aboutsummaryrefslogtreecommitdiff
path: root/drivers/staging/lustre/lustre/llite/dir.c
blob: 407718a0026f56071a9667ebe7088f158eba013e (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
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
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
/*
 * GPL HEADER START
 *
 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License version 2 only,
 * as published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it will be useful, but
 * WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * General Public License version 2 for more details (a copy is included
 * in the LICENSE file that accompanied this code).
 *
 * You should have received a copy of the GNU General Public License
 * version 2 along with this program; If not, see
 * http://www.sun.com/software/products/lustre/docs/GPLv2.pdf
 *
 * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara,
 * CA 95054 USA or visit www.sun.com if you need additional information or
 * have any questions.
 *
 * GPL HEADER END
 */
/*
 * Copyright (c) 2002, 2010, Oracle and/or its affiliates. All rights reserved.
 * Use is subject to license terms.
 *
 * Copyright (c) 2011, 2012, Intel Corporation.
 */
/*
 * This file is part of Lustre, http://www.lustre.org/
 * Lustre is a trademark of Sun Microsystems, Inc.
 *
 * lustre/llite/dir.c
 *
 * Directory code for lustre client.
 */

#include <linux/fs.h>
#include <linux/pagemap.h>
#include <linux/mm.h>
#include <asm/uaccess.h>
#include <linux/buffer_head.h>   /* for wait_on_buffer */
#include <linux/pagevec.h>
#include <linux/prefetch.h>

#define DEBUG_SUBSYSTEM S_LLITE

#include "../include/obd_support.h"
#include "../include/obd_class.h"
#include "../include/lustre_lib.h"
#include "../include/lustre/lustre_idl.h"
#include "../include/lustre_lite.h"
#include "../include/lustre_dlm.h"
#include "../include/lustre_fid.h"
#include "llite_internal.h"

/*
 * (new) readdir implementation overview.
 *
 * Original lustre readdir implementation cached exact copy of raw directory
 * pages on the client. These pages were indexed in client page cache by
 * logical offset in the directory file. This design, while very simple and
 * intuitive had some inherent problems:
 *
 *     . it implies that byte offset to the directory entry serves as a
 *     telldir(3)/seekdir(3) cookie, but that offset is not stable: in
 *     ext3/htree directory entries may move due to splits, and more
 *     importantly,
 *
 *     . it is incompatible with the design of split directories for cmd3,
 *     that assumes that names are distributed across nodes based on their
 *     hash, and so readdir should be done in hash order.
 *
 * New readdir implementation does readdir in hash order, and uses hash of a
 * file name as a telldir/seekdir cookie. This led to number of complications:
 *
 *     . hash is not unique, so it cannot be used to index cached directory
 *     pages on the client (note, that it requires a whole pageful of hash
 *     collided entries to cause two pages to have identical hashes);
 *
 *     . hash is not unique, so it cannot, strictly speaking, be used as an
 *     entry cookie. ext3/htree has the same problem and lustre implementation
 *     mimics their solution: seekdir(hash) positions directory at the first
 *     entry with the given hash.
 *
 * Client side.
 *
 * 0. caching
 *
 * Client caches directory pages using hash of the first entry as an index. As
 * noted above hash is not unique, so this solution doesn't work as is:
 * special processing is needed for "page hash chains" (i.e., sequences of
 * pages filled with entries all having the same hash value).
 *
 * First, such chains have to be detected. To this end, server returns to the
 * client the hash of the first entry on the page next to one returned. When
 * client detects that this hash is the same as hash of the first entry on the
 * returned page, page hash collision has to be handled. Pages in the
 * hash chain, except first one, are termed "overflow pages".
 *
 * Solution to index uniqueness problem is to not cache overflow
 * pages. Instead, when page hash collision is detected, all overflow pages
 * from emerging chain are immediately requested from the server and placed in
 * a special data structure (struct ll_dir_chain). This data structure is used
 * by ll_readdir() to process entries from overflow pages. When readdir
 * invocation finishes, overflow pages are discarded. If page hash collision
 * chain weren't completely processed, next call to readdir will again detect
 * page hash collision, again read overflow pages in, process next portion of
 * entries and again discard the pages. This is not as wasteful as it looks,
 * because, given reasonable hash, page hash collisions are extremely rare.
 *
 * 1. directory positioning
 *
 * When seekdir(hash) is called, original
 *
 *
 *
 *
 *
 *
 *
 *
 * Server.
 *
 * identification of and access to overflow pages
 *
 * page format
 *
 * Page in MDS_READPAGE RPC is packed in LU_PAGE_SIZE, and each page contains
 * a header lu_dirpage which describes the start/end hash, and whether this
 * page is empty (contains no dir entry) or hash collide with next page.
 * After client receives reply, several pages will be integrated into dir page
 * in PAGE_CACHE_SIZE (if PAGE_CACHE_SIZE greater than LU_PAGE_SIZE), and the
 * lu_dirpage for this integrated page will be adjusted. See
 * lmv_adjust_dirpages().
 *
 */

/* returns the page unlocked, but with a reference */
static int ll_dir_filler(void *_hash, struct page *page0)
{
	struct inode *inode = page0->mapping->host;
	int hash64 = ll_i2sbi(inode)->ll_flags & LL_SBI_64BIT_HASH;
	struct obd_export *exp = ll_i2sbi(inode)->ll_md_exp;
	struct ptlrpc_request *request;
	struct mdt_body *body;
	struct md_op_data *op_data;
	__u64 hash = *((__u64 *)_hash);
	struct page **page_pool;
	struct page *page;
	struct lu_dirpage *dp;
	int max_pages = ll_i2sbi(inode)->ll_md_brw_size >> PAGE_CACHE_SHIFT;
	int nrdpgs = 0; /* number of pages read actually */
	int npages;
	int i;
	int rc;

	CDEBUG(D_VFSTRACE, "VFS Op:inode=%lu/%u(%p) hash %llu\n",
	       inode->i_ino, inode->i_generation, inode, hash);

	LASSERT(max_pages > 0 && max_pages <= MD_MAX_BRW_PAGES);

	page_pool = kcalloc(max_pages, sizeof(page), GFP_NOFS);
	if (page_pool) {
		page_pool[0] = page0;
	} else {
		page_pool = &page0;
		max_pages = 1;
	}
	for (npages = 1; npages < max_pages; npages++) {
		page = page_cache_alloc_cold(inode->i_mapping);
		if (!page)
			break;
		page_pool[npages] = page;
	}

	op_data = ll_prep_md_op_data(NULL, inode, NULL, NULL, 0, 0,
				     LUSTRE_OPC_ANY, NULL);
	op_data->op_npages = npages;
	op_data->op_offset = hash;
	rc = md_readpage(exp, op_data, page_pool, &request);
	ll_finish_md_op_data(op_data);
	if (rc == 0) {
		body = req_capsule_server_get(&request->rq_pill, &RMF_MDT_BODY);
		/* Checked by mdc_readpage() */
		LASSERT(body != NULL);

		if (body->valid & OBD_MD_FLSIZE)
			cl_isize_write(inode, body->size);

		nrdpgs = (request->rq_bulk->bd_nob_transferred+PAGE_CACHE_SIZE-1)
			 >> PAGE_CACHE_SHIFT;
		SetPageUptodate(page0);
	}
	unlock_page(page0);
	ptlrpc_req_finished(request);

	CDEBUG(D_VFSTRACE, "read %d/%d pages\n", nrdpgs, npages);

	ll_pagevec_init(&lru_pvec, 0);
	for (i = 1; i < npages; i++) {
		unsigned long offset;
		int ret;

		page = page_pool[i];

		if (rc < 0 || i >= nrdpgs) {
			page_cache_release(page);
			continue;
		}

		SetPageUptodate(page);

		dp = kmap(page);
		hash = le64_to_cpu(dp->ldp_hash_start);
		kunmap(page);

		offset = hash_x_index(hash, hash64);

		prefetchw(&page->flags);
		ret = add_to_page_cache_lru(page, inode->i_mapping, offset,
					    GFP_KERNEL);
		if (ret == 0) {
			unlock_page(page);
			if (ll_pagevec_add(&lru_pvec, page) == 0)
				ll_pagevec_lru_add_file(&lru_pvec);
		} else {
			CDEBUG(D_VFSTRACE, "page %lu add to page cache failed: %d\n",
			       offset, ret);
		}
		page_cache_release(page);
	}
	ll_pagevec_lru_add_file(&lru_pvec);

	if (page_pool != &page0)
		OBD_FREE(page_pool, sizeof(struct page *) * max_pages);
	return rc;
}

static void ll_check_page(struct inode *dir, struct page *page)
{
	/* XXX: check page format later */
	SetPageChecked(page);
}

void ll_release_page(struct page *page, int remove)
{
	kunmap(page);
	if (remove) {
		lock_page(page);
		if (likely(page->mapping != NULL))
			truncate_complete_page(page->mapping, page);
		unlock_page(page);
	}
	page_cache_release(page);
}

