @wbuf. This means that
 * the last piece of the node won't reach the flash media immediately if it
 * does not take whole max. write unit (@c->max_write_size). Instead, the node
 * will sit in RAM until the write-buffer is synchronized (e.g., by timer, or
 * because more data are appended to the write-buffer).
 *
 * This function returns zero in case of success and a negative error code in
 * case of failure. If the node cannot be written because there is no more
 * space in this logical eraseblock, %-ENOSPC is returned.
 */
int ubifs_wbuf_write_nolock(struct ubifs_wbuf *wbuf, void *buf, int len)
{
	struct ubifs_info *c = wbuf->c;
	int err, written, n, aligned_len = ALIGN(len, 8);

	dbg_io("%d bytes (%s) to jhead %s wbuf at LEB %d:%d", len,
	       dbg_ntype(((struct ubifs_ch *)buf)->node_type),
	       dbg_jhead(wbuf->jhead), wbuf->lnum, wbuf->offs + wbuf->used);
	ubifs_assert(len > 0 && wbuf->lnum >= 0 && wbuf->lnum < c->leb_cnt);
	ubifs_assert(wbuf->offs >= 0 && wbuf->offs % c->min_io_size == 0);
	ubifs_assert(!(wbuf->offs & 7) && wbuf->offs <= c->leb_size);
	ubifs_assert(wbuf->avail > 0 && wbuf->avail <= wbuf->size);
	ubifs_assert(wbuf->size >= c->min_io_size);
	ubifs_assert(wbuf->size <= c->max_write_size);
	ubifs_assert(wbuf->size % c->min_io_size == 0);
	ubifs_assert(mutex_is_locked(&wbuf->io_mutex));
	ubifs_assert(!c->ro_media && !c->ro_mount);
	ubifs_assert(!c->space_fixup);
	if (c->leb_size - wbuf->offs >= c->max_write_size)
		ubifs_assert(!((wbuf->offs + wbuf->size) % c->max_write_size));

	if (c->leb_size - wbuf->offs - wbuf->used < aligned_len) {
		err = -ENOSPC;
		goto out;
	}

	cancel_wbuf_timer_nolock(wbuf);

	if (c->ro_error)
		return -EROFS;

	if (aligned_len <= wbuf->avail) {
		/*
		 * The node is not very large and fits entirely within
		 * write-buffer.
		 */
		memcpy(wbuf->buf + wbuf->used, buf, len);

		if (aligned_len == wbuf->avail) {
			dbg_io("flush jhead %s wbuf to LEB %d:%d",
			       dbg_jhead(wbuf->jhead), wbuf->lnum, wbuf->offs);
			err = ubifs_leb_write(c, wbuf->lnum, wbuf->buf,
					      wbuf->offs, wbuf->size,
					      wbuf->dtype);
			if (err)
				goto out;

			spin_lock(&wbuf->lock);
			wbuf->offs += wbuf->size;
			if (c->leb_size - wbuf->offs >= c->max_write_size)
				wbuf->size = c->max_write_size;
			else
				wbuf->size = c->leb_size - wbuf->offs;
			wbuf->avail = wbuf->size;
			wbuf->used = 0;
			wbuf->next_ino = 0;
			spin_unlock(&wbuf->lock);
		} else {
			spin_lock(&wbuf->lock);
			wbuf->avail -= aligned_len;
			wbuf->used += aligned_len;
			spin_unlock(&wbuf->lock);
		}

		goto exit;
	}

	written = 0;

	if (wbuf->used) {
		/*
		 * The node is large enough and does not fit entirely within
		 * current available space. We have to fill and flush
		 * write-buffer and switch to the next max. write unit.
		 */
		dbg_io("flush jhead %s wbuf to LEB %d:%d",
		       dbg_jhead(wbuf->jhead), wbuf->lnum, wbuf->offs);
		memcpy(wbuf->buf + wbuf->used, buf, wbuf->avail);
		err = ubifs_leb_write(c, wbuf->lnum, wbuf->buf, wbuf->offs,
				      wbuf->size, wbuf->dtype);
		if (err)
			goto out;

		wbuf->offs += wbuf->size;
		len -= wbuf->avail;
		aligned_len -= wbuf->avail;
		written += wbuf->avail;
	} else if (wbuf->offs & (c->max_write_size - 1)) {
		/*
		 * The write-buffer offset is not aligned to
		 * @c->max_write_size and @wbuf->size is less than
		 * @c->max_write_size. Write @wbuf->size bytes to make sure the
		 * following writes are done in optimal @c->max_write_size
		 * chunks.
		 */
		dbg_io("write %d bytes to LEB %d:%d",
		       wbuf->size, wbuf->lnum, wbuf->offs);
		err = ubifs_leb_write(c, wbuf->lnum, buf, wbuf->offs,
				      wbuf->size, wbuf->dtype);
		if (err)
			goto out;

		wbuf->offs += wbuf->size;
		len -= wbuf->size;
		aligned_len -= wbuf->size;
		written += wbuf->size;
	}

