Free Electrons

Embedded Linux Experts

   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
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
#include "dm.h"
#include "persistent-data/dm-transaction-manager.h"
#include "persistent-data/dm-bitset.h"
#include "persistent-data/dm-space-map.h"

#include <linux/dm-io.h>
#include <linux/dm-kcopyd.h>
#include <linux/init.h>
#include <linux/mempool.h>
#include <linux/module.h>
#include <linux/slab.h>
#include <linux/vmalloc.h>

#define DM_MSG_PREFIX "era"

#define SUPERBLOCK_LOCATION 0
#define SUPERBLOCK_MAGIC 2126579579
#define SUPERBLOCK_CSUM_XOR 146538381
#define MIN_ERA_VERSION 1
#define MAX_ERA_VERSION 1
#define INVALID_WRITESET_ROOT SUPERBLOCK_LOCATION
#define MIN_BLOCK_SIZE 8

/*----------------------------------------------------------------
 * Writeset
 *--------------------------------------------------------------*/
struct writeset_metadata {
	uint32_t nr_bits;
	dm_block_t root;
};

struct writeset {
	struct writeset_metadata md;

	/*
	 * An in core copy of the bits to save constantly doing look ups on
	 * disk.
	 */
	unsigned long *bits;
};

/*
 * This does not free off the on disk bitset as this will normally be done
 * after digesting into the era array.
 */
static void writeset_free(struct writeset *ws)
{
	vfree(ws->bits);
}

static int setup_on_disk_bitset(struct dm_disk_bitset *info,
				unsigned nr_bits, dm_block_t *root)
{
	int r;

	r = dm_bitset_empty(info, root);
	if (r)
		return r;

	return dm_bitset_resize(info, *root, 0, nr_bits, false, root);
}

static size_t bitset_size(unsigned nr_bits)
{
	return sizeof(unsigned long) * dm_div_up(nr_bits, BITS_PER_LONG);
}

/*
 * Allocates memory for the in core bitset.
 */
static int writeset_alloc(struct writeset *ws, dm_block_t nr_blocks)
{
	ws->md.nr_bits = nr_blocks;
	ws->md.root = INVALID_WRITESET_ROOT;
	ws->bits = vzalloc(bitset_size(nr_blocks));
	if (!ws->bits) {
		DMERR("%s: couldn't allocate in memory bitset", __func__);
		return -ENOMEM;
	}

	return 0;
}

/*
 * Wipes the in-core bitset, and creates a new on disk bitset.
 */
static int writeset_init(struct dm_disk_bitset *info, struct writeset *ws)
{
	int r;

	memset(ws->bits, 0, bitset_size(ws->md.nr_bits));

	r = setup_on_disk_bitset(info, ws->md.nr_bits, &ws->md.root);
	if (r) {
		DMERR("%s: setup_on_disk_bitset failed", __func__);
		return r;
	}

	return 0;
}

static bool writeset_marked(struct writeset *ws, dm_block_t block)
{
	return test_bit(block, ws->bits);
}

static int writeset_marked_on_disk(struct dm_disk_bitset *info,
				   struct writeset_metadata *m, dm_block_t block,
				   bool *result)
{
	dm_block_t old = m->root;

	/*
	 * The bitset was flushed when it was archived, so we know there'll
	 * be no change to the root.
	 */
	int r = dm_bitset_test_bit(info, m->root, block, &m->root, result);
	if (r) {
		DMERR("%s: dm_bitset_test_bit failed", __func__);
		return r;
	}

	BUG_ON(m->root != old);

	return r;
}

/*
 * Returns < 0 on error, 0 if the bit wasn't previously set, 1 if it was.
 */
static int writeset_test_and_set(struct dm_disk_bitset *info,
				 struct writeset *ws, uint32_t block)
{
	int r;

	if (!test_and_set_bit(block, ws->bits)) {
		r = dm_bitset_set_bit(info, ws->md.root, block, &ws->md.root);
		if (r) {
			/* FIXME: fail mode */
			return r;
		}

		return 0;
	}

	return 1;
}

/*----------------------------------------------------------------
 * On disk metadata layout
 *--------------------------------------------------------------*/
#define SPACE_MAP_ROOT_SIZE 128
#define UUID_LEN 16

struct writeset_disk {
	__le32 nr_bits;
	__le64 root;
} __packed;

struct superblock_disk {
	__le32 csum;
	__le32 flags;
	__le64 blocknr;

	__u8 uuid[UUID_LEN];
	__le64 magic;
	__le32 version;

	__u8 metadata_space_map_root[SPACE_MAP_ROOT_SIZE];

	__le32 data_block_size;
	__le32 metadata_block_size;
	__le32 nr_blocks;

	__le32 current_era;
	struct writeset_disk current_writeset;

	/*
	 * Only these two fields are valid within the metadata snapshot.
	 */
	__le64 writeset_tree_root;
	__le64 era_array_root;

	__le64 metadata_snap;
} __packed;

/*----------------------------------------------------------------
 * Superblock validation
 *--------------------------------------------------------------*/
static void sb_prepare_for_write(struct dm_block_validator *v,
				 struct dm_block *b,
				 size_t sb_block_size)
{
	struct superblock_disk *disk = dm_block_data(b);

	disk->blocknr = cpu_to_le64(dm_block_location(b));
	disk->csum = cpu_to_le32(dm_bm_checksum(&disk->flags,
						sb_block_size - sizeof(__le32),
						SUPERBLOCK_CSUM_XOR));
}

static int check_metadata_version(struct superblock_disk *disk)
{
	uint32_t metadata_version = le32_to_cpu(disk->version);
	if (metadata_version < MIN_ERA_VERSION || metadata_version > MAX_ERA_VERSION) {
		DMERR("Era metadata version %u found, but only versions between %u and %u supported.",
		      metadata_version, MIN_ERA_VERSION, MAX_ERA_VERSION);
		return -EINVAL;
	}

	return 0;
}

static int sb_check(struct dm_block_validator *v,
		    struct dm_block *b,
		    size_t sb_block_size)
{
	struct superblock_disk *disk = dm_block_data(b);
	__le32 csum_le;