/*
 * Find, kmap and return page that contains given hash.
 */
static struct page *ll_dir_page_locate(struct inode *dir, __u64 *hash,
				       __u64 *start, __u64 *end)
{
	int hash64 = ll_i2sbi(dir)->ll_flags & LL_SBI_64BIT_HASH;
	struct address_space *mapping = dir->i_mapping;
	/*
	 * Complement of hash is used as an index so that
	 * radix_tree_gang_lookup() can be used to find a page with starting
	 * hash _smaller_ than one we are looking for.
	 */
	unsigned long offset = hash_x_index(*hash, hash64);
	struct page *page;
	int found;

	spin_lock_irq(&mapping->tree_lock);
	found = radix_tree_gang_lookup(&mapping->page_tree,
				       (void **)&page, offset, 1);
	if (found > 0) {
		struct lu_dirpage *dp;

		page_cache_get(page);
		spin_unlock_irq(&mapping->tree_lock);
		/*
		 * In contrast to find_lock_page() we are sure that directory
		 * page cannot be truncated (while DLM lock is held) and,
		 * hence, can avoid restart.
		 *
		 * In fact, page cannot be locked here at all, because
		 * ll_dir_filler() does synchronous io.
		 */
		wait_on_page_locked(page);
		if (PageUptodate(page)) {
			dp = kmap(page);
			if (BITS_PER_LONG == 32 && hash64) {
				*start = le64_to_cpu(dp->ldp_hash_start) >> 32;
				*end   = le64_to_cpu(dp->ldp_hash_end) >> 32;
				*hash  = *hash >> 32;
			} else {
				*start = le64_to_cpu(dp->ldp_hash_start);
				*end   = le64_to_cpu(dp->ldp_hash_end);
			}
			LASSERTF(*start <= *hash, "start = %#llx,end = %#llx,hash = %#llx\n",
				 *start, *end, *hash);
			CDEBUG(D_VFSTRACE, "page %lu [%llu %llu], hash %llu\n",
			       offset, *start, *end, *hash);
			if (*hash > *end) {
				ll_release_page(page, 0);
				page = NULL;
			} else if (*end != *start && *hash == *end) {
				/*
				 * upon hash collision, remove this page,
				 * otherwise put page reference, and
				 * ll_get_dir_page() will issue RPC to fetch
				 * the page we want.
				 */
				ll_release_page(page,
				    le32_to_cpu(dp->ldp_flags) & LDF_COLLIDE);
				page = NULL;
			}
		} else {
			page_cache_release(page);
			page = ERR_PTR(-EIO);
		}

	} else {
		spin_unlock_irq(&mapping->tree_lock);
		page = NULL;
	}
	return page;
}

struct page *ll_get_dir_page(struct inode *dir, __u64 hash,
			     struct ll_dir_chain *chain)
{
	ldlm_policy_data_t policy = {.l_inodebits = {MDS_INODELOCK_UPDATE} };
	struct address_space *mapping = dir->i_mapping;
	struct lustre_handle lockh;
	struct lu_dirpage *dp;
	struct page *page;
	ldlm_mode_t mode;
	int rc;
	__u64 start = 0;
	__u64 end = 0;
	__u64 lhash = hash;
	struct ll_inode_info *lli = ll_i2info(dir);
	int hash64 = ll_i2sbi(dir)->ll_flags & LL_SBI_64BIT_HASH;

	mode = LCK_PR;
	rc = md_lock_match(ll_i2sbi(dir)->ll_md_exp, LDLM_FL_BLOCK_GRANTED,
			   ll_inode2fid(dir), LDLM_IBITS, &policy, mode, &lockh);
	if (!rc) {
		struct ldlm_enqueue_info einfo = {
			.ei_type = LDLM_IBITS,
			.ei_mode = mode,
			.ei_cb_bl = ll_md_blocking_ast,
			.ei_cb_cp = ldlm_completion_ast,
		};
		struct lookup_intent it = { .it_op = IT_READDIR };
		struct ptlrpc_request *request;
		struct md_op_data *op_data;

		op_data = ll_prep_md_op_data(NULL, dir, dir, NULL, 0, 0,
		LUSTRE_OPC_ANY, NULL);
		if (IS_ERR(op_data))
			return (void *)op_data;

		rc = md_enqueue(ll_i2sbi(dir)->ll_md_exp, &einfo, &it,
				op_data, &lockh, NULL, 0, NULL, 0);

		ll_finish_md_op_data(op_data);

		request = (struct ptlrpc_request *)it.d.lustre.it_data;
		if (request)
			ptlrpc_req_finished(request);
		if (rc < 0) {
			CERROR("lock enqueue: "DFID" at %llu: rc %d\n",
				PFID(ll_inode2fid(dir)), hash, rc);
			return ERR_PTR(rc);
		}

		CDEBUG(D_INODE, "setting lr_lvb_inode to inode %p (%lu/%u)\n",
		       dir, dir->i_ino, dir->i_generation);
		md_set_lock_data(ll_i2sbi(dir)->ll_md_exp,
				 &it.d.lustre.it_lock_handle, dir, NULL);
	} else {
		/* for cross-ref object, l_ast_data of the lock may not be set,
		 * we reset it here */
		md_set_lock_data(ll_i2sbi(dir)->ll_md_exp, &lockh.cookie,
				 dir, NULL);
	}
	ldlm_lock_dump_handle(D_OTHER, &lockh);

	mutex_lock(&lli->lli_readdir_mutex);
	page = ll_dir_page_locate(dir, &lhash, &start, &end);
	if (IS_ERR(page)) {
		CERROR("dir page locate: "DFID" at %llu: rc %ld\n",
		       PFID(ll_inode2fid(dir)), lhash, PTR_ERR(page));
		goto out_unlock;
	} else if (page != NULL) {
		/*
		 * XXX nikita: not entirely correct handling of a corner case:
		 * suppose hash chain of entries with hash value HASH crosses
		 * border between pages P0 and P1. First both P0 and P1 are
		 * cached, seekdir() is called for some entry from the P0 part
		 * of the chain. Later P0 goes out of cache. telldir(HASH)
		 * happens and finds P1, as it starts with matching hash
		 * value. Remaining entries from P0 part of the chain are
		 * skipped. (Is that really a bug?)
		 *
		 * Possible solutions: 0. don't cache P1 is such case, handle
		 * it as an "overflow" page. 1. invalidate all pages at
		 * once. 2. use HASH|1 as an index for P1.
		 */
		goto hash_collision;
	}

	page = read_cache_page(mapping, hash_x_index(hash, hash64),
			       ll_dir_filler, &lhash);
	if (IS_ERR(page)) {
		CERROR("read cache page: "DFID" at %llu: rc %ld\n",
		       PFID(ll_inode2fid(dir)), hash, PTR_ERR(page));
		goto out_unlock;
	}

	wait_on_page_locked(page);
	(void)kmap(page);
	if (!PageUptodate(page)) {
		CERROR("page not updated: "DFID" at %llu: rc %d\n",
		       PFID(ll_inode2fid(dir)), hash, -5);
		goto fail;
	}
	if (!PageChecked(page))
		ll_check_page(dir, page);
	if (PageError(page)) {
		CERROR("page error: "DFID" at %llu: rc %d\n",
		       PFID(ll_inode2fid(dir)), hash, -5);
		goto fail;
	}
hash_collision:
	dp = page_address(page);
	if (BITS_PER_LONG == 32 && hash64) {
		start = le64_to_cpu(dp->ldp_hash_start) >> 32;
		end   = le64_to_cpu(dp->ldp_hash_end) >> 32;
		lhash = hash >> 32;
	} else {
		start = le64_to_cpu(dp->ldp_hash_start);
		end   = le64_to_cpu(dp->ldp_hash_end);
		lhash = hash;
	}
	if (end == start) {
		LASSERT(start == lhash);
		CWARN("Page-wide hash collision: %llu\n", end);
		if (BITS_PER_LONG == 32 && hash64)
			CWARN("Real page-wide hash collision at [%llu %llu] with hash %llu\n",
			      le64_to_cpu(dp->ldp_hash_start),
			      le64_to_cpu(dp->ldp_hash_end), hash);
		/*
		 * Fetch whole overflow chain...
		 *
		 * XXX not yet.
		 */
		goto fail;
	}
out_unlock:
	mutex_unlock(&lli->lli_readdir_mutex);
	ldlm_lock_decref(&lockh, mode);
	return page;

fail:
	ll_release_page(page, 1);
	page = ERR_PTR(-EIO);
	goto out_unlock;
}

int ll_dir_read(struct inode *inode, struct dir_context *ctx)
{
	struct ll_inode_info *info       = ll_i2info(inode);
	struct ll_sb_info    *sbi	= ll_i2sbi(inode);
	__u64		   pos		= ctx->pos;
	int		   api32      = ll_need_32bit_api(sbi);
	int		   hash64     = sbi->ll_flags & LL_SBI_64BIT_HASH;
	struct page	  *page;
	struct ll_dir_chain   chain;
	int		   done = 0;
	int		   rc = 0;

	ll_dir_chain_init(&chain);

	page = ll_get_dir_page(inode, pos, &chain);

	while (rc == 0 && !done) {
		struct lu_dirpage *dp;
		struct lu_dirent  *ent;

		if (!IS_ERR(page)) {
			/*
			 * If page is empty (end of directory is reached),
			 * use this value.
			 */
			__u64 hash = MDS_DIR_END_OFF;
			__u64 next;

			dp = page_address(page);
			for (ent = lu_dirent_start(dp); ent != NULL && !done;
			     ent = lu_dirent_next(ent)) {
				__u16	  type;
				int	    namelen;
				struct lu_fid  fid;
				__u64	  lhash;
				__u64	  ino;