	/*
	 * The remaining data may take more whole max. write units, so write the
	 * remains multiple to max. write unit size directly to the flash media.
	 * We align node length to 8-byte boundary because we anyway flash wbuf
	 * if the remaining space is less than 8 bytes.
	 */
	n = aligned_len >> c->max_write_shift;
	if (n) {
		n <<= c->max_write_shift;
		dbg_io("write %d bytes to LEB %d:%d", n, wbuf->lnum,
		       wbuf->offs);
		err = ubifs_leb_write(c, wbuf->lnum, buf + written,
				      wbuf->offs, n, wbuf->dtype);
		if (err)
			goto out;
		wbuf->offs += n;
		aligned_len -= n;
		len -= n;
		written += n;
	}

	spin_lock(&wbuf->lock);
	if (aligned_len)
		/*
		 * And now we have what's left and what does not take whole
		 * max. write unit, so write it to the write-buffer and we are
		 * done.
		 */
		memcpy(wbuf->buf, buf + written, len);

	if (c->leb_size - wbuf->offs >= c->max_write_size)
		wbuf->size = c->max_write_size;
	else
		wbuf->size = c->leb_size - wbuf->offs;
	wbuf->avail = wbuf->size - aligned_len;
	wbuf->used = aligned_len;
	wbuf->next_ino = 0;
	spin_unlock(&wbuf->lock);

exit:
	if (wbuf->sync_callback) {
		int free = c->leb_size - wbuf->offs - wbuf->used;

		err = wbuf->sync_callback(c, wbuf->lnum, free, 0);
		if (err)
			goto out;
	}

	if (wbuf->used)
		new_wbuf_timer_nolock(wbuf);

	return 0;

out:
	ubifs_err("cannot write %d bytes to LEB %d:%d, error %d",
		  len, wbuf->lnum, wbuf->offs, err);
	dbg_dump_node(c, buf);
	dbg_dump_stack();
	dbg_dump_leb(c, wbuf->lnum);
	return err;
}

/**
 * ubifs_write_node - write node to the media.
 * @c: UBIFS file-system description object
 * @buf: the node to write
 * @len: node length
 * @lnum: logical eraseblock number
 * @offs: offset within the logical eraseblock
 * @dtype: node life-time hint (%UBI_LONGTERM, %UBI_SHORTTERM, %UBI_UNKNOWN)
 *
 * This function automatically fills node magic number, assigns sequence
 * number, and calculates node CRC checksum. The length of the @buf buffer has
 * to be aligned to the minimal I/O unit size. This function automatically
 * appends padding node and padding bytes if needed. Returns zero in case of
 * success and a negative error code in case of failure.
 */
int ubifs_write_node(struct ubifs_info *c, void *buf, int len, int lnum,
		     int offs, int dtype)
{
	int err, buf_len = ALIGN(len, c->min_io_size);

	dbg_io("LEB %d:%d, %s, length %d (aligned %d)",
	       lnum, offs, dbg_ntype(((struct ubifs_ch *)buf)->node_type), len,
	       buf_len);
	ubifs_assert(lnum >= 0 && lnum < c->leb_cnt && offs >= 0);
	ubifs_assert(offs % c->min_io_size == 0 && offs < c->leb_size);
	ubifs_assert(!c->ro_media && !c->ro_mount);
	ubifs_assert(!c->space_fixup);

	if (c->ro_error)
		return -EROFS;

	ubifs_prepare_node(c, buf, len, 1);
	err = ubifs_leb_write(c, lnum, buf, offs, buf_len, dtype);
	if (err)
		dbg_dump_node(c, buf);

	return err;
}

/**
 * ubifs_read_node_wbuf - read node from the media or write-buffer.
 * @wbuf: wbuf to check for un-written data
 * @buf: buffer to read to
 * @type: node type
 * @len: node length
 * @lnum: logical eraseblock number
 * @offs: offset within the logical eraseblock
 *
 * This function reads a node of known type and length, checks it and stores
 * in @buf. If the node partially or fully sits in the write-buffer, this
 * function takes data from the buffer, otherwise it reads the flash media.
 * Returns zero in case of success, %-EUCLEAN if CRC mismatched and a negative
 * error code in case of failure.
 */
int ubifs_read_node_wbuf(struct ubifs_wbuf *wbuf, void *buf, int type, int len,
			 int lnum, int offs)
{
	const struct ubifs_info *c = wbuf->c;
	int err, rlen, overlap;
	struct ubifs_ch *ch = buf;

