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
|
/*
* 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) 2010, 2012, Intel Corporation.
*/
/*
* This file is part of Lustre, http://www.lustre.org/
* Lustre is a trademark of Sun Microsystems, Inc.
*
* lustre/ldlm/ldlm_resource.c
*
* Author: Phil Schwan <phil@clusterfs.com>
* Author: Peter Braam <braam@clusterfs.com>
*/
#define DEBUG_SUBSYSTEM S_LDLM
#include "../include/lustre_dlm.h"
#include "../include/lustre_fid.h"
#include "../include/obd_class.h"
#include "ldlm_internal.h"
struct kmem_cache *ldlm_resource_slab, *ldlm_lock_slab;
int ldlm_srv_namespace_nr = 0;
int ldlm_cli_namespace_nr = 0;
struct mutex ldlm_srv_namespace_lock;
LIST_HEAD(ldlm_srv_namespace_list);
struct mutex ldlm_cli_namespace_lock;
/* Client Namespaces that have active resources in them.
* Once all resources go away, ldlm_poold moves such namespaces to the
* inactive list */
LIST_HEAD(ldlm_cli_active_namespace_list);
/* Client namespaces that don't have any locks in them */
LIST_HEAD(ldlm_cli_inactive_namespace_list);
static struct dentry *ldlm_debugfs_dir;
static struct dentry *ldlm_ns_debugfs_dir;
struct dentry *ldlm_svc_debugfs_dir;
/* during debug dump certain amount of granted locks for one resource to avoid
* DDOS. */
unsigned int ldlm_dump_granted_max = 256;
static ssize_t
lprocfs_wr_dump_ns(struct file *file, const char __user *buffer,
size_t count, loff_t *off)
{
ldlm_dump_all_namespaces(LDLM_NAMESPACE_SERVER, D_DLMTRACE);
ldlm_dump_all_namespaces(LDLM_NAMESPACE_CLIENT, D_DLMTRACE);
return count;
}
LPROC_SEQ_FOPS_WR_ONLY(ldlm, dump_ns);
LPROC_SEQ_FOPS_RW_TYPE(ldlm_rw, uint);
static struct lprocfs_vars ldlm_debugfs_list[] = {
{ "dump_namespaces", &ldlm_dump_ns_fops, NULL, 0222 },
{ "dump_granted_max", &ldlm_rw_uint_fops, &ldlm_dump_granted_max },
{ NULL }
};
int ldlm_debugfs_setup(void)
{
int rc;
ldlm_debugfs_dir = ldebugfs_register(OBD_LDLM_DEVICENAME,
debugfs_lustre_root,
NULL, NULL);
if (IS_ERR_OR_NULL(ldlm_debugfs_dir)) {
CERROR("LProcFS failed in ldlm-init\n");
rc = ldlm_debugfs_dir ? PTR_ERR(ldlm_debugfs_dir) : -ENOMEM;
goto err;
}
ldlm_ns_debugfs_dir = ldebugfs_register("namespaces",
ldlm_debugfs_dir,
NULL, NULL);
if (IS_ERR_OR_NULL(ldlm_ns_debugfs_dir)) {
CERROR("LProcFS failed in ldlm-init\n");
rc = ldlm_ns_debugfs_dir ? PTR_ERR(ldlm_ns_debugfs_dir)
: -ENOMEM;
goto err_type;
}
ldlm_svc_debugfs_dir = ldebugfs_register("services",
ldlm_debugfs_dir,
NULL, NULL);
if (IS_ERR_OR_NULL(ldlm_svc_debugfs_dir)) {
CERROR("LProcFS failed in ldlm-init\n");
rc = ldlm_svc_debugfs_dir ? PTR_ERR(ldlm_svc_debugfs_dir)
: -ENOMEM;
goto err_ns;
}
rc = ldebugfs_add_vars(ldlm_debugfs_dir, ldlm_debugfs_list, NULL);
return 0;
err_ns:
ldebugfs_remove(&ldlm_ns_debugfs_dir);
err_type:
ldebugfs_remove(&ldlm_debugfs_dir);
err:
ldlm_svc_debugfs_dir = NULL;
ldlm_ns_debugfs_dir = NULL;
ldlm_debugfs_dir = NULL;
return rc;
}
void ldlm_debugfs_cleanup(void)
{
if (!IS_ERR_OR_NULL(ldlm_svc_debugfs_dir))
ldebugfs_remove(&ldlm_svc_debugfs_dir);
if (!IS_ERR_OR_NULL(ldlm_ns_debugfs_dir))
ldebugfs_remove(&ldlm_ns_debugfs_dir);
if (!IS_ERR_OR_NULL(ldlm_debugfs_dir))
ldebugfs_remove(&ldlm_debugfs_dir);
ldlm_svc_debugfs_dir = NULL;
ldlm_ns_debugfs_dir = NULL;
ldlm_debugfs_dir = NULL;
}
static ssize_t resource_count_show(struct kobject *kobj, struct attribute *attr,
char *buf)
{
struct ldlm_namespace *ns = container_of(kobj, struct ldlm_namespace,
ns_kobj);
__u64 res = 0;
struct cfs_hash_bd bd;
int i;
/* result is not strictly consistent */
cfs_hash_for_each_bucket(ns->ns_rs_hash, &bd, i)
res += cfs_hash_bd_count_get(&bd);
return sprintf(buf, "%lld\n", res);
}
LUSTRE_RO_ATTR(resource_count);
static ssize_t lock_count_show(struct kobject *kobj, struct attribute *attr,
char *buf)
{
struct ldlm_namespace *ns = container_of(kobj, struct ldlm_namespace,
ns_kobj);
__u64 locks;
locks = lprocfs_stats_collector(ns->ns_stats, LDLM_NSS_LOCKS,
LPROCFS_FIELDS_FLAGS_SUM);
return sprintf(buf, "%lld\n", locks);
}
LUSTRE_RO_ATTR(lock_count);
static ssize_t lock_unused_count_show(struct kobject *kobj,
struct attribute *attr,
char *buf)
{
struct ldlm_namespace *ns = container_of(kobj, struct ldlm_namespace,
ns_kobj);
return sprintf(buf, "%d\n", ns->ns_nr_unused);
}
LUSTRE_RO_ATTR(lock_unused_count);
static ssize_t lru_size_show(struct kobject *kobj, struct attribute *attr,
char *buf)
{
struct ldlm_namespace *ns = container_of(kobj, struct ldlm_namespace,
ns_kobj);
__u32 *nr = &ns->ns_max_unused;
if (ns_connect_lru_resize(ns))
nr = &ns->ns_nr_unused;
return sprintf(buf, "%u", *nr);