				/*
				 * XXX: implement correct swabbing here.
				 */

				hash = le64_to_cpu(ent->lde_hash);
				if (hash < pos)
					/*
					 * Skip until we find target hash
					 * value.
					 */
					continue;

				namelen = le16_to_cpu(ent->lde_namelen);
				if (namelen == 0)
					/*
					 * Skip dummy record.
					 */
					continue;

				if (api32 && hash64)
					lhash = hash >> 32;
				else
					lhash = hash;
				fid_le_to_cpu(&fid, &ent->lde_fid);
				ino = cl_fid_build_ino(&fid, api32);
				type = ll_dirent_type_get(ent);
				ctx->pos = lhash;
				/* For 'll_nfs_get_name_filldir()', it will try
				 * to access the 'ent' through its 'lde_name',
				 * so the parameter 'name' for 'ctx->actor()'
				 * must be part of the 'ent'.
				 */
				done = !dir_emit(ctx, ent->lde_name,
						 namelen, ino, type);
			}
			next = le64_to_cpu(dp->ldp_hash_end);
			if (!done) {
				pos = next;
				if (pos == MDS_DIR_END_OFF) {
					/*
					 * End of directory reached.
					 */
					done = 1;
					ll_release_page(page, 0);
				} else if (1 /* chain is exhausted*/) {
					/*
					 * Normal case: continue to the next
					 * page.
					 */
					ll_release_page(page,
					    le32_to_cpu(dp->ldp_flags) &
							LDF_COLLIDE);
					next = pos;
					page = ll_get_dir_page(inode, pos,
							       &chain);
				} else {
					/*
					 * go into overflow page.
					 */
					LASSERT(le32_to_cpu(dp->ldp_flags) &
						LDF_COLLIDE);
					ll_release_page(page, 1);
				}
			} else {
				pos = hash;
				ll_release_page(page, 0);
			}
		} else {
			rc = PTR_ERR(page);
			CERROR("error reading dir "DFID" at %lu: rc %d\n",
			       PFID(&info->lli_fid), (unsigned long)pos, rc);
		}
	}

	ctx->pos = pos;
	ll_dir_chain_fini(&chain);
	return rc;
}

static int ll_readdir(struct file *filp, struct dir_context *ctx)
{
	struct inode		*inode	= file_inode(filp);
	struct ll_file_data	*lfd	= LUSTRE_FPRIVATE(filp);
	struct ll_sb_info	*sbi	= ll_i2sbi(inode);
	int			hash64	= sbi->ll_flags & LL_SBI_64BIT_HASH;
	int			api32	= ll_need_32bit_api(sbi);
	int			rc;

	CDEBUG(D_VFSTRACE, "VFS Op:inode=%lu/%u(%p) pos %lu/%llu 32bit_api %d\n",
	       inode->i_ino, inode->i_generation,
	       inode, (unsigned long)lfd->lfd_pos, i_size_read(inode), api32);

	if (lfd->lfd_pos == MDS_DIR_END_OFF) {
		/*
		 * end-of-file.
		 */
		rc = 0;
		goto out;
	}

	ctx->pos = lfd->lfd_pos;
	rc = ll_dir_read(inode, ctx);
	lfd->lfd_pos = ctx->pos;
	if (ctx->pos == MDS_DIR_END_OFF) {
		if (api32)
			ctx->pos = LL_DIR_END_OFF_32BIT;
		else
			ctx->pos = LL_DIR_END_OFF;
	} else {
		if (api32 && hash64)
			ctx->pos >>= 32;
	}
	filp->f_version = inode->i_version;

out:
	if (!rc)
		ll_stats_ops_tally(sbi, LPROC_LL_READDIR, 1);

	return rc;
}

static int ll_send_mgc_param(struct obd_export *mgc, char *string)
{
	struct mgs_send_param *msp;
	int rc = 0;

	msp = kzalloc(sizeof(*msp), GFP_NOFS);
	if (!msp)
		return -ENOMEM;

	strncpy(msp->mgs_param, string, MGS_PARAM_MAXLEN);
	rc = obd_set_info_async(NULL, mgc, sizeof(KEY_SET_INFO), KEY_SET_INFO,
				sizeof(struct mgs_send_param), msp, NULL);
	if (rc)
		CERROR("Failed to set parameter: %d\n", rc);
	OBD_FREE_PTR(msp);

	return rc;
}

int ll_dir_setdirstripe(struct inode *dir, struct lmv_user_md *lump,
			char *filename)
{
	struct ptlrpc_request *request = NULL;
	struct md_op_data *op_data;
	struct ll_sb_info *sbi = ll_i2sbi(dir);
	int mode;
	int err;

	mode = (0755 & (S_IRWXUGO|S_ISVTX) & ~current->fs->umask) | S_IFDIR;
	op_data = ll_prep_md_op_data(NULL, dir, NULL, filename,
				     strlen(filename), mode, LUSTRE_OPC_MKDIR,
				     lump);
	if (IS_ERR(op_data)) {
		err = PTR_ERR(op_data);
		goto err_exit;
	}

	op_data->op_cli_flags |= CLI_SET_MEA;
	err = md_create(sbi->ll_md_exp, op_data, lump, sizeof(*lump), mode,
			from_kuid(&init_user_ns, current_fsuid()),
			from_kgid(&init_user_ns, current_fsgid()),
			cfs_curproc_cap_pack(), 0, &request);
	ll_finish_md_op_data(op_data);
	if (err)
		goto err_exit;
err_exit:
	ptlrpc_req_finished(request);
	return err;
}

int ll_dir_setstripe(struct inode *inode, struct lov_user_md *lump,
		     int set_default)
{
	struct ll_sb_info *sbi = ll_i2sbi(inode);
	struct md_op_data *op_data;
	struct ptlrpc_request *req = NULL;
	int rc = 0;
	struct lustre_sb_info *lsi = s2lsi(inode->i_sb);
	struct obd_device *mgc = lsi->lsi_mgc;
	int lum_size;

	if (lump != NULL) {
		/*
		 * This is coming from userspace, so should be in
		 * local endian.  But the MDS would like it in little
		 * endian, so we swab it before we send it.
		 */
		switch (lump->lmm_magic) {
		case LOV_USER_MAGIC_V1: {
			if (lump->lmm_magic != cpu_to_le32(LOV_USER_MAGIC_V1))
				lustre_swab_lov_user_md_v1(lump);
			lum_size = sizeof(struct lov_user_md_v1);
			break;
		}
		case LOV_USER_MAGIC_V3: {
			if (lump->lmm_magic != cpu_to_le32(LOV_USER_MAGIC_V3))
				lustre_swab_lov_user_md_v3(
					(struct lov_user_md_v3 *)lump);
			lum_size = sizeof(struct lov_user_md_v3);
			break;
		}
		default: {
			CDEBUG(D_IOCTL, "bad userland LOV MAGIC: %#08x != %#08x nor %#08x\n",
			       lump->lmm_magic, LOV_USER_MAGIC_V1,
			       LOV_USER_MAGIC_V3);
			return -EINVAL;
		}
		}
	} else {
		lum_size = sizeof(struct lov_user_md_v1);
	}

	op_data = ll_prep_md_op_data(NULL, inode, NULL, NULL, 0, 0,
				     LUSTRE_OPC_ANY, NULL);
	if (IS_ERR(op_data))
		return PTR_ERR(op_data);

	if (lump != NULL && lump->lmm_magic == cpu_to_le32(LMV_USER_MAGIC))
		op_data->op_cli_flags |= CLI_SET_MEA;

