4 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
6 * This program is free software; you can redistribute it and/or modify
7 * it under the terms of the GNU General Public License version 2 only,
8 * as published by the Free Software Foundation.
10 * This program is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 * GNU General Public License version 2 for more details. A copy is
14 * included in the COPYING file that accompanied this code.
16 * You should have received a copy of the GNU General Public License
17 * along with this program; if not, write to the Free Software
18 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
23 * Copyright (c) 2013, Intel Corporation.
26 * lustre/lfsck/lfsck_layout.c
28 * Author: Fan, Yong <fan.yong@intel.com>
32 # define EXPORT_SYMTAB
34 #define DEBUG_SUBSYSTEM S_LFSCK
36 #include <linux/bitops.h>
37 #include <linux/rbtree.h>
39 #include <lustre/lustre_idl.h>
40 #include <lu_object.h>
41 #include <dt_object.h>
42 #include <lustre_linkea.h>
43 #include <lustre_fid.h>
44 #include <lustre_lib.h>
45 #include <lustre_net.h>
46 #include <lustre/lustre_user.h>
47 #include <md_object.h>
48 #include <obd_class.h>
50 #include "lfsck_internal.h"
52 #define LFSCK_LAYOUT_MAGIC 0xB173AE14
54 static const char lfsck_layout_name[] = "lfsck_layout";
56 struct lfsck_layout_seq {
57 struct list_head lls_list;
60 __u64 lls_lastid_known;
61 struct dt_object *lls_lastid_obj;
62 unsigned int lls_dirty:1;
65 struct lfsck_layout_slave_target {
66 /* link into lfsck_layout_slave_data::llsd_master_list. */
67 struct list_head llst_list;
68 /* The position for next record in the rbtree for iteration. */
69 struct lu_fid llst_fid;
70 /* Dummy hash for iteration against the rbtree. */
77 struct lfsck_layout_slave_data {
78 /* list for lfsck_layout_seq */
79 struct list_head llsd_seq_list;
81 /* list for the masters involve layout verification. */
82 struct list_head llsd_master_list;
85 struct dt_object *llsd_rb_obj;
86 struct rb_root llsd_rb_root;
87 rwlock_t llsd_rb_lock;
88 unsigned int llsd_rbtree_valid:1;
91 struct lfsck_layout_object {
92 struct dt_object *llo_obj;
93 struct lu_attr llo_attr;
98 struct lfsck_layout_req {
99 struct list_head llr_list;
100 struct lfsck_layout_object *llr_parent;
101 struct dt_object *llr_child;
103 __u32 llr_lov_idx; /* offset in LOV EA */
106 struct lfsck_layout_master_data {
107 spinlock_t llmd_lock;
108 struct list_head llmd_req_list;
110 /* list for the ost targets involve layout verification. */
111 struct list_head llmd_ost_list;
113 /* list for the ost targets in phase1 scanning. */
114 struct list_head llmd_ost_phase1_list;
116 /* list for the ost targets in phase1 scanning. */
117 struct list_head llmd_ost_phase2_list;
119 /* list for the mdt targets involve layout verification. */
120 struct list_head llmd_mdt_list;
122 /* list for the mdt targets in phase1 scanning. */
123 struct list_head llmd_mdt_phase1_list;
125 /* list for the mdt targets in phase1 scanning. */
126 struct list_head llmd_mdt_phase2_list;
128 struct ptlrpc_thread llmd_thread;
129 __u32 llmd_touch_gen;
131 int llmd_assistant_status;
132 int llmd_post_result;
133 unsigned int llmd_to_post:1,
134 llmd_to_double_scan:1,
135 llmd_in_double_scan:1,
139 struct lfsck_layout_slave_async_args {
140 struct obd_export *llsaa_exp;
141 struct lfsck_component *llsaa_com;
142 struct lfsck_layout_slave_target *llsaa_llst;
145 static struct lfsck_layout_object *
146 lfsck_layout_object_init(const struct lu_env *env, struct dt_object *obj,
149 struct lfsck_layout_object *llo;
154 return ERR_PTR(-ENOMEM);
156 rc = dt_attr_get(env, obj, &llo->llo_attr, BYPASS_CAPA);
163 lu_object_get(&obj->do_lu);
165 /* The gen can be used to check whether some others have changed the
166 * file layout after LFSCK pre-fetching but before real verification. */
168 atomic_set(&llo->llo_ref, 1);
174 lfsck_layout_llst_put(struct lfsck_layout_slave_target *llst)
176 if (atomic_dec_and_test(&llst->llst_ref)) {
177 LASSERT(list_empty(&llst->llst_list));
184 lfsck_layout_llst_add(struct lfsck_layout_slave_data *llsd, __u32 index)
186 struct lfsck_layout_slave_target *llst;
187 struct lfsck_layout_slave_target *tmp;
194 INIT_LIST_HEAD(&llst->llst_list);
196 llst->llst_index = index;
197 atomic_set(&llst->llst_ref, 1);
199 spin_lock(&llsd->llsd_lock);
200 list_for_each_entry(tmp, &llsd->llsd_master_list, llst_list) {
201 if (tmp->llst_index == index) {
207 list_add_tail(&llst->llst_list, &llsd->llsd_master_list);
208 spin_unlock(&llsd->llsd_lock);
217 lfsck_layout_llst_del(struct lfsck_layout_slave_data *llsd,
218 struct lfsck_layout_slave_target *llst)
222 spin_lock(&llsd->llsd_lock);
223 if (!list_empty(&llst->llst_list)) {
224 list_del_init(&llst->llst_list);
227 spin_unlock(&llsd->llsd_lock);
230 lfsck_layout_llst_put(llst);
233 static inline struct lfsck_layout_slave_target *
234 lfsck_layout_llst_find_and_del(struct lfsck_layout_slave_data *llsd,
235 __u32 index, bool unlink)
237 struct lfsck_layout_slave_target *llst;
239 spin_lock(&llsd->llsd_lock);
240 list_for_each_entry(llst, &llsd->llsd_master_list, llst_list) {
241 if (llst->llst_index == index) {
243 list_del_init(&llst->llst_list);
245 atomic_inc(&llst->llst_ref);
246 spin_unlock(&llsd->llsd_lock);
251 spin_unlock(&llsd->llsd_lock);
256 static inline void lfsck_layout_object_put(const struct lu_env *env,
