aboutsummaryrefslogtreecommitdiff
path: root/fs/ocfs2/file.c
blob: b32cdb3bf7c5abe7eeb5f9f9ac0f93049a52a258 (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
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
/* -*- mode: c; c-basic-offset: 8; -*-
 * vim: noexpandtab sw=8 ts=8 sts=0:
 *
 * file.c
 *
 * File open, close, extend, truncate
 *
 * Copyright (C) 2002, 2004 Oracle.  All rights reserved.
 *
 * 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.
 *
 * 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., 59 Temple Place - Suite 330,
 * Boston, MA 021110-1307, USA.
 */

#include <linux/capability.h>
#include <linux/fs.h>
#include <linux/types.h>
#include <linux/slab.h>
#include <linux/highmem.h>
#include <linux/pagemap.h>
#include <linux/uio.h>
#include <linux/sched.h>
#include <linux/pipe_fs_i.h>

#define MLOG_MASK_PREFIX ML_INODE
#include <cluster/masklog.h>

#include "ocfs2.h"

#include "alloc.h"
#include "aops.h"
#include "dir.h"
#include "dlmglue.h"
#include "extent_map.h"
#include "file.h"
#include "sysfile.h"
#include "inode.h"
#include "ioctl.h"
#include "journal.h"
#include "mmap.h"
#include "suballoc.h"
#include "super.h"

#include "buffer_head_io.h"

static int ocfs2_sync_inode(struct inode *inode)
{
	filemap_fdatawrite(inode->i_mapping);
	return sync_mapping_buffers(inode->i_mapping);
}

static int ocfs2_file_open(struct inode *inode, struct file *file)
{
	int status;
	int mode = file->f_flags;
	struct ocfs2_inode_info *oi = OCFS2_I(inode);

	mlog_entry("(0x%p, 0x%p, '%.*s')\n", inode, file,
		   file->f_dentry->d_name.len, file->f_dentry->d_name.name);

	spin_lock(&oi->ip_lock);

	/* Check that the inode hasn't been wiped from disk by another
	 * node. If it hasn't then we're safe as long as we hold the
	 * spin lock until our increment of open count. */
	if (OCFS2_I(inode)->ip_flags & OCFS2_INODE_DELETED) {
		spin_unlock(&oi->ip_lock);

		status = -ENOENT;
		goto leave;
	}

	if (mode & O_DIRECT)
		oi->ip_flags |= OCFS2_INODE_OPEN_DIRECT;

	oi->ip_open_count++;
	spin_unlock(&oi->ip_lock);
	status = 0;
leave:
	mlog_exit(status);
	return status;
}

static int ocfs2_file_release(struct inode *inode, struct file *file)
{
	struct ocfs2_inode_info *oi = OCFS2_I(inode);

	mlog_entry("(0x%p, 0x%p, '%.*s')\n", inode, file,
		       file->f_dentry->d_name.len,
		       file->f_dentry->d_name.name);

	spin_lock(&oi->ip_lock);
	if (!--oi->ip_open_count)
		oi->ip_flags &= ~OCFS2_INODE_OPEN_DIRECT;
	spin_unlock(&oi->ip_lock);

	mlog_exit(0);

	return 0;
}

static int ocfs2_sync_file(struct file *file,
			   struct dentry *dentry,
			   int datasync)
{
	int err = 0;
	journal_t *journal;
	struct inode *inode = dentry->d_inode;
	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);

	mlog_entry("(0x%p, 0x%p, %d, '%.*s')\n", file, dentry, datasync,
		   dentry->d_name.len, dentry->d_name.name);

	err = ocfs2_sync_inode(dentry->d_inode);
	if (err)
		goto bail;

	journal = osb->journal->j_journal;
	err = journal_force_commit(journal);

bail:
	mlog_exit(err);

	return (err < 0) ? -EIO : 0;
}

int ocfs2_set_inode_size(handle_t *handle,
			 struct inode *inode,
			 struct buffer_head *fe_bh,
			 u64 new_i_size)
{
	int status;

	mlog_entry_void();
	i_size_write(inode, new_i_size);
	inode->i_blocks = ocfs2_align_bytes_to_sectors(new_i_size);
	inode->i_ctime = inode->i_mtime = CURRENT_TIME;

	status = ocfs2_mark_inode_dirty(handle, inode, fe_bh);
	if (status < 0) {
		mlog_errno(status);
		goto bail;
	}

bail:
	mlog_exit(status);
	return status;
}

static int ocfs2_simple_size_update(struct inode *inode,
				    struct buffer_head *di_bh,
				    u64 new_i_size)
{
	int ret;
	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
	handle_t *handle = NULL;

	handle = ocfs2_start_trans(osb, OCFS2_INODE_UPDATE_CREDITS);
	if (handle == NULL) {
		ret = -ENOMEM;
		mlog_errno(ret);
		goto out;
	}

	ret = ocfs2_set_inode_size(handle, inode, di_bh,
				   new_i_size);
	if (ret < 0)
		mlog_errno(ret);

	ocfs2_commit_trans(osb, handle);
out:
	return ret;
}

static int ocfs2_orphan_for_truncate(struct ocfs2_super *osb,
				     struct inode *inode,
				     struct buffer_head *fe_bh,
				     u64 new_i_size)
{
	int status;
	handle_t *handle;

	mlog_entry_void();