	/* swabbing is done in lov_setstripe() on server side */
	rc = md_setattr(sbi->ll_md_exp, op_data, lump, lum_size,
			NULL, 0, &req, NULL);
	ll_finish_md_op_data(op_data);
	ptlrpc_req_finished(req);
	if (rc) {
		if (rc != -EPERM && rc != -EACCES)
			CERROR("mdc_setattr fails: rc = %d\n", rc);
	}

	/* In the following we use the fact that LOV_USER_MAGIC_V1 and
	 LOV_USER_MAGIC_V3 have the same initial fields so we do not
	 need to make the distinction between the 2 versions */
	if (set_default && mgc->u.cli.cl_mgc_mgsexp) {
		char *param = NULL;
		char *buf;

		param = kzalloc(MGS_PARAM_MAXLEN, GFP_NOFS);
		if (!param) {
			rc = -ENOMEM;
			goto end;
		}

		buf = param;
		/* Get fsname and assume devname to be -MDT0000. */
		ll_get_fsname(inode->i_sb, buf, MTI_NAME_MAXLEN);
		strcat(buf, "-MDT0000.lov");
		buf += strlen(buf);

		/* Set root stripesize */
		sprintf(buf, ".stripesize=%u",
			lump ? le32_to_cpu(lump->lmm_stripe_size) : 0);
		rc = ll_send_mgc_param(mgc->u.cli.cl_mgc_mgsexp, param);
		if (rc)
			goto end;

		/* Set root stripecount */
		sprintf(buf, ".stripecount=%hd",
			lump ? le16_to_cpu(lump->lmm_stripe_count) : 0);
		rc = ll_send_mgc_param(mgc->u.cli.cl_mgc_mgsexp, param);
		if (rc)
			goto end;

		/* Set root stripeoffset */
		sprintf(buf, ".stripeoffset=%hd",
			lump ? le16_to_cpu(lump->lmm_stripe_offset) :
			(typeof(lump->lmm_stripe_offset))(-1));
		rc = ll_send_mgc_param(mgc->u.cli.cl_mgc_mgsexp, param);

end:
		if (param != NULL)
			OBD_FREE(param, MGS_PARAM_MAXLEN);
	}
	return rc;
}

int ll_dir_getstripe(struct inode *inode, struct lov_mds_md **lmmp,
		     int *lmm_size, struct ptlrpc_request **request)
{
	struct ll_sb_info *sbi = ll_i2sbi(inode);
	struct mdt_body   *body;
	struct lov_mds_md *lmm = NULL;
	struct ptlrpc_request *req = NULL;
	int rc, lmmsize;
	struct md_op_data *op_data;

	rc = ll_get_default_mdsize(sbi, &lmmsize);
	if (rc)
		return rc;

	op_data = ll_prep_md_op_data(NULL, inode, NULL, NULL,
				     0, lmmsize, LUSTRE_OPC_ANY,
				     NULL);
	if (IS_ERR(op_data))
		return PTR_ERR(op_data);

	op_data->op_valid = OBD_MD_FLEASIZE | OBD_MD_FLDIREA;
	rc = md_getattr(sbi->ll_md_exp, op_data, &req);
	ll_finish_md_op_data(op_data);
	if (rc < 0) {
		CDEBUG(D_INFO, "md_getattr failed on inode %lu/%u: rc %d\n",
		       inode->i_ino,
		       inode->i_generation, rc);
		goto out;
	}

	body = req_capsule_server_get(&req->rq_pill, &RMF_MDT_BODY);
	LASSERT(body != NULL);

	lmmsize = body->eadatasize;

	if (!(body->valid & (OBD_MD_FLEASIZE | OBD_MD_FLDIREA)) ||
	    lmmsize == 0) {
		rc = -ENODATA;
		goto out;
	}

	lmm = req_capsule_server_sized_get(&req->rq_pill,
					   &RMF_MDT_MD, lmmsize);
	LASSERT(lmm != NULL);

	/*
	 * This is coming from the MDS, so is probably in
	 * little endian.  We convert it to host endian before
	 * passing it to userspace.
	 */
	/* We don't swab objects for directories */
	switch (le32_to_cpu(lmm->lmm_magic)) {
	case LOV_MAGIC_V1:
		if (LOV_MAGIC != cpu_to_le32(LOV_MAGIC))
			lustre_swab_lov_user_md_v1((struct lov_user_md_v1 *)lmm);
		break;
	case LOV_MAGIC_V3:
		if (LOV_MAGIC != cpu_to_le32(LOV_MAGIC))
			lustre_swab_lov_user_md_v3((struct lov_user_md_v3 *)lmm);
		break;
	default:
		CERROR("unknown magic: %lX\n", (unsigned long)lmm->lmm_magic);
		rc = -EPROTO;
	}
out:
	*lmmp = lmm;
	*lmm_size = lmmsize;
	*request = req;
	return rc;
}

/*
 *  Get MDT index for the inode.
 */
int ll_get_mdt_idx(struct inode *inode)
{
	struct ll_sb_info *sbi = ll_i2sbi(inode);
	struct md_op_data *op_data;
	int rc, mdtidx;

	op_data = ll_prep_md_op_data(NULL, inode, NULL, NULL, 0,
				     0, LUSTRE_OPC_ANY, NULL);
	if (IS_ERR(op_data))
		return PTR_ERR(op_data);

	op_data->op_flags |= MF_GET_MDT_IDX;
	rc = md_getattr(sbi->ll_md_exp, op_data, NULL);
	mdtidx = op_data->op_mds;
	ll_finish_md_op_data(op_data);
	if (rc < 0) {
		CDEBUG(D_INFO, "md_getattr_name: %d\n", rc);
		return rc;
	}
	return mdtidx;
}

/**
 * Generic handler to do any pre-copy work.
 *
 * It send a first hsm_progress (with extent length == 0) to coordinator as a
 * first information for it that real work has started.
 *
 * Moreover, for a ARCHIVE request, it will sample the file data version and
 * store it in \a copy.
 *
 * \return 0 on success.
 */
static int ll_ioc_copy_start(struct super_block *sb, struct hsm_copy *copy)
{
	struct ll_sb_info		*sbi = ll_s2sbi(sb);
	struct hsm_progress_kernel	 hpk;
	int				 rc;

	/* Forge a hsm_progress based on data from copy. */
	hpk.hpk_fid = copy->hc_hai.hai_fid;
	hpk.hpk_cookie = copy->hc_hai.hai_cookie;
	hpk.hpk_extent.offset = copy->hc_hai.hai_extent.offset;
	hpk.hpk_extent.length = 0;
	hpk.hpk_flags = 0;
	hpk.hpk_errval = 0;
	hpk.hpk_data_version = 0;


	/* For archive request, we need to read the current file version. */
	if (copy->hc_hai.hai_action == HSMA_ARCHIVE) {
		struct inode	*inode;
		__u64		 data_version = 0;

		/* Get inode for this fid */
		inode = search_inode_for_lustre(sb, &copy->hc_hai.hai_fid);
		if (IS_ERR(inode)) {
			hpk.hpk_flags |= HP_FLAG_RETRY;
			/* hpk_errval is >= 0 */
			hpk.hpk_errval = -PTR_ERR(inode);
			rc = PTR_ERR(inode);
			goto progress;
		}

		/* Read current file data version */
		rc = ll_data_version(inode, &data_version, 1);
		iput(inode);
		if (rc != 0) {
			CDEBUG(D_HSM, "Could not read file data version of "
				      DFID" (rc = %d). Archive request (%#llx) could not be done.\n",
				      PFID(&copy->hc_hai.hai_fid), rc,
				      copy->hc_hai.hai_cookie);
			hpk.hpk_flags |= HP_FLAG_RETRY;
			/* hpk_errval must be >= 0 */
			hpk.hpk_errval = -rc;
			goto progress;
		}

		/* Store it the hsm_copy for later copytool use.
		 * Always modified even if no lsm. */
		copy->hc_data_version = data_version;
	}

progress:
	rc = obd_iocontrol(LL_IOC_HSM_PROGRESS, sbi->ll_md_exp, sizeof(hpk),
			   &hpk, NULL);

	return rc;
}

/**
 * Generic handler to do any post-copy work.
 *
 * It will send the last hsm_progress update to coordinator to inform it
 * that copy is finished and whether it was successful or not.
 *
 * Moreover,
 * - for ARCHIVE request, it will sample the file data version and compare it
 *   with the version saved in ll_ioc_copy_start(). If they do not match, copy
 *   will be considered as failed.
 * - for RESTORE request, it will sample the file data version and send it to
 *   coordinator which is useful if the file was imported as 'released'.
 *
 * \return 0 on success.
 */
static int ll_ioc_copy_end(struct super_block *sb, struct hsm_copy *copy)
{
	struct ll_sb_info		*sbi = ll_s2sbi(sb);
	struct hsm_progress_kernel	 hpk;
	int				 rc;