	if (dm_block_location(b) != le64_to_cpu(disk->blocknr)) {
		DMERR("sb_check failed: blocknr %llu: wanted %llu",
		      le64_to_cpu(disk->blocknr),
		      (unsigned long long)dm_block_location(b));
		return -ENOTBLK;
	}

	if (le64_to_cpu(disk->magic) != SUPERBLOCK_MAGIC) {
		DMERR("sb_check failed: magic %llu: wanted %llu",
		      le64_to_cpu(disk->magic),
		      (unsigned long long) SUPERBLOCK_MAGIC);
		return -EILSEQ;
	}

	csum_le = cpu_to_le32(dm_bm_checksum(&disk->flags,
					     sb_block_size - sizeof(__le32),
					     SUPERBLOCK_CSUM_XOR));
	if (csum_le != disk->csum) {
		DMERR("sb_check failed: csum %u: wanted %u",
		      le32_to_cpu(csum_le), le32_to_cpu(disk->csum));
		return -EILSEQ;
	}

	return check_metadata_version(disk);
}

static struct dm_block_validator sb_validator = {
	.name = "superblock",
	.prepare_for_write = sb_prepare_for_write,
	.check = sb_check
};

/*----------------------------------------------------------------
 * Low level metadata handling
 *--------------------------------------------------------------*/
#define DM_ERA_METADATA_BLOCK_SIZE 4096
#define DM_ERA_METADATA_CACHE_SIZE 64
#define ERA_MAX_CONCURRENT_LOCKS 5

struct era_metadata {
	struct block_device *bdev;
	struct dm_block_manager *bm;
	struct dm_space_map *sm;
	struct dm_transaction_manager *tm;

	dm_block_t block_size;
	uint32_t nr_blocks;

	uint32_t current_era;

	/*
	 * We preallocate 2 writesets.  When an era rolls over we
	 * switch between them. This means the allocation is done at
	 * preresume time, rather than on the io path.
	 */
	struct writeset writesets[2];
	struct writeset *current_writeset;

	dm_block_t writeset_tree_root;
	dm_block_t era_array_root;

	struct dm_disk_bitset bitset_info;
	struct dm_btree_info writeset_tree_info;
	struct dm_array_info era_array_info;

	dm_block_t metadata_snap;

	/*
	 * A flag that is set whenever a writeset has been archived.
	 */
	bool archived_writesets;

	/*
	 * Reading the space map root can fail, so we read it into this
	 * buffer before the superblock is locked and updated.
	 */
	__u8 metadata_space_map_root[SPACE_MAP_ROOT_SIZE];
};

static int superblock_read_lock(struct era_metadata *md,
				struct dm_block **sblock)
{
	return dm_bm_read_lock(md->bm, SUPERBLOCK_LOCATION,
			       &sb_validator, sblock);
}

static int superblock_lock_zero(struct era_metadata *md,
				struct dm_block **sblock)
{
	return dm_bm_write_lock_zero(md->bm, SUPERBLOCK_LOCATION,
				     &sb_validator, sblock);
}

static int superblock_lock(struct era_metadata *md,
			   struct dm_block **sblock)
{
	return dm_bm_write_lock(md->bm, SUPERBLOCK_LOCATION,
				&sb_validator, sblock);
}

/* FIXME: duplication with cache and thin */
static int superblock_all_zeroes(struct dm_block_manager *bm, bool *result)
{
	int r;
	unsigned i;
	struct dm_block *b;
	__le64 *data_le, zero = cpu_to_le64(0);
	unsigned sb_block_size = dm_bm_block_size(bm) / sizeof(__le64);

	/*
	 * We can't use a validator here - it may be all zeroes.
	 */
	r = dm_bm_read_lock(bm, SUPERBLOCK_LOCATION, NULL, &b);
	if (r)
		return r;

	data_le = dm_block_data(b);
	*result = true;
	for (i = 0; i < sb_block_size; i++) {
		if (data_le[i] != zero) {
			*result = false;
			break;
		}
	}

	return dm_bm_unlock(b);
}

/*----------------------------------------------------------------*/

static void ws_pack(const struct writeset_metadata *core, struct writeset_disk *disk)
{
	disk->nr_bits = cpu_to_le32(core->nr_bits);
	disk->root = cpu_to_le64(core->root);
}

static void ws_unpack(const struct writeset_disk *disk, struct writeset_metadata *core)
{
	core->nr_bits = le32_to_cpu(disk->nr_bits);
	core->root = le64_to_cpu(disk->root);
}

static void ws_inc(void *context, const void *value)
{
	struct era_metadata *md = context;
	struct writeset_disk ws_d;
	dm_block_t b;

	memcpy(&ws_d, value, sizeof(ws_d));
	b = le64_to_cpu(ws_d.root);

	dm_tm_inc(md->tm, b);
}

static void ws_dec(void *context, const void *value)
{
	struct era_metadata *md = context;
	struct writeset_disk ws_d;
	dm_block_t b;

	memcpy(&ws_d, value, sizeof(ws_d));
	b = le64_to_cpu(ws_d.root);

	dm_bitset_del(&md->bitset_info, b);
}

static int ws_eq(void *context, const void *value1, const void *value2)
{
	return !memcmp(value1, value2, sizeof(struct writeset_metadata));
}

/*----------------------------------------------------------------*/

static void setup_writeset_tree_info(struct era_metadata *md)
{
	struct dm_btree_value_type *vt = &md->writeset_tree_info.value_type;
	md->writeset_tree_info.tm = md->tm;
	md->writeset_tree_info.levels = 1;
	vt->context = md;
	vt->size = sizeof(struct writeset_disk);
	vt->inc = ws_inc;
	vt->dec = ws_dec;
	vt->equal = ws_eq;
}

static void setup_era_array_info(struct era_metadata *md)

