aboutsummaryrefslogtreecommitdiff
path: root/fs/btrfs/extent-tree.c
blob: 2937fd9aba747e5b2556ed2422368f6e220b0f9a (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
#include <linux/module.h>
#include "ctree.h"
#include "disk-io.h"
#include "print-tree.h"
#include "transaction.h"

static int find_free_extent(struct btrfs_trans_handle *trans, struct btrfs_root
			    *orig_root, u64 num_blocks, u64 search_start, u64
			    search_end, struct btrfs_key *ins, int data);
static int finish_current_insert(struct btrfs_trans_handle *trans, struct
				 btrfs_root *extent_root);
static int del_pending_extents(struct btrfs_trans_handle *trans, struct
			       btrfs_root *extent_root);

static struct btrfs_block_group_cache *lookup_block_group(struct
							  btrfs_fs_info *info,
							  u64 blocknr)
{
	struct btrfs_block_group_cache *block_group;
	int ret;

	ret = radix_tree_gang_lookup(&info->block_group_radix,
				     (void **)&block_group,
				     blocknr, 1);
	if (ret) {
		if (block_group->key.objectid <= blocknr && blocknr <
		    block_group->key.objectid + block_group->key.offset)
			return block_group;
	}
	ret = radix_tree_gang_lookup(&info->block_group_data_radix,
				     (void **)&block_group,
				     blocknr, 1);
	if (ret) {
		if (block_group->key.objectid <= blocknr && blocknr <
		    block_group->key.objectid + block_group->key.offset)
			return block_group;
	}
printk("lookup_block_group fails for blocknr %Lu\n", blocknr);
	return NULL;
}

struct btrfs_block_group_cache *btrfs_find_block_group(struct btrfs_root *root,
						 struct btrfs_block_group_cache
						 *hint, u64 search_start,
						 int data)
{
	struct btrfs_block_group_cache *cache[8];
	struct btrfs_block_group_cache *found_group = NULL;
	struct btrfs_fs_info *info = root->fs_info;
	struct radix_tree_root *radix;
	u64 used;
	u64 last = 0;
	u64 hint_last;
	int i;
	int ret;
	int full_search = 0;

	if (data)
		radix = &info->block_group_data_radix;
	else
		radix = &info->block_group_radix;

	if (search_start) {
		struct btrfs_block_group_cache *shint;
		shint = lookup_block_group(info, search_start);
		if (shint->data == data) {
			used = btrfs_block_group_used(&shint->item);
			if (used + shint->pinned <
			    (shint->key.offset * 8) / 10) {
				return shint;
			}
		}
	}
	if (hint && hint->data == data) {
		used = btrfs_block_group_used(&hint->item);
		if (used + hint->pinned < (hint->key.offset * 8) / 10) {
			return hint;
		}
		if (used >= (hint->key.offset * 8) / 10) {
			radix_tree_tag_clear(radix,
					     hint->key.objectid +
					     hint->key.offset - 1,
					     BTRFS_BLOCK_GROUP_AVAIL);
		}
		last = hint->key.offset * 2;
		if (hint->key.objectid >= last)
			last = max(search_start, hint->key.objectid - last);
		else
			last = hint->key.objectid + hint->key.offset;
		hint_last = last;
	} else {
		hint_last = search_start;
		last = search_start;
	}
	while(1) {
		ret = radix_tree_gang_lookup_tag(radix, (void **)cache,
						 last, ARRAY_SIZE(cache),
						 BTRFS_BLOCK_GROUP_AVAIL);
		if (!ret)
			break;
		for (i = 0; i < ret; i++) {
			last = cache[i]->key.objectid +
				cache[i]->key.offset;
			used = btrfs_block_group_used(&cache[i]->item);
			if (used + cache[i]->pinned <
			    (cache[i]->key.offset * 8) / 10) {
				found_group = cache[i];
				goto found;
			}
			if (used >= (cache[i]->key.offset * 8) / 10) {
				radix_tree_tag_clear(radix,
						     cache[i]->key.objectid +
						     cache[i]->key.offset - 1,
						     BTRFS_BLOCK_GROUP_AVAIL);
			}
		}
	}
	last = hint_last;
again:
	while(1) {
		ret = radix_tree_gang_lookup(radix, (void **)cache,
					     last, ARRAY_SIZE(cache));
		if (!ret)
			break;
		for (i = 0; i < ret; i++) {
			last = cache[i]->key.objectid +
				cache[i]->key.offset;
			used = btrfs_block_group_used(&cache[i]->item);
			if (used + cache[i]->pinned < cache[i]->key.offset) {
				found_group = cache[i];
				goto found;
			}
			if (used >= cache[i]->key.offset) {
				radix_tree_tag_clear(radix,
						     cache[i]->key.objectid +
						     cache[i]->key.offset - 1,
						     BTRFS_BLOCK_GROUP_AVAIL);
			}
		}
	}
	if (!full_search) {
		last = search_start;
		full_search = 1;
		goto again;
	}
	if (!found_group) {
		ret = radix_tree_gang_lookup(radix,
					     (void **)&found_group, 0, 1);
		BUG_ON(ret != 1);
	}
found:
	return found_group;
}