257 struct lfsck_layout_object *llo)
259 if (atomic_dec_and_test(&llo->llo_ref)) {
260 lfsck_object_put(env, llo->llo_obj);
265 static struct lfsck_layout_req *
266 lfsck_layout_req_init(struct lfsck_layout_object *parent,
267 struct dt_object *child, __u32 ost_idx, __u32 lov_idx)
269 struct lfsck_layout_req *llr;
273 return ERR_PTR(-ENOMEM);
275 INIT_LIST_HEAD(&llr->llr_list);
276 atomic_inc(&parent->llo_ref);
277 llr->llr_parent = parent;
278 llr->llr_child = child;
279 llr->llr_ost_idx = ost_idx;
280 llr->llr_lov_idx = lov_idx;
285 static inline void lfsck_layout_req_fini(const struct lu_env *env,
286 struct lfsck_layout_req *llr)
288 lu_object_put(env, &llr->llr_child->do_lu);
289 lfsck_layout_object_put(env, llr->llr_parent);
293 static inline bool lfsck_layout_req_empty(struct lfsck_layout_master_data *llmd)
297 spin_lock(&llmd->llmd_lock);
298 if (list_empty(&llmd->llmd_req_list))
300 spin_unlock(&llmd->llmd_lock);
305 static int lfsck_layout_get_lovea(const struct lu_env *env,
306 struct dt_object *obj,
307 struct lu_buf *buf, ssize_t *buflen)
312 rc = dt_xattr_get(env, obj, buf, XATTR_NAME_LOV, BYPASS_CAPA);
314 rc = dt_xattr_get(env, obj, &LU_BUF_NULL, XATTR_NAME_LOV,
319 lu_buf_realloc(buf, rc);
321 *buflen = buf->lb_len;
323 if (buf->lb_buf == NULL)
335 if (unlikely(buf->lb_buf == NULL)) {
336 lu_buf_alloc(buf, rc);
338 *buflen = buf->lb_len;
340 if (buf->lb_buf == NULL)
349 static int lfsck_layout_verify_header(struct lov_mds_md_v1 *lmm)
354 magic = le32_to_cpu(lmm->lmm_magic);
355 /* If magic crashed, keep it there. Sometime later, during OST-object
356 * orphan handling, if some OST-object(s) back-point to it, it can be
357 * verified and repaired. */
358 if (magic != LOV_MAGIC_V1 && magic != LOV_MAGIC_V3)
361 patten = le32_to_cpu(lmm->lmm_pattern);
362 /* XXX: currently, we only support LOV_PATTERN_RAID0. */
363 if (patten != LOV_PATTERN_RAID0)
369 #define LFSCK_RBTREE_BITMAP_SIZE PAGE_CACHE_SIZE
370 #define LFSCK_RBTREE_BITMAP_WIDTH (LFSCK_RBTREE_BITMAP_SIZE << 3)
371 #define LFSCK_RBTREE_BITMAP_MASK (LFSCK_RBTREE_BITMAP_SIZE - 1)
373 struct lfsck_rbtree_node {
374 struct rb_node lrn_node;
377 atomic_t lrn_known_count;
378 atomic_t lrn_accessed_count;
379 void *lrn_known_bitmap;
380 void *lrn_accessed_bitmap;
383 static inline int lfsck_rbtree_cmp(struct lfsck_rbtree_node *lrn,
384 __u64 seq, __u32 oid)
386 if (seq < lrn->lrn_seq)
389 if (seq > lrn->lrn_seq)
392 if (oid < lrn->lrn_first_oid)
395 if (oid >= lrn->lrn_first_oid + LFSCK_RBTREE_BITMAP_WIDTH)
401 /* The caller should hold llsd->llsd_rb_lock. */
402 static struct lfsck_rbtree_node *
403 lfsck_rbtree_search(struct lfsck_layout_slave_data *llsd,
404 const struct lu_fid *fid, bool *exact)
406 struct rb_node *node = llsd->llsd_rb_root.rb_node;
407 struct rb_node *prev = NULL;
408 struct lfsck_rbtree_node *lrn = NULL;
414 while (node != NULL) {
416 lrn = rb_entry(node, struct lfsck_rbtree_node, lrn_node);
417 rc = lfsck_rbtree_cmp(lrn, fid_seq(fid), fid_oid(fid));
419 node = node->rb_left;
421 node = node->rb_right;
429 /* If there is no exactly matched one, then to the next valid one. */
432 /* The rbtree is empty. */
439 node = rb_next(prev);
441 /* The end of the rbtree. */
445 lrn = rb_entry(node, struct lfsck_rbtree_node, lrn_node);
450 static struct lfsck_rbtree_node *lfsck_rbtree_new(const struct lu_env *env,
451 const struct lu_fid *fid)
453 struct lfsck_rbtree_node *lrn;
457 return ERR_PTR(-ENOMEM);
459 OBD_ALLOC(lrn->lrn_known_bitmap, LFSCK_RBTREE_BITMAP_SIZE);
460 if (lrn->lrn_known_bitmap == NULL) {
463 return ERR_PTR(-ENOMEM);
466 OBD_ALLOC(lrn->lrn_accessed_bitmap, LFSCK_RBTREE_BITMAP_SIZE);
467 if (lrn->lrn_accessed_bitmap == NULL) {
468 OBD_FREE(lrn->lrn_known_bitmap, LFSCK_RBTREE_BITMAP_SIZE);
471 return ERR_PTR(-ENOMEM);
474 rb_init_node(&lrn->lrn_node);
475 lrn->lrn_seq = fid_seq(fid);
476 lrn->lrn_first_oid = fid_oid(fid) & ~LFSCK_RBTREE_BITMAP_MASK;
477 atomic_set(&lrn->lrn_known_count, 0);
478 atomic_set(&lrn->lrn_accessed_count, 0);
483 static void lfsck_rbtree_free(struct lfsck_rbtree_node *lrn)
485 OBD_FREE(lrn->lrn_accessed_bitmap, LFSCK_RBTREE_BITMAP_SIZE);
486 OBD_FREE(lrn->lrn_known_bitmap, LFSCK_RBTREE_BITMAP_SIZE);
490 /* The caller should hold lock. */
491 static struct lfsck_rbtree_node *
492 lfsck_rbtree_insert(struct lfsck_layout_slave_data *llsd,
493 struct lfsck_rbtree_node *lrn)
495 struct rb_node **pos = &(llsd->llsd_rb_root.rb_node);
496 struct rb_node *parent = NULL;
497 struct lfsck_rbtree_node *tmp;
502 tmp = rb_entry(*pos, struct lfsck_rbtree_node, lrn_node);
503 rc = lfsck_rbtree_cmp(tmp, lrn->lrn_seq, lrn->lrn_first_oid);
505 pos = &((*pos)->rb_left);
507 pos = &((*pos)->rb_right);
512 rb_link_node(&lrn->lrn_node, parent, pos);
513 rb_insert_color(&lrn->lrn_node, &llsd->llsd_rb_root);
518 extern const struct dt_index_operations lfsck_orphan_index_ops;
520 static int lfsck_rbtree_setup(const struct lu_env *env,
521 struct lfsck_component *com)
523 struct lu_fid *fid = &lfsck_env_info(env)->lti_fid;
524 struct lfsck_instance *lfsck = com->lc_lfsck;
525 struct dt_device *dev = lfsck->li_bottom;
526 struct lfsck_layout_slave_data *llsd = com->lc_data;
527 struct dt_object *obj;
529 fid->f_seq = FID_SEQ_LAYOUT_RBTREE;
530 fid->f_oid = lfsck_dev_idx(dev);
532 obj = dt_locate(env, dev, fid);
534 RETURN(PTR_ERR(obj));
536 /* Generate an in-RAM object to stand for the layout rbtree.