	/* If you modify the logic here, also check llapi_hsm_copy_end(). */
	/* Take care: copy->hc_hai.hai_action, len, gid and data are not
	 * initialized if copy_end was called with copy == NULL.
	 */

	/* Forge a hsm_progress based on data from copy. */
	hpk.hpk_fid = copy->hc_hai.hai_fid;
	hpk.hpk_cookie = copy->hc_hai.hai_cookie;
	hpk.hpk_extent = copy->hc_hai.hai_extent;
	hpk.hpk_flags = copy->hc_flags | HP_FLAG_COMPLETED;
	hpk.hpk_errval = copy->hc_errval;
	hpk.hpk_data_version = 0;

	/* For archive request, we need to check the file data was not changed.
	 *
	 * For restore request, we need to send the file data version, this is
	 * useful when the file was created using hsm_import.
	 */
	if (((copy->hc_hai.hai_action == HSMA_ARCHIVE) ||
	     (copy->hc_hai.hai_action == HSMA_RESTORE)) &&
	    (copy->hc_errval == 0)) {
		struct inode	*inode;
		__u64		 data_version = 0;

		/* Get lsm for this fid */
		inode = search_inode_for_lustre(sb, &copy->hc_hai.hai_fid);
		if (IS_ERR(inode)) {
			hpk.hpk_flags |= HP_FLAG_RETRY;
			/* hpk_errval must be >= 0 */
			hpk.hpk_errval = -PTR_ERR(inode);
			rc = PTR_ERR(inode);
			goto progress;
		}

		rc = ll_data_version(inode, &data_version,
				     copy->hc_hai.hai_action == HSMA_ARCHIVE);
		iput(inode);
		if (rc) {
			CDEBUG(D_HSM, "Could not read file data version. Request could not be confirmed.\n");
			if (hpk.hpk_errval == 0)
				hpk.hpk_errval = -rc;
			goto progress;
		}

		/* Store it the hsm_copy for later copytool use.
		 * Always modified even if no lsm. */
		hpk.hpk_data_version = data_version;

		/* File could have been stripped during archiving, so we need
		 * to check anyway. */
		if ((copy->hc_hai.hai_action == HSMA_ARCHIVE) &&
		    (copy->hc_data_version != data_version)) {
			CDEBUG(D_HSM, "File data version mismatched. File content was changed during archiving. "
			       DFID", start:%#llx current:%#llx\n",
			       PFID(&copy->hc_hai.hai_fid),
			       copy->hc_data_version, data_version);
			/* File was changed, send error to cdt. Do not ask for
			 * retry because if a file is modified frequently,
			 * the cdt will loop on retried archive requests.
			 * The policy engine will ask for a new archive later
			 * when the file will not be modified for some tunable
			 * time */
			/* we do not notify caller */
			hpk.hpk_flags &= ~HP_FLAG_RETRY;
			/* hpk_errval must be >= 0 */
			hpk.hpk_errval = EBUSY;
		}

	}

progress:
	rc = obd_iocontrol(LL_IOC_HSM_PROGRESS, sbi->ll_md_exp, sizeof(hpk),
			   &hpk, NULL);

	return rc;
}


static int copy_and_ioctl(int cmd, struct obd_export *exp,
			  const void __user *data, size_t size)
{
	void *copy;
	int rc;

	copy = kzalloc(size, GFP_NOFS);
	if (!copy)
		return -ENOMEM;

	if (copy_from_user(copy, data, size)) {
		rc = -EFAULT;
		goto out;
	}

	rc = obd_iocontrol(cmd, exp, size, copy, NULL);
out:
	OBD_FREE(copy, size);

	return rc;
}

static int quotactl_ioctl(struct ll_sb_info *sbi, struct if_quotactl *qctl)
{
	int cmd = qctl->qc_cmd;
	int type = qctl->qc_type;
	int id = qctl->qc_id;
	int valid = qctl->qc_valid;
	int rc = 0;

	switch (cmd) {
	case LUSTRE_Q_INVALIDATE:
	case LUSTRE_Q_FINVALIDATE:
	case Q_QUOTAON:
	case Q_QUOTAOFF:
	case Q_SETQUOTA:
	case Q_SETINFO:
		if (!capable(CFS_CAP_SYS_ADMIN) ||
		    sbi->ll_flags & LL_SBI_RMT_CLIENT)
			return -EPERM;
		break;
	case Q_GETQUOTA:
		if (((type == USRQUOTA &&
		      !uid_eq(current_euid(), make_kuid(&init_user_ns, id))) ||
		     (type == GRPQUOTA &&
		      !in_egroup_p(make_kgid(&init_user_ns, id)))) &&
		    (!capable(CFS_CAP_SYS_ADMIN) ||
		     sbi->ll_flags & LL_SBI_RMT_CLIENT))
			return -EPERM;
		break;
	case Q_GETINFO:
		break;
	default:
		CERROR("unsupported quotactl op: %#x\n", cmd);
		return -ENOTTY;
	}

	if (valid != QC_GENERAL) {
		if (sbi->ll_flags & LL_SBI_RMT_CLIENT)
			return -EOPNOTSUPP;

		if (cmd == Q_GETINFO)
			qctl->qc_cmd = Q_GETOINFO;
		else if (cmd == Q_GETQUOTA)
			qctl->qc_cmd = Q_GETOQUOTA;
		else
			return -EINVAL;

		switch (valid) {
		case QC_MDTIDX:
			rc = obd_iocontrol(OBD_IOC_QUOTACTL, sbi->ll_md_exp,
					   sizeof(*qctl), qctl, NULL);
			break;
		case QC_OSTIDX:
			rc = obd_iocontrol(OBD_IOC_QUOTACTL, sbi->ll_dt_exp,
					   sizeof(*qctl), qctl, NULL);
			break;
		case QC_UUID:
			rc = obd_iocontrol(OBD_IOC_QUOTACTL, sbi->ll_md_exp,
					   sizeof(*qctl), qctl, NULL);
			if (rc == -EAGAIN)
				rc = obd_iocontrol(OBD_IOC_QUOTACTL,
						   sbi->ll_dt_exp,
						   sizeof(*qctl), qctl, NULL);
			break;
		default:
			rc = -EINVAL;
			break;
		}

		if (rc)
			return rc;

		qctl->qc_cmd = cmd;
	} else {
		struct obd_quotactl *oqctl;

		oqctl = kzalloc(sizeof(*oqctl), GFP_NOFS);
		if (!oqctl)
			return -ENOMEM;

		QCTL_COPY(oqctl, qctl);
		rc = obd_quotactl(sbi->ll_md_exp, oqctl);
		if (rc) {
			if (rc != -EALREADY && cmd == Q_QUOTAON) {
				oqctl->qc_cmd = Q_QUOTAOFF;
				obd_quotactl(sbi->ll_md_exp, oqctl);
			}
			OBD_FREE_PTR(oqctl);
			return rc;
		}
		/* If QIF_SPACE is not set, client should collect the
		 * space usage from OSSs by itself */
		if (cmd == Q_GETQUOTA &&
		    !(oqctl->qc_dqblk.dqb_valid & QIF_SPACE) &&
		    !oqctl->qc_dqblk.dqb_curspace) {
			struct obd_quotactl *oqctl_tmp;

			oqctl_tmp = kzalloc(sizeof(*oqctl_tmp), GFP_NOFS);
			if (!oqctl_tmp) {
				rc = -ENOMEM;
				goto out;
			}

			oqctl_tmp->qc_cmd = Q_GETOQUOTA;
			oqctl_tmp->qc_id = oqctl->qc_id;
			oqctl_tmp->qc_type = oqctl->qc_type;

			/* collect space usage from OSTs */
			oqctl_tmp->qc_dqblk.dqb_curspace = 0;
			rc = obd_quotactl(sbi->ll_dt_exp, oqctl_tmp);
			if (!rc || rc == -EREMOTEIO) {
				oqctl->qc_dqblk.dqb_curspace =
					oqctl_tmp->qc_dqblk.dqb_curspace;
				oqctl->qc_dqblk.dqb_valid |= QIF_SPACE;
			}

			/* collect space & inode usage from MDTs */
			oqctl_tmp->qc_dqblk.dqb_curspace = 0;
			oqctl_tmp->qc_dqblk.dqb_curinodes = 0;
			rc = obd_quotactl(sbi->ll_md_exp, oqctl_tmp);
			if (!rc || rc == -EREMOTEIO) {
				oqctl->qc_dqblk.dqb_curspace +=
					oqctl_tmp->qc_dqblk.dqb_curspace;
				oqctl->qc_dqblk.dqb_curinodes =
					oqctl_tmp->qc_dqblk.dqb_curinodes;
				oqctl->qc_dqblk.dqb_valid |= QIF_INODES;
			} else {
				oqctl->qc_dqblk.dqb_valid &= ~QIF_SPACE;
			}