int btrfs_inc_extent_ref(struct btrfs_trans_handle *trans,
				struct btrfs_root *root,
				u64 blocknr, u64 num_blocks)
{
	struct btrfs_path *path;
	int ret;
	struct btrfs_key key;
	struct btrfs_leaf *l;
	struct btrfs_extent_item *item;
	struct btrfs_key ins;
	u32 refs;

	find_free_extent(trans, root->fs_info->extent_root, 0, 0, (u64)-1,
			 &ins, 0);
	path = btrfs_alloc_path();
	BUG_ON(!path);
	btrfs_init_path(path);
	key.objectid = blocknr;
	key.flags = 0;
	btrfs_set_key_type(&key, BTRFS_EXTENT_ITEM_KEY);
	key.offset = num_blocks;
	ret = btrfs_search_slot(trans, root->fs_info->extent_root, &key, path,
				0, 1);
	if (ret != 0) {
printk("can't find block %Lu %Lu\n", blocknr, num_blocks);
		BUG();
	}
	BUG_ON(ret != 0);
	l = btrfs_buffer_leaf(path->nodes[0]);
	item = btrfs_item_ptr(l, path->slots[0], struct btrfs_extent_item);
	refs = btrfs_extent_refs(item);
	btrfs_set_extent_refs(item, refs + 1);
	btrfs_mark_buffer_dirty(path->nodes[0]);

	btrfs_release_path(root->fs_info->extent_root, path);
	btrfs_free_path(path);
	finish_current_insert(trans, root->fs_info->extent_root);
	del_pending_extents(trans, root->fs_info->extent_root);
	return 0;
}

static int lookup_extent_ref(struct btrfs_trans_handle *trans,
			     struct btrfs_root *root, u64 blocknr,
			     u64 num_blocks, u32 *refs)
{
	struct btrfs_path *path;
	int ret;
	struct btrfs_key key;
	struct btrfs_leaf *l;
	struct btrfs_extent_item *item;

	path = btrfs_alloc_path();
	btrfs_init_path(path);
	key.objectid = blocknr;
	key.offset = num_blocks;
	key.flags = 0;
	btrfs_set_key_type(&key, BTRFS_EXTENT_ITEM_KEY);
	ret = btrfs_search_slot(trans, root->fs_info->extent_root, &key, path,
				0, 0);
	if (ret != 0)
		BUG();
	l = btrfs_buffer_leaf(path->nodes[0]);
	item = btrfs_item_ptr(l, path->slots[0], struct btrfs_extent_item);
	*refs = btrfs_extent_refs(item);
	btrfs_release_path(root->fs_info->extent_root, path);
	btrfs_free_path(path);
	return 0;
}

int btrfs_inc_root_ref(struct btrfs_trans_handle *trans,
		       struct btrfs_root *root)
{
	return btrfs_inc_extent_ref(trans, root, bh_blocknr(root->node), 1);
}

int btrfs_inc_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root,
		  struct buffer_head *buf)
{
	u64 blocknr;
	struct btrfs_node *buf_node;
	struct btrfs_leaf *buf_leaf;
	struct btrfs_disk_key *key;
	struct btrfs_file_extent_item *fi;
	int i;
	int leaf;
	int ret;

	if (!root->ref_cows)
		return 0;
	buf_node = btrfs_buffer_node(buf);
	leaf = btrfs_is_leaf(buf_node);
	buf_leaf = btrfs_buffer_leaf(buf);
	for (i = 0; i < btrfs_header_nritems(&buf_node->header); i++) {
		if (leaf) {
			key = &buf_leaf->items[i].key;
			if (btrfs_disk_key_type(key) != BTRFS_EXTENT_DATA_KEY)
				continue;
			fi = btrfs_item_ptr(buf_leaf, i,
					    struct btrfs_file_extent_item);
			if (btrfs_file_extent_type(fi) ==
			    BTRFS_FILE_EXTENT_INLINE)
				continue;
			ret = btrfs_inc_extent_ref(trans, root,
				    btrfs_file_extent_disk_blocknr(fi),
				    btrfs_file_extent_disk_num_blocks(fi));
			BUG_ON(ret);
		} else {
			blocknr = btrfs_node_blockptr(buf_node, i);
			ret = btrfs_inc_extent_ref(trans, root, blocknr, 1);
			BUG_ON(ret);
		}
	}
	return 0;
}

static int write_one_cache_group(struct btrfs_trans_handle *trans,
				 struct btrfs_root *root,
				 struct btrfs_path *path,
				 struct btrfs_block_group_cache *cache)
{
	int ret;
	int pending_ret;
	struct btrfs_root *extent_root = root->fs_info->extent_root;
	struct btrfs_block_group_item *bi;
	struct btrfs_key ins;

	find_free_extent(trans, extent_root, 0, 0, (u64)-1, &ins, 0);
	ret = btrfs_search_slot(trans, extent_root, &cache->key, path, 0, 1);
	BUG_ON(ret);
	bi = btrfs_item_ptr(btrfs_buffer_leaf(path->nodes[0]), path->slots[0],
			    struct btrfs_block_group_item);
	memcpy(bi, &cache->item, sizeof(*bi));
	mark_buffer_dirty(path->nodes[0]);
	btrfs_release_path(extent_root, path);

	finish_current_insert(trans, extent_root);
	pending_ret = del_pending_extents(trans, extent_root);
	if (ret)
		return ret;
	if (pending_ret)
		return pending_ret;
	if (cache->data)
		cache->last_alloc = cache->first_free;
	return 0;