537 * Scanning the layout rbtree will be via the iteration over
538 * the object. In the future, the rbtree may be written onto
539 * disk with the object.
541 * Mark the object to be as exist. */
542 obj->do_lu.lo_header->loh_attr |= LOHA_EXISTS;
543 obj->do_index_ops = &lfsck_orphan_index_ops;
544 llsd->llsd_rb_obj = obj;
545 llsd->llsd_rbtree_valid = 1;
546 dev->dd_record_fid_accessed = 1;
551 static void lfsck_rbtree_cleanup(const struct lu_env *env,
552 struct lfsck_component *com)
554 struct lfsck_instance *lfsck = com->lc_lfsck;
555 struct lfsck_layout_slave_data *llsd = com->lc_data;
556 struct rb_node *node = rb_first(&llsd->llsd_rb_root);
557 struct rb_node *next;
558 struct lfsck_rbtree_node *lrn;
560 lfsck->li_bottom->dd_record_fid_accessed = 0;
561 /* Invalid the rbtree, then no others will use it. */
562 write_lock(&llsd->llsd_rb_lock);
563 llsd->llsd_rbtree_valid = 0;
564 write_unlock(&llsd->llsd_rb_lock);
566 while (node != NULL) {
567 next = rb_next(node);
568 lrn = rb_entry(node, struct lfsck_rbtree_node, lrn_node);
569 rb_erase(node, &llsd->llsd_rb_root);
570 lfsck_rbtree_free(lrn);
574 if (llsd->llsd_rb_obj != NULL) {
575 lu_object_put(env, &llsd->llsd_rb_obj->do_lu);
576 llsd->llsd_rb_obj = NULL;
580 static void lfsck_rbtree_update_bitmap(const struct lu_env *env,
581 struct lfsck_component *com,
582 const struct lu_fid *fid,
585 struct lfsck_layout_slave_data *llsd = com->lc_data;
586 struct lfsck_rbtree_node *lrn;
592 CDEBUG(D_LFSCK, "%s: update bitmap for "DFID"\n",
593 lfsck_lfsck2name(com->lc_lfsck), PFID(fid));
595 if (unlikely(!fid_is_sane(fid) || fid_is_last_id(fid)))
598 if (!fid_is_idif(fid) && !fid_is_norm(fid))
601 read_lock(&llsd->llsd_rb_lock);
602 if (!llsd->llsd_rbtree_valid)
603 GOTO(unlock, rc = 0);
605 lrn = lfsck_rbtree_search(llsd, fid, NULL);
607 struct lfsck_rbtree_node *tmp;
611 read_unlock(&llsd->llsd_rb_lock);
612 tmp = lfsck_rbtree_new(env, fid);
614 GOTO(out, rc = PTR_ERR(tmp));
617 write_lock(&llsd->llsd_rb_lock);
618 if (!llsd->llsd_rbtree_valid) {
619 lfsck_rbtree_free(tmp);
620 GOTO(unlock, rc = 0);
623 lrn = lfsck_rbtree_insert(llsd, tmp);
625 lfsck_rbtree_free(tmp);
628 idx = fid_oid(fid) & LFSCK_RBTREE_BITMAP_MASK;
629 /* Any accessed object must be a known object. */
630 if (!test_and_set_bit(idx, lrn->lrn_known_bitmap))
631 atomic_inc(&lrn->lrn_known_count);
632 if (accessed && !test_and_set_bit(idx, lrn->lrn_accessed_bitmap))
633 atomic_inc(&lrn->lrn_accessed_count);
635 GOTO(unlock, rc = 0);
639 write_unlock(&llsd->llsd_rb_lock);
641 read_unlock(&llsd->llsd_rb_lock);
643 if (rc != 0 && accessed) {
644 struct lfsck_layout *lo = com->lc_file_ram;
646 CERROR("%s: Fail to update object accessed bitmap, will cause "
647 "incorrect LFSCK OST-object handling, so disable it to "
648 "cancel orphan handling for related device. rc = %d.\n",
649 lfsck_lfsck2name(com->lc_lfsck), rc);
650 lo->ll_flags |= LF_INCOMPLETE;
651 lfsck_rbtree_cleanup(env, com);
655 static inline bool is_dummy_lov_ost_data(struct lov_ost_data_v1 *obj)
657 if (fid_is_zero(&obj->l_ost_oi.oi_fid) &&
658 obj->l_ost_gen == 0 && obj->l_ost_idx == 0)
664 static void lfsck_layout_le_to_cpu(struct lfsck_layout *des,
665 const struct lfsck_layout *src)
669 des->ll_magic = le32_to_cpu(src->ll_magic);
670 des->ll_status = le32_to_cpu(src->ll_status);
671 des->ll_flags = le32_to_cpu(src->ll_flags);
672 des->ll_success_count = le32_to_cpu(src->ll_success_count);
673 des->ll_run_time_phase1 = le32_to_cpu(src->ll_run_time_phase1);
674 des->ll_run_time_phase2 = le32_to_cpu(src->ll_run_time_phase2);
675 des->ll_time_last_complete = le64_to_cpu(src->ll_time_last_complete);
676 des->ll_time_latest_start = le64_to_cpu(src->ll_time_latest_start);
677 des->ll_time_last_checkpoint =
678 le64_to_cpu(src->ll_time_last_checkpoint);
679 des->ll_pos_latest_start = le64_to_cpu(src->ll_pos_latest_start);
680 des->ll_pos_last_checkpoint = le64_to_cpu(src->ll_pos_last_checkpoint);
681 des->ll_pos_first_inconsistent =
682 le64_to_cpu(src->ll_pos_first_inconsistent);
683 des->ll_objs_checked_phase1 = le64_to_cpu(src->ll_objs_checked_phase1);
684 des->ll_objs_failed_phase1 = le64_to_cpu(src->ll_objs_failed_phase1);
685 des->ll_objs_checked_phase2 = le64_to_cpu(src->ll_objs_checked_phase2);
686 des->ll_objs_failed_phase2 = le64_to_cpu(src->ll_objs_failed_phase2);
687 for (i = 0; i < LLIT_MAX; i++)
688 des->ll_objs_repaired[i] =
689 le64_to_cpu(src->ll_objs_repaired[i]);
690 des->ll_objs_skipped = le64_to_cpu(src->ll_objs_skipped);
693 static void lfsck_layout_cpu_to_le(struct lfsck_layout *des,
694 const struct lfsck_layout *src)
698 des->ll_magic = cpu_to_le32(src->ll_magic);
699 des->ll_status = cpu_to_le32(src->ll_status);
700 des->ll_flags = cpu_to_le32(src->ll_flags);
701 des->ll_success_count = cpu_to_le32(src->ll_success_count);
702 des->ll_run_time_phase1 = cpu_to_le32(src->ll_run_time_phase1);
703 des->ll_run_time_phase2 = cpu_to_le32(src->ll_run_time_phase2);
704 des->ll_time_last_complete = cpu_to_le64(src->ll_time_last_complete);
705 des->ll_time_latest_start = cpu_to_le64(src->ll_time_latest_start);
706 des->ll_time_last_checkpoint =
707 cpu_to_le64(src->ll_time_last_checkpoint);
708 des->ll_pos_latest_start = cpu_to_le64(src->ll_pos_latest_start);
709 des->ll_pos_last_checkpoint = cpu_to_le64(src->ll_pos_last_checkpoint);
710 des->ll_pos_first_inconsistent =
711 cpu_to_le64(src->ll_pos_first_inconsistent);
712 des->ll_objs_checked_phase1 = cpu_to_le64(src->ll_objs_checked_phase1);
713 des->ll_objs_failed_phase1 = cpu_to_le64(src->ll_objs_failed_phase1);
714 des->ll_objs_checked_phase2 = cpu_to_le64(src->ll_objs_checked_phase2);
715 des->ll_objs_failed_phase2 = cpu_to_le64(src->ll_objs_failed_phase2);
716 for (i = 0; i < LLIT_MAX; i++)
717 des->ll_objs_repaired[i] =
718 cpu_to_le64(src->ll_objs_repaired[i]);
719 des->ll_objs_skipped = cpu_to_le64(src->ll_objs_skipped);
723 * \retval +ve: the lfsck_layout is broken, the caller should reset it.