{
	struct dm_btree_value_type vt;
	vt.context = NULL;
	vt.size = sizeof(__le32);
	vt.inc = NULL;
	vt.dec = NULL;
	vt.equal = NULL;

	dm_array_info_init(&md->era_array_info, md->tm, &vt);
}

static void setup_infos(struct era_metadata *md)
{
	dm_disk_bitset_init(md->tm, &md->bitset_info);
	setup_writeset_tree_info(md);
	setup_era_array_info(md);
}

/*----------------------------------------------------------------*/

static int create_fresh_metadata(struct era_metadata *md)
{
	int r;

	r = dm_tm_create_with_sm(md->bm, SUPERBLOCK_LOCATION,
				 &md->tm, &md->sm);
	if (r < 0) {
		DMERR("dm_tm_create_with_sm failed");
		return r;
	}

	setup_infos(md);

	r = dm_btree_empty(&md->writeset_tree_info, &md->writeset_tree_root);
	if (r) {
		DMERR("couldn't create new writeset tree");
		goto bad;
	}

	r = dm_array_empty(&md->era_array_info, &md->era_array_root);
	if (r) {
		DMERR("couldn't create era array");
		goto bad;
	}

	return 0;

bad:
	dm_sm_destroy(md->sm);
	dm_tm_destroy(md->tm);

	return r;
}

static int save_sm_root(struct era_metadata *md)
{
	int r;
	size_t metadata_len;

	r = dm_sm_root_size(md->sm, &metadata_len);
	if (r < 0)
		return r;

	return dm_sm_copy_root(md->sm, &md->metadata_space_map_root,
			       metadata_len);
}

static void copy_sm_root(struct era_metadata *md, struct superblock_disk *disk)
{
	memcpy(&disk->metadata_space_map_root,
	       &md->metadata_space_map_root,
	       sizeof(md->metadata_space_map_root));
}

/*
 * Writes a superblock, including the static fields that don't get updated
 * with every commit (possible optimisation here).  'md' should be fully
 * constructed when this is called.
 */
static void prepare_superblock(struct era_metadata *md, struct superblock_disk *disk)
{
	disk->magic = cpu_to_le64(SUPERBLOCK_MAGIC);
	disk->flags = cpu_to_le32(0ul);

	/* FIXME: can't keep blanking the uuid (uuid is currently unused though) */
	memset(disk->uuid, 0, sizeof(disk->uuid));
	disk->version = cpu_to_le32(MAX_ERA_VERSION);

	copy_sm_root(md, disk);

	disk->data_block_size = cpu_to_le32(md->block_size);
	disk->metadata_block_size = cpu_to_le32(DM_ERA_METADATA_BLOCK_SIZE >> SECTOR_SHIFT);
	disk->nr_blocks = cpu_to_le32(md->nr_blocks);
	disk->current_era = cpu_to_le32(md->current_era);

	ws_pack(&md->current_writeset->md, &disk->current_writeset);
	disk->writeset_tree_root = cpu_to_le64(md->writeset_tree_root);
	disk->era_array_root = cpu_to_le64(md->era_array_root);
	disk->metadata_snap = cpu_to_le64(md->metadata_snap);
}

static int write_superblock(struct era_metadata *md)
{
	int r;
	struct dm_block *sblock;
	struct superblock_disk *disk;

	r = save_sm_root(md);
	if (r) {
		DMERR("%s: save_sm_root failed", __func__);
		return r;
	}

	r = superblock_lock_zero(md, &sblock);
	if (r)
		return r;

	disk = dm_block_data(sblock);
	prepare_superblock(md, disk);

	return dm_tm_commit(md->tm, sblock);
}

/*
 * Assumes block_size and the infos are set.
 */
static int format_metadata(struct era_metadata *md)
{
	int r;

	r = create_fresh_metadata(md);
	if (r)
		return r;

	r = write_superblock(md);
	if (r) {
		dm_sm_destroy(md->sm);
		dm_tm_destroy(md->tm);
		return r;
	}

	return 0;
}

static int open_metadata(struct era_metadata *md)
{
	int r;
	struct dm_block *sblock;
	struct superblock_disk *disk;

	r = superblock_read_lock(md, &sblock);
	if (r) {
		DMERR("couldn't read_lock superblock");
		return r;
	}

	disk = dm_block_data(sblock);
	r = dm_tm_open_with_sm(md->bm, SUPERBLOCK_LOCATION,
			       disk->metadata_space_map_root,
			       sizeof(disk->metadata_space_map_root),
			       &md->tm, &md->sm);
	if (r) {
		DMERR("dm_tm_open_with_sm failed");
		goto bad;
	}

	setup_infos(md);

	md->block_size = le32_to_cpu(disk->data_block_size);
	md->nr_blocks = le32_to_cpu(disk->nr_blocks);
	md->current_era = le32_to_cpu(disk->current_era);

	md->writeset_tree_root = le64_to_cpu(disk->writeset_tree_root);
	md->era_array_root = le64_to_cpu(disk->era_array_root);
	md->metadata_snap = le64_to_cpu(disk->metadata_snap);
	md->archived_writesets = true;

	return dm_bm_unlock(sblock);

bad:
	dm_bm_unlock(sblock);
	return r;
}

static int open_or_format_metadata(struct era_metadata *md,
				   bool may_format)
{
	int r;
	bool unformatted = false;

	r = superblock_all_zeroes(md->bm, &unformatted);
	if (r)
		return r;

	if (unformatted)
		return may_format ? format_metadata(md) : -EPERM;

	return open_metadata(md);
}

static int create_persistent_data_objects(struct era_metadata *md,
					  bool may_format)
{
	int r;

	md->bm = dm_block_manager_create(md->bdev, DM_ERA_METADATA_BLOCK_SIZE,
					 DM_ERA_METADATA_CACHE_SIZE,
					 ERA_MAX_CONCURRENT_LOCKS);
	if (IS_ERR(md->bm)) {
		DMERR("could not create block manager");
		return PTR_ERR(md->bm);
	}

	r = open_or_format_metadata(md, may_format);
	if (r)
		dm_block_manager_destroy(md->bm);

	return r;
}

static void destroy_persistent_data_objects(struct era_metadata *md)
{
	dm_sm_destroy(md->sm);
	dm_tm_destroy(md->tm);
	dm_block_manager_destroy(md->bm);
}