			OBD_FREE_PTR(oqctl_tmp);
		}
out:
		QCTL_COPY(qctl, oqctl);
		OBD_FREE_PTR(oqctl);
	}

	return rc;
}

static char *
ll_getname(const char __user *filename)
{
	int ret = 0, len;
	char *tmp = __getname();

	if (!tmp)
		return ERR_PTR(-ENOMEM);

	len = strncpy_from_user(tmp, filename, PATH_MAX);
	if (len == 0)
		ret = -ENOENT;
	else if (len > PATH_MAX)
		ret = -ENAMETOOLONG;

	if (ret) {
		__putname(tmp);
		tmp =  ERR_PTR(ret);
	}
	return tmp;
}

#define ll_putname(filename) __putname(filename)

static long ll_dir_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
{
	struct inode *inode = file_inode(file);
	struct ll_sb_info *sbi = ll_i2sbi(inode);
	struct obd_ioctl_data *data;
	int rc = 0;

	CDEBUG(D_VFSTRACE, "VFS Op:inode=%lu/%u(%p), cmd=%#x\n",
	       inode->i_ino, inode->i_generation, inode, cmd);

	/* asm-ppc{,64} declares TCGETS, et. al. as type 't' not 'T' */
	if (_IOC_TYPE(cmd) == 'T' || _IOC_TYPE(cmd) == 't') /* tty ioctls */
		return -ENOTTY;

	ll_stats_ops_tally(ll_i2sbi(inode), LPROC_LL_IOCTL, 1);
	switch (cmd) {
	case FSFILT_IOC_GETFLAGS:
	case FSFILT_IOC_SETFLAGS:
		return ll_iocontrol(inode, file, cmd, arg);
	case FSFILT_IOC_GETVERSION_OLD:
	case FSFILT_IOC_GETVERSION:
		return put_user(inode->i_generation, (int *)arg);
	/* We need to special case any other ioctls we want to handle,
	 * to send them to the MDS/OST as appropriate and to properly
	 * network encode the arg field.
	case FSFILT_IOC_SETVERSION_OLD:
	case FSFILT_IOC_SETVERSION:
	*/
	case LL_IOC_GET_MDTIDX: {
		int mdtidx;

		mdtidx = ll_get_mdt_idx(inode);
		if (mdtidx < 0)
			return mdtidx;

		if (put_user((int)mdtidx, (int *)arg))
			return -EFAULT;

		return 0;
	}
	case IOC_MDC_LOOKUP: {
		struct ptlrpc_request *request = NULL;
		int namelen, len = 0;
		char *buf = NULL;
		char *filename;
		struct md_op_data *op_data;

		rc = obd_ioctl_getdata(&buf, &len, (void *)arg);
		if (rc)
			return rc;
		data = (void *)buf;

		filename = data->ioc_inlbuf1;
		namelen = strlen(filename);

		if (namelen < 1) {
			CDEBUG(D_INFO, "IOC_MDC_LOOKUP missing filename\n");
			rc = -EINVAL;
			goto out_free;
		}

		op_data = ll_prep_md_op_data(NULL, inode, NULL, filename, namelen,
					     0, LUSTRE_OPC_ANY, NULL);
		if (IS_ERR(op_data)) {
			rc = PTR_ERR(op_data);
			goto out_free;
		}

		op_data->op_valid = OBD_MD_FLID;
		rc = md_getattr_name(sbi->ll_md_exp, op_data, &request);
		ll_finish_md_op_data(op_data);
		if (rc < 0) {
			CDEBUG(D_INFO, "md_getattr_name: %d\n", rc);
			goto out_free;
		}
		ptlrpc_req_finished(request);
out_free:
		obd_ioctl_freedata(buf, len);
		return rc;
	}
	case LL_IOC_LMV_SETSTRIPE: {
		struct lmv_user_md  *lum;
		char		*buf = NULL;
		char		*filename;
		int		 namelen = 0;
		int		 lumlen = 0;
		int		 len;
		int		 rc;

		rc = obd_ioctl_getdata(&buf, &len, (void *)arg);
		if (rc)
			return rc;

		data = (void *)buf;
		if (data->ioc_inlbuf1 == NULL || data->ioc_inlbuf2 == NULL ||
		    data->ioc_inllen1 == 0 || data->ioc_inllen2 == 0) {
			rc = -EINVAL;
			goto lmv_out_free;
		}

		filename = data->ioc_inlbuf1;
		namelen = data->ioc_inllen1;

		if (namelen < 1) {
			CDEBUG(D_INFO, "IOC_MDC_LOOKUP missing filename\n");
			rc = -EINVAL;
			goto lmv_out_free;
		}
		lum = (struct lmv_user_md *)data->ioc_inlbuf2;
		lumlen = data->ioc_inllen2;

		if (lum->lum_magic != LMV_USER_MAGIC ||
		    lumlen != sizeof(*lum)) {
			CERROR("%s: wrong lum magic %x or size %d: rc = %d\n",
			       filename, lum->lum_magic, lumlen, -EFAULT);
			rc = -EINVAL;
			goto lmv_out_free;
		}

		/**
		 * ll_dir_setdirstripe will be used to set dir stripe
		 *  mdc_create--->mdt_reint_create (with dirstripe)
		 */
		rc = ll_dir_setdirstripe(inode, lum, filename);
lmv_out_free:
		obd_ioctl_freedata(buf, len);
		return rc;

	}
	case LL_IOC_LOV_SETSTRIPE: {
		struct lov_user_md_v3 lumv3;
		struct lov_user_md_v1 *lumv1 = (struct lov_user_md_v1 *)&lumv3;
		struct lov_user_md_v1 *lumv1p = (struct lov_user_md_v1 *)arg;
		struct lov_user_md_v3 *lumv3p = (struct lov_user_md_v3 *)arg;

		int set_default = 0;

		LASSERT(sizeof(lumv3) == sizeof(*lumv3p));
		LASSERT(sizeof(lumv3.lmm_objects[0]) ==
			sizeof(lumv3p->lmm_objects[0]));
		/* first try with v1 which is smaller than v3 */
		if (copy_from_user(lumv1, lumv1p, sizeof(*lumv1)))
			return -EFAULT;

		if (lumv1->lmm_magic == LOV_USER_MAGIC_V3) {
			if (copy_from_user(&lumv3, lumv3p, sizeof(lumv3)))
				return -EFAULT;
		}

		if (is_root_inode(inode))
			set_default = 1;

		/* in v1 and v3 cases lumv1 points to data */
		rc = ll_dir_setstripe(inode, lumv1, set_default);

		return rc;
	}
	case LL_IOC_LMV_GETSTRIPE: {
		struct lmv_user_md *lump = (struct lmv_user_md *)arg;
		struct lmv_user_md lum;
		struct lmv_user_md *tmp;
		int lum_size;
		int rc = 0;
		int mdtindex;

		if (copy_from_user(&lum, lump, sizeof(struct lmv_user_md)))
			return -EFAULT;

		if (lum.lum_magic != LMV_MAGIC_V1)
			return -EINVAL;

		lum_size = lmv_user_md_size(1, LMV_MAGIC_V1);
		tmp = kzalloc(lum_size, GFP_NOFS);
		if (!tmp) {
			rc = -ENOMEM;
			goto free_lmv;
		}

		*tmp = lum;
		tmp->lum_type = LMV_STRIPE_TYPE;
		tmp->lum_stripe_count = 1;
		mdtindex = ll_get_mdt_idx(inode);
		if (mdtindex < 0) {
			rc = -ENOMEM;
			goto free_lmv;
		}

		tmp->lum_stripe_offset = mdtindex;
		tmp->lum_objects[0].lum_mds = mdtindex;
		memcpy(&tmp->lum_objects[0].lum_fid, ll_inode2fid(inode),
		       sizeof(struct lu_fid));
		if (copy_to_user((void *)arg, tmp, lum_size)) {
			rc = -EFAULT;
			goto free_lmv;
		}
free_lmv:
		if (tmp)
			OBD_FREE(tmp, lum_size);
		return rc;
	}
	case LL_IOC_REMOVE_ENTRY: {
		char		*filename = NULL;
		int		 namelen = 0;
		int		 rc;