}
static ssize_t lru_size_store(struct kobject *kobj, struct attribute *attr,
const char *buffer, size_t count)
{
struct ldlm_namespace *ns = container_of(kobj, struct ldlm_namespace,
ns_kobj);
unsigned long tmp;
int lru_resize;
int err;
if (strncmp(buffer, "clear", 5) == 0) {
CDEBUG(D_DLMTRACE,
"dropping all unused locks from namespace %s\n",
ldlm_ns_name(ns));
if (ns_connect_lru_resize(ns)) {
int canceled, unused = ns->ns_nr_unused;
/* Try to cancel all @ns_nr_unused locks. */
canceled = ldlm_cancel_lru(ns, unused, 0,
LDLM_CANCEL_PASSED);
if (canceled < unused) {
CDEBUG(D_DLMTRACE,
"not all requested locks are canceled, requested: %d, canceled: %d\n",
unused,
canceled);
return -EINVAL;
}
} else {
tmp = ns->ns_max_unused;
ns->ns_max_unused = 0;
ldlm_cancel_lru(ns, 0, 0, LDLM_CANCEL_PASSED);
ns->ns_max_unused = tmp;
}
return count;
}
err = kstrtoul(buffer, 10, &tmp);
if (err != 0) {
CERROR("lru_size: invalid value written\n");
return -EINVAL;
}
lru_resize = (tmp == 0);
if (ns_connect_lru_resize(ns)) {
if (!lru_resize)
ns->ns_max_unused = (unsigned int)tmp;
if (tmp > ns->ns_nr_unused)
tmp = ns->ns_nr_unused;
tmp = ns->ns_nr_unused - tmp;
CDEBUG(D_DLMTRACE,
"changing namespace %s unused locks from %u to %u\n",
ldlm_ns_name(ns), ns->ns_nr_unused,
(unsigned int)tmp);
ldlm_cancel_lru(ns, tmp, LCF_ASYNC, LDLM_CANCEL_PASSED);
if (!lru_resize) {
CDEBUG(D_DLMTRACE,
"disable lru_resize for namespace %s\n",
ldlm_ns_name(ns));
ns->ns_connect_flags &= ~OBD_CONNECT_LRU_RESIZE;
}
} else {
CDEBUG(D_DLMTRACE,
"changing namespace %s max_unused from %u to %u\n",
ldlm_ns_name(ns), ns->ns_max_unused,
(unsigned int)tmp);
ns->ns_max_unused = (unsigned int)tmp;
ldlm_cancel_lru(ns, 0, LCF_ASYNC, LDLM_CANCEL_PASSED);
/* Make sure that LRU resize was originally supported before
* turning it on here. */
if (lru_resize &&
(ns->ns_orig_connect_flags & OBD_CONNECT_LRU_RESIZE)) {
CDEBUG(D_DLMTRACE,
"enable lru_resize for namespace %s\n",
ldlm_ns_name(ns));
ns->ns_connect_flags |= OBD_CONNECT_LRU_RESIZE;
}
}
return count;
}
LUSTRE_RW_ATTR(lru_size);
static ssize_t lru_max_age_show(struct kobject *kobj, struct attribute *attr,
char *buf)
{
struct ldlm_namespace *ns = container_of(kobj, struct ldlm_namespace,
ns_kobj);
return sprintf(buf, "%u", ns->ns_max_age);
}
static ssize_t lru_max_age_store(struct kobject *kobj, struct attribute *attr,
const char *buffer, size_t count)
{
struct ldlm_namespace *ns = container_of(kobj, struct ldlm_namespace,
ns_kobj);
unsigned long tmp;
int err;
err = kstrtoul(buffer, 10, &tmp);
if (err != 0)
return -EINVAL;
ns->ns_max_age = tmp;
return count;
}
LUSTRE_RW_ATTR(lru_max_age);
static ssize_t early_lock_cancel_show(struct kobject *kobj,
struct attribute *attr,
char *buf)
{
struct ldlm_namespace *ns = container_of(kobj, struct ldlm_namespace,
ns_kobj);
return sprintf(buf, "%d\n", ns_connect_cancelset(ns));
}
static ssize_t early_lock_cancel_store(struct kobject *kobj,
struct attribute *attr,
const char *buffer,
size_t count)
{
struct ldlm_namespace *ns = container_of(kobj, struct ldlm_namespace,
ns_kobj);
unsigned long supp = -1;
int rc;
rc = kstrtoul(buffer, 10, &supp);
if (rc < 0)
return rc;
if (supp == 0)
ns->ns_connect_flags &= ~OBD_CONNECT_CANCELSET;
else if (ns->ns_orig_connect_flags & OBD_CONNECT_CANCELSET)
ns->ns_connect_flags |= OBD_CONNECT_CANCELSET;
return count;
}
LUSTRE_RW_ATTR(early_lock_cancel);
/* These are for namespaces in /sys/fs/lustre/ldlm/namespaces/ */
static struct attribute *ldlm_ns_attrs[] = {
&lustre_attr_resource_count.attr,
&lustre_attr_lock_count.attr,
&lustre_attr_lock_unused_count.attr,
&lustre_attr_lru_size.attr,
&lustre_attr_lru_max_age.attr,
&lustre_attr_early_lock_cancel.attr,
NULL,
};
static void ldlm_ns_release(struct kobject *kobj)
{
struct ldlm_namespace *ns = container_of(kobj, struct ldlm_namespace,
ns_kobj);
complete(&ns->ns_kobj_unregister);
}
static struct kobj_type ldlm_ns_ktype = {
.default_attrs = ldlm_ns_attrs,
.sysfs_ops = &lustre_sysfs_ops,
.release = ldlm_ns_release,
};
static void ldlm_namespace_debugfs_unregister(struct ldlm_namespace *ns)
{
if (IS_ERR_OR_NULL(ns->ns_debugfs_entry))
CERROR("dlm namespace %s has no procfs dir?\n",
ldlm_ns_name(ns));
else
ldebugfs_remove(&ns->ns_debugfs_entry);
if (ns->ns_stats != NULL)
lprocfs_free_stats(&ns->ns_stats);
}
void ldlm_namespace_sysfs_unregister(struct ldlm_namespace *ns)
{
kobject_put(&ns->ns_kobj);
wait_for_completion(&ns->ns_kobj_unregister);
}
int ldlm_namespace_sysfs_register(struct ldlm_namespace *ns)
{
int err;
ns->ns_kobj.kset = ldlm_ns_kset;