	/* TODO: This needs to actually orphan the inode in this
	 * transaction. */

	handle = ocfs2_start_trans(osb, OCFS2_INODE_UPDATE_CREDITS);
	if (IS_ERR(handle)) {
		status = PTR_ERR(handle);
		mlog_errno(status);
		goto out;
	}

	status = ocfs2_set_inode_size(handle, inode, fe_bh, new_i_size);
	if (status < 0)
		mlog_errno(status);

	ocfs2_commit_trans(osb, handle);
out:
	mlog_exit(status);
	return status;
}

static int ocfs2_truncate_file(struct inode *inode,
			       struct buffer_head *di_bh,
			       u64 new_i_size)
{
	int status = 0;
	struct ocfs2_dinode *fe = NULL;
	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
	struct ocfs2_truncate_context *tc = NULL;

	mlog_entry("(inode = %llu, new_i_size = %llu\n",
		   (unsigned long long)OCFS2_I(inode)->ip_blkno,
		   (unsigned long long)new_i_size);

	truncate_inode_pages(inode->i_mapping, new_i_size);

	fe = (struct ocfs2_dinode *) di_bh->b_data;
	if (!OCFS2_IS_VALID_DINODE(fe)) {
		OCFS2_RO_ON_INVALID_DINODE(inode->i_sb, fe);
		status = -EIO;
		goto bail;
	}

	mlog_bug_on_msg(le64_to_cpu(fe->i_size) != i_size_read(inode),
			"Inode %llu, inode i_size = %lld != di "
			"i_size = %llu, i_flags = 0x%x\n",
			(unsigned long long)OCFS2_I(inode)->ip_blkno,
			i_size_read(inode),
			(unsigned long long)le64_to_cpu(fe->i_size),
			le32_to_cpu(fe->i_flags));

	if (new_i_size > le64_to_cpu(fe->i_size)) {
		mlog(0, "asked to truncate file with size (%llu) to size (%llu)!\n",
		     (unsigned long long)le64_to_cpu(fe->i_size),
		     (unsigned long long)new_i_size);
		status = -EINVAL;
		mlog_errno(status);
		goto bail;
	}

	mlog(0, "inode %llu, i_size = %llu, new_i_size = %llu\n",
	     (unsigned long long)le64_to_cpu(fe->i_blkno),
	     (unsigned long long)le64_to_cpu(fe->i_size),
	     (unsigned long long)new_i_size);

	/* lets handle the simple truncate cases before doing any more
	 * cluster locking. */
	if (new_i_size == le64_to_cpu(fe->i_size))
		goto bail;

	/* This forces other nodes to sync and drop their pages. Do
	 * this even if we have a truncate without allocation change -
	 * ocfs2 cluster sizes can be much greater than page size, so
	 * we have to truncate them anyway.  */
	status = ocfs2_data_lock(inode, 1);
	if (status < 0) {
		mlog_errno(status);
		goto bail;
	}
	ocfs2_data_unlock(inode, 1);

	if (le32_to_cpu(fe->i_clusters) ==
	    ocfs2_clusters_for_bytes(osb->sb, new_i_size)) {
		mlog(0, "fe->i_clusters = %u, so we do a simple truncate\n",
		     fe->i_clusters);
		/* No allocation change is required, so lets fast path
		 * this truncate. */
		status = ocfs2_simple_size_update(inode, di_bh, new_i_size);
		if (status < 0)
			mlog_errno(status);
		goto bail;
	}

	/* alright, we're going to need to do a full blown alloc size
	 * change. Orphan the inode so that recovery can complete the
	 * truncate if necessary. This does the task of marking
	 * i_size. */
	status = ocfs2_orphan_for_truncate(osb, inode, di_bh, new_i_size);
	if (status < 0) {
		mlog_errno(status);
		goto bail;
	}

	status = ocfs2_prepare_truncate(osb, inode, di_bh, &tc);
	if (status < 0) {
		mlog_errno(status);
		goto bail;
	}

	status = ocfs2_commit_truncate(osb, inode, di_bh, tc);
	if (status < 0) {
		mlog_errno(status);
		goto bail;
	}

	/* TODO: orphan dir cleanup here. */
bail:

	mlog_exit(status);
	return status;
}

/*
 * extend allocation only here.
 * we'll update all the disk stuff, and oip->alloc_size
 *
 * expect stuff to be locked, a transaction started and enough data /
 * metadata reservations in the contexts.
 *
 * Will return -EAGAIN, and a reason if a restart is needed.
 * If passed in, *reason will always be set, even in error.
 */
int ocfs2_do_extend_allocation(struct ocfs2_super *osb,
			       struct inode *inode,
			       u32 clusters_to_add,
			       struct buffer_head *fe_bh,
			       handle_t *handle,
			       struct ocfs2_alloc_context *data_ac,
			       struct ocfs2_alloc_context *meta_ac,
			       enum ocfs2_alloc_restarted *reason_ret)
{
	int status = 0;
	int free_extents;
	struct ocfs2_dinode *fe = (struct ocfs2_dinode *) fe_bh->b_data;
	enum ocfs2_alloc_restarted reason = RESTART_NONE;
	u32 bit_off, num_bits;
	u64 block;