/*
 * This waits until all era_map threads have picked up the new filter.
 */
static void swap_writeset(struct era_metadata *md, struct writeset *new_writeset)
{
	rcu_assign_pointer(md->current_writeset, new_writeset);
	synchronize_rcu();
}

/*----------------------------------------------------------------
 * Writesets get 'digested' into the main era array.
 *
 * We're using a coroutine here so the worker thread can do the digestion,
 * thus avoiding synchronisation of the metadata.  Digesting a whole
 * writeset in one go would cause too much latency.
 *--------------------------------------------------------------*/
struct digest {
	uint32_t era;
	unsigned nr_bits, current_bit;
	struct writeset_metadata writeset;
	__le32 value;
	struct dm_disk_bitset info;

	int (*step)(struct era_metadata *, struct digest *);
};

static int metadata_digest_lookup_writeset(struct era_metadata *md,
					   struct digest *d);

static int metadata_digest_remove_writeset(struct era_metadata *md,
					   struct digest *d)
{
	int r;
	uint64_t key = d->era;

	r = dm_btree_remove(&md->writeset_tree_info, md->writeset_tree_root,
			    &key, &md->writeset_tree_root);
	if (r) {
		DMERR("%s: dm_btree_remove failed", __func__);
		return r;
	}

	d->step = metadata_digest_lookup_writeset;
	return 0;
}

#define INSERTS_PER_STEP 100

static int metadata_digest_transcribe_writeset(struct era_metadata *md,
					       struct digest *d)
{
	int r;
	bool marked;
	unsigned b, e = min(d->current_bit + INSERTS_PER_STEP, d->nr_bits);

	for (b = d->current_bit; b < e; b++) {
		r = writeset_marked_on_disk(&d->info, &d->writeset, b, &marked);
		if (r) {
			DMERR("%s: writeset_marked_on_disk failed", __func__);
			return r;
		}

		if (!marked)
			continue;

		__dm_bless_for_disk(&d->value);
		r = dm_array_set_value(&md->era_array_info, md->era_array_root,
				       b, &d->value, &md->era_array_root);
		if (r) {
			DMERR("%s: dm_array_set_value failed", __func__);
			return r;
		}
	}

	if (b == d->nr_bits)
		d->step = metadata_digest_remove_writeset;
	else
		d->current_bit = b;

	return 0;
}

static int metadata_digest_lookup_writeset(struct era_metadata *md,
					   struct digest *d)
{
	int r;
	uint64_t key;
	struct writeset_disk disk;

	r = dm_btree_find_lowest_key(&md->writeset_tree_info,
				     md->writeset_tree_root, &key);
	if (r < 0)
		return r;

	d->era = key;

	r = dm_btree_lookup(&md->writeset_tree_info,
			    md->writeset_tree_root, &key, &disk);
	if (r) {
		if (r == -ENODATA) {
			d->step = NULL;
			return 0;
		}

		DMERR("%s: dm_btree_lookup failed", __func__);
		return r;
	}

	ws_unpack(&disk, &d->writeset);
	d->value = cpu_to_le32(key);

	d->nr_bits = min(d->writeset.nr_bits, md->nr_blocks);
	d->current_bit = 0;
	d->step = metadata_digest_transcribe_writeset;

	return 0;
}

static int metadata_digest_start(struct era_metadata *md, struct digest *d)
{
	if (d->step)
		return 0;

	memset(d, 0, sizeof(*d));

	/*
	 * We initialise another bitset info to avoid any caching side
	 * effects with the previous one.
	 */
	dm_disk_bitset_init(md->tm, &d->info);
	d->step = metadata_digest_lookup_writeset;

	return 0;
}

/*----------------------------------------------------------------
 * High level metadata interface.  Target methods should use these, and not
 * the lower level ones.
 *--------------------------------------------------------------*/
static struct era_metadata *metadata_open(struct block_device *bdev,
					  sector_t block_size,
					  bool may_format)
{
	int r;
	struct era_metadata *md = kzalloc(sizeof(*md), GFP_KERNEL);

	if (!md)
		return NULL;

	md->bdev = bdev;
	md->block_size = block_size;

	md->writesets[0].md.root = INVALID_WRITESET_ROOT;
	md->writesets[1].md.root = INVALID_WRITESET_ROOT;
	md->current_writeset = &md->writesets[0];

	r = create_persistent_data_objects(md, may_format);
	if (r) {
		kfree(md);
		return ERR_PTR(r);
	}

	return md;
}

static void metadata_close(struct era_metadata *md)
{
	destroy_persistent_data_objects(md);
	kfree(md);
}

static bool valid_nr_blocks(dm_block_t n)
{
	/*
	 * dm_bitset restricts us to 2^32.  test_bit & co. restrict us
	 * further to 2^31 - 1
	 */
	return n < (1ull << 31);
}

static int metadata_resize(struct era_metadata *md, void *arg)
{
	int r;
	dm_block_t *new_size = arg;
	__le32 value;

	if (!valid_nr_blocks(*new_size)) {
		DMERR("Invalid number of origin blocks %llu",
		      (unsigned long long) *new_size);
		return -EINVAL;
	}

	writeset_free(&md->writesets[0]);
	writeset_free(&md->writesets[1]);

	r = writeset_alloc(&md->writesets[0], *new_size);
	if (r) {
		DMERR("%s: writeset_alloc failed for writeset 0", __func__);
		return r;
	}

	r = writeset_alloc(&md->writesets[1], *new_size);
	if (r) {
		DMERR("%s: writeset_alloc failed for writeset 1", __func__);
		return r;
	}

	value = cpu_to_le32(0u);
	__dm_bless_for_disk(&value);
	r = dm_array_resize(&md->era_array_info, md->era_array_root,
			    md->nr_blocks, *new_size,
			    &value, &md->era_array_root);
	if (r) {
		DMERR("%s: dm_array_resize failed", __func__);
		return r;
	}