724 * \retval 0: succeed.
725 * \retval -ve: failed cases.
727 static int lfsck_layout_load(const struct lu_env *env,
728 struct lfsck_component *com)
730 struct lfsck_layout *lo = com->lc_file_ram;
731 const struct dt_body_operations *dbo = com->lc_obj->do_body_ops;
732 ssize_t size = com->lc_file_size;
736 rc = dbo->dbo_read(env, com->lc_obj,
737 lfsck_buf_get(env, com->lc_file_disk, size), &pos,
742 CWARN("%s: failed to load lfsck_layout: rc = %d\n",
743 lfsck_lfsck2name(com->lc_lfsck), rc);
745 } else if (rc != size) {
746 CWARN("%s: crashed lfsck_layout, to be reset: rc = %d\n",
747 lfsck_lfsck2name(com->lc_lfsck), rc);
751 lfsck_layout_le_to_cpu(lo, com->lc_file_disk);
752 if (lo->ll_magic != LFSCK_LAYOUT_MAGIC) {
753 CWARN("%s: invalid lfsck_layout magic %#x != %#x, "
754 "to be reset\n", lfsck_lfsck2name(com->lc_lfsck),
755 lo->ll_magic, LFSCK_LAYOUT_MAGIC);
762 static int lfsck_layout_store(const struct lu_env *env,
763 struct lfsck_component *com)
765 struct dt_object *obj = com->lc_obj;
766 struct lfsck_instance *lfsck = com->lc_lfsck;
767 struct lfsck_layout *lo = com->lc_file_disk;
768 struct thandle *handle;
769 ssize_t size = com->lc_file_size;
774 lfsck_layout_cpu_to_le(lo, com->lc_file_ram);
775 handle = dt_trans_create(env, lfsck->li_bottom);
776 if (IS_ERR(handle)) {
777 rc = PTR_ERR(handle);
778 CERROR("%s: fail to create trans for storing lfsck_layout: "
779 "rc = %d\n", lfsck_lfsck2name(lfsck), rc);
783 rc = dt_declare_record_write(env, obj, size, pos, handle);
785 CERROR("%s: fail to declare trans for storing lfsck_layout(1): "
786 "rc = %d\n", lfsck_lfsck2name(lfsck), rc);
790 rc = dt_trans_start_local(env, lfsck->li_bottom, handle);
792 CERROR("%s: fail to start trans for storing lfsck_layout: "
793 "rc = %d\n", lfsck_lfsck2name(lfsck), rc);
797 rc = dt_record_write(env, obj, lfsck_buf_get(env, lo, size), &pos,
800 CERROR("%s: fail to store lfsck_layout(1): size = %d, "
801 "rc = %d\n", lfsck_lfsck2name(lfsck), (int)size, rc);
806 dt_trans_stop(env, lfsck->li_bottom, handle);
811 static int lfsck_layout_init(const struct lu_env *env,
812 struct lfsck_component *com)
814 struct lfsck_layout *lo = com->lc_file_ram;
817 memset(lo, 0, com->lc_file_size);
818 lo->ll_magic = LFSCK_LAYOUT_MAGIC;
819 lo->ll_status = LS_INIT;
820 down_write(&com->lc_sem);
821 rc = lfsck_layout_store(env, com);
822 up_write(&com->lc_sem);
827 static int fid_is_for_ostobj(const struct lu_env *env, struct dt_device *dt,
828 struct dt_object *obj, const struct lu_fid *fid)
830 struct seq_server_site *ss = lu_site2seq(dt->dd_lu_dev.ld_site);
831 struct lu_seq_range range = { 0 };
832 struct lustre_mdt_attrs *lma;
835 fld_range_set_any(&range);
836 rc = fld_server_lookup(env, ss->ss_server_fld, fid_seq(fid), &range);
838 if (fld_range_is_ost(&range))
844 lma = &lfsck_env_info(env)->lti_lma;
845 rc = dt_xattr_get(env, obj, lfsck_buf_get(env, lma, sizeof(*lma)),
846 XATTR_NAME_LMA, BYPASS_CAPA);
847 if (rc == sizeof(*lma)) {
848 lustre_lma_swab(lma);
850 return lma->lma_compat & LMAC_FID_ON_OST ? 1 : 0;
853 rc = dt_xattr_get(env, obj, &LU_BUF_NULL, XATTR_NAME_FID, BYPASS_CAPA);
858 static struct lfsck_layout_seq *
859 lfsck_layout_seq_lookup(struct lfsck_layout_slave_data *llsd, __u64 seq)
861 struct lfsck_layout_seq *lls;
863 list_for_each_entry(lls, &llsd->llsd_seq_list, lls_list) {
864 if (lls->lls_seq == seq)
867 if (lls->lls_seq > seq)
875 lfsck_layout_seq_insert(struct lfsck_layout_slave_data *llsd,
876 struct lfsck_layout_seq *lls)
878 struct lfsck_layout_seq *tmp;
879 struct list_head *pos = &llsd->llsd_seq_list;
881 list_for_each_entry(tmp, &llsd->llsd_seq_list, lls_list) {
882 if (lls->lls_seq < tmp->lls_seq) {
883 pos = &tmp->lls_list;
887 list_add_tail(&lls->lls_list, pos);
891 lfsck_layout_lastid_create(const struct lu_env *env,
892 struct lfsck_instance *lfsck,
893 struct dt_object *obj)
895 struct lfsck_thread_info *info = lfsck_env_info(env);
896 struct lu_attr *la = &info->lti_la;
897 struct dt_object_format *dof = &info->lti_dof;
898 struct lfsck_bookmark *bk = &lfsck->li_bookmark_ram;
899 struct dt_device *dt = lfsck->li_bottom;
906 CDEBUG(D_LFSCK, "To create LAST_ID for <seq> "LPX64"\n",
907 fid_seq(lfsck_dto2fid(obj)));
909 if (bk->lb_param & LPF_DRYRUN)
912 memset(la, 0, sizeof(*la));
913 la->la_mode = S_IFREG | S_IRUGO | S_IWUSR;
914 la->la_valid = LA_MODE | LA_UID | LA_GID;
915 dof->dof_type = dt_mode_to_dft(S_IFREG);
917 th = dt_trans_create(env, dt);
919 RETURN(rc = PTR_ERR(th));
921 rc = dt_declare_create(env, obj, la, NULL, dof, th);
925 rc = dt_declare_record_write(env, obj, sizeof(lastid), pos, th);
929 rc = dt_trans_start_local(env, dt, th);
933 dt_write_lock(env, obj, 0);