	BUG_ON(!clusters_to_add);

	free_extents = ocfs2_num_free_extents(osb, inode, fe);
	if (free_extents < 0) {
		status = free_extents;
		mlog_errno(status);
		goto leave;
	}

	/* there are two cases which could cause us to EAGAIN in the
	 * we-need-more-metadata case:
	 * 1) we haven't reserved *any*
	 * 2) we are so fragmented, we've needed to add metadata too
	 *    many times. */
	if (!free_extents && !meta_ac) {
		mlog(0, "we haven't reserved any metadata!\n");
		status = -EAGAIN;
		reason = RESTART_META;
		goto leave;
	} else if ((!free_extents)
		   && (ocfs2_alloc_context_bits_left(meta_ac)
		       < ocfs2_extend_meta_needed(fe))) {
		mlog(0, "filesystem is really fragmented...\n");
		status = -EAGAIN;
		reason = RESTART_META;
		goto leave;
	}

	status = ocfs2_claim_clusters(osb, handle, data_ac, 1,
				      &bit_off, &num_bits);
	if (status < 0) {
		if (status != -ENOSPC)
			mlog_errno(status);
		goto leave;
	}

	BUG_ON(num_bits > clusters_to_add);

	/* reserve our write early -- insert_extent may update the inode */
	status = ocfs2_journal_access(handle, inode, fe_bh,
				      OCFS2_JOURNAL_ACCESS_WRITE);
	if (status < 0) {
		mlog_errno(status);
		goto leave;
	}

	block = ocfs2_clusters_to_blocks(osb->sb, bit_off);
	mlog(0, "Allocating %u clusters at block %u for inode %llu\n",
	     num_bits, bit_off, (unsigned long long)OCFS2_I(inode)->ip_blkno);
	status = ocfs2_insert_extent(osb, handle, inode, fe_bh, block,
				     num_bits, meta_ac);
	if (status < 0) {
		mlog_errno(status);
		goto leave;
	}

	le32_add_cpu(&fe->i_clusters, num_bits);
	spin_lock(&OCFS2_I(inode)->ip_lock);
	OCFS2_I(inode)->ip_clusters = le32_to_cpu(fe->i_clusters);
	spin_unlock(&OCFS2_I(inode)->ip_lock);

	status = ocfs2_journal_dirty(handle, fe_bh);
	if (status < 0) {
		mlog_errno(status);
		goto leave;
	}

	clusters_to_add -= num_bits;

	if (clusters_to_add) {
		mlog(0, "need to alloc once more, clusters = %u, wanted = "
		     "%u\n", fe->i_clusters, clusters_to_add);
		status = -EAGAIN;
		reason = RESTART_TRANS;
	}

leave:
	mlog_exit(status);
	if (reason_ret)
		*reason_ret = reason;
	return status;
}

static int ocfs2_extend_allocation(struct inode *inode,
				   u32 clusters_to_add)
{
	int status = 0;
	int restart_func = 0;
	int drop_alloc_sem = 0;
	int credits, num_free_extents;
	u32 prev_clusters;
	struct buffer_head *bh = NULL;
	struct ocfs2_dinode *fe = NULL;
	handle_t *handle = NULL;
	struct ocfs2_alloc_context *data_ac = NULL;
	struct ocfs2_alloc_context *meta_ac = NULL;
	enum ocfs2_alloc_restarted why;
	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);

	mlog_entry("(clusters_to_add = %u)\n", clusters_to_add);

	status = ocfs2_read_block(osb, OCFS2_I(inode)->ip_blkno, &bh,
				  OCFS2_BH_CACHED, inode);
	if (status < 0) {
		mlog_errno(status);
		goto leave;
	}

	fe = (struct ocfs2_dinode *) bh->b_data;
	if (!OCFS2_IS_VALID_DINODE(fe)) {
		OCFS2_RO_ON_INVALID_DINODE(inode->i_sb, fe);
		status = -EIO;
		goto leave;
	}

restart_all:
	BUG_ON(le32_to_cpu(fe->i_clusters) != OCFS2_I(inode)->ip_clusters);

	mlog(0, "extend inode %llu, i_size = %lld, fe->i_clusters = %u, "
	     "clusters_to_add = %u\n",
	     (unsigned long long)OCFS2_I(inode)->ip_blkno, i_size_read(inode),
	     fe->i_clusters, clusters_to_add);

	num_free_extents = ocfs2_num_free_extents(osb,
						  inode,
						  fe);
	if (num_free_extents < 0) {
		status = num_free_extents;
		mlog_errno(status);
		goto leave;
	}

	if (!num_free_extents) {
		status = ocfs2_reserve_new_metadata(osb, fe, &meta_ac);
		if (status < 0) {
			if (status != -ENOSPC)
				mlog_errno(status);
			goto leave;
		}
	}

	status = ocfs2_reserve_clusters(osb, clusters_to_add, &data_ac);
	if (status < 0) {
		if (status != -ENOSPC)
			mlog_errno(status);
		goto leave;
	}