}

static int write_dirty_block_radix(struct btrfs_trans_handle *trans,
				   struct btrfs_root *root,
				   struct radix_tree_root *radix)
{
	struct btrfs_block_group_cache *cache[8];
	int ret;
	int err = 0;
	int werr = 0;
	int i;
	struct btrfs_path *path;

	path = btrfs_alloc_path();
	if (!path)
		return -ENOMEM;

	while(1) {
		ret = radix_tree_gang_lookup_tag(radix, (void **)cache,
						 0, ARRAY_SIZE(cache),
						 BTRFS_BLOCK_GROUP_DIRTY);
		if (!ret)
			break;
		for (i = 0; i < ret; i++) {
			radix_tree_tag_clear(radix, cache[i]->key.objectid +
					     cache[i]->key.offset - 1,
					     BTRFS_BLOCK_GROUP_DIRTY);
			err = write_one_cache_group(trans, root,
						    path, cache[i]);
			if (err)
				werr = err;
		}
	}
	btrfs_free_path(path);
	return werr;
}

int btrfs_write_dirty_block_groups(struct btrfs_trans_handle *trans,
				   struct btrfs_root *root)
{
	int ret;
	int ret2;
	ret = write_dirty_block_radix(trans, root,
				      &root->fs_info->block_group_radix);
	ret2 = write_dirty_block_radix(trans, root,
				      &root->fs_info->block_group_data_radix);
	if (ret)
		return ret;
	if (ret2)
		return ret2;
	return 0;
}

static int update_block_group(struct btrfs_trans_handle *trans,
			      struct btrfs_root *root,
			      u64 blocknr, u64 num, int alloc)
{
	struct btrfs_block_group_cache *cache;
	struct btrfs_fs_info *info = root->fs_info;
	struct radix_tree_root *radix;
	u64 total = num;
	u64 old_val;
	u64 block_in_group;
	int ret;
	if (num != 1)
		radix = &info->block_group_data_radix;
	else
		radix = &info->block_group_radix;
	while(total) {
		ret = radix_tree_gang_lookup(radix, (void **)&cache,
					     blocknr, 1);
		if (!ret) {
			printk(KERN_CRIT "blocknr %Lu lookup failed\n",
			       blocknr);
			return -1;
		}
		block_in_group = blocknr - cache->key.objectid;
		if (block_in_group > cache->key.offset || cache->key.objectid >
		    blocknr) {
			if (radix == &info->block_group_data_radix)
				radix = &info->block_group_radix;
			else
				radix = &info->block_group_data_radix;
			ret = radix_tree_gang_lookup(radix, (void **)&cache,
						     blocknr, 1);
			if (!ret) {
				printk(KERN_CRIT "blocknr %Lu lookup failed\n",
				       blocknr);
				return -1;
			}
			block_in_group = blocknr - cache->key.objectid;
			if (block_in_group > cache->key.offset ||
			    cache->key.objectid > blocknr) {
				BUG();
			}
		}
		WARN_ON(block_in_group > cache->key.offset);
		radix_tree_tag_set(radix, cache->key.objectid +
				   cache->key.offset - 1,
				   BTRFS_BLOCK_GROUP_DIRTY);

		old_val = btrfs_block_group_used(&cache->item);
		num = min(total, cache->key.offset - block_in_group);
		total -= num;
		blocknr += num;
		if (alloc) {
			old_val += num;
			if (blocknr > cache->last_alloc)
				cache->last_alloc = blocknr;
		} else {
			old_val -= num;
			if (blocknr < cache->first_free)
				cache->first_free = blocknr;
		}
		btrfs_set_block_group_used(&cache->item, old_val);
	}
	return 0;
}

static int try_remove_page(struct address_space *mapping, unsigned long index)
{
	int ret;
	ret = invalidate_mapping_pages(mapping, index, index);
	return ret;
}

int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans, struct
			       btrfs_root *root)
{
	unsigned long gang[8];
	struct inode *btree_inode = root->fs_info->btree_inode;
	struct btrfs_block_group_cache *block_group;
	u64 first = 0;
	int ret;
	int i;
	struct radix_tree_root *pinned_radix = &root->fs_info->pinned_radix;

	while(1) {
		ret = find_first_radix_bit(pinned_radix, gang,
					   ARRAY_SIZE(gang));
		if (!ret)
			break;
		if (!first)
			first = gang[0];
		for (i = 0; i < ret; i++) {
			clear_radix_bit(pinned_radix, gang[i]);
			block_group = lookup_block_group(root->fs_info,
							 gang[i]);
			if (block_group) {
				WARN_ON(block_group->pinned == 0);
				block_group->pinned--;
				if (gang[i] < block_group->last_alloc)
					block_group->last_alloc = gang[i];
			}
			try_remove_page(btree_inode->i_mapping,
					gang[i] << (PAGE_CACHE_SHIFT -
						    btree_inode->i_blkbits));
		}
	}
	return 0;
}