init_completion(&ns->ns_kobj_unregister);
err = kobject_init_and_add(&ns->ns_kobj, &ldlm_ns_ktype, NULL,
"%s", ldlm_ns_name(ns));
ns->ns_stats = lprocfs_alloc_stats(LDLM_NSS_LAST, 0);
if (ns->ns_stats == NULL) {
kobject_put(&ns->ns_kobj);
return -ENOMEM;
}
lprocfs_counter_init(ns->ns_stats, LDLM_NSS_LOCKS,
LPROCFS_CNTR_AVGMINMAX, "locks", "locks");
return err;
}
static int ldlm_namespace_debugfs_register(struct ldlm_namespace *ns)
{
struct dentry *ns_entry;
if (!IS_ERR_OR_NULL(ns->ns_debugfs_entry)) {
ns_entry = ns->ns_debugfs_entry;
} else {
ns_entry = debugfs_create_dir(ldlm_ns_name(ns),
ldlm_ns_debugfs_dir);
if (ns_entry == NULL)
return -ENOMEM;
ns->ns_debugfs_entry = ns_entry;
}
return 0;
}
#undef MAX_STRING_SIZE
static unsigned ldlm_res_hop_hash(struct cfs_hash *hs,
const void *key, unsigned mask)
{
const struct ldlm_res_id *id = key;
unsigned val = 0;
unsigned i;
for (i = 0; i < RES_NAME_SIZE; i++)
val += id->name[i];
return val & mask;
}
static unsigned ldlm_res_hop_fid_hash(struct cfs_hash *hs,
const void *key, unsigned mask)
{
const struct ldlm_res_id *id = key;
struct lu_fid fid;
__u32 hash;
__u32 val;
fid.f_seq = id->name[LUSTRE_RES_ID_SEQ_OFF];
fid.f_oid = (__u32)id->name[LUSTRE_RES_ID_VER_OID_OFF];
fid.f_ver = (__u32)(id->name[LUSTRE_RES_ID_VER_OID_OFF] >> 32);
hash = fid_flatten32(&fid);
hash += (hash >> 4) + (hash << 12); /* mixing oid and seq */
if (id->name[LUSTRE_RES_ID_HSH_OFF] != 0) {
val = id->name[LUSTRE_RES_ID_HSH_OFF];
hash += (val >> 5) + (val << 11);
} else {
val = fid_oid(&fid);
}
hash = hash_long(hash, hs->hs_bkt_bits);
/* give me another random factor */
hash -= hash_long((unsigned long)hs, val % 11 + 3);
hash <<= hs->hs_cur_bits - hs->hs_bkt_bits;
hash |= ldlm_res_hop_hash(hs, key, CFS_HASH_NBKT(hs) - 1);
return hash & mask;
}
static void *ldlm_res_hop_key(struct hlist_node *hnode)
{
struct ldlm_resource *res;
res = hlist_entry(hnode, struct ldlm_resource, lr_hash);
return &res->lr_name;
}
static int ldlm_res_hop_keycmp(const void *key, struct hlist_node *hnode)
{
struct ldlm_resource *res;
res = hlist_entry(hnode, struct ldlm_resource, lr_hash);
return ldlm_res_eq((const struct ldlm_res_id *)key,
(const struct ldlm_res_id *)&res->lr_name);
}
static void *ldlm_res_hop_object(struct hlist_node *hnode)
{
return hlist_entry(hnode, struct ldlm_resource, lr_hash);
}
static void ldlm_res_hop_get_locked(struct cfs_hash *hs,
struct hlist_node *hnode)
{
struct ldlm_resource *res;
res = hlist_entry(hnode, struct ldlm_resource, lr_hash);
ldlm_resource_getref(res);
}
static void ldlm_res_hop_put_locked(struct cfs_hash *hs,
struct hlist_node *hnode)
{
struct ldlm_resource *res;
res = hlist_entry(hnode, struct ldlm_resource, lr_hash);
/* cfs_hash_for_each_nolock is the only chance we call it */
ldlm_resource_putref_locked(res);
}
static void ldlm_res_hop_put(struct cfs_hash *hs, struct hlist_node *hnode)
{
struct ldlm_resource *res;
res = hlist_entry(hnode, struct ldlm_resource, lr_hash);
ldlm_resource_putref(res);
}
cfs_hash_ops_t ldlm_ns_hash_ops = {
.hs_hash = ldlm_res_hop_hash,
.hs_key = ldlm_res_hop_key,
.hs_keycmp = ldlm_res_hop_keycmp,
.hs_keycpy = NULL,
.hs_object = ldlm_res_hop_object,
.hs_get = ldlm_res_hop_get_locked,
.hs_put_locked = ldlm_res_hop_put_locked,
.hs_put = ldlm_res_hop_put
};
cfs_hash_ops_t ldlm_ns_fid_hash_ops = {
.hs_hash = ldlm_res_hop_fid_hash,
.hs_key = ldlm_res_hop_key,
.hs_keycmp = ldlm_res_hop_keycmp,
.hs_keycpy = NULL,
.hs_object = ldlm_res_hop_object,
.hs_get = ldlm_res_hop_get_locked,
.hs_put_locked = ldlm_res_hop_put_locked,
.hs_put = ldlm_res_hop_put
};
struct ldlm_ns_hash_def {
ldlm_ns_type_t nsd_type;
/** hash bucket bits */
unsigned nsd_bkt_bits;
/** hash bits */
unsigned nsd_all_bits;
/** hash operations */
cfs_hash_ops_t *nsd_hops;
};
struct ldlm_ns_hash_def ldlm_ns_hash_defs[] = {
{
.nsd_type = LDLM_NS_TYPE_MDC,
.nsd_bkt_bits = 11,
.nsd_all_bits = 16,
.nsd_hops = &ldlm_ns_fid_hash_ops,
},
{
.nsd_type = LDLM_NS_TYPE_MDT,
.nsd_bkt_bits = 14,
.nsd_all_bits = 21,
.nsd_hops = &ldlm_ns_fid_hash_ops,
},
{
.nsd_type = LDLM_NS_TYPE_OSC,
.nsd_bkt_bits = 8,
.nsd_all_bits = 12,
.nsd_hops = &ldlm_ns_hash_ops,
},
{
.nsd_type = LDLM_NS_TYPE_OST,
.nsd_bkt_bits = 11,
.nsd_all_bits = 17,
.nsd_hops = &ldlm_ns_hash_ops,
},
{
.nsd_type = LDLM_NS_TYPE_MGC,
.nsd_bkt_bits = 4,
.nsd_all_bits = 4,
.nsd_hops = &ldlm_ns_hash_ops,
},
{
.nsd_type = LDLM_NS_TYPE_MGT,
.nsd_bkt_bits = 4,
.nsd_all_bits = 4,
.nsd_hops = &ldlm_ns_hash_ops,
},
{
.nsd_type = LDLM_NS_TYPE_UNKNOWN,
},
};
/**
* Create and initialize new empty namespace.