	/* blocks peope in read/write from reading our allocation
	 * until we're done changing it. We depend on i_mutex to block
	 * other extend/truncate calls while we're here. Ordering wrt
	 * start_trans is important here -- always do it before! */
	down_write(&OCFS2_I(inode)->ip_alloc_sem);
	drop_alloc_sem = 1;

	credits = ocfs2_calc_extend_credits(osb->sb, fe, clusters_to_add);
	handle = ocfs2_start_trans(osb, credits);
	if (IS_ERR(handle)) {
		status = PTR_ERR(handle);
		handle = NULL;
		mlog_errno(status);
		goto leave;
	}

restarted_transaction:
	/* reserve a write to the file entry early on - that we if we
	 * run out of credits in the allocation path, we can still
	 * update i_size. */
	status = ocfs2_journal_access(handle, inode, bh,
				      OCFS2_JOURNAL_ACCESS_WRITE);
	if (status < 0) {
		mlog_errno(status);
		goto leave;
	}

	prev_clusters = OCFS2_I(inode)->ip_clusters;

	status = ocfs2_do_extend_allocation(osb,
					    inode,
					    clusters_to_add,
					    bh,
					    handle,
					    data_ac,
					    meta_ac,
					    &why);
	if ((status < 0) && (status != -EAGAIN)) {
		if (status != -ENOSPC)
			mlog_errno(status);
		goto leave;
	}

	status = ocfs2_journal_dirty(handle, bh);
	if (status < 0) {
		mlog_errno(status);
		goto leave;
	}

	spin_lock(&OCFS2_I(inode)->ip_lock);
	clusters_to_add -= (OCFS2_I(inode)->ip_clusters - prev_clusters);
	spin_unlock(&OCFS2_I(inode)->ip_lock);

	if (why != RESTART_NONE && clusters_to_add) {
		if (why == RESTART_META) {
			mlog(0, "restarting function.\n");
			restart_func = 1;
		} else {
			BUG_ON(why != RESTART_TRANS);

			mlog(0, "restarting transaction.\n");
			/* TODO: This can be more intelligent. */
			credits = ocfs2_calc_extend_credits(osb->sb,
							    fe,
							    clusters_to_add);
			status = ocfs2_extend_trans(handle, credits);
			if (status < 0) {
				/* handle still has to be committed at
				 * this point. */
				status = -ENOMEM;
				mlog_errno(status);
				goto leave;
			}
			goto restarted_transaction;
		}
	}

	mlog(0, "fe: i_clusters = %u, i_size=%llu\n",
	     fe->i_clusters, (unsigned long long)fe->i_size);
	mlog(0, "inode: ip_clusters=%u, i_size=%lld\n",
	     OCFS2_I(inode)->ip_clusters, i_size_read(inode));

leave:
	if (drop_alloc_sem) {
		up_write(&OCFS2_I(inode)->ip_alloc_sem);
		drop_alloc_sem = 0;
	}
	if (handle) {
		ocfs2_commit_trans(osb, handle);
		handle = NULL;
	}
	if (data_ac) {
		ocfs2_free_alloc_context(data_ac);
		data_ac = NULL;
	}
	if (meta_ac) {
		ocfs2_free_alloc_context(meta_ac);
		meta_ac = NULL;
	}
	if ((!status) && restart_func) {
		restart_func = 0;
		goto restart_all;
	}
	if (bh) {
		brelse(bh);
		bh = NULL;
	}

	mlog_exit(status);
	return status;
}

/* Some parts of this taken from generic_cont_expand, which turned out
 * to be too fragile to do exactly what we need without us having to
 * worry about recursive locking in ->prepare_write() and
 * ->commit_write(). */
static int ocfs2_write_zero_page(struct inode *inode,
				 u64 size)
{
	struct address_space *mapping = inode->i_mapping;
	struct page *page;
	unsigned long index;
	unsigned int offset;
	handle_t *handle = NULL;
	int ret;

	offset = (size & (PAGE_CACHE_SIZE-1)); /* Within page */
	/* ugh.  in prepare/commit_write, if from==to==start of block, we 
	** skip the prepare.  make sure we never send an offset for the start
	** of a block
	*/
	if ((offset & (inode->i_sb->s_blocksize - 1)) == 0) {
		offset++;
	}
	index = size >> PAGE_CACHE_SHIFT;

	page = grab_cache_page(mapping, index);
	if (!page) {
		ret = -ENOMEM;
		mlog_errno(ret);
		goto out;
	}

	ret = ocfs2_prepare_write_nolock(inode, page, offset, offset);
	if (ret < 0) {
		mlog_errno(ret);
		goto out_unlock;
	}

	if (ocfs2_should_order_data(inode)) {
		handle = ocfs2_start_walk_page_trans(inode, page, offset,
						     offset);
		if (IS_ERR(handle)) {
			ret = PTR_ERR(handle);
			handle = NULL;
			goto out_unlock;
		}
	}