	dbg_io("LEB %d:%d, %s, length %d, jhead %s", lnum, offs,
	       dbg_ntype(type), len, dbg_jhead(wbuf->jhead));
	ubifs_assert(wbuf && lnum >= 0 && lnum < c->leb_cnt && offs >= 0);
	ubifs_assert(!(offs & 7) && offs < c->leb_size);
	ubifs_assert(type >= 0 && type < UBIFS_NODE_TYPES_CNT);

	spin_lock(&wbuf->lock);
	overlap = (lnum == wbuf->lnum && offs + len > wbuf->offs);
	if (!overlap) {
		/* We may safely unlock the write-buffer and read the data */
		spin_unlock(&wbuf->lock);
		return ubifs_read_node(c, buf, type, len, lnum, offs);
	}

	/* Don't read under wbuf */
	rlen = wbuf->offs - offs;
	if (rlen < 0)
		rlen = 0;

	/* Copy the rest from the write-buffer */
	memcpy(buf + rlen, wbuf->buf + offs + rlen - wbuf->offs, len - rlen);
	spin_unlock(&wbuf->lock);

	if (rlen > 0) {
		/* Read everything that goes before write-buffer */
		err = ubifs_leb_read(c, lnum, buf, offs, rlen, 0);
		if (err && err != -EBADMSG)
			return err;
	}

	if (type != ch->node_type) {
		ubifs_err("bad node type (%d but expected %d)",
			  ch->node_type, type);
		goto out;
	}

	err = ubifs_check_node(c, buf, lnum, offs, 0, 0);
	if (err) {
		ubifs_err("expected node type %d", type);
		return err;
	}

	rlen = le32_to_cpu(ch->len);
	if (rlen != len) {
		ubifs_err("bad node length %d, expected %d", rlen, len);
		goto out;
	}

	return 0;

out:
	ubifs_err("bad node at LEB %d:%d", lnum, offs);
	dbg_dump_node(c, buf);
	dbg_dump_stack();
	return -EINVAL;
}

/**
 * ubifs_read_node - read node.
 * @c: UBIFS file-system description object
 * @buf: buffer to read to
 * @type: node type
 * @len: node length (not aligned)
 * @lnum: logical eraseblock number
 * @offs: offset within the logical eraseblock
 *
 * This function reads a node of known type and and length, checks it and
 * stores in @buf. Returns zero in case of success, %-EUCLEAN if CRC mismatched
 * and a negative error code in case of failure.
 */
int ubifs_read_node(const struct ubifs_info *c, void *buf, int type, int len,
		    int lnum, int offs)
{
	int err, l;
	struct ubifs_ch *ch = buf;

	dbg_io("LEB %d:%d, %s, length %d", lnum, offs, dbg_ntype(type), len);
	ubifs_assert(lnum >= 0 && lnum < c->leb_cnt && offs >= 0);
	ubifs_assert(len >= UBIFS_CH_SZ && offs + len <= c->leb_size);
	ubifs_assert(!(offs & 7) && offs < c->leb_size);
	ubifs_assert(type >= 0 && type < UBIFS_NODE_TYPES_CNT);

	err = ubifs_leb_read(c, lnum, buf, offs, len, 0);
	if (err && err != -EBADMSG)
		return err;

	if (type != ch->node_type) {
		ubifs_err("bad node type (%d but expected %d)",
			  ch->node_type, type);
		goto out;
	}

	err = ubifs_check_node(c, buf, lnum, offs, 0, 0);
	if (err) {
		ubifs_err("expected node type %d", type);
		return err;
	}

	l = le32_to_cpu(ch->len);
	if (l != len) {
		ubifs_err("bad node length %d, expected %d", l, len);
		goto out;
	}

	return 0;

out:
	ubifs_err("bad node at LEB %d:%d, LEB mapping status %d", lnum, offs,
		  ubi_is_mapped(c->ubi, lnum));
	dbg_dump_node(c, buf);
	dbg_dump_stack();
	return -EINVAL;
}