934 if (likely(!dt_object_exists(obj))) {
935 rc = dt_create(env, obj, la, NULL, dof, th);
937 rc = dt_record_write(env, obj,
938 lfsck_buf_get(env, &lastid, sizeof(lastid)),
941 dt_write_unlock(env, obj);
946 dt_trans_stop(env, dt, th);
952 lfsck_layout_lastid_reload(const struct lu_env *env,
953 struct lfsck_component *com,
954 struct lfsck_layout_seq *lls)
960 dt_read_lock(env, lls->lls_lastid_obj, 0);
961 rc = dt_record_read(env, lls->lls_lastid_obj,
962 lfsck_buf_get(env, &lastid, sizeof(lastid)), &pos);
963 dt_read_unlock(env, lls->lls_lastid_obj);
964 if (unlikely(rc != 0))
967 lastid = le64_to_cpu(lastid);
968 if (lastid < lls->lls_lastid_known) {
969 struct lfsck_instance *lfsck = com->lc_lfsck;
970 struct lfsck_layout *lo = com->lc_file_ram;
972 lls->lls_lastid = lls->lls_lastid_known;
974 if (!(lo->ll_flags & LF_CRASHED_LASTID)) {
975 LASSERT(lfsck->li_out_notify != NULL);
977 lfsck->li_out_notify(env, lfsck->li_out_notify_data,
978 LE_LASTID_REBUILDING);
979 lo->ll_flags |= LF_CRASHED_LASTID;
981 } else if (lastid >= lls->lls_lastid) {
982 lls->lls_lastid = lastid;
990 lfsck_layout_lastid_store(const struct lu_env *env,
991 struct lfsck_component *com)
993 struct lfsck_instance *lfsck = com->lc_lfsck;
994 struct lfsck_bookmark *bk = &lfsck->li_bookmark_ram;
995 struct dt_device *dt = lfsck->li_bottom;
996 struct lfsck_layout_slave_data *llsd = com->lc_data;
997 struct lfsck_layout_seq *lls;
1003 list_for_each_entry(lls, &llsd->llsd_seq_list, lls_list) {
1006 /* XXX: Add the code back if we really found related
1007 * inconsistent cases in the future. */
1009 if (!lls->lls_dirty) {
1010 /* In OFD, before the pre-creation, the LAST_ID
1011 * file will be updated firstly, which may hide
1012 * some potential crashed cases. For example:
1014 * The old obj1's ID is higher than old LAST_ID
1015 * but lower than the new LAST_ID, but the LFSCK
1016 * have not touch the obj1 until the OFD updated
1017 * the LAST_ID. So the LFSCK does not regard it
1018 * as crashed case. But when OFD does not create
1019 * successfully, it will set the LAST_ID as the
1020 * real created objects' ID, then LFSCK needs to
1021 * found related inconsistency. */
1022 rc = lfsck_layout_lastid_reload(env, com, lls);
1023 if (likely(!lls->lls_dirty))
1028 CDEBUG(D_LFSCK, "To sync the LAST_ID for <seq> "LPX64
1029 " as <oid> "LPU64"\n", lls->lls_seq, lls->lls_lastid);
1031 if (bk->lb_param & LPF_DRYRUN) {
1036 th = dt_trans_create(env, dt);
1039 CERROR("%s: (1) failed to store "LPX64": rc = %d\n",
1040 lfsck_lfsck2name(com->lc_lfsck),
1045 rc = dt_declare_record_write(env, lls->lls_lastid_obj,
1046 sizeof(lastid), pos, th);
1050 rc = dt_trans_start_local(env, dt, th);
1054 lastid = cpu_to_le64(lls->lls_lastid);
1055 dt_write_lock(env, lls->lls_lastid_obj, 0);
1056 rc = dt_record_write(env, lls->lls_lastid_obj,
1057 lfsck_buf_get(env, &lastid,
1058 sizeof(lastid)), &pos, th);
1059 dt_write_unlock(env, lls->lls_lastid_obj);
1064 dt_trans_stop(env, dt, th);
1067 CERROR("%s: (2) failed to store "LPX64": rc = %d\n",
1068 lfsck_lfsck2name(com->lc_lfsck),
1077 lfsck_layout_lastid_load(const struct lu_env *env,
1078 struct lfsck_component *com,
1079 struct lfsck_layout_seq *lls)
1081 struct lfsck_instance *lfsck = com->lc_lfsck;
1082 struct lfsck_layout *lo = com->lc_file_ram;
1083 struct lu_fid *fid = &lfsck_env_info(env)->lti_fid;
1084 struct dt_object *obj;
1089 lu_last_id_fid(fid, lls->lls_seq, lfsck_dev_idx(lfsck->li_bottom));
1090 obj = dt_locate(env, lfsck->li_bottom, fid);
1092 RETURN(PTR_ERR(obj));
1094 /* LAST_ID crashed, to be rebuilt */
1095 if (!dt_object_exists(obj)) {
1096 if (!(lo->ll_flags & LF_CRASHED_LASTID)) {
1097 LASSERT(lfsck->li_out_notify != NULL);
1099 lfsck->li_out_notify(env, lfsck->li_out_notify_data,
1100 LE_LASTID_REBUILDING);
1101 lo->ll_flags |= LF_CRASHED_LASTID;
1103 if (OBD_FAIL_CHECK(OBD_FAIL_LFSCK_DELAY4) &&
1105 struct l_wait_info lwi = LWI_TIMEOUT(
1106 cfs_time_seconds(cfs_fail_val),
1109 up_write(&com->lc_sem);
1110 l_wait_event(lfsck->li_thread.t_ctl_waitq,
1111 !thread_is_running(&lfsck->li_thread),
1113 down_write(&com->lc_sem);
1117 rc = lfsck_layout_lastid_create(env, lfsck, obj);
1119 dt_read_lock(env, obj, 0);
1120 rc = dt_read(env, obj,
1121 lfsck_buf_get(env, &lls->lls_lastid, sizeof(__u64)),
1123 dt_read_unlock(env, obj);
1124 if (rc != 0 && rc != sizeof(__u64))
1125 GOTO(out, rc = (rc > 0 ? -EFAULT : rc));
1127 if (rc == 0 && !(lo->ll_flags & LF_CRASHED_LASTID)) {
1128 LASSERT(lfsck->li_out_notify != NULL);
1130 lfsck->li_out_notify(env, lfsck->li_out_notify_data,
1131 LE_LASTID_REBUILDING);
1132 lo->ll_flags |= LF_CRASHED_LASTID;
1135 lls->lls_lastid = le64_to_cpu(lls->lls_lastid);