	/* must not update i_size! */
	ret = block_commit_write(page, offset, offset);
	if (ret < 0)
		mlog_errno(ret);
	else
		ret = 0;

	if (handle)
		ocfs2_commit_trans(OCFS2_SB(inode->i_sb), handle);
out_unlock:
	unlock_page(page);
	page_cache_release(page);
out:
	return ret;
}

static int ocfs2_zero_extend(struct inode *inode,
			     u64 zero_to_size)
{
	int ret = 0;
	u64 start_off;
	struct super_block *sb = inode->i_sb;

	start_off = ocfs2_align_bytes_to_blocks(sb, i_size_read(inode));
	while (start_off < zero_to_size) {
		ret = ocfs2_write_zero_page(inode, start_off);
		if (ret < 0) {
			mlog_errno(ret);
			goto out;
		}

		start_off += sb->s_blocksize;

		/*
		 * Very large extends have the potential to lock up
		 * the cpu for extended periods of time.
		 */
		cond_resched();
	}

out:
	return ret;
}

/* 
 * A tail_to_skip value > 0 indicates that we're being called from
 * ocfs2_file_aio_write(). This has the following implications:
 *
 * - we don't want to update i_size
 * - di_bh will be NULL, which is fine because it's only used in the
 *   case where we want to update i_size.
 * - ocfs2_zero_extend() will then only be filling the hole created
 *   between i_size and the start of the write.
 */
static int ocfs2_extend_file(struct inode *inode,
			     struct buffer_head *di_bh,
			     u64 new_i_size,
			     size_t tail_to_skip)
{
	int ret = 0;
	u32 clusters_to_add;

	BUG_ON(!tail_to_skip && !di_bh);

	/* setattr sometimes calls us like this. */
	if (new_i_size == 0)
		goto out;

	if (i_size_read(inode) == new_i_size)
  		goto out;
	BUG_ON(new_i_size < i_size_read(inode));

	clusters_to_add = ocfs2_clusters_for_bytes(inode->i_sb, new_i_size) - 
		OCFS2_I(inode)->ip_clusters;

	/* 
	 * protect the pages that ocfs2_zero_extend is going to be
	 * pulling into the page cache.. we do this before the
	 * metadata extend so that we don't get into the situation
	 * where we've extended the metadata but can't get the data
	 * lock to zero.
	 */
	ret = ocfs2_data_lock(inode, 1);
	if (ret < 0) {
		mlog_errno(ret);
		goto out;
	}

	if (clusters_to_add) {
		ret = ocfs2_extend_allocation(inode, clusters_to_add);
		if (ret < 0) {
			mlog_errno(ret);
			goto out_unlock;
		}
	}

	/*
	 * Call this even if we don't add any clusters to the tree. We
	 * still need to zero the area between the old i_size and the
	 * new i_size.
	 */
	ret = ocfs2_zero_extend(inode, (u64)new_i_size - tail_to_skip);
	if (ret < 0) {
		mlog_errno(ret);
		goto out_unlock;
	}

	if (!tail_to_skip) {
		/* We're being called from ocfs2_setattr() which wants
		 * us to update i_size */
		ret = ocfs2_simple_size_update(inode, di_bh, new_i_size);
		if (ret < 0)
			mlog_errno(ret);
	}

out_unlock:
	ocfs2_data_unlock(inode, 1);

out:
	return ret;
}

int ocfs2_setattr(struct dentry *dentry, struct iattr *attr)
{
	int status = 0, size_change;
	struct inode *inode = dentry->d_inode;
	struct super_block *sb = inode->i_sb;
	struct ocfs2_super *osb = OCFS2_SB(sb);
	struct buffer_head *bh = NULL;
	handle_t *handle = NULL;

	mlog_entry("(0x%p, '%.*s')\n", dentry,
	           dentry->d_name.len, dentry->d_name.name);

	if (attr->ia_valid & ATTR_MODE)
		mlog(0, "mode change: %d\n", attr->ia_mode);
	if (attr->ia_valid & ATTR_UID)
		mlog(0, "uid change: %d\n", attr->ia_uid);
	if (attr->ia_valid & ATTR_GID)
		mlog(0, "gid change: %d\n", attr->ia_gid);
	if (attr->ia_valid & ATTR_SIZE)
		mlog(0, "size change...\n");
	if (attr->ia_valid & (ATTR_ATIME | ATTR_MTIME | ATTR_CTIME))
		mlog(0, "time change...\n");

#define OCFS2_VALID_ATTRS (ATTR_ATIME | ATTR_MTIME | ATTR_CTIME | ATTR_SIZE \
			   | ATTR_GID | ATTR_UID | ATTR_MODE)
	if (!(attr->ia_valid & OCFS2_VALID_ATTRS)) {
		mlog(0, "can't handle attrs: 0x%x\n", attr->ia_valid);
		return 0;
	}

	status = inode_change_ok(inode, attr);
	if (status)
		return status;

	size_change = S_ISREG(inode->i_mode) && attr->ia_valid & ATTR_SIZE;
	if (size_change) {
		status = ocfs2_rw_lock(inode, 1);
		if (status < 0) {
			mlog_errno(status);
			goto bail;
		}
	}

	status = ocfs2_meta_lock(inode, &bh, 1);
	if (status < 0) {
		if (status != -ENOENT)
			mlog_errno(status);
		goto bail_unlock_rw;
	}