static int finish_current_insert(struct btrfs_trans_handle *trans, struct
				 btrfs_root *extent_root)
{
	struct btrfs_key ins;
	struct btrfs_extent_item extent_item;
	int i;
	int ret;
	u64 super_blocks_used;
	struct btrfs_fs_info *info = extent_root->fs_info;

	btrfs_set_extent_refs(&extent_item, 1);
	ins.offset = 1;
	ins.flags = 0;
	btrfs_set_key_type(&ins, BTRFS_EXTENT_ITEM_KEY);
	btrfs_set_extent_owner(&extent_item, extent_root->root_key.objectid);

	for (i = 0; i < extent_root->fs_info->extent_tree_insert_nr; i++) {
		ins.objectid = extent_root->fs_info->extent_tree_insert[i];
		super_blocks_used = btrfs_super_blocks_used(info->disk_super);
		btrfs_set_super_blocks_used(info->disk_super,
					    super_blocks_used + 1);
		ret = btrfs_insert_item(trans, extent_root, &ins, &extent_item,
					sizeof(extent_item));
		BUG_ON(ret);
	}
	extent_root->fs_info->extent_tree_insert_nr = 0;
	extent_root->fs_info->extent_tree_prealloc_nr = 0;
	return 0;
}

static int pin_down_block(struct btrfs_root *root, u64 blocknr, int pending)
{
	int err;
	struct btrfs_header *header;
	struct buffer_head *bh;

	if (!pending) {
		bh = btrfs_find_tree_block(root, blocknr);
		if (bh) {
			if (buffer_uptodate(bh)) {
				u64 transid =
				    root->fs_info->running_transaction->transid;
				header = btrfs_buffer_header(bh);
				if (btrfs_header_generation(header) ==
				    transid) {
					btrfs_block_release(root, bh);
					return 0;
				}
			}
			btrfs_block_release(root, bh);
		}
		err = set_radix_bit(&root->fs_info->pinned_radix, blocknr);
		if (!err) {
			struct btrfs_block_group_cache *cache;
			cache = lookup_block_group(root->fs_info, blocknr);
			if (cache)
				cache->pinned++;
		}
	} else {
		err = set_radix_bit(&root->fs_info->pending_del_radix, blocknr);
	}
	BUG_ON(err < 0);
	return 0;
}

/*
 * remove an extent from the root, returns 0 on success
 */
static int __free_extent(struct btrfs_trans_handle *trans, struct btrfs_root
			 *root, u64 blocknr, u64 num_blocks, int pin)
{
	struct btrfs_path *path;
	struct btrfs_key key;
	struct btrfs_fs_info *info = root->fs_info;
	struct btrfs_root *extent_root = info->extent_root;
	int ret;
	struct btrfs_extent_item *ei;
	struct btrfs_key ins;
	u32 refs;

	key.objectid = blocknr;
	key.flags = 0;
	btrfs_set_key_type(&key, BTRFS_EXTENT_ITEM_KEY);
	key.offset = num_blocks;

	find_free_extent(trans, root, 0, 0, (u64)-1, &ins, 0);
	path = btrfs_alloc_path();
	BUG_ON(!path);
	btrfs_init_path(path);

	ret = btrfs_search_slot(trans, extent_root, &key, path, -1, 1);
	if (ret) {
		printk("failed to find %Lu\n", key.objectid);
		btrfs_print_tree(extent_root, extent_root->node);
		printk("failed to find %Lu\n", key.objectid);
		BUG();
	}
	ei = btrfs_item_ptr(btrfs_buffer_leaf(path->nodes[0]), path->slots[0],
			    struct btrfs_extent_item);
	BUG_ON(ei->refs == 0);
	refs = btrfs_extent_refs(ei) - 1;
	btrfs_set_extent_refs(ei, refs);
	btrfs_mark_buffer_dirty(path->nodes[0]);
	if (refs == 0) {
		u64 super_blocks_used;

		if (pin) {
			ret = pin_down_block(root, blocknr, 0);
			BUG_ON(ret);
		}

		super_blocks_used = btrfs_super_blocks_used(info->disk_super);
		btrfs_set_super_blocks_used(info->disk_super,
					    super_blocks_used - num_blocks);
		ret = btrfs_del_item(trans, extent_root, path);
		if (ret)
			BUG();
		ret = update_block_group(trans, root, blocknr, num_blocks, 0);
		BUG_ON(ret);
	}
	btrfs_release_path(extent_root, path);
	btrfs_free_path(path);
	finish_current_insert(trans, extent_root);
	return ret;
}