1143 lfsck_object_put(env, obj);
1145 lls->lls_lastid_obj = obj;
1150 static int lfsck_layout_master_async_interpret(const struct lu_env *env,
1151 struct ptlrpc_request *req,
1154 struct lfsck_async_interpret_args *laia = args;
1155 struct lfsck_component *com = laia->laia_com;
1156 struct lfsck_layout_master_data *llmd = com->lc_data;
1157 struct lfsck_tgt_descs *ltds = laia->laia_ltds;
1158 struct lfsck_tgt_desc *ltd = laia->laia_ltd;
1159 struct lfsck_request *lr = laia->laia_lr;
1161 switch (lr->lr_event) {
1164 struct lfsck_layout *lo = com->lc_file_ram;
1166 CERROR("%s: fail to notify %s %x for layout start: "
1167 "rc = %d\n", lfsck_lfsck2name(com->lc_lfsck),
1168 (lr->lr_flags & LEF_TO_OST) ? "OST" : "MDT",
1169 ltd->ltd_index, rc);
1170 lo->ll_flags |= LF_INCOMPLETE;
1174 spin_lock(<ds->ltd_lock);
1175 if (ltd->ltd_dead || ltd->ltd_layout_done) {
1176 spin_unlock(<ds->ltd_lock);
1180 if (lr->lr_flags & LEF_TO_OST) {
1181 if (list_empty(<d->ltd_layout_list))
1182 list_add_tail(<d->ltd_layout_list,
1183 &llmd->llmd_ost_list);
1184 if (list_empty(<d->ltd_layout_phase_list))
1185 list_add_tail(<d->ltd_layout_phase_list,
1186 &llmd->llmd_ost_phase1_list);
1188 if (list_empty(<d->ltd_layout_list))
1189 list_add_tail(<d->ltd_layout_list,
1190 &llmd->llmd_mdt_list);
1191 if (list_empty(<d->ltd_layout_phase_list))
1192 list_add_tail(<d->ltd_layout_phase_list,
1193 &llmd->llmd_mdt_phase1_list);
1195 spin_unlock(<ds->ltd_lock);
1198 case LE_PHASE1_DONE:
1199 case LE_PHASE2_DONE:
1201 if (rc != 0 && rc != -EALREADY)
1202 CWARN("%s: fail to notify %s %x for layout: "
1203 "event = %d, rc = %d\n",
1204 lfsck_lfsck2name(com->lc_lfsck),
1205 (lr->lr_flags & LEF_TO_OST) ? "OST" : "MDT",
1206 ltd->ltd_index, lr->lr_event, rc);
1209 struct lfsck_reply *reply;
1212 spin_lock(<ds->ltd_lock);
1213 list_del_init(<d->ltd_layout_phase_list);
1214 list_del_init(<d->ltd_layout_list);
1215 spin_unlock(<ds->ltd_lock);
1219 reply = req_capsule_server_get(&req->rq_pill,
1221 if (reply == NULL) {
1223 CERROR("%s: invalid return value: rc = %d\n",
1224 lfsck_lfsck2name(com->lc_lfsck), rc);
1225 spin_lock(<ds->ltd_lock);
1226 list_del_init(<d->ltd_layout_phase_list);
1227 list_del_init(<d->ltd_layout_list);
1228 spin_unlock(<ds->ltd_lock);
1232 switch (reply->lr_status) {
1233 case LS_SCANNING_PHASE1:
1235 case LS_SCANNING_PHASE2:
1236 spin_lock(<ds->ltd_lock);
1237 list_del_init(<d->ltd_layout_phase_list);
1238 if (ltd->ltd_dead || ltd->ltd_layout_done) {
1239 spin_unlock(<ds->ltd_lock);
1243 if (lr->lr_flags & LEF_TO_OST)
1244 list_add_tail(<d->ltd_layout_phase_list,
1245 &llmd->llmd_ost_phase2_list);
1247 list_add_tail(<d->ltd_layout_phase_list,
1248 &llmd->llmd_mdt_phase2_list);
1249 spin_unlock(<ds->ltd_lock);
1252 spin_lock(<ds->ltd_lock);
1253 list_del_init(<d->ltd_layout_phase_list);
1254 list_del_init(<d->ltd_layout_list);
1255 spin_unlock(<ds->ltd_lock);
1261 CERROR("%s: unexpected event: rc = %d\n",
1262 lfsck_lfsck2name(com->lc_lfsck), lr->lr_event);
1266 if (!laia->laia_shared) {
1268 lfsck_component_put(env, com);
1274 static int lfsck_layout_master_query_others(const struct lu_env *env,
1275 struct lfsck_component *com)
1277 struct lfsck_thread_info *info = lfsck_env_info(env);
1278 struct lfsck_request *lr = &info->lti_lr;
1279 struct lfsck_async_interpret_args *laia = &info->lti_laia;
1280 struct lfsck_instance *lfsck = com->lc_lfsck;
1281 struct lfsck_layout_master_data *llmd = com->lc_data;
1282 struct ptlrpc_request_set *set;
1283 struct lfsck_tgt_descs *ltds;
1284 struct lfsck_tgt_desc *ltd;
1285 struct list_head *head;
1290 set = ptlrpc_prep_set();
1294 llmd->llmd_touch_gen++;
1295 memset(lr, 0, sizeof(*lr));
1296 lr->lr_index = lfsck_dev_idx(lfsck->li_bottom);
1297 lr->lr_event = LE_QUERY;
1298 lr->lr_active = LT_LAYOUT;
1299 laia->laia_com = com;
1301 laia->laia_shared = 0;
1303 if (!list_empty(&llmd->llmd_mdt_phase1_list)) {
1304 ltds = &lfsck->li_mdt_descs;
1306 head = &llmd->llmd_mdt_phase1_list;
1310 ltds = &lfsck->li_ost_descs;
1311 lr->lr_flags = LEF_TO_OST;
1312 head = &llmd->llmd_ost_phase1_list;
1315 laia->laia_ltds = ltds;
1316 spin_lock(<ds->ltd_lock);
1317 while (!list_empty(head)) {
1318 ltd = list_entry(head->next,
1319 struct lfsck_tgt_desc,
1320 ltd_layout_phase_list);
1321 if (ltd->ltd_layout_gen == llmd->llmd_touch_gen)
1324 ltd->ltd_layout_gen = llmd->llmd_touch_gen;
1325 list_del(<d->ltd_layout_phase_list);
1326 list_add_tail(<d->ltd_layout_phase_list, head);
1327 atomic_inc(<d->ltd_ref);
1328 laia->laia_ltd = ltd;
1329 spin_unlock(<ds->ltd_lock);
1330 rc = lfsck_async_request(env, ltd->ltd_exp, lr, set,
1331 lfsck_layout_master_async_interpret,