		/* Here is a little hack to avoid sending REINT_RMENTRY to
		 * unsupported server, which might crash the server(LU-2730),
		 * Because both LVB_TYPE and REINT_RMENTRY will be supported
		 * on 2.4, we use OBD_CONNECT_LVB_TYPE to detect whether the
		 * server will support REINT_RMENTRY XXX*/
		if (!(exp_connect_flags(sbi->ll_md_exp) & OBD_CONNECT_LVB_TYPE))
			return -ENOTSUPP;

		filename = ll_getname((const char *)arg);
		if (IS_ERR(filename))
			return PTR_ERR(filename);

		namelen = strlen(filename);
		if (namelen < 1) {
			rc = -EINVAL;
			goto out_rmdir;
		}

		rc = ll_rmdir_entry(inode, filename, namelen);
out_rmdir:
		if (filename)
			ll_putname(filename);
		return rc;
	}
	case LL_IOC_LOV_SWAP_LAYOUTS:
		return -EPERM;
	case LL_IOC_OBD_STATFS:
		return ll_obd_statfs(inode, (void *)arg);
	case LL_IOC_LOV_GETSTRIPE:
	case LL_IOC_MDC_GETINFO:
	case IOC_MDC_GETFILEINFO:
	case IOC_MDC_GETFILESTRIPE: {
		struct ptlrpc_request *request = NULL;
		struct lov_user_md *lump;
		struct lov_mds_md *lmm = NULL;
		struct mdt_body *body;
		char *filename = NULL;
		int lmmsize;

		if (cmd == IOC_MDC_GETFILEINFO ||
		    cmd == IOC_MDC_GETFILESTRIPE) {
			filename = ll_getname((const char *)arg);
			if (IS_ERR(filename))
				return PTR_ERR(filename);

			rc = ll_lov_getstripe_ea_info(inode, filename, &lmm,
						      &lmmsize, &request);
		} else {
			rc = ll_dir_getstripe(inode, &lmm, &lmmsize, &request);
		}

		if (request) {
			body = req_capsule_server_get(&request->rq_pill,
						      &RMF_MDT_BODY);
			LASSERT(body != NULL);
		} else {
			goto out_req;
		}

		if (rc < 0) {
			if (rc == -ENODATA && (cmd == IOC_MDC_GETFILEINFO ||
					       cmd == LL_IOC_MDC_GETINFO)) {
				rc = 0;
				goto skip_lmm;
			} else
				goto out_req;
		}

		if (cmd == IOC_MDC_GETFILESTRIPE ||
		    cmd == LL_IOC_LOV_GETSTRIPE) {
			lump = (struct lov_user_md *)arg;
		} else {
			struct lov_user_mds_data *lmdp;
			lmdp = (struct lov_user_mds_data *)arg;
			lump = &lmdp->lmd_lmm;
		}
		if (copy_to_user(lump, lmm, lmmsize)) {
			if (copy_to_user(lump, lmm, sizeof(*lump))) {
				rc = -EFAULT;
				goto out_req;
			}
			rc = -EOVERFLOW;
		}
skip_lmm:
		if (cmd == IOC_MDC_GETFILEINFO || cmd == LL_IOC_MDC_GETINFO) {
			struct lov_user_mds_data *lmdp;
			lstat_t st = { 0 };

			st.st_dev     = inode->i_sb->s_dev;
			st.st_mode    = body->mode;
			st.st_nlink   = body->nlink;
			st.st_uid     = body->uid;
			st.st_gid     = body->gid;
			st.st_rdev    = body->rdev;
			st.st_size    = body->size;
			st.st_blksize = PAGE_CACHE_SIZE;
			st.st_blocks  = body->blocks;
			st.st_atime   = body->atime;
			st.st_mtime   = body->mtime;
			st.st_ctime   = body->ctime;
			st.st_ino     = inode->i_ino;

			lmdp = (struct lov_user_mds_data *)arg;
			if (copy_to_user(&lmdp->lmd_st, &st, sizeof(st))) {
				rc = -EFAULT;
				goto out_req;
			}
		}

out_req:
		ptlrpc_req_finished(request);
		if (filename)
			ll_putname(filename);
		return rc;
	}
	case IOC_LOV_GETINFO: {
		struct lov_user_mds_data *lumd;
		struct lov_stripe_md *lsm;
		struct lov_user_md *lum;
		struct lov_mds_md *lmm;
		int lmmsize;
		lstat_t st;

		lumd = (struct lov_user_mds_data *)arg;
		lum = &lumd->lmd_lmm;

		rc = ll_get_max_mdsize(sbi, &lmmsize);
		if (rc)
			return rc;

		OBD_ALLOC_LARGE(lmm, lmmsize);
		if (lmm == NULL)
			return -ENOMEM;
		if (copy_from_user(lmm, lum, lmmsize)) {
			rc = -EFAULT;
			goto free_lmm;
		}

		switch (lmm->lmm_magic) {
		case LOV_USER_MAGIC_V1:
			if (LOV_USER_MAGIC_V1 == cpu_to_le32(LOV_USER_MAGIC_V1))
				break;
			/* swab objects first so that stripes num will be sane */
			lustre_swab_lov_user_md_objects(
				((struct lov_user_md_v1 *)lmm)->lmm_objects,
				((struct lov_user_md_v1 *)lmm)->lmm_stripe_count);
			lustre_swab_lov_user_md_v1((struct lov_user_md_v1 *)lmm);
			break;
		case LOV_USER_MAGIC_V3:
			if (LOV_USER_MAGIC_V3 == cpu_to_le32(LOV_USER_MAGIC_V3))
				break;
			/* swab objects first so that stripes num will be sane */
			lustre_swab_lov_user_md_objects(
				((struct lov_user_md_v3 *)lmm)->lmm_objects,
				((struct lov_user_md_v3 *)lmm)->lmm_stripe_count);
			lustre_swab_lov_user_md_v3((struct lov_user_md_v3 *)lmm);
			break;
		default:
			rc = -EINVAL;
			goto free_lmm;
		}

		rc = obd_unpackmd(sbi->ll_dt_exp, &lsm, lmm, lmmsize);
		if (rc < 0) {
			rc = -ENOMEM;
			goto free_lmm;
		}

		/* Perform glimpse_size operation. */
		memset(&st, 0, sizeof(st));

		rc = ll_glimpse_ioctl(sbi, lsm, &st);
		if (rc)
			goto free_lsm;

		if (copy_to_user(&lumd->lmd_st, &st, sizeof(st))) {
			rc = -EFAULT;
			goto free_lsm;
		}

free_lsm:
		obd_free_memmd(sbi->ll_dt_exp, &lsm);
free_lmm:
		OBD_FREE_LARGE(lmm, lmmsize);
		return rc;
	}
	case OBD_IOC_LLOG_CATINFO: {
		return -EOPNOTSUPP;
	}
	case OBD_IOC_QUOTACHECK: {
		struct obd_quotactl *oqctl;
		int error = 0;

		if (!capable(CFS_CAP_SYS_ADMIN) ||
		    sbi->ll_flags & LL_SBI_RMT_CLIENT)
			return -EPERM;

		oqctl = kzalloc(sizeof(*oqctl), GFP_NOFS);
		if (!oqctl)
			return -ENOMEM;
		oqctl->qc_type = arg;
		rc = obd_quotacheck(sbi->ll_md_exp, oqctl);
		if (rc < 0) {
			CDEBUG(D_INFO, "md_quotacheck failed: rc %d\n", rc);
			error = rc;
		}

		rc = obd_quotacheck(sbi->ll_dt_exp, oqctl);
		if (rc < 0)
			CDEBUG(D_INFO, "obd_quotacheck failed: rc %d\n", rc);

		OBD_FREE_PTR(oqctl);
		return error ?: rc;
	}
	case OBD_IOC_POLL_QUOTACHECK: {
		struct if_quotacheck *check;

		if (!capable(CFS_CAP_SYS_ADMIN) ||
		    sbi->ll_flags & LL_SBI_RMT_CLIENT)
			return -EPERM;

		check = kzalloc(sizeof(*check), GFP_NOFS);
		if (!check)
			return -ENOMEM;

		rc = obd_iocontrol(cmd, sbi->ll_md_exp, 0, (void *)check,
				   NULL);
		if (rc) {
			CDEBUG(D_QUOTA, "mdc ioctl %d failed: %d\n", cmd, rc);
			if (copy_to_user((void *)arg, check,
					     sizeof(*check)))
				CDEBUG(D_QUOTA, "copy_to_user failed\n");
			goto out_poll;
		}

		rc = obd_iocontrol(cmd, sbi->ll_dt_exp, 0, (void *)check,
				   NULL);
		if (rc) {
			CDEBUG(D_QUOTA, "osc ioctl %d failed: %d\n", cmd, rc);
			if (copy_to_user((void *)arg, check,
					     sizeof(*check)))
				CDEBUG(D_QUOTA, "copy_to_user failed\n");
			goto out_poll;
		}
out_poll:
		OBD_FREE_PTR(check);
		return rc;
	}
	case LL_IOC_QUOTACTL: {
		struct if_quotactl *qctl;

		qctl = kzalloc(sizeof(*qctl), GFP_NOFS);
		if (!qctl)
			return -ENOMEM;

		if (copy_from_user(qctl, (void *)arg, sizeof(*qctl))) {
			rc = -EFAULT;
			goto out_quotactl;
		}

		rc = quotactl_ioctl(sbi, qctl);

		if (rc == 0 && copy_to_user((void *)arg, qctl, sizeof(*qctl)))
			rc = -EFAULT;

out_quotactl:
		OBD_FREE_PTR(qctl);
		return rc;
	}
	case OBD_IOC_GETDTNAME:
	case OBD_IOC_GETMDNAME:
		return ll_get_obd_name(inode, cmd, arg);
	case LL_IOC_FLUSHCTX:
		return ll_flush_ctx(inode);
#ifdef CONFIG_FS_POSIX_ACL
	case LL_IOC_RMTACL: {
	    if (sbi->ll_flags & LL_SBI_RMT_CLIENT && is_root_inode(inode)) {
		struct ll_file_data *fd = LUSTRE_FPRIVATE(file);