*/
struct ldlm_namespace *ldlm_namespace_new(struct obd_device *obd, char *name,
ldlm_side_t client,
ldlm_appetite_t apt,
ldlm_ns_type_t ns_type)
{
struct ldlm_namespace *ns = NULL;
struct ldlm_ns_bucket *nsb;
struct ldlm_ns_hash_def *nsd;
struct cfs_hash_bd bd;
int idx;
int rc;
LASSERT(obd != NULL);
rc = ldlm_get_ref();
if (rc) {
CERROR("ldlm_get_ref failed: %d\n", rc);
return NULL;
}
for (idx = 0;; idx++) {
nsd = &ldlm_ns_hash_defs[idx];
if (nsd->nsd_type == LDLM_NS_TYPE_UNKNOWN) {
CERROR("Unknown type %d for ns %s\n", ns_type, name);
goto out_ref;
}
if (nsd->nsd_type == ns_type)
break;
}
ns = kzalloc(sizeof(*ns), GFP_NOFS);
if (!ns)
goto out_ref;
ns->ns_rs_hash = cfs_hash_create(name,
nsd->nsd_all_bits, nsd->nsd_all_bits,
nsd->nsd_bkt_bits, sizeof(*nsb),
CFS_HASH_MIN_THETA,
CFS_HASH_MAX_THETA,
nsd->nsd_hops,
CFS_HASH_DEPTH |
CFS_HASH_BIGNAME |
CFS_HASH_SPIN_BKTLOCK |
CFS_HASH_NO_ITEMREF);
if (ns->ns_rs_hash == NULL)
goto out_ns;
cfs_hash_for_each_bucket(ns->ns_rs_hash, &bd, idx) {
nsb = cfs_hash_bd_extra_get(ns->ns_rs_hash, &bd);
at_init(&nsb->nsb_at_estimate, ldlm_enqueue_min, 0);
nsb->nsb_namespace = ns;
}
ns->ns_obd = obd;
ns->ns_appetite = apt;
ns->ns_client = client;
INIT_LIST_HEAD(&ns->ns_list_chain);
INIT_LIST_HEAD(&ns->ns_unused_list);
spin_lock_init(&ns->ns_lock);
atomic_set(&ns->ns_bref, 0);
init_waitqueue_head(&ns->ns_waitq);
ns->ns_max_parallel_ast = LDLM_DEFAULT_PARALLEL_AST_LIMIT;
ns->ns_nr_unused = 0;
ns->ns_max_unused = LDLM_DEFAULT_LRU_SIZE;
ns->ns_max_age = LDLM_DEFAULT_MAX_ALIVE;
ns->ns_orig_connect_flags = 0;
ns->ns_connect_flags = 0;
ns->ns_stopping = 0;
rc = ldlm_namespace_sysfs_register(ns);
if (rc != 0) {
CERROR("Can't initialize ns sysfs, rc %d\n", rc);
goto out_hash;
}
rc = ldlm_namespace_debugfs_register(ns);
if (rc != 0) {
CERROR("Can't initialize ns proc, rc %d\n", rc);
goto out_sysfs;
}
idx = ldlm_namespace_nr_read(client);
rc = ldlm_pool_init(&ns->ns_pool, ns, idx, client);
if (rc) {
CERROR("Can't initialize lock pool, rc %d\n", rc);
goto out_proc;
}
ldlm_namespace_register(ns, client);
return ns;
out_proc:
ldlm_namespace_debugfs_unregister(ns);
out_sysfs:
ldlm_namespace_sysfs_unregister(ns);
ldlm_namespace_cleanup(ns, 0);
out_hash:
cfs_hash_putref(ns->ns_rs_hash);
out_ns:
kfree(ns);
out_ref:
ldlm_put_ref();
return NULL;
}
EXPORT_SYMBOL(ldlm_namespace_new);
extern struct ldlm_lock *ldlm_lock_get(struct ldlm_lock *lock);
/**
* Cancel and destroy all locks on a resource.
*
* If flags contains FL_LOCAL_ONLY, don't try to tell the server, just
* clean up. This is currently only used for recovery, and we make
* certain assumptions as a result--notably, that we shouldn't cancel
* locks with refs.