	if (size_change && attr->ia_size != i_size_read(inode)) {
		if (i_size_read(inode) > attr->ia_size)
			status = ocfs2_truncate_file(inode, bh, attr->ia_size);
		else
			status = ocfs2_extend_file(inode, bh, attr->ia_size, 0);
		if (status < 0) {
			if (status != -ENOSPC)
				mlog_errno(status);
			status = -ENOSPC;
			goto bail_unlock;
		}
	}

	handle = ocfs2_start_trans(osb, OCFS2_INODE_UPDATE_CREDITS);
	if (IS_ERR(handle)) {
		status = PTR_ERR(handle);
		mlog_errno(status);
		goto bail_unlock;
	}

	status = inode_setattr(inode, attr);
	if (status < 0) {
		mlog_errno(status);
		goto bail_commit;
	}

	status = ocfs2_mark_inode_dirty(handle, inode, bh);
	if (status < 0)
		mlog_errno(status);

bail_commit:
	ocfs2_commit_trans(osb, handle);
bail_unlock:
	ocfs2_meta_unlock(inode, 1);
bail_unlock_rw:
	if (size_change)
		ocfs2_rw_unlock(inode, 1);
bail:
	if (bh)
		brelse(bh);

	mlog_exit(status);
	return status;
}

int ocfs2_getattr(struct vfsmount *mnt,
		  struct dentry *dentry,
		  struct kstat *stat)
{
	struct inode *inode = dentry->d_inode;
	struct super_block *sb = dentry->d_inode->i_sb;
	struct ocfs2_super *osb = sb->s_fs_info;
	int err;

	mlog_entry_void();

	err = ocfs2_inode_revalidate(dentry);
	if (err) {
		if (err != -ENOENT)
			mlog_errno(err);
		goto bail;
	}

	generic_fillattr(inode, stat);

	/* We set the blksize from the cluster size for performance */
	stat->blksize = osb->s_clustersize;

bail:
	mlog_exit(err);

	return err;
}

static int ocfs2_write_remove_suid(struct inode *inode)
{
	int ret;
	struct buffer_head *bh = NULL;
	struct ocfs2_inode_info *oi = OCFS2_I(inode);
	handle_t *handle;
	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
	struct ocfs2_dinode *di;

	mlog_entry("(Inode %llu, mode 0%o)\n",
		   (unsigned long long)oi->ip_blkno, inode->i_mode);

	handle = ocfs2_start_trans(osb, OCFS2_INODE_UPDATE_CREDITS);
	if (handle == NULL) {
		ret = -ENOMEM;
		mlog_errno(ret);
		goto out;
	}

	ret = ocfs2_read_block(osb, oi->ip_blkno, &bh, OCFS2_BH_CACHED, inode);
	if (ret < 0) {
		mlog_errno(ret);
		goto out_trans;
	}

	ret = ocfs2_journal_access(handle, inode, bh,
				   OCFS2_JOURNAL_ACCESS_WRITE);
	if (ret < 0) {
		mlog_errno(ret);
		goto out_bh;
	}

	inode->i_mode &= ~S_ISUID;
	if ((inode->i_mode & S_ISGID) && (inode->i_mode & S_IXGRP))
		inode->i_mode &= ~S_ISGID;

	di = (struct ocfs2_dinode *) bh->b_data;
	di->i_mode = cpu_to_le16(inode->i_mode);

	ret = ocfs2_journal_dirty(handle, bh);
	if (ret < 0)
		mlog_errno(ret);
out_bh:
	brelse(bh);
out_trans:
	ocfs2_commit_trans(osb, handle);
out:
	mlog_exit(ret);
	return ret;
}

static int ocfs2_prepare_inode_for_write(struct dentry *dentry,
					 loff_t *ppos,
					 size_t count,
					 int appending)
{
	int ret = 0, meta_level = appending;
	struct inode *inode = dentry->d_inode;
	u32 clusters;
	loff_t newsize, saved_pos;

	/* 
	 * We sample i_size under a read level meta lock to see if our write
	 * is extending the file, if it is we back off and get a write level
	 * meta lock.
	 */
	for(;;) {
		ret = ocfs2_meta_lock(inode, NULL, meta_level);
		if (ret < 0) {
			meta_level = -1;
			mlog_errno(ret);
			goto out;
		}

		/* Clear suid / sgid if necessary. We do this here
		 * instead of later in the write path because
		 * remove_suid() calls ->setattr without any hint that
		 * we may have already done our cluster locking. Since
		 * ocfs2_setattr() *must* take cluster locks to
		 * proceeed, this will lead us to recursively lock the
		 * inode. There's also the dinode i_size state which
		 * can be lost via setattr during extending writes (we
		 * set inode->i_size at the end of a write. */
		if (should_remove_suid(dentry)) {
			if (meta_level == 0) {
				ocfs2_meta_unlock(inode, meta_level);
				meta_level = 1;
				continue;
			}

			ret = ocfs2_write_remove_suid(inode);
			if (ret < 0) {
				mlog_errno(ret);
				goto out_unlock;
			}
		}