		LASSERT(fd != NULL);
		rc = rct_add(&sbi->ll_rct, current_pid(), arg);
		if (!rc)
			fd->fd_flags |= LL_FILE_RMTACL;
		return rc;
	    } else
		return 0;
	}
#endif
	case LL_IOC_GETOBDCOUNT: {
		int count, vallen;
		struct obd_export *exp;

		if (copy_from_user(&count, (int *)arg, sizeof(int)))
			return -EFAULT;

		/* get ost count when count is zero, get mdt count otherwise */
		exp = count ? sbi->ll_md_exp : sbi->ll_dt_exp;
		vallen = sizeof(count);
		rc = obd_get_info(NULL, exp, sizeof(KEY_TGT_COUNT),
				  KEY_TGT_COUNT, &vallen, &count, NULL);
		if (rc) {
			CERROR("get target count failed: %d\n", rc);
			return rc;
		}

		if (copy_to_user((int *)arg, &count, sizeof(int)))
			return -EFAULT;

		return 0;
	}
	case LL_IOC_PATH2FID:
		if (copy_to_user((void *)arg, ll_inode2fid(inode),
				     sizeof(struct lu_fid)))
			return -EFAULT;
		return 0;
	case LL_IOC_GET_CONNECT_FLAGS: {
		return obd_iocontrol(cmd, sbi->ll_md_exp, 0, NULL, (void *)arg);
	}
	case OBD_IOC_CHANGELOG_SEND:
	case OBD_IOC_CHANGELOG_CLEAR:
		rc = copy_and_ioctl(cmd, sbi->ll_md_exp, (void *)arg,
				    sizeof(struct ioc_changelog));
		return rc;
	case OBD_IOC_FID2PATH:
		return ll_fid2path(inode, (void *)arg);
	case LL_IOC_HSM_REQUEST: {
		struct hsm_user_request	*hur;
		ssize_t			 totalsize;

		hur = kzalloc(sizeof(*hur), GFP_NOFS);
		if (!hur)
			return -ENOMEM;

		/* We don't know the true size yet; copy the fixed-size part */
		if (copy_from_user(hur, (void *)arg, sizeof(*hur))) {
			OBD_FREE_PTR(hur);
			return -EFAULT;
		}

		/* Compute the whole struct size */
		totalsize = hur_len(hur);
		OBD_FREE_PTR(hur);
		if (totalsize < 0)
			return -E2BIG;

		/* Final size will be more than double totalsize */
		if (totalsize >= MDS_MAXREQSIZE / 3)
			return -E2BIG;

		OBD_ALLOC_LARGE(hur, totalsize);
		if (hur == NULL)
			return -ENOMEM;

		/* Copy the whole struct */
		if (copy_from_user(hur, (void *)arg, totalsize)) {
			OBD_FREE_LARGE(hur, totalsize);
			return -EFAULT;
		}

		if (hur->hur_request.hr_action == HUA_RELEASE) {
			const struct lu_fid *fid;
			struct inode *f;
			int i;

			for (i = 0; i < hur->hur_request.hr_itemcount; i++) {
				fid = &hur->hur_user_item[i].hui_fid;
				f = search_inode_for_lustre(inode->i_sb, fid);
				if (IS_ERR(f)) {
					rc = PTR_ERR(f);
					break;
				}

				rc = ll_hsm_release(f);
				iput(f);
				if (rc != 0)
					break;
			}
		} else {
			rc = obd_iocontrol(cmd, ll_i2mdexp(inode), totalsize,
					   hur, NULL);
		}

		OBD_FREE_LARGE(hur, totalsize);

		return rc;
	}
	case LL_IOC_HSM_PROGRESS: {
		struct hsm_progress_kernel	hpk;
		struct hsm_progress		hp;

		if (copy_from_user(&hp, (void *)arg, sizeof(hp)))
			return -EFAULT;

		hpk.hpk_fid = hp.hp_fid;
		hpk.hpk_cookie = hp.hp_cookie;
		hpk.hpk_extent = hp.hp_extent;
		hpk.hpk_flags = hp.hp_flags;
		hpk.hpk_errval = hp.hp_errval;
		hpk.hpk_data_version = 0;

		/* File may not exist in Lustre; all progress
		 * reported to Lustre root */
		rc = obd_iocontrol(cmd, sbi->ll_md_exp, sizeof(hpk), &hpk,
				   NULL);
		return rc;
	}
	case LL_IOC_HSM_CT_START:
		rc = copy_and_ioctl(cmd, sbi->ll_md_exp, (void *)arg,
				    sizeof(struct lustre_kernelcomm));
		return rc;

	case LL_IOC_HSM_COPY_START: {
		struct hsm_copy	*copy;
		int		 rc;

		copy = kzalloc(sizeof(*copy), GFP_NOFS);
		if (!copy)
			return -ENOMEM;
		if (copy_from_user(copy, (char *)arg, sizeof(*copy))) {
			OBD_FREE_PTR(copy);
			return -EFAULT;
		}

		rc = ll_ioc_copy_start(inode->i_sb, copy);
		if (copy_to_user((char *)arg, copy, sizeof(*copy)))
			rc = -EFAULT;

		OBD_FREE_PTR(copy);
		return rc;
	}
	case LL_IOC_HSM_COPY_END: {
		struct hsm_copy	*copy;
		int		 rc;

		copy = kzalloc(sizeof(*copy), GFP_NOFS);
		if (!copy)
			return -ENOMEM;
		if (copy_from_user(copy, (char *)arg, sizeof(*copy))) {
			OBD_FREE_PTR(copy);
			return -EFAULT;
		}

		rc = ll_ioc_copy_end(inode->i_sb, copy);
		if (copy_to_user((char *)arg, copy, sizeof(*copy)))
			rc = -EFAULT;

		OBD_FREE_PTR(copy);
		return rc;
	}
	default:
		return obd_iocontrol(cmd, sbi->ll_dt_exp, 0, NULL, (void *)arg);
	}
}

static loff_t ll_dir_seek(struct file *file, loff_t offset, int origin)
{
	struct inode *inode = file->f_mapping->host;
	struct ll_file_data *fd = LUSTRE_FPRIVATE(file);
	struct ll_sb_info *sbi = ll_i2sbi(inode);
	int api32 = ll_need_32bit_api(sbi);
	loff_t ret = -EINVAL;

	mutex_lock(&inode->i_mutex);
	switch (origin) {
		case SEEK_SET:
			break;
		case SEEK_CUR:
			offset += file->f_pos;
			break;
		case SEEK_END:
			if (offset > 0)
				goto out;
			if (api32)
				offset += LL_DIR_END_OFF_32BIT;
			else
				offset += LL_DIR_END_OFF;
			break;
		default:
			goto out;
	}

	if (offset >= 0 &&
	    ((api32 && offset <= LL_DIR_END_OFF_32BIT) ||
	     (!api32 && offset <= LL_DIR_END_OFF))) {
		if (offset != file->f_pos) {
			if ((api32 && offset == LL_DIR_END_OFF_32BIT) ||
			    (!api32 && offset == LL_DIR_END_OFF))
				fd->lfd_pos = MDS_DIR_END_OFF;
			else if (api32 && sbi->ll_flags & LL_SBI_64BIT_HASH)
				fd->lfd_pos = offset << 32;
			else
				fd->lfd_pos = offset;
			file->f_pos = offset;
			file->f_version = 0;
		}
		ret = offset;
	}
	goto out;

out:
	mutex_unlock(&inode->i_mutex);
	return ret;
}

static int ll_dir_open(struct inode *inode, struct file *file)
{
	return ll_file_open(inode, file);
}

static int ll_dir_release(struct inode *inode, struct file *file)
{
	return ll_file_release(inode, file);
}

const struct file_operations ll_dir_operations = {
	.llseek   = ll_dir_seek,
	.open     = ll_dir_open,
	.release  = ll_dir_release,
	.read     = generic_read_dir,
	.iterate  = ll_readdir,
	.unlocked_ioctl   = ll_dir_ioctl,
	.fsync    = ll_fsync,
};