1334 CERROR("%s: fail to query %s %x for layout: rc = %d\n",
1335 lfsck_lfsck2name(lfsck),
1336 (lr->lr_flags & LEF_TO_OST) ? "OST" : "MDT",
1337 ltd->ltd_index, rc);
1341 spin_lock(<ds->ltd_lock);
1343 spin_unlock(<ds->ltd_lock);
1345 rc = ptlrpc_set_wait(set);
1347 ptlrpc_set_destroy(set);
1351 if (!(lr->lr_flags & LEF_TO_OST) &&
1352 list_empty(&llmd->llmd_mdt_phase1_list))
1355 ptlrpc_set_destroy(set);
1357 RETURN(rc1 != 0 ? rc1 : rc);
1361 lfsck_layout_master_to_orphan(struct lfsck_layout_master_data *llmd)
1363 return list_empty(&llmd->llmd_mdt_phase1_list) &&
1364 (!list_empty(&llmd->llmd_ost_phase2_list) ||
1365 list_empty(&llmd->llmd_ost_phase1_list));
1368 static int lfsck_layout_master_notify_others(const struct lu_env *env,
1369 struct lfsck_component *com,
1370 struct lfsck_request *lr)
1372 struct lfsck_thread_info *info = lfsck_env_info(env);
1373 struct lfsck_async_interpret_args *laia = &info->lti_laia;
1374 struct lfsck_instance *lfsck = com->lc_lfsck;
1375 struct lfsck_layout_master_data *llmd = com->lc_data;
1376 struct lfsck_layout *lo = com->lc_file_ram;
1377 struct lfsck_bookmark *bk = &lfsck->li_bookmark_ram;
1378 struct ptlrpc_request_set *set;
1379 struct lfsck_tgt_descs *ltds;
1380 struct lfsck_tgt_desc *ltd;
1381 struct lfsck_tgt_desc *next;
1382 struct list_head *head;
1387 set = ptlrpc_prep_set();
1391 lr->lr_index = lfsck_dev_idx(lfsck->li_bottom);
1392 lr->lr_active = LT_LAYOUT;
1393 laia->laia_com = com;
1395 laia->laia_shared = 0;
1396 switch (lr->lr_event) {
1398 /* Notify OSTs firstly, then handle other MDTs if needed. */
1399 ltds = &lfsck->li_ost_descs;
1400 laia->laia_ltds = ltds;
1401 down_read(<ds->ltd_rw_sem);
1402 cfs_foreach_bit(ltds->ltd_tgts_bitmap, idx) {
1403 ltd = lfsck_tgt_get(ltds, idx);
1404 LASSERT(ltd != NULL);
1406 laia->laia_ltd = ltd;
1407 ltd->ltd_layout_done = 0;
1408 rc = lfsck_async_request(env, ltd->ltd_exp, lr, set,
1409 lfsck_layout_master_async_interpret,
1410 laia, LFSCK_NOTIFY);
1412 CERROR("%s: fail to notify %s %x for layout "
1414 lfsck_lfsck2name(lfsck),
1415 (lr->lr_flags & LEF_TO_OST) ? "OST" :
1418 lo->ll_flags |= LF_INCOMPLETE;
1421 up_read(<ds->ltd_rw_sem);
1424 rc = ptlrpc_set_wait(set);
1426 ptlrpc_set_destroy(set);
1430 if (!(bk->lb_param & LPF_ALL_TGT))
1433 /* link other MDT targets locallly. */
1434 spin_lock(<ds->ltd_lock);
1435 cfs_foreach_bit(ltds->ltd_tgts_bitmap, idx) {
1436 ltd = LTD_TGT(ltds, idx);
1437 LASSERT(ltd != NULL);
1439 if (!list_empty(<d->ltd_layout_list))
1442 list_add_tail(<d->ltd_layout_list,
1443 &llmd->llmd_mdt_list);
1444 list_add_tail(<d->ltd_layout_phase_list,
1445 &llmd->llmd_mdt_phase1_list);
1447 spin_unlock(<ds->ltd_lock);
1450 case LE_PHASE2_DONE:
1451 case LE_PEER_EXIT: {
1452 /* Handle other MDTs firstly if needed, then notify the OSTs. */
1453 if (bk->lb_param & LPF_ALL_TGT) {
1454 head = &llmd->llmd_mdt_list;
1455 ltds = &lfsck->li_mdt_descs;
1456 if (lr->lr_event == LE_STOP) {
1457 /* unlink other MDT targets locallly. */
1458 spin_lock(<ds->ltd_lock);
1459 list_for_each_entry_safe(ltd, next, head,
1461 list_del_init(<d->ltd_layout_phase_list);
1462 list_del_init(<d->ltd_layout_list);
1464 spin_unlock(<ds->ltd_lock);
1466 lr->lr_flags |= LEF_TO_OST;
1467 head = &llmd->llmd_ost_list;
1468 ltds = &lfsck->li_ost_descs;
1470 lr->lr_flags &= ~LEF_TO_OST;
1473 lr->lr_flags |= LEF_TO_OST;
1474 head = &llmd->llmd_ost_list;
1475 ltds = &lfsck->li_ost_descs;
1479 laia->laia_ltds = ltds;
1480 spin_lock(<ds->ltd_lock);
1481 while (!list_empty(head)) {
1482 ltd = list_entry(head->next, struct lfsck_tgt_desc,
1484 if (!list_empty(<d->ltd_layout_phase_list))
1485 list_del_init(<d->ltd_layout_phase_list);
1486 list_del_init(<d->ltd_layout_list);
1487 atomic_inc(<d->ltd_ref);
1488 laia->laia_ltd = ltd;
1489 spin_unlock(<ds->ltd_lock);
1490 rc = lfsck_async_request(env, ltd->ltd_exp, lr, set,
1491 lfsck_layout_master_async_interpret,
1492 laia, LFSCK_NOTIFY);
1494 CERROR("%s: fail to notify %s %x for layout "
1495 "stop/phase2: rc = %d\n",
1496 lfsck_lfsck2name(lfsck),
1497 (lr->lr_flags & LEF_TO_OST) ? "OST" :
1498 "MDT", ltd->ltd_index, rc);
1501 spin_lock(<ds->ltd_lock);
1503 spin_unlock(<ds->ltd_lock);
1505 rc = ptlrpc_set_wait(set);
1507 ptlrpc_set_destroy(set);
1511 if (!(lr->lr_flags & LEF_TO_OST)) {
1512 lr->lr_flags |= LEF_TO_OST;
1513 head = &llmd->llmd_ost_list;
1514 ltds = &lfsck->li_ost_descs;
1519 case LE_PHASE1_DONE:
1520 llmd->llmd_touch_gen++;
1521 ltds = &lfsck->li_mdt_descs;
1522 laia->laia_ltds = ltds;
1523 spin_lock(<ds->ltd_lock);
1524 while (!list_empty(&llmd->llmd_mdt_phase1_list)) {
1525 ltd = list_entry(llmd->llmd_mdt_phase1_list.next,
1526 struct lfsck_tgt_desc,
1527 ltd_layout_phase_list);