*/
static void cleanup_resource(struct ldlm_resource *res, struct list_head *q,
__u64 flags)
{
struct list_head *tmp;
int rc = 0, client = ns_is_client(ldlm_res_to_ns(res));
bool local_only = !!(flags & LDLM_FL_LOCAL_ONLY);
do {
struct ldlm_lock *lock = NULL;
/* First, we look for non-cleaned-yet lock
* all cleaned locks are marked by CLEANED flag. */
lock_res(res);
list_for_each(tmp, q) {
lock = list_entry(tmp, struct ldlm_lock,
l_res_link);
if (lock->l_flags & LDLM_FL_CLEANED) {
lock = NULL;
continue;
}
LDLM_LOCK_GET(lock);
lock->l_flags |= LDLM_FL_CLEANED;
break;
}
if (lock == NULL) {
unlock_res(res);
break;
}
/* Set CBPENDING so nothing in the cancellation path
* can match this lock. */
lock->l_flags |= LDLM_FL_CBPENDING;
lock->l_flags |= LDLM_FL_FAILED;
lock->l_flags |= flags;
/* ... without sending a CANCEL message for local_only. */
if (local_only)
lock->l_flags |= LDLM_FL_LOCAL_ONLY;
if (local_only && (lock->l_readers || lock->l_writers)) {
/* This is a little bit gross, but much better than the
* alternative: pretend that we got a blocking AST from
* the server, so that when the lock is decref'd, it
* will go away ... */
unlock_res(res);
LDLM_DEBUG(lock, "setting FL_LOCAL_ONLY");
if (lock->l_completion_ast)
lock->l_completion_ast(lock, 0, NULL);
LDLM_LOCK_RELEASE(lock);
continue;
}
if (client) {
struct lustre_handle lockh;
unlock_res(res);
ldlm_lock2handle(lock, &lockh);
rc = ldlm_cli_cancel(&lockh, LCF_ASYNC);
if (rc)
CERROR("ldlm_cli_cancel: %d\n", rc);
} else {
ldlm_resource_unlink_lock(lock);
unlock_res(res);
LDLM_DEBUG(lock, "Freeing a lock still held by a client node");
ldlm_lock_destroy(lock);
}
LDLM_LOCK_RELEASE(lock);
} while (1);
}
static int ldlm_resource_clean(struct cfs_hash *hs, struct cfs_hash_bd *bd,
struct hlist_node *hnode, void *arg)
{
struct ldlm_resource *res = cfs_hash_object(hs, hnode);
__u64 flags = *(__u64 *)arg;
cleanup_resource(res, &res->lr_granted, flags);
cleanup_resource(res, &res->lr_converting, flags);
cleanup_resource(res, &res->lr_waiting, flags);
return 0;
}
static int ldlm_resource_complain(struct cfs_hash *hs, struct cfs_hash_bd *bd,
struct hlist_node *hnode, void *arg)
{
struct ldlm_resource *res = cfs_hash_object(hs, hnode);
lock_res(res);
CERROR("%s: namespace resource "DLDLMRES
" (%p) refcount nonzero (%d) after lock cleanup; forcing cleanup.\n",
ldlm_ns_name(ldlm_res_to_ns(res)), PLDLMRES(res), res,
atomic_read(&res->lr_refcount) - 1);
ldlm_resource_dump(D_ERROR, res);
unlock_res(res);
return 0;
}
/**
* Cancel and destroy all locks in the namespace.
*
* Typically used during evictions when server notified client that it was
* evicted and all of its state needs to be destroyed.
* Also used during shutdown.
*/
int ldlm_namespace_cleanup(struct ldlm_namespace *ns, __u64 flags)
{
if (ns == NULL) {
CDEBUG(D_INFO, "NULL ns, skipping cleanup\n");
return ELDLM_OK;
}
cfs_hash_for_each_nolock(ns->ns_rs_hash, ldlm_resource_clean, &flags);
cfs_hash_for_each_nolock(ns->ns_rs_hash, ldlm_resource_complain, NULL);
return ELDLM_OK;
}
EXPORT_SYMBOL(ldlm_namespace_cleanup);
/**
* Attempts to free namespace.
*
* Only used when namespace goes away, like during an unmount.
*/
static int __ldlm_namespace_free(struct ldlm_namespace *ns, int force)
{
/* At shutdown time, don't call the cancellation callback */
ldlm_namespace_cleanup(ns, force ? LDLM_FL_LOCAL_ONLY : 0);
if (atomic_read(&ns->ns_bref) > 0) {
struct l_wait_info lwi = LWI_INTR(LWI_ON_SIGNAL_NOOP, NULL);
int rc;
CDEBUG(D_DLMTRACE,
"dlm namespace %s free waiting on refcount %d\n",
ldlm_ns_name(ns), atomic_read(&ns->ns_bref));
force_wait:
if (force)
lwi = LWI_TIMEOUT(obd_timeout * HZ / 4, NULL, NULL);
rc = l_wait_event(ns->ns_waitq,
atomic_read(&ns->ns_bref) == 0, &lwi);
/* Forced cleanups should be able to reclaim all references,
* so it's safe to wait forever... we can't leak locks... */
if (force && rc == -ETIMEDOUT) {
LCONSOLE_ERROR("Forced cleanup waiting for %s namespace with %d resources in use, (rc=%d)\n",
ldlm_ns_name(ns),
atomic_read(&ns->ns_bref), rc);
goto force_wait;
}
if (atomic_read(&ns->ns_bref)) {
LCONSOLE_ERROR("Cleanup waiting for %s namespace with %d resources in use, (rc=%d)\n",
ldlm_ns_name(ns),
atomic_read(&ns->ns_bref), rc);
return ELDLM_NAMESPACE_EXISTS;
}
CDEBUG(D_DLMTRACE, "dlm namespace %s free done waiting\n",
ldlm_ns_name(ns));
}
return ELDLM_OK;
}
/**
* Performs various cleanups for passed \a ns to make it drop refc and be
* ready for freeing. Waits for refc == 0.
*
* The following is done:
* (0) Unregister \a ns from its list to make inaccessible for potential
* users like pools thread and others;
* (1) Clear all locks in \a ns.
*/
void ldlm_namespace_free_prior(struct ldlm_namespace *ns,
struct obd_import *imp,
int force)
{
int rc;
if (!ns)
return;
spin_lock(&ns->ns_lock);
ns->ns_stopping = 1;
spin_unlock(&ns->ns_lock);
/*
* Can fail with -EINTR when force == 0 in which case try harder.
*/
rc = __ldlm_namespace_free(ns, force);
if (rc != ELDLM_OK) {
if (imp) {
ptlrpc_disconnect_import(imp, 0);
ptlrpc_invalidate_import(imp);
}
/*
* With all requests dropped and the import inactive
* we are guaranteed all reference will be dropped.
*/
rc = __ldlm_namespace_free(ns, 1);
LASSERT(rc == 0);
}
}
/**
* Performs freeing memory structures related to \a ns. This is only done
* when ldlm_namespce_free_prior() successfully removed all resources
* referencing \a ns and its refc == 0.
*/
void ldlm_namespace_free_post(struct ldlm_namespace *ns)
{
if (!ns)
return;
/* Make sure that nobody can find this ns in its list. */
ldlm_namespace_unregister(ns, ns->ns_client);
/* Fini pool _before_ parent proc dir is removed. This is important as
* ldlm_pool_fini() removes own proc dir which is child to @dir.
* Removing it after @dir may cause oops. */
ldlm_pool_fini(&ns->ns_pool);
ldlm_namespace_debugfs_unregister(ns);
cfs_hash_putref(ns->ns_rs_hash);
/* Namespace \a ns should be not on list at this time, otherwise
* this will cause issues related to using freed \a ns in poold
* thread. */
LASSERT(list_empty(&ns->ns_list_chain));
kfree(ns);
ldlm_put_ref();
}
/**
* Cleanup the resource, and free namespace.