		/* work on a copy of ppos until we're sure that we won't have
		 * to recalculate it due to relocking. */
		if (appending) {
			saved_pos = i_size_read(inode);
			mlog(0, "O_APPEND: inode->i_size=%llu\n", saved_pos);
		} else {
			saved_pos = *ppos;
		}
		newsize = count + saved_pos;

		mlog(0, "pos=%lld newsize=%lld cursize=%lld\n",
		     (long long) saved_pos, (long long) newsize,
		     (long long) i_size_read(inode));

		/* No need for a higher level metadata lock if we're
		 * never going past i_size. */
		if (newsize <= i_size_read(inode))
			break;

		if (meta_level == 0) {
			ocfs2_meta_unlock(inode, meta_level);
			meta_level = 1;
			continue;
		}

		spin_lock(&OCFS2_I(inode)->ip_lock);
		clusters = ocfs2_clusters_for_bytes(inode->i_sb, newsize) -
			OCFS2_I(inode)->ip_clusters;
		spin_unlock(&OCFS2_I(inode)->ip_lock);

		mlog(0, "Writing at EOF, may need more allocation: "
		     "i_size = %lld, newsize = %lld, need %u clusters\n",
		     (long long) i_size_read(inode), (long long) newsize,
		     clusters);

		/* We only want to continue the rest of this loop if
		 * our extend will actually require more
		 * allocation. */
		if (!clusters)
			break;

		ret = ocfs2_extend_file(inode, NULL, newsize, count);
		if (ret < 0) {
			if (ret != -ENOSPC)
				mlog_errno(ret);
			goto out_unlock;
		}
		break;
	}

	if (appending)
		*ppos = saved_pos;

out_unlock:
	ocfs2_meta_unlock(inode, meta_level);

out:
	return ret;
}

static ssize_t ocfs2_file_aio_write(struct kiocb *iocb,
				    const struct iovec *iov,
				    unsigned long nr_segs,
				    loff_t pos)
{
	int ret, rw_level, have_alloc_sem = 0;
	struct file *filp = iocb->ki_filp;
	struct inode *inode = filp->f_dentry->d_inode;
	int appending = filp->f_flags & O_APPEND ? 1 : 0;

	mlog_entry("(0x%p, %u, '%.*s')\n", filp,
		   (unsigned int)nr_segs,
		   filp->f_dentry->d_name.len,
		   filp->f_dentry->d_name.name);

	/* happy write of zero bytes */
	if (iocb->ki_left == 0)
		return 0;

	mutex_lock(&inode->i_mutex);
	/* to match setattr's i_mutex -> i_alloc_sem -> rw_lock ordering */
	if (filp->f_flags & O_DIRECT) {
		have_alloc_sem = 1;
		down_read(&inode->i_alloc_sem);
	}

	/* concurrent O_DIRECT writes are allowed */
	rw_level = (filp->f_flags & O_DIRECT) ? 0 : 1;
	ret = ocfs2_rw_lock(inode, rw_level);
	if (ret < 0) {
		rw_level = -1;
		mlog_errno(ret);
		goto out;
	}

	ret = ocfs2_prepare_inode_for_write(filp->f_dentry, &iocb->ki_pos,
					    iocb->ki_left, appending);
	if (ret < 0) {
		mlog_errno(ret);
		goto out;
	}

	/* communicate with ocfs2_dio_end_io */
	ocfs2_iocb_set_rw_locked(iocb);

	ret = generic_file_aio_write_nolock(iocb, iov, nr_segs, iocb->ki_pos);

	/* buffered aio wouldn't have proper lock coverage today */
	BUG_ON(ret == -EIOCBQUEUED && !(filp->f_flags & O_DIRECT));

	/* 
	 * deep in g_f_a_w_n()->ocfs2_direct_IO we pass in a ocfs2_dio_end_io
	 * function pointer which is called when o_direct io completes so that
	 * it can unlock our rw lock.  (it's the clustered equivalent of
	 * i_alloc_sem; protects truncate from racing with pending ios).
	 * Unfortunately there are error cases which call end_io and others
	 * that don't.  so we don't have to unlock the rw_lock if either an
	 * async dio is going to do it in the future or an end_io after an
	 * error has already done it.
	 */
	if (ret == -EIOCBQUEUED || !ocfs2_iocb_is_rw_locked(iocb)) {
		rw_level = -1;
		have_alloc_sem = 0;
	}

out:
	if (have_alloc_sem)
		up_read(&inode->i_alloc_sem);
	if (rw_level != -1) 
		ocfs2_rw_unlock(inode, rw_level);
	mutex_unlock(&inode->i_mutex);

	mlog_exit(ret);
	return ret;
}

static ssize_t ocfs2_file_splice_write(struct pipe_inode_info *pipe,
				       struct file *out,
				       loff_t *ppos,
				       size_t len,
				       unsigned int flags)
{
	int ret;
	struct inode *inode = out->f_dentry->d_inode;

	mlog_entry("(0x%p, 0x%p, %u, '%.*s')\n", out, pipe,
		   (unsigned int)len,
		   out->f_dentry->d_name.len,
		   out->f_dentry->d_name.name);

	inode_double_lock(inode, pipe->inode);

	ret = ocfs2_rw_lock(inode, 1);
	if (ret < 0) {
		mlog_errno(ret);
		goto out;
	}

	ret = ocfs2_prepare_inode_for_write(out->f_dentry, ppos, len, 0);
	if (ret < 0) {
		mlog_errno(ret);
		goto out_unlock;
	}