/**
 * ubifs_wbuf_init - initialize write-buffer.
 * @c: UBIFS file-system description object
 * @wbuf: write-buffer to initialize
 *
 * This function initializes write-buffer. Returns zero in case of success
 * %-ENOMEM in case of failure.
 */
int ubifs_wbuf_init(struct ubifs_info *c, struct ubifs_wbuf *wbuf)
{
	size_t size;

	wbuf->buf = kmalloc(c->max_write_size, GFP_KERNEL);
	if (!wbuf->buf)
		return -ENOMEM;

	size = (c->max_write_size / UBIFS_CH_SZ + 1) * sizeof(ino_t);
	wbuf->inodes = kmalloc(size, GFP_KERNEL);
	if (!wbuf->inodes) {
		kfree(wbuf->buf);
		wbuf->buf = NULL;
		return -ENOMEM;
	}

	wbuf->used = 0;
	wbuf->lnum = wbuf->offs = -1;
	/*
	 * If the LEB starts at the max. write size aligned address, then
	 * write-buffer size has to be set to @c->max_write_size. Otherwise,
	 * set it to something smaller so that it ends at the closest max.
	 * write size boundary.
	 */
	size = c->max_write_size - (c->leb_start % c->max_write_size);
	wbuf->avail = wbuf->size = size;
	wbuf->dtype = UBI_UNKNOWN;
	wbuf->sync_callback = NULL;
	mutex_init(&wbuf->io_mutex);
	spin_lock_init(&wbuf->lock);
	wbuf->c = c;
	wbuf->next_ino = 0;

	hrtimer_init(&wbuf->timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
	wbuf->timer.function = wbuf_timer_callback_nolock;
	wbuf->softlimit = ktime_set(WBUF_TIMEOUT_SOFTLIMIT, 0);
	wbuf->delta = WBUF_TIMEOUT_HARDLIMIT - WBUF_TIMEOUT_SOFTLIMIT;
	wbuf->delta *= 1000000000ULL;
	ubifs_assert(wbuf->delta <= ULONG_MAX);
	return 0;
}

/**
 * ubifs_wbuf_add_ino_nolock - add an inode number into the wbuf inode array.
 * @wbuf: the write-buffer where to add
 * @inum: the inode number
 *
 * This function adds an inode number to the inode array of the write-buffer.
 */
void ubifs_wbuf_add_ino_nolock(struct ubifs_wbuf *wbuf, ino_t inum)
{
	if (!wbuf->buf)
		/* NOR flash or something similar */
		return;

	spin_lock(&wbuf->lock);
	if (wbuf->used)
		wbuf->inodes[wbuf->next_ino++] = inum;
	spin_unlock(&wbuf->lock);
}

/**
 * wbuf_has_ino - returns if the wbuf contains data from the inode.
 * @wbuf: the write-buffer
 * @inum: the inode number
 *
 * This function returns with %1 if the write-buffer contains some data from the
 * given inode otherwise it returns with %0.
 */
static int wbuf_has_ino(struct ubifs_wbuf *wbuf, ino_t inum)
{
	int i, ret = 0;

	spin_lock(&wbuf->lock);
	for (i = 0; i < wbuf->next_ino; i++)
		if (inum == wbuf->inodes[i]) {
			ret = 1;
			break;
		}
	spin_unlock(&wbuf->lock);

	return ret;
}

/**
 * ubifs_sync_wbufs_by_inode - synchronize write-buffers for an inode.
 * @c: UBIFS file-system description object
 * @inode: inode to synchronize
 *
 * This function synchronizes write-buffers which contain nodes belonging to
 * @inode. Returns zero in case of success and a negative error code in case of
 * failure.
 */
int ubifs_sync_wbufs_by_inode(struct ubifs_info *c, struct inode *inode)
{
	int i, err = 0;

	for (i = 0; i < c->jhead_cnt; i++) {
		struct ubifs_wbuf *wbuf = &c->jheads[i].wbuf;

		if (i == GCHD)
			/*
			 * GC head is special, do not look at it. Even if the
			 * head contains something related to this inode, it is
			 * a _copy_ of corresponding on-flash node which sits
			 * somewhere else.
			 */
			continue;

		if (!wbuf_has_ino(wbuf, inode->i_ino))
			continue;

		mutex_lock_nested(&wbuf->io_mutex, wbuf->jhead);
		if (wbuf_has_ino(wbuf, inode->i_ino))
			err = ubifs_wbuf_sync_nolock(wbuf);
		mutex_unlock(&wbuf->io_mutex);

		if (err) {
			ubifs_ro_mode(c, err);
			return err;
		}
	}
	return 0;
}