/*
 * find all the blocks marked as pending in the radix tree and remove
 * them from the extent map
 */
static int del_pending_extents(struct btrfs_trans_handle *trans, struct
			       btrfs_root *extent_root)
{
	int ret;
	int wret;
	int err = 0;
	unsigned long gang[4];
	int i;
	struct radix_tree_root *pending_radix;
	struct radix_tree_root *pinned_radix;
	struct btrfs_block_group_cache *cache;

	pending_radix = &extent_root->fs_info->pending_del_radix;
	pinned_radix = &extent_root->fs_info->pinned_radix;

	while(1) {
		ret = find_first_radix_bit(pending_radix, gang,
					   ARRAY_SIZE(gang));
		if (!ret)
			break;
		for (i = 0; i < ret; i++) {
			wret = set_radix_bit(pinned_radix, gang[i]);
			if (wret == 0) {
				cache = lookup_block_group(extent_root->fs_info,
							   gang[i]);
				if (cache)
					cache->pinned++;
			}
			if (wret < 0) {
				printk(KERN_CRIT "set_radix_bit, err %d\n",
				       wret);
				BUG_ON(wret < 0);
			}
			wret = clear_radix_bit(pending_radix, gang[i]);
			BUG_ON(wret);
			wret = __free_extent(trans, extent_root,
					     gang[i], 1, 0);
			if (wret)
				err = wret;
		}
	}
	return err;
}

/*
 * remove an extent from the root, returns 0 on success
 */
int btrfs_free_extent(struct btrfs_trans_handle *trans, struct btrfs_root
		      *root, u64 blocknr, u64 num_blocks, int pin)
{
	struct btrfs_root *extent_root = root->fs_info->extent_root;
	int pending_ret;
	int ret;

	if (root == extent_root) {
		pin_down_block(root, blocknr, 1);
		return 0;
	}
	ret = __free_extent(trans, root, blocknr, num_blocks, pin);
	pending_ret = del_pending_extents(trans, root->fs_info->extent_root);
	return ret ? ret : pending_ret;
}

/*
 * walks the btree of allocated extents and find a hole of a given size.
 * The key ins is changed to record the hole:
 * ins->objectid == block start
 * ins->flags = BTRFS_EXTENT_ITEM_KEY
 * ins->offset == number of blocks
 * Any available blocks before search_start are skipped.
 */
static int find_free_extent(struct btrfs_trans_handle *trans, struct btrfs_root
			    *orig_root, u64 num_blocks, u64 search_start, u64
			    search_end, struct btrfs_key *ins, int data)
{
	struct btrfs_path *path;
	struct btrfs_key key;
	int ret;
	u64 hole_size = 0;
	int slot = 0;
	u64 last_block = 0;
	u64 test_block;
	u64 orig_search_start = search_start;
	int start_found;
	struct btrfs_leaf *l;
	struct btrfs_root * root = orig_root->fs_info->extent_root;
	struct btrfs_fs_info *info = root->fs_info;
	int total_needed = num_blocks;
	int total_found = 0;
	int fill_prealloc = 0;
	int level;
	struct btrfs_block_group_cache *block_group;
	int full_scan = 0;

	path = btrfs_alloc_path();
	ins->flags = 0;
	btrfs_set_key_type(ins, BTRFS_EXTENT_ITEM_KEY);

	level = btrfs_header_level(btrfs_buffer_header(root->node));
	if (num_blocks == 0) {
		fill_prealloc = 1;
		num_blocks = 1;
		total_needed = (min(level + 1, BTRFS_MAX_LEVEL) + 2) * 3;
	}
	if (search_start) {
		block_group = lookup_block_group(info, search_start);
		block_group = btrfs_find_block_group(root, block_group,
						     search_start, data);
	} else {
		block_group = btrfs_find_block_group(root,
						     trans->block_group, 0,
						     data);
	}

check_failed:
	if (block_group->data != data)
		WARN_ON(1);
	if (block_group->last_alloc > search_start)
		search_start = block_group->last_alloc;
	btrfs_init_path(path);
	ins->objectid = search_start;
	ins->offset = 0;
	start_found = 0;
	ret = btrfs_search_slot(trans, root, ins, path, 0, 0);
	if (ret < 0)
		goto error;

	if (path->slots[0] > 0)
		path->slots[0]--;