	md->nr_blocks = *new_size;
	return 0;
}

static int metadata_era_archive(struct era_metadata *md)
{
	int r;
	uint64_t keys[1];
	struct writeset_disk value;

	r = dm_bitset_flush(&md->bitset_info, md->current_writeset->md.root,
			    &md->current_writeset->md.root);
	if (r) {
		DMERR("%s: dm_bitset_flush failed", __func__);
		return r;
	}

	ws_pack(&md->current_writeset->md, &value);
	md->current_writeset->md.root = INVALID_WRITESET_ROOT;

	keys[0] = md->current_era;
	__dm_bless_for_disk(&value);
	r = dm_btree_insert(&md->writeset_tree_info, md->writeset_tree_root,
			    keys, &value, &md->writeset_tree_root);
	if (r) {
		DMERR("%s: couldn't insert writeset into btree", __func__);
		/* FIXME: fail mode */
		return r;
	}

	md->archived_writesets = true;

	return 0;
}

static struct writeset *next_writeset(struct era_metadata *md)
{
	return (md->current_writeset == &md->writesets[0]) ?
		&md->writesets[1] : &md->writesets[0];
}

static int metadata_new_era(struct era_metadata *md)
{
	int r;
	struct writeset *new_writeset = next_writeset(md);

	r = writeset_init(&md->bitset_info, new_writeset);
	if (r) {
		DMERR("%s: writeset_init failed", __func__);
		return r;
	}

	swap_writeset(md, new_writeset);
	md->current_era++;

	return 0;
}

static int metadata_era_rollover(struct era_metadata *md)
{
	int r;

	if (md->current_writeset->md.root != INVALID_WRITESET_ROOT) {
		r = metadata_era_archive(md);
		if (r) {
			DMERR("%s: metadata_archive_era failed", __func__);
			/* FIXME: fail mode? */
			return r;
		}
	}

	r = metadata_new_era(md);
	if (r) {
		DMERR("%s: new era failed", __func__);
		/* FIXME: fail mode */
		return r;
	}

	return 0;
}

static bool metadata_current_marked(struct era_metadata *md, dm_block_t block)
{
	bool r;
	struct writeset *ws;

	rcu_read_lock();
	ws = rcu_dereference(md->current_writeset);
	r = writeset_marked(ws, block);
	rcu_read_unlock();

	return r;
}

static int metadata_commit(struct era_metadata *md)
{
	int r;
	struct dm_block *sblock;

	if (md->current_writeset->md.root != SUPERBLOCK_LOCATION) {
		r = dm_bitset_flush(&md->bitset_info, md->current_writeset->md.root,
				    &md->current_writeset->md.root);
		if (r) {
			DMERR("%s: bitset flush failed", __func__);
			return r;
		}
	}

	r = save_sm_root(md);
	if (r) {
		DMERR("%s: save_sm_root failed", __func__);
		return r;
	}

	r = dm_tm_pre_commit(md->tm);
	if (r) {
		DMERR("%s: pre commit failed", __func__);
		return r;
	}

	r = superblock_lock(md, &sblock);
	if (r) {
		DMERR("%s: superblock lock failed", __func__);
		return r;
	}

	prepare_superblock(md, dm_block_data(sblock));

	return dm_tm_commit(md->tm, sblock);
}

static int metadata_checkpoint(struct era_metadata *md)
{
	/*
	 * For now we just rollover, but later I want to put a check in to
	 * avoid this if the filter is still pretty fresh.
	 */
	return metadata_era_rollover(md);
}

/*
 * Metadata snapshots allow userland to access era data.
 */
static int metadata_take_snap(struct era_metadata *md)
{
	int r, inc;
	struct dm_block *clone;

	if (md->metadata_snap != SUPERBLOCK_LOCATION) {
		DMERR("%s: metadata snapshot already exists", __func__);
		return -EINVAL;
	}

	r = metadata_era_rollover(md);
	if (r) {
		DMERR("%s: era rollover failed", __func__);
		return r;
	}

	r = metadata_commit(md);
	if (r) {
		DMERR("%s: pre commit failed", __func__);
		return r;
	}

	r = dm_sm_inc_block(md->sm, SUPERBLOCK_LOCATION);
	if (r) {
		DMERR("%s: couldn't increment superblock", __func__);
		return r;
	}

	r = dm_tm_shadow_block(md->tm, SUPERBLOCK_LOCATION,
			       &sb_validator, &clone, &inc);
	if (r) {
		DMERR("%s: couldn't shadow superblock", __func__);
		dm_sm_dec_block(md->sm, SUPERBLOCK_LOCATION);
		return r;
	}
	BUG_ON(!inc);

	r = dm_sm_inc_block(md->sm, md->writeset_tree_root);
	if (r) {
		DMERR("%s: couldn't inc writeset tree root", __func__);
		dm_tm_unlock(md->tm, clone);
		return r;
	}

	r = dm_sm_inc_block(md->sm, md->era_array_root);
	if (r) {
		DMERR("%s: couldn't inc era tree root", __func__);
		dm_sm_dec_block(md->sm, md->writeset_tree_root);
		dm_tm_unlock(md->tm, clone);
		return r;
	}

	md->metadata_snap = dm_block_location(clone);

	r = dm_tm_unlock(md->tm, clone);
	if (r) {
		DMERR("%s: couldn't unlock clone", __func__);
		md->metadata_snap = SUPERBLOCK_LOCATION;
		return r;
	}

	return 0;
}

static int metadata_drop_snap(struct era_metadata *md)
{
	int r;
	dm_block_t location;
	struct dm_block *clone;
	struct superblock_disk *disk;

	if (md->metadata_snap == SUPERBLOCK_LOCATION) {
		DMERR("%s: no snap to drop", __func__);
		return -EINVAL;
	}

	r = dm_tm_read_lock(md->tm, md->metadata_snap, &sb_validator, &clone);
	if (r) {
		DMERR("%s: couldn't read lock superblock clone", __func__);
		return r;
	}