* bug 12864:
* Deadlock issue:
* proc1: destroy import
* class_disconnect_export(grab cl_sem) ->
* -> ldlm_namespace_free ->
* -> lprocfs_remove(grab _lprocfs_lock).
* proc2: read proc info
* lprocfs_fops_read(grab _lprocfs_lock) ->
* -> osc_rd_active, etc(grab cl_sem).
*
* So that I have to split the ldlm_namespace_free into two parts - the first
* part ldlm_namespace_free_prior is used to cleanup the resource which is
* being used; the 2nd part ldlm_namespace_free_post is used to unregister the
* lprocfs entries, and then free memory. It will be called w/o cli->cl_sem
* held.
*/
void ldlm_namespace_free(struct ldlm_namespace *ns,
struct obd_import *imp,
int force)
{
ldlm_namespace_free_prior(ns, imp, force);
ldlm_namespace_free_post(ns);
}
EXPORT_SYMBOL(ldlm_namespace_free);
void ldlm_namespace_get(struct ldlm_namespace *ns)
{
atomic_inc(&ns->ns_bref);
}
EXPORT_SYMBOL(ldlm_namespace_get);
/* This is only for callers that care about refcount */
int ldlm_namespace_get_return(struct ldlm_namespace *ns)
{
return atomic_inc_return(&ns->ns_bref);
}
void ldlm_namespace_put(struct ldlm_namespace *ns)
{
if (atomic_dec_and_lock(&ns->ns_bref, &ns->ns_lock)) {
wake_up(&ns->ns_waitq);
spin_unlock(&ns->ns_lock);
}
}
EXPORT_SYMBOL(ldlm_namespace_put);
/** Register \a ns in the list of namespaces */
void ldlm_namespace_register(struct ldlm_namespace *ns, ldlm_side_t client)
{
mutex_lock(ldlm_namespace_lock(client));
LASSERT(list_empty(&ns->ns_list_chain));
list_add(&ns->ns_list_chain, ldlm_namespace_inactive_list(client));
ldlm_namespace_nr_inc(client);
mutex_unlock(ldlm_namespace_lock(client));
}
/** Unregister \a ns from the list of namespaces. */
void ldlm_namespace_unregister(struct ldlm_namespace *ns, ldlm_side_t client)
{
mutex_lock(ldlm_namespace_lock(client));
LASSERT(!list_empty(&ns->ns_list_chain));
/* Some asserts and possibly other parts of the code are still
* using list_empty(&ns->ns_list_chain). This is why it is
* important to use list_del_init() here. */
list_del_init(&ns->ns_list_chain);
ldlm_namespace_nr_dec(client);
mutex_unlock(ldlm_namespace_lock(client));
}
/** Should be called with ldlm_namespace_lock(client) taken. */
void ldlm_namespace_move_to_active_locked(struct ldlm_namespace *ns,
ldlm_side_t client)
{
LASSERT(!list_empty(&ns->ns_list_chain));
LASSERT(mutex_is_locked(ldlm_namespace_lock(client)));
list_move_tail(&ns->ns_list_chain, ldlm_namespace_list(client));
}
/** Should be called with ldlm_namespace_lock(client) taken. */
void ldlm_namespace_move_to_inactive_locked(struct ldlm_namespace *ns,
ldlm_side_t client)
{
LASSERT(!list_empty(&ns->ns_list_chain));
LASSERT(mutex_is_locked(ldlm_namespace_lock(client)));
list_move_tail(&ns->ns_list_chain,
ldlm_namespace_inactive_list(client));
}
/** Should be called with ldlm_namespace_lock(client) taken. */
struct ldlm_namespace *ldlm_namespace_first_locked(ldlm_side_t client)
{
LASSERT(mutex_is_locked(ldlm_namespace_lock(client)));
LASSERT(!list_empty(ldlm_namespace_list(client)));
return container_of(ldlm_namespace_list(client)->next,
struct ldlm_namespace, ns_list_chain);
}
/** Create and initialize new resource. */
static struct ldlm_resource *ldlm_resource_new(void)
{
struct ldlm_resource *res;
int idx;
OBD_SLAB_ALLOC_PTR_GFP(res, ldlm_resource_slab, GFP_NOFS);
if (res == NULL)
return NULL;
INIT_LIST_HEAD(&res->lr_granted);
INIT_LIST_HEAD(&res->lr_converting);
INIT_LIST_HEAD(&res->lr_waiting);
/* Initialize interval trees for each lock mode. */
for (idx = 0; idx < LCK_MODE_NUM; idx++) {
res->lr_itree[idx].lit_size = 0;
res->lr_itree[idx].lit_mode = 1 << idx;
res->lr_itree[idx].lit_root = NULL;
}
atomic_set(&res->lr_refcount, 1);
spin_lock_init(&res->lr_lock);
lu_ref_init(&res->lr_reference);
/* The creator of the resource must unlock the mutex after LVB
* initialization. */
mutex_init(&res->lr_lvb_mutex);
mutex_lock(&res->lr_lvb_mutex);
return res;
}
/**
* Return a reference to resource with given name, creating it if necessary.