	/* ok, we're done with i_size and alloc work */
	ret = generic_file_splice_write_nolock(pipe, out, ppos, len, flags);

out_unlock:
	ocfs2_rw_unlock(inode, 1);
out:
	inode_double_unlock(inode, pipe->inode);

	mlog_exit(ret);
	return ret;
}

static ssize_t ocfs2_file_splice_read(struct file *in,
				      loff_t *ppos,
				      struct pipe_inode_info *pipe,
				      size_t len,
				      unsigned int flags)
{
	int ret = 0;
	struct inode *inode = in->f_dentry->d_inode;

	mlog_entry("(0x%p, 0x%p, %u, '%.*s')\n", in, pipe,
		   (unsigned int)len,
		   in->f_dentry->d_name.len,
		   in->f_dentry->d_name.name);

	/*
	 * See the comment in ocfs2_file_aio_read()
	 */
	ret = ocfs2_meta_lock(inode, NULL, 0);
	if (ret < 0) {
		mlog_errno(ret);
		goto bail;
	}
	ocfs2_meta_unlock(inode, 0);

	ret = generic_file_splice_read(in, ppos, pipe, len, flags);

bail:
	mlog_exit(ret);
	return ret;
}

static ssize_t ocfs2_file_aio_read(struct kiocb *iocb,
				   const struct iovec *iov,
				   unsigned long nr_segs,
				   loff_t pos)
{
	int ret = 0, rw_level = -1, have_alloc_sem = 0;
	struct file *filp = iocb->ki_filp;
	struct inode *inode = filp->f_dentry->d_inode;

	mlog_entry("(0x%p, %u, '%.*s')\n", filp,
		   (unsigned int)nr_segs,
		   filp->f_dentry->d_name.len,
		   filp->f_dentry->d_name.name);

	if (!inode) {
		ret = -EINVAL;
		mlog_errno(ret);
		goto bail;
	}

	/* 
	 * buffered reads protect themselves in ->readpage().  O_DIRECT reads
	 * need locks to protect pending reads from racing with truncate.
	 */
	if (filp->f_flags & O_DIRECT) {
		down_read(&inode->i_alloc_sem);
		have_alloc_sem = 1;

		ret = ocfs2_rw_lock(inode, 0);
		if (ret < 0) {
			mlog_errno(ret);
			goto bail;
		}
		rw_level = 0;
		/* communicate with ocfs2_dio_end_io */
		ocfs2_iocb_set_rw_locked(iocb);
	}

	/*
	 * We're fine letting folks race truncates and extending
	 * writes with read across the cluster, just like they can
	 * locally. Hence no rw_lock during read.
	 * 
	 * Take and drop the meta data lock to update inode fields
	 * like i_size. This allows the checks down below
	 * generic_file_aio_read() a chance of actually working. 
	 */
	ret = ocfs2_meta_lock(inode, NULL, 0);
	if (ret < 0) {
		mlog_errno(ret);
		goto bail;
	}
	ocfs2_meta_unlock(inode, 0);

	ret = generic_file_aio_read(iocb, iov, nr_segs, iocb->ki_pos);
	if (ret == -EINVAL)
		mlog(ML_ERROR, "generic_file_aio_read returned -EINVAL\n");

	/* buffered aio wouldn't have proper lock coverage today */
	BUG_ON(ret == -EIOCBQUEUED && !(filp->f_flags & O_DIRECT));

	/* see ocfs2_file_aio_write */
	if (ret == -EIOCBQUEUED || !ocfs2_iocb_is_rw_locked(iocb)) {
		rw_level = -1;
		have_alloc_sem = 0;
	}

bail:
	if (have_alloc_sem)
		up_read(&inode->i_alloc_sem);
	if (rw_level != -1) 
		ocfs2_rw_unlock(inode, rw_level);
	mlog_exit(ret);

	return ret;
}

struct inode_operations ocfs2_file_iops = {
	.setattr	= ocfs2_setattr,
	.getattr	= ocfs2_getattr,
};

struct inode_operations ocfs2_special_file_iops = {
	.setattr	= ocfs2_setattr,
	.getattr	= ocfs2_getattr,
};

const struct file_operations ocfs2_fops = {
	.read		= do_sync_read,
	.write		= do_sync_write,
	.sendfile	= generic_file_sendfile,
	.mmap		= ocfs2_mmap,
	.fsync		= ocfs2_sync_file,
	.release	= ocfs2_file_release,
	.open		= ocfs2_file_open,
	.aio_read	= ocfs2_file_aio_read,
	.aio_write	= ocfs2_file_aio_write,
	.ioctl		= ocfs2_ioctl,
	.splice_read	= ocfs2_file_splice_read,
	.splice_write	= ocfs2_file_splice_write,
};

const struct file_operations ocfs2_dops = {
	.read		= generic_read_dir,
	.readdir	= ocfs2_readdir,
	.fsync		= ocfs2_sync_file,
	.ioctl		= ocfs2_ioctl,
};