	/*
	 * Whatever happens now we'll commit with no record of the metadata
	 * snap.
	 */
	md->metadata_snap = SUPERBLOCK_LOCATION;

	disk = dm_block_data(clone);
	r = dm_btree_del(&md->writeset_tree_info,
			 le64_to_cpu(disk->writeset_tree_root));
	if (r) {
		DMERR("%s: error deleting writeset tree clone", __func__);
		dm_tm_unlock(md->tm, clone);
		return r;
	}

	r = dm_array_del(&md->era_array_info, le64_to_cpu(disk->era_array_root));
	if (r) {
		DMERR("%s: error deleting era array clone", __func__);
		dm_tm_unlock(md->tm, clone);
		return r;
	}

	location = dm_block_location(clone);
	dm_tm_unlock(md->tm, clone);

	return dm_sm_dec_block(md->sm, location);
}

struct metadata_stats {
	dm_block_t used;
	dm_block_t total;
	dm_block_t snap;
	uint32_t era;
};

static int metadata_get_stats(struct era_metadata *md, void *ptr)
{
	int r;
	struct metadata_stats *s = ptr;
	dm_block_t nr_free, nr_total;

	r = dm_sm_get_nr_free(md->sm, &nr_free);
	if (r) {
		DMERR("dm_sm_get_nr_free returned %d", r);
		return r;
	}

	r = dm_sm_get_nr_blocks(md->sm, &nr_total);
	if (r) {
		DMERR("dm_pool_get_metadata_dev_size returned %d", r);
		return r;
	}

	s->used = nr_total - nr_free;
	s->total = nr_total;
	s->snap = md->metadata_snap;
	s->era = md->current_era;

	return 0;
}

/*----------------------------------------------------------------*/

struct era {
	struct dm_target *ti;
	struct dm_target_callbacks callbacks;

	struct dm_dev *metadata_dev;
	struct dm_dev *origin_dev;

	dm_block_t nr_blocks;
	uint32_t sectors_per_block;
	int sectors_per_block_shift;
	struct era_metadata *md;

	struct workqueue_struct *wq;
	struct work_struct worker;

	spinlock_t deferred_lock;
	struct bio_list deferred_bios;

	spinlock_t rpc_lock;
	struct list_head rpc_calls;

	struct digest digest;
	atomic_t suspended;
};

struct rpc {
	struct list_head list;

	int (*fn0)(struct era_metadata *);
	int (*fn1)(struct era_metadata *, void *);
	void *arg;
	int result;

	struct completion complete;
};

/*----------------------------------------------------------------
 * Remapping.
 *---------------------------------------------------------------*/
static bool block_size_is_power_of_two(struct era *era)
{
	return era->sectors_per_block_shift >= 0;
}

static dm_block_t get_block(struct era *era, struct bio *bio)
{
	sector_t block_nr = bio->bi_iter.bi_sector;

	if (!block_size_is_power_of_two(era))
		(void) sector_div(block_nr, era->sectors_per_block);
	else
		block_nr >>= era->sectors_per_block_shift;

	return block_nr;
}

static void remap_to_origin(struct era *era, struct bio *bio)
{
	bio->bi_bdev = era->origin_dev->bdev;
}

/*----------------------------------------------------------------
 * Worker thread
 *--------------------------------------------------------------*/
static void wake_worker(struct era *era)
{
	if (!atomic_read(&era->suspended))
		queue_work(era->wq, &era->worker);
}

static void process_old_eras(struct era *era)
{
	int r;

	if (!era->digest.step)
		return;

	r = era->digest.step(era->md, &era->digest);
	if (r < 0) {
		DMERR("%s: digest step failed, stopping digestion", __func__);
		era->digest.step = NULL;

	} else if (era->digest.step)
		wake_worker(era);
}

static void process_deferred_bios(struct era *era)
{
	int r;
	struct bio_list deferred_bios, marked_bios;
	struct bio *bio;
	bool commit_needed = false;
	bool failed = false;

	bio_list_init(&deferred_bios);
	bio_list_init(&marked_bios);

	spin_lock(&era->deferred_lock);
	bio_list_merge(&deferred_bios, &era->deferred_bios);
	bio_list_init(&era->deferred_bios);
	spin_unlock(&era->deferred_lock);

	while ((bio = bio_list_pop(&deferred_bios))) {
		r = writeset_test_and_set(&era->md->bitset_info,
					  era->md->current_writeset,
					  get_block(era, bio));
		if (r < 0) {
			/*
			 * This is bad news, we need to rollback.
			 * FIXME: finish.
			 */
			failed = true;

		} else if (r == 0)
			commit_needed = true;

		bio_list_add(&marked_bios, bio);
	}

	if (commit_needed) {
		r = metadata_commit(era->md);
		if (r)
			failed = true;
	}

	if (failed)
		while ((bio = bio_list_pop(&marked_bios)))
			bio_io_error(bio);
	else
		while ((bio = bio_list_pop(&marked_bios)))
			generic_make_request(bio);
}

static void process_rpc_calls(struct era *era)
{
	int r;
	bool need_commit = false;
	struct list_head calls;
	struct rpc *rpc, *tmp;

	INIT_LIST_HEAD(&calls);
	spin_lock(&era->rpc_lock);
	list_splice_init(&era->rpc_calls, &calls);
	spin_unlock(&era->rpc_lock);

	list_for_each_entry_safe(rpc, tmp, &calls, list) {
		rpc->result = rpc->fn0 ? rpc->fn0(era->md) : rpc->fn1(era->md, rpc->arg);
		need_commit = true;
	}

	if (need_commit) {
		r = metadata_commit(era->md);
		if (r)
			list_for_each_entry_safe(rpc, tmp, &calls, list)
				rpc->result = r;
	}

	list_for_each_entry_safe(rpc, tmp, &calls, list)
		complete(&rpc->complete);
}

static void kick_off_digest(struct era *era)
{
	if (era->md->archived_writesets) {
		era->md->archived_writesets = false;
		metadata_digest_start(era->md, &era->digest);
	}
}

static void do_work(struct work_struct *ws)
{
	struct era *era = container_of(ws, struct era, worker);

	kick_off_digest(era);
	process_old_eras(era);
	process_deferred_bios(era);
	process_rpc_calls(era);
}