* Args: namespace with ns_lock unlocked
* Locks: takes and releases NS hash-lock and res->lr_lock
* Returns: referenced, unlocked ldlm_resource or NULL
*/
struct ldlm_resource *
ldlm_resource_get(struct ldlm_namespace *ns, struct ldlm_resource *parent,
const struct ldlm_res_id *name, ldlm_type_t type, int create)
{
struct hlist_node *hnode;
struct ldlm_resource *res;
struct cfs_hash_bd bd;
__u64 version;
int ns_refcount = 0;
LASSERT(ns != NULL);
LASSERT(parent == NULL);
LASSERT(ns->ns_rs_hash != NULL);
LASSERT(name->name[0] != 0);
cfs_hash_bd_get_and_lock(ns->ns_rs_hash, (void *)name, &bd, 0);
hnode = cfs_hash_bd_lookup_locked(ns->ns_rs_hash, &bd, (void *)name);
if (hnode != NULL) {
cfs_hash_bd_unlock(ns->ns_rs_hash, &bd, 0);
res = hlist_entry(hnode, struct ldlm_resource, lr_hash);
/* Synchronize with regard to resource creation. */
if (ns->ns_lvbo && ns->ns_lvbo->lvbo_init) {
mutex_lock(&res->lr_lvb_mutex);
mutex_unlock(&res->lr_lvb_mutex);
}
if (unlikely(res->lr_lvb_len < 0)) {
ldlm_resource_putref(res);
res = NULL;
}
return res;
}
version = cfs_hash_bd_version_get(&bd);
cfs_hash_bd_unlock(ns->ns_rs_hash, &bd, 0);
if (create == 0)
return NULL;
LASSERTF(type >= LDLM_MIN_TYPE && type < LDLM_MAX_TYPE,
"type: %d\n", type);
res = ldlm_resource_new();
if (!res)
return NULL;
res->lr_ns_bucket = cfs_hash_bd_extra_get(ns->ns_rs_hash, &bd);
res->lr_name = *name;
res->lr_type = type;
res->lr_most_restr = LCK_NL;
cfs_hash_bd_lock(ns->ns_rs_hash, &bd, 1);
hnode = (version == cfs_hash_bd_version_get(&bd)) ? NULL :
cfs_hash_bd_lookup_locked(ns->ns_rs_hash, &bd, (void *)name);
if (hnode != NULL) {
/* Someone won the race and already added the resource. */
cfs_hash_bd_unlock(ns->ns_rs_hash, &bd, 1);
/* Clean lu_ref for failed resource. */
lu_ref_fini(&res->lr_reference);
/* We have taken lr_lvb_mutex. Drop it. */
mutex_unlock(&res->lr_lvb_mutex);
OBD_SLAB_FREE(res, ldlm_resource_slab, sizeof(*res));
res = hlist_entry(hnode, struct ldlm_resource, lr_hash);
/* Synchronize with regard to resource creation. */
if (ns->ns_lvbo && ns->ns_lvbo->lvbo_init) {
mutex_lock(&res->lr_lvb_mutex);
mutex_unlock(&res->lr_lvb_mutex);
}
if (unlikely(res->lr_lvb_len < 0)) {
ldlm_resource_putref(res);
res = NULL;
}
return res;
}
/* We won! Let's add the resource. */
cfs_hash_bd_add_locked(ns->ns_rs_hash, &bd, &res->lr_hash);
if (cfs_hash_bd_count_get(&bd) == 1)
ns_refcount = ldlm_namespace_get_return(ns);
cfs_hash_bd_unlock(ns->ns_rs_hash, &bd, 1);
if (ns->ns_lvbo && ns->ns_lvbo->lvbo_init) {
int rc;
OBD_FAIL_TIMEOUT(OBD_FAIL_LDLM_CREATE_RESOURCE, 2);
rc = ns->ns_lvbo->lvbo_init(res);
if (rc < 0) {
CERROR("%s: lvbo_init failed for resource %#llx:%#llx: rc = %d\n",
ns->ns_obd->obd_name, name->name[0],
name->name[1], rc);
kfree(res->lr_lvb_data);
res->lr_lvb_data = NULL;
res->lr_lvb_len = rc;
mutex_unlock(&res->lr_lvb_mutex);
ldlm_resource_putref(res);
return NULL;
}
}
/* We create resource with locked lr_lvb_mutex. */
mutex_unlock(&res->lr_lvb_mutex);
/* Let's see if we happened to be the very first resource in this
* namespace. If so, and this is a client namespace, we need to move
* the namespace into the active namespaces list to be patrolled by
* the ldlm_poold. */
if (ns_is_client(ns) && ns_refcount == 1) {
mutex_lock(ldlm_namespace_lock(LDLM_NAMESPACE_CLIENT));
ldlm_namespace_move_to_active_locked(ns, LDLM_NAMESPACE_CLIENT);
mutex_unlock(ldlm_namespace_lock(LDLM_NAMESPACE_CLIENT));
}
return res;
}
EXPORT_SYMBOL(ldlm_resource_get);
struct ldlm_resource *ldlm_resource_getref(struct ldlm_resource *res)
{
LASSERT(res != NULL);
LASSERT(res != LP_POISON);
atomic_inc(&res->lr_refcount);
CDEBUG(D_INFO, "getref res: %p count: %d\n", res,
atomic_read(&res->lr_refcount));
return res;
}
static void __ldlm_resource_putref_final(struct cfs_hash_bd *bd,
struct ldlm_resource *res)
{
struct ldlm_ns_bucket *nsb = res->lr_ns_bucket;
if (!list_empty(&res->lr_granted)) {
ldlm_resource_dump(D_ERROR, res);
LBUG();
}
if (!list_empty(&res->lr_converting)) {
ldlm_resource_dump(D_ERROR, res);
LBUG();
}
if (!list_empty(&res->lr_waiting)) {
ldlm_resource_dump(D_ERROR, res);
LBUG();
}
cfs_hash_bd_del_locked(nsb->nsb_namespace->ns_rs_hash,
bd, &res->lr_hash);
lu_ref_fini(&res->lr_reference);
if (cfs_hash_bd_count_get(bd) == 0)
ldlm_namespace_put(nsb->nsb_namespace);
}
/* Returns 1 if the resource was freed, 0 if it remains. */
int ldlm_resource_putref(struct ldlm_resource *res)
{
struct ldlm_namespace *ns = ldlm_res_to_ns(res);
struct cfs_hash_bd bd;
LASSERT_ATOMIC_GT_LT(&res->lr_refcount, 0, LI_POISON);
CDEBUG(D_INFO, "putref res: %p count: %d\n",
res, atomic_read(&res->lr_refcount) - 1);
cfs_hash_bd_get(ns->ns_rs_hash, &res->lr_name, &bd);
if (cfs_hash_bd_dec_and_lock(ns->ns_rs_hash, &bd, &res->lr_refcount)) {
__ldlm_resource_putref_final(&bd, res);
cfs_hash_bd_unlock(ns->ns_rs_hash, &bd, 1);
if (ns->ns_lvbo && ns->ns_lvbo->lvbo_free)
ns->ns_lvbo->lvbo_free(res);
OBD_SLAB_FREE(res, ldlm_resource_slab, sizeof(*res));
return 1;
}
return 0;
}
EXPORT_SYMBOL(ldlm_resource_putref);
/* Returns 1 if the resource was freed, 0 if it remains. */
int ldlm_resource_putref_locked(struct ldlm_resource *res)
{
struct ldlm_namespace *ns = ldlm_res_to_ns(res);
LASSERT_ATOMIC_GT_LT(&res->lr_refcount, 0, LI_POISON);
CDEBUG(D_INFO, "putref res: %p count: %d\n",
res, atomic_read(&res->lr_refcount) - 1);
if (atomic_dec_and_test(&res->lr_refcount)) {
struct cfs_hash_bd bd;
cfs_hash_bd_get(ldlm_res_to_ns(res)->ns_rs_hash,
&res->lr_name, &bd);
__ldlm_resource_putref_final(&bd, res);
cfs_hash_bd_unlock(ns->ns_rs_hash, &bd, 1);
/* NB: ns_rs_hash is created with CFS_HASH_NO_ITEMREF,
* so we should never be here while calling cfs_hash_del,
* cfs_hash_for_each_nolock is the only case we can get
* here, which is safe to release cfs_hash_bd_lock.