	while (1) {
		l = btrfs_buffer_leaf(path->nodes[0]);
		slot = path->slots[0];
		if (slot >= btrfs_header_nritems(&l->header)) {
			if (fill_prealloc) {
				info->extent_tree_prealloc_nr = 0;
				total_found = 0;
			}
			ret = btrfs_next_leaf(root, path);
			if (ret == 0)
				continue;
			if (ret < 0)
				goto error;
			if (!start_found) {
				ins->objectid = search_start;
				ins->offset = (u64)-1 - search_start;
				start_found = 1;
				goto check_pending;
			}
			ins->objectid = last_block > search_start ?
					last_block : search_start;
			ins->offset = (u64)-1 - ins->objectid;
			goto check_pending;
		}
		btrfs_disk_key_to_cpu(&key, &l->items[slot].key);
		if (btrfs_key_type(&key) != BTRFS_EXTENT_ITEM_KEY)
			goto next;
		if (key.objectid >= search_start) {
			if (start_found) {
				if (last_block < search_start)
					last_block = search_start;
				hole_size = key.objectid - last_block;
				if (hole_size >= num_blocks) {
					ins->objectid = last_block;
					ins->offset = hole_size;
					goto check_pending;
				}
			}
		}
		start_found = 1;
		last_block = key.objectid + key.offset;
		if (last_block >= block_group->key.objectid +
		    block_group->key.offset) {
			btrfs_release_path(root, path);
			search_start = block_group->key.objectid +
				block_group->key.offset * 2;
			goto new_group;
		}
next:
		path->slots[0]++;
	}
	// FIXME -ENOSPC
check_pending:
	/* we have to make sure we didn't find an extent that has already
	 * been allocated by the map tree or the original allocation
	 */
	btrfs_release_path(root, path);
	BUG_ON(ins->objectid < search_start);
	if (ins->objectid >= btrfs_super_total_blocks(info->disk_super)) {
		if (full_scan)
			return -ENOSPC;
		search_start = orig_search_start;
		full_scan = 1;
		goto new_group;
	}
	for (test_block = ins->objectid;
	     test_block < ins->objectid + num_blocks; test_block++) {
		if (test_radix_bit(&info->pinned_radix, test_block)) {
			search_start = test_block + 1;
			goto new_group;
		}
	}
	if (!fill_prealloc && info->extent_tree_insert_nr) {
		u64 last =
		  info->extent_tree_insert[info->extent_tree_insert_nr - 1];
		if (ins->objectid + num_blocks >
		    info->extent_tree_insert[0] &&
		    ins->objectid <= last) {
			search_start = last + 1;
			WARN_ON(1);
			goto new_group;
		}
	}
	if (!fill_prealloc && info->extent_tree_prealloc_nr) {
		u64 first =
		  info->extent_tree_prealloc[info->extent_tree_prealloc_nr - 1];
		if (ins->objectid + num_blocks > first &&
		    ins->objectid <= info->extent_tree_prealloc[0]) {
			search_start = info->extent_tree_prealloc[0] + 1;
			WARN_ON(1);
			goto new_group;
		}
	}
	if (fill_prealloc) {
		int nr;
		test_block = ins->objectid;
		while(test_block < ins->objectid + ins->offset &&
		      total_found < total_needed) {
			nr = total_needed - total_found - 1;
			BUG_ON(nr < 0);
			info->extent_tree_prealloc[nr] = test_block;
			total_found++;
			test_block++;
		}
		if (total_found < total_needed) {
			search_start = test_block;
			goto new_group;
		}
		info->extent_tree_prealloc_nr = total_found;
	}
	block_group = lookup_block_group(info, ins->objectid);
	if (block_group) {
		block_group->last_alloc = ins->objectid;
		if (!data)
			trans->block_group = block_group;
	}
	ins->offset = num_blocks;
	btrfs_free_path(path);
	return 0;

new_group:
	if (search_start >= btrfs_super_total_blocks(info->disk_super)) {
		search_start = orig_search_start;
		full_scan = 1;
	}
	block_group = lookup_block_group(info, search_start);
	if (!full_scan)
		block_group = btrfs_find_block_group(root, block_group,
						     search_start, data);
	goto check_failed;

error:
	btrfs_release_path(root, path);
	btrfs_free_path(path);
	return ret;
}
/*
 * finds a free extent and does all the dirty work required for allocation
 * returns the key for the extent through ins, and a tree buffer for
 * the first block of the extent through buf.
 *
 * returns 0 if everything worked, non-zero otherwise.
 */
int btrfs_alloc_extent(struct btrfs_trans_handle *trans,
		       struct btrfs_root *root, u64 owner,
		       u64 num_blocks, u64 search_start,
		       u64 search_end, struct btrfs_key *ins, int data)
{
	int ret;
	int pending_ret;
	u64 super_blocks_used;
	struct btrfs_fs_info *info = root->fs_info;
	struct btrfs_root *extent_root = info->extent_root;
	struct btrfs_extent_item extent_item;
	struct btrfs_key prealloc_key;

	btrfs_set_extent_refs(&extent_item, 1);
	btrfs_set_extent_owner(&extent_item, owner);

	if (root == extent_root) {
		int nr;
		BUG_ON(info->extent_tree_prealloc_nr == 0);
		BUG_ON(num_blocks != 1);
		ins->offset = 1;
		info->extent_tree_prealloc_nr--;
		nr = info->extent_tree_prealloc_nr;
		ins->objectid = info->extent_tree_prealloc[nr];
		info->extent_tree_insert[info->extent_tree_insert_nr++] =
			ins->objectid;
		ret = update_block_group(trans, root,
					 ins->objectid, ins->offset, 1);
		BUG_ON(ret);
		return 0;
	}
	/* do the real allocation */
	ret = find_free_extent(trans, root, num_blocks, search_start,
			       search_end, ins, data);
	if (ret)
		return ret;

	/* then do prealloc for the extent tree */
	ret = find_free_extent(trans, root, 0, ins->objectid + ins->offset,
			       search_end, &prealloc_key, 0);
	if (ret)
		return ret;

	super_blocks_used = btrfs_super_blocks_used(info->disk_super);
	btrfs_set_super_blocks_used(info->disk_super, super_blocks_used +
				    num_blocks);
	ret = btrfs_insert_item(trans, extent_root, ins, &extent_item,
				sizeof(extent_item));

	finish_current_insert(trans, extent_root);
	pending_ret = del_pending_extents(trans, extent_root);
	if (ret)
		return ret;
	if (pending_ret)
		return pending_ret;
	ret = update_block_group(trans, root, ins->objectid, ins->offset, 1);
	return 0;
}