static void defer_bio(struct era *era, struct bio *bio)
{
	spin_lock(&era->deferred_lock);
	bio_list_add(&era->deferred_bios, bio);
	spin_unlock(&era->deferred_lock);

	wake_worker(era);
}

/*
 * Make an rpc call to the worker to change the metadata.
 */
static int perform_rpc(struct era *era, struct rpc *rpc)
{
	rpc->result = 0;
	init_completion(&rpc->complete);

	spin_lock(&era->rpc_lock);
	list_add(&rpc->list, &era->rpc_calls);
	spin_unlock(&era->rpc_lock);

	wake_worker(era);
	wait_for_completion(&rpc->complete);

	return rpc->result;
}

static int in_worker0(struct era *era, int (*fn)(struct era_metadata *))
{
	struct rpc rpc;
	rpc.fn0 = fn;
	rpc.fn1 = NULL;

	return perform_rpc(era, &rpc);
}

static int in_worker1(struct era *era,
		      int (*fn)(struct era_metadata *, void *), void *arg)
{
	struct rpc rpc;
	rpc.fn0 = NULL;
	rpc.fn1 = fn;
	rpc.arg = arg;

	return perform_rpc(era, &rpc);
}

static void start_worker(struct era *era)
{
	atomic_set(&era->suspended, 0);
}

static void stop_worker(struct era *era)
{
	atomic_set(&era->suspended, 1);
	flush_workqueue(era->wq);
}

/*----------------------------------------------------------------
 * Target methods
 *--------------------------------------------------------------*/
static int dev_is_congested(struct dm_dev *dev, int bdi_bits)
{
	struct request_queue *q = bdev_get_queue(dev->bdev);
	return bdi_congested(&q->backing_dev_info, bdi_bits);
}

static int era_is_congested(struct dm_target_callbacks *cb, int bdi_bits)
{
	struct era *era = container_of(cb, struct era, callbacks);
	return dev_is_congested(era->origin_dev, bdi_bits);
}

static void era_destroy(struct era *era)
{
	if (era->md)
		metadata_close(era->md);

	if (era->wq)
		destroy_workqueue(era->wq);

	if (era->origin_dev)
		dm_put_device(era->ti, era->origin_dev);

	if (era->metadata_dev)
		dm_put_device(era->ti, era->metadata_dev);

	kfree(era);
}

static dm_block_t calc_nr_blocks(struct era *era)
{
	return dm_sector_div_up(era->ti->len, era->sectors_per_block);
}

static bool valid_block_size(dm_block_t block_size)
{
	bool greater_than_zero = block_size > 0;
	bool multiple_of_min_block_size = (block_size & (MIN_BLOCK_SIZE - 1)) == 0;

	return greater_than_zero && multiple_of_min_block_size;
}

/*
 * <metadata dev> <data dev> <data block size (sectors)>
 */
static int era_ctr(struct dm_target *ti, unsigned argc, char **argv)
{
	int r;
	char dummy;
	struct era *era;
	struct era_metadata *md;

	if (argc != 3) {
		ti->error = "Invalid argument count";
		return -EINVAL;
	}

	era = kzalloc(sizeof(*era), GFP_KERNEL);
	if (!era) {
		ti->error = "Error allocating era structure";
		return -ENOMEM;
	}

	era->ti = ti;

	r = dm_get_device(ti, argv[0], FMODE_READ | FMODE_WRITE, &era->metadata_dev);
	if (r) {
		ti->error = "Error opening metadata device";
		era_destroy(era);
		return -EINVAL;
	}

	r = dm_get_device(ti, argv[1], FMODE_READ | FMODE_WRITE, &era->origin_dev);
	if (r) {
		ti->error = "Error opening data device";
		era_destroy(era);
		return -EINVAL;
	}

	r = sscanf(argv[2], "%u%c", &era->sectors_per_block, &dummy);
	if (r != 1) {
		ti->error = "Error parsing block size";
		era_destroy(era);
		return -EINVAL;
	}

	r = dm_set_target_max_io_len(ti, era->sectors_per_block);
	if (r) {
		ti->error = "could not set max io len";
		era_destroy(era);
		return -EINVAL;
	}

	if (!valid_block_size(era->sectors_per_block)) {
		ti->error = "Invalid block size";
		era_destroy(era);
		return -EINVAL;
	}
	if (era->sectors_per_block & (era->sectors_per_block - 1))
		era->sectors_per_block_shift = -1;
	else
		era->sectors_per_block_shift = __ffs(era->sectors_per_block);

	md = metadata_open(era->metadata_dev->bdev, era->sectors_per_block, true);
	if (IS_ERR(md)) {
		ti->error = "Error reading metadata";
		era_destroy(era);
		return PTR_ERR(md);
	}
	era->md = md;

	era->nr_blocks = calc_nr_blocks(era);

	r = metadata_resize(era->md, &era->nr_blocks);
	if (r) {
		ti->error = "couldn't resize metadata";
		era_destroy(era);
		return -ENOMEM;
	}

	era->wq = alloc_ordered_workqueue("dm-" DM_MSG_PREFIX, WQ_MEM_RECLAIM);
	if (!era->wq) {
		ti->error = "could not create workqueue for metadata object";
		era_destroy(era);
		return -ENOMEM;
	}
	INIT_WORK(&era->worker, do_work);

	spin_lock_init(&era->deferred_lock);
	bio_list_init(&era->deferred_bios);

	spin_lock_init(&era->rpc_lock);
	INIT_LIST_HEAD(&era->rpc_calls);

	ti->private = era;
	ti->num_flush_bios = 1;
	ti->flush_supported = true;

	ti->num_discard_bios = 1;
	ti->discards_supported = true;
	era->callbacks.congested_fn = era_is_congested;
	dm_table_add_target_callbacks(ti->table, &era->callbacks);

	return 0;
}

static void era_dtr(struct dm_target *ti)
{
	era_destroy(ti->private);
}

static int era_map(struct dm_target *ti, struct bio *bio)
{
	struct era *era = ti->private;
	dm_block_t block = get_block(era, bio);