*/
if (ns->ns_lvbo && ns->ns_lvbo->lvbo_free)
ns->ns_lvbo->lvbo_free(res);
OBD_SLAB_FREE(res, ldlm_resource_slab, sizeof(*res));
cfs_hash_bd_lock(ns->ns_rs_hash, &bd, 1);
return 1;
}
return 0;
}
/**
* Add a lock into a given resource into specified lock list.
*/
void ldlm_resource_add_lock(struct ldlm_resource *res, struct list_head *head,
struct ldlm_lock *lock)
{
check_res_locked(res);
LDLM_DEBUG(lock, "About to add this lock:\n");
if (lock->l_flags & LDLM_FL_DESTROYED) {
CDEBUG(D_OTHER, "Lock destroyed, not adding to resource\n");
return;
}
LASSERT(list_empty(&lock->l_res_link));
list_add_tail(&lock->l_res_link, head);
}
/**
* Insert a lock into resource after specified lock.
*
* Obtain resource description from the lock we are inserting after.
*/
void ldlm_resource_insert_lock_after(struct ldlm_lock *original,
struct ldlm_lock *new)
{
struct ldlm_resource *res = original->l_resource;
check_res_locked(res);
ldlm_resource_dump(D_INFO, res);
LDLM_DEBUG(new, "About to insert this lock after %p:\n", original);
if (new->l_flags & LDLM_FL_DESTROYED) {
CDEBUG(D_OTHER, "Lock destroyed, not adding to resource\n");
goto out;
}
LASSERT(list_empty(&new->l_res_link));
list_add(&new->l_res_link, &original->l_res_link);
out:;
}
void ldlm_resource_unlink_lock(struct ldlm_lock *lock)
{
int type = lock->l_resource->lr_type;
check_res_locked(lock->l_resource);
if (type == LDLM_IBITS || type == LDLM_PLAIN)
ldlm_unlink_lock_skiplist(lock);
else if (type == LDLM_EXTENT)
ldlm_extent_unlink_lock(lock);
list_del_init(&lock->l_res_link);
}
EXPORT_SYMBOL(ldlm_resource_unlink_lock);
void ldlm_res2desc(struct ldlm_resource *res, struct ldlm_resource_desc *desc)
{
desc->lr_type = res->lr_type;
desc->lr_name = res->lr_name;
}
/**
* Print information about all locks in all namespaces on this node to debug
* log.
*/
void ldlm_dump_all_namespaces(ldlm_side_t client, int level)
{
struct list_head *tmp;
if (!((libcfs_debug | D_ERROR) & level))
return;
mutex_lock(ldlm_namespace_lock(client));
list_for_each(tmp, ldlm_namespace_list(client)) {
struct ldlm_namespace *ns;
ns = list_entry(tmp, struct ldlm_namespace, ns_list_chain);
ldlm_namespace_dump(level, ns);
}
mutex_unlock(ldlm_namespace_lock(client));
}
EXPORT_SYMBOL(ldlm_dump_all_namespaces);
static int ldlm_res_hash_dump(struct cfs_hash *hs, struct cfs_hash_bd *bd,
struct hlist_node *hnode, void *arg)
{
struct ldlm_resource *res = cfs_hash_object(hs, hnode);
int level = (int)(unsigned long)arg;
lock_res(res);
ldlm_resource_dump(level, res);
unlock_res(res);
return 0;
}
/**
* Print information about all locks in this namespace on this node to debug
* log.
*/
void ldlm_namespace_dump(int level, struct ldlm_namespace *ns)
{
if (!((libcfs_debug | D_ERROR) & level))
return;
CDEBUG(level, "--- Namespace: %s (rc: %d, side: %s)\n",
ldlm_ns_name(ns), atomic_read(&ns->ns_bref),
ns_is_client(ns) ? "client" : "server");
if (time_before(cfs_time_current(), ns->ns_next_dump))
return;
cfs_hash_for_each_nolock(ns->ns_rs_hash,
ldlm_res_hash_dump,
(void *)(unsigned long)level);
spin_lock(&ns->ns_lock);
ns->ns_next_dump = cfs_time_shift(10);
spin_unlock(&ns->ns_lock);
}
EXPORT_SYMBOL(ldlm_namespace_dump);
/**
* Print information about all locks in this resource to debug log.
*/
void ldlm_resource_dump(int level, struct ldlm_resource *res)
{
struct ldlm_lock *lock;
unsigned int granted = 0;
CLASSERT(RES_NAME_SIZE == 4);
if (!((libcfs_debug | D_ERROR) & level))
return;
CDEBUG(level, "--- Resource: "DLDLMRES" (%p) refcount = %d\n",
PLDLMRES(res), res, atomic_read(&res->lr_refcount));
if (!list_empty(&res->lr_granted)) {
CDEBUG(level, "Granted locks (in reverse order):\n");
list_for_each_entry_reverse(lock, &res->lr_granted,
l_res_link) {
LDLM_DEBUG_LIMIT(level, lock, "###");
if (!(level & D_CANTMASK) &&
++granted > ldlm_dump_granted_max) {
CDEBUG(level, "only dump %d granted locks to avoid DDOS.\n",
granted);
break;
}
}
}
if (!list_empty(&res->lr_converting)) {
CDEBUG(level, "Converting locks:\n");
list_for_each_entry(lock, &res->lr_converting, l_res_link)
LDLM_DEBUG_LIMIT(level, lock, "###");
}
if (!list_empty(&res->lr_waiting)) {
CDEBUG(level, "Waiting locks:\n");
list_for_each_entry(lock, &res->lr_waiting, l_res_link)
LDLM_DEBUG_LIMIT(level, lock, "###");
}
}
|