/*
 * helper function to allocate a block for a given tree
 * returns the tree buffer or NULL.
 */
struct buffer_head *btrfs_alloc_free_block(struct btrfs_trans_handle *trans,
					   struct btrfs_root *root, u64 hint)
{
	struct btrfs_key ins;
	int ret;
	struct buffer_head *buf;

	ret = btrfs_alloc_extent(trans, root, root->root_key.objectid,
				 1, hint, (unsigned long)-1, &ins, 0);
	if (ret) {
		BUG();
		return NULL;
	}
	BUG_ON(ret);
	buf = btrfs_find_create_tree_block(root, ins.objectid);
	set_buffer_uptodate(buf);
	set_buffer_checked(buf);
	set_radix_bit(&trans->transaction->dirty_pages, buf->b_page->index);
	return buf;
}

static int drop_leaf_ref(struct btrfs_trans_handle *trans,
			 struct btrfs_root *root, struct buffer_head *cur)
{
	struct btrfs_disk_key *key;
	struct btrfs_leaf *leaf;
	struct btrfs_file_extent_item *fi;
	int i;
	int nritems;
	int ret;

	BUG_ON(!btrfs_is_leaf(btrfs_buffer_node(cur)));
	leaf = btrfs_buffer_leaf(cur);
	nritems = btrfs_header_nritems(&leaf->header);
	for (i = 0; i < nritems; i++) {
		key = &leaf->items[i].key;
		if (btrfs_disk_key_type(key) != BTRFS_EXTENT_DATA_KEY)
			continue;
		fi = btrfs_item_ptr(leaf, i, struct btrfs_file_extent_item);
		if (btrfs_file_extent_type(fi) == BTRFS_FILE_EXTENT_INLINE)
			continue;
		/*
		 * FIXME make sure to insert a trans record that
		 * repeats the snapshot del on crash
		 */
		ret = btrfs_free_extent(trans, root,
					btrfs_file_extent_disk_blocknr(fi),
					btrfs_file_extent_disk_num_blocks(fi),
					0);
		BUG_ON(ret);
	}
	return 0;
}

/*
 * helper function for drop_snapshot, this walks down the tree dropping ref
 * counts as it goes.
 */
static int walk_down_tree(struct btrfs_trans_handle *trans, struct btrfs_root
			  *root, struct btrfs_path *path, int *level)
{
	struct buffer_head *next;
	struct buffer_head *cur;
	u64 blocknr;
	int ret;
	u32 refs;

	WARN_ON(*level < 0);
	WARN_ON(*level >= BTRFS_MAX_LEVEL);
	ret = lookup_extent_ref(trans, root, bh_blocknr(path->nodes[*level]),
			       1, &refs);
	BUG_ON(ret);
	if (refs > 1)
		goto out;
	/*
	 * walk down to the last node level and free all the leaves
	 */
	while(*level >= 0) {
		WARN_ON(*level < 0);
		WARN_ON(*level >= BTRFS_MAX_LEVEL);
		cur = path->nodes[*level];
		if (btrfs_header_level(btrfs_buffer_header(cur)) != *level)
			WARN_ON(1);
		if (path->slots[*level] >=
		    btrfs_header_nritems(btrfs_buffer_header(cur)))
			break;
		if (*level == 0) {
			ret = drop_leaf_ref(trans, root, cur);
			BUG_ON(ret);
			break;
		}
		blocknr = btrfs_node_blockptr(btrfs_buffer_node(cur),
					      path->slots[*level]);
		ret = lookup_extent_ref(trans, root, blocknr, 1, &refs);
		BUG_ON(ret);
		if (refs != 1) {
			path->slots[*level]++;
			ret = btrfs_free_extent(trans, root, blocknr, 1, 1);
			BUG_ON(ret);
			continue;
		}
		next = read_tree_block(root, blocknr);
		WARN_ON(*level <= 0);
		if (path->nodes[*level-1])
			btrfs_block_release(root, path->nodes[*level-1]);
		path->nodes[*level-1] = next;
		*level = btrfs_header_level(btrfs_buffer_header(next));
		path->slots[*level] = 0;
	}
out:
	WARN_ON(*level < 0);
	WARN_ON(*level >= BTRFS_MAX_LEVEL);
	ret = btrfs_free_extent(trans, root,
				bh_blocknr(path->nodes[*level]), 1, 1);
	btrfs_block_release(root, path->nodes[*level]);
	path->nodes[*level] = NULL;
	*level += 1;
	BUG_ON(ret);
	return 0;
}