	/*
	 * All bios get remapped to the origin device.  We do this now, but
	 * it may not get issued until later.  Depending on whether the
	 * block is marked in this era.
	 */
	remap_to_origin(era, bio);

	/*
	 * REQ_FLUSH bios carry no data, so we're not interested in them.
	 */
	if (!(bio->bi_rw & REQ_FLUSH) &&
	    (bio_data_dir(bio) == WRITE) &&
	    !metadata_current_marked(era->md, block)) {
		defer_bio(era, bio);
		return DM_MAPIO_SUBMITTED;
	}

	return DM_MAPIO_REMAPPED;
}

static void era_postsuspend(struct dm_target *ti)
{
	int r;
	struct era *era = ti->private;

	r = in_worker0(era, metadata_era_archive);
	if (r) {
		DMERR("%s: couldn't archive current era", __func__);
		/* FIXME: fail mode */
	}

	stop_worker(era);
}

static int era_preresume(struct dm_target *ti)
{
	int r;
	struct era *era = ti->private;
	dm_block_t new_size = calc_nr_blocks(era);

	if (era->nr_blocks != new_size) {
		r = in_worker1(era, metadata_resize, &new_size);
		if (r)
			return r;

		era->nr_blocks = new_size;
	}

	start_worker(era);

	r = in_worker0(era, metadata_new_era);
	if (r) {
		DMERR("%s: metadata_era_rollover failed", __func__);
		return r;
	}

	return 0;
}

/*
 * Status format:
 *
 * <metadata block size> <#used metadata blocks>/<#total metadata blocks>
 * <current era> <held metadata root | '-'>
 */
static void era_status(struct dm_target *ti, status_type_t type,
		       unsigned status_flags, char *result, unsigned maxlen)
{
	int r;
	struct era *era = ti->private;
	ssize_t sz = 0;
	struct metadata_stats stats;
	char buf[BDEVNAME_SIZE];

	switch (type) {
	case STATUSTYPE_INFO:
		r = in_worker1(era, metadata_get_stats, &stats);
		if (r)
			goto err;

		DMEMIT("%u %llu/%llu %u",
		       (unsigned) (DM_ERA_METADATA_BLOCK_SIZE >> SECTOR_SHIFT),
		       (unsigned long long) stats.used,
		       (unsigned long long) stats.total,
		       (unsigned) stats.era);

		if (stats.snap != SUPERBLOCK_LOCATION)
			DMEMIT(" %llu", stats.snap);
		else
			DMEMIT(" -");
		break;

	case STATUSTYPE_TABLE:
		format_dev_t(buf, era->metadata_dev->bdev->bd_dev);
		DMEMIT("%s ", buf);
		format_dev_t(buf, era->origin_dev->bdev->bd_dev);
		DMEMIT("%s %u", buf, era->sectors_per_block);
		break;
	}

	return;

err:
	DMEMIT("Error");
}

static int era_message(struct dm_target *ti, unsigned argc, char **argv)
{
	struct era *era = ti->private;

	if (argc != 1) {
		DMERR("incorrect number of message arguments");
		return -EINVAL;
	}

	if (!strcasecmp(argv[0], "checkpoint"))
		return in_worker0(era, metadata_checkpoint);

	if (!strcasecmp(argv[0], "take_metadata_snap"))
		return in_worker0(era, metadata_take_snap);

	if (!strcasecmp(argv[0], "drop_metadata_snap"))
		return in_worker0(era, metadata_drop_snap);

	DMERR("unsupported message '%s'", argv[0]);
	return -EINVAL;
}

static sector_t get_dev_size(struct dm_dev *dev)
{
	return i_size_read(dev->bdev->bd_inode) >> SECTOR_SHIFT;
}

static int era_iterate_devices(struct dm_target *ti,
			       iterate_devices_callout_fn fn, void *data)
{
	struct era *era = ti->private;
	return fn(ti, era->origin_dev, 0, get_dev_size(era->origin_dev), data);
}

static int era_merge(struct dm_target *ti, struct bvec_merge_data *bvm,
		     struct bio_vec *biovec, int max_size)
{
	struct era *era = ti->private;
	struct request_queue *q = bdev_get_queue(era->origin_dev->bdev);

	if (!q->merge_bvec_fn)
		return max_size;

	bvm->bi_bdev = era->origin_dev->bdev;

	return min(max_size, q->merge_bvec_fn(q, bvm, biovec));
}

static void era_io_hints(struct dm_target *ti, struct queue_limits *limits)
{
	struct era *era = ti->private;
	uint64_t io_opt_sectors = limits->io_opt >> SECTOR_SHIFT;

	/*
	 * If the system-determined stacked limits are compatible with the
	 * era device's blocksize (io_opt is a factor) do not override them.
	 */
	if (io_opt_sectors < era->sectors_per_block ||
	    do_div(io_opt_sectors, era->sectors_per_block)) {
		blk_limits_io_min(limits, 0);
		blk_limits_io_opt(limits, era->sectors_per_block << SECTOR_SHIFT);
	}
}

/*----------------------------------------------------------------*/

static struct target_type era_target = {
	.name = "era",
	.version = {1, 0, 0},
	.module = THIS_MODULE,
	.ctr = era_ctr,
	.dtr = era_dtr,
	.map = era_map,
	.postsuspend = era_postsuspend,
	.preresume = era_preresume,
	.status = era_status,
	.message = era_message,
	.iterate_devices = era_iterate_devices,
	.merge = era_merge,
	.io_hints = era_io_hints
};

static int __init dm_era_init(void)
{
	int r;

	r = dm_register_target(&era_target);
	if (r) {
		DMERR("era target registration failed: %d", r);
		return r;
	}

	return 0;
}

static void __exit dm_era_exit(void)
{
	dm_unregister_target(&era_target);
}

module_init(dm_era_init);
module_exit(dm_era_exit);

MODULE_DESCRIPTION(DM_NAME " era target");
MODULE_AUTHOR("Joe Thornber <ejt@redhat.com>");
MODULE_LICENSE("GPL");