/*
 * helper for dropping snapshots.  This walks back up the tree in the path
 * to find the first node higher up where we haven't yet gone through
 * all the slots
 */
static int walk_up_tree(struct btrfs_trans_handle *trans, struct btrfs_root
			*root, struct btrfs_path *path, int *level)
{
	int i;
	int slot;
	int ret;
	for(i = *level; i < BTRFS_MAX_LEVEL - 1 && path->nodes[i]; i++) {
		slot = path->slots[i];
		if (slot < btrfs_header_nritems(
		    btrfs_buffer_header(path->nodes[i])) - 1) {
			path->slots[i]++;
			*level = i;
			return 0;
		} else {
			ret = btrfs_free_extent(trans, root,
						bh_blocknr(path->nodes[*level]),
						1, 1);
			BUG_ON(ret);
			btrfs_block_release(root, path->nodes[*level]);
			path->nodes[*level] = NULL;
			*level = i + 1;
		}
	}
	return 1;
}

/*
 * drop the reference count on the tree rooted at 'snap'.  This traverses
 * the tree freeing any blocks that have a ref count of zero after being
 * decremented.
 */
int btrfs_drop_snapshot(struct btrfs_trans_handle *trans, struct btrfs_root
			*root, struct buffer_head *snap)
{
	int ret = 0;
	int wret;
	int level;
	struct btrfs_path *path;
	int i;
	int orig_level;

	path = btrfs_alloc_path();
	BUG_ON(!path);
	btrfs_init_path(path);

	level = btrfs_header_level(btrfs_buffer_header(snap));
	orig_level = level;
	path->nodes[level] = snap;
	path->slots[level] = 0;
	while(1) {
		wret = walk_down_tree(trans, root, path, &level);
		if (wret > 0)
			break;
		if (wret < 0)
			ret = wret;

		wret = walk_up_tree(trans, root, path, &level);
		if (wret > 0)
			break;
		if (wret < 0)
			ret = wret;
		btrfs_btree_balance_dirty(root);
	}
	for (i = 0; i <= orig_level; i++) {
		if (path->nodes[i]) {
			btrfs_block_release(root, path->nodes[i]);
		}
	}
	btrfs_free_path(path);
	return ret;
}

static int free_block_group_radix(struct radix_tree_root *radix)
{
	int ret;
	struct btrfs_block_group_cache *cache[8];
	int i;

	while(1) {
		ret = radix_tree_gang_lookup(radix, (void **)cache, 0,
					     ARRAY_SIZE(cache));
		if (!ret)
			break;
		for (i = 0; i < ret; i++) {
			radix_tree_delete(radix, cache[i]->key.objectid +
					  cache[i]->key.offset - 1);
			kfree(cache[i]);
		}
	}
	return 0;
}

int btrfs_free_block_groups(struct btrfs_fs_info *info)
{
	int ret;
	int ret2;

	ret = free_block_group_radix(&info->block_group_radix);
	ret2 = free_block_group_radix(&info->block_group_data_radix);
	if (ret)
		return ret;
	if (ret2)
		return ret2;
	return 0;
}

int btrfs_read_block_groups(struct btrfs_root *root)
{
	struct btrfs_path *path;
	int ret;
	int err = 0;
	struct btrfs_block_group_item *bi;
	struct btrfs_block_group_cache *cache;
	struct btrfs_fs_info *info = root->fs_info;
	struct radix_tree_root *radix;
	struct btrfs_key key;
	struct btrfs_key found_key;
	struct btrfs_leaf *leaf;
	u64 group_size_blocks = BTRFS_BLOCK_GROUP_SIZE / root->blocksize;
	u64 used;
	u64 nr = 0;

	root = info->extent_root;
	key.objectid = 0;
	key.offset = group_size_blocks;
	key.flags = 0;
	btrfs_set_key_type(&key, BTRFS_BLOCK_GROUP_ITEM_KEY);

	path = btrfs_alloc_path();
	if (!path)
		return -ENOMEM;

	while(1) {
		ret = btrfs_search_slot(NULL, info->extent_root,
					&key, path, 0, 0);
		if (ret != 0) {
			err = ret;
			break;
		}
		leaf = btrfs_buffer_leaf(path->nodes[0]);
		btrfs_disk_key_to_cpu(&found_key,
				      &leaf->items[path->slots[0]].key);
		cache = kmalloc(sizeof(*cache), GFP_NOFS);
		if (!cache) {
			err = -1;
			break;
		}
		bi = btrfs_item_ptr(leaf, path->slots[0],
				    struct btrfs_block_group_item);
		memcpy(&cache->item, bi, sizeof(*bi));
		memcpy(&cache->key, &found_key, sizeof(found_key));
		cache->last_alloc = cache->key.objectid;
		cache->first_free = cache->key.objectid;
		cache->pinned = 0;
		cache->data = (nr & 1);
		key.objectid = found_key.objectid + found_key.offset;
		btrfs_release_path(root, path);
		if (nr & 1)
			radix = &info->block_group_data_radix;
		else
			radix = &info->block_group_radix;
		ret = radix_tree_insert(radix, found_key.objectid +
					found_key.offset - 1,
					(void *)cache);
		BUG_ON(ret);
		used = btrfs_block_group_used(bi);
		if (used < (key.offset * 8) / 10) {
			radix_tree_tag_set(radix, found_key.objectid +
					   found_key.offset - 1,
					   BTRFS_BLOCK_GROUP_AVAIL);
		}
		if (key.objectid >=
		    btrfs_super_total_blocks(info->disk_super))
			break;
		nr++;
	}

	btrfs_free_path(path);
	return 0;
}