1528 if (ltd->ltd_layout_gen == llmd->llmd_touch_gen)
1531 ltd->ltd_layout_gen = llmd->llmd_touch_gen;
1532 list_del_init(<d->ltd_layout_phase_list);
1533 list_add_tail(<d->ltd_layout_phase_list,
1534 &llmd->llmd_mdt_phase1_list);
1535 atomic_inc(<d->ltd_ref);
1536 laia->laia_ltd = ltd;
1537 spin_unlock(<ds->ltd_lock);
1538 rc = lfsck_async_request(env, ltd->ltd_exp, lr, set,
1539 lfsck_layout_master_async_interpret,
1540 laia, LFSCK_NOTIFY);
1542 CERROR("%s: fail to notify MDT %x for layout "
1543 "phase1 done: rc = %d\n",
1544 lfsck_lfsck2name(lfsck),
1545 ltd->ltd_index, rc);
1548 spin_lock(<ds->ltd_lock);
1550 spin_unlock(<ds->ltd_lock);
1553 CERROR("%s: unexpected LFSCK event: rc = %d\n",
1554 lfsck_lfsck2name(lfsck), lr->lr_event);
1559 rc = ptlrpc_set_wait(set);
1560 ptlrpc_set_destroy(set);
1565 static int lfsck_layout_double_scan_result(const struct lu_env *env,
1566 struct lfsck_component *com,
1569 struct lfsck_instance *lfsck = com->lc_lfsck;
1570 struct lfsck_layout *lo = com->lc_file_ram;
1571 struct lfsck_bookmark *bk = &lfsck->li_bookmark_ram;
1573 down_write(&com->lc_sem);
1575 lo->ll_run_time_phase2 += cfs_duration_sec(cfs_time_current() +
1576 HALF_SEC - lfsck->li_time_last_checkpoint);
1577 lo->ll_time_last_checkpoint = cfs_time_current_sec();
1578 lo->ll_objs_checked_phase2 += com->lc_new_checked;
1581 com->lc_journal = 0;
1582 if (lo->ll_flags & LF_INCOMPLETE)
1583 lo->ll_status = LS_PARTIAL;
1585 lo->ll_status = LS_COMPLETED;
1586 if (!(bk->lb_param & LPF_DRYRUN))
1587 lo->ll_flags &= ~(LF_SCANNED_ONCE | LF_INCONSISTENT);
1588 lo->ll_time_last_complete = lo->ll_time_last_checkpoint;
1589 lo->ll_success_count++;
1590 } else if (rc == 0) {
1591 lo->ll_status = lfsck->li_status;
1592 if (lo->ll_status == 0)
1593 lo->ll_status = LS_STOPPED;
1595 lo->ll_status = LS_FAILED;
1598 if (lo->ll_status != LS_PAUSED) {
1599 spin_lock(&lfsck->li_lock);
1600 list_del_init(&com->lc_link);
1601 list_add_tail(&com->lc_link, &lfsck->li_list_idle);
1602 spin_unlock(&lfsck->li_lock);
1605 rc = lfsck_layout_store(env, com);
1607 up_write(&com->lc_sem);
1612 static int lfsck_layout_lock(const struct lu_env *env,
1613 struct lfsck_component *com,
1614 struct dt_object *obj,
1615 struct lustre_handle *lh, __u64 bits)
1617 struct lfsck_thread_info *info = lfsck_env_info(env);
1618 ldlm_policy_data_t *policy = &info->lti_policy;
1619 struct ldlm_res_id *resid = &info->lti_resid;
1620 struct lfsck_instance *lfsck = com->lc_lfsck;
1621 __u64 flags = LDLM_FL_ATOMIC_CB;
1624 LASSERT(lfsck->li_namespace != NULL);
1626 memset(policy, 0, sizeof(*policy));
1627 policy->l_inodebits.bits = bits;
1628 fid_build_reg_res_name(lfsck_dto2fid(obj), resid);
1629 rc = ldlm_cli_enqueue_local(lfsck->li_namespace, resid, LDLM_IBITS,
1630 policy, LCK_EX, &flags, ldlm_blocking_ast,
1631 ldlm_completion_ast, NULL, NULL, 0,
1632 LVB_T_NONE, NULL, lh);
1633 if (rc == ELDLM_OK) {
1636 memset(lh, 0, sizeof(*lh));
1643 static void lfsck_layout_unlock(struct lustre_handle *lh)
1645 if (lustre_handle_is_used(lh)) {
1646 ldlm_lock_decref(lh, LCK_EX);
1647 memset(lh, 0, sizeof(*lh));
1651 static int lfsck_layout_trans_stop(const struct lu_env *env,
1652 struct dt_device *dev,
1653 struct thandle *handle, int result)
1657 handle->th_result = result;
1658 rc = dt_trans_stop(env, dev, handle);
1668 * \retval +1: repaired
1669 * \retval 0: did nothing
1670 * \retval -ve: on error
1672 static int lfsck_layout_refill_lovea(const struct lu_env *env,
1673 struct thandle *handle,
1674 struct dt_object *parent,
1675 struct lu_fid *cfid,
1677 struct lov_ost_data_v1 *slot,
1678 int fl, __u32 ost_idx)
1680 struct ost_id *oi = &lfsck_env_info(env)->lti_oi;
1683 fid_to_ostid(cfid, oi);
1684 ostid_cpu_to_le(oi, &slot->l_ost_oi);
1685 slot->l_ost_gen = cpu_to_le32(0);
1686 slot->l_ost_idx = cpu_to_le32(ost_idx);
1687 rc = dt_xattr_set(env, parent, buf, XATTR_NAME_LOV, fl, handle,
1696 * \retval +1: repaired
1697 * \retval 0: did nothing
1698 * \retval -ve: on error
1700 static int lfsck_layout_extend_lovea(const struct lu_env *env,
1701 struct thandle *handle,
1702 struct dt_object *parent,
1703 struct lu_fid *cfid,
1704 struct lu_buf *buf, int fl,
1705 __u32 ost_idx, __u32 ea_off)
1707 struct lov_mds_md_v1 *lmm = buf->lb_buf;
1708 struct lov_ost_data_v1 *objs;
1712 if (fl == LU_XATTR_CREATE) {
1713 LASSERT(buf->lb_len == lov_mds_md_size(ea_off + 1,
1716 memset(lmm, 0, buf->lb_len);
1717 lmm->lmm_magic = cpu_to_le32(LOV_MAGIC_V1);
1718 /* XXX: currently, we only support LOV_PATTERN_RAID0. */
1719 lmm->lmm_pattern = cpu_to_le32(LOV_PATTERN_RAID0);
1720 fid_to_lmm_oi(lfsck_dto2fid(parent), &lmm->lmm_oi);
1721 lmm_oi_cpu_to_le(&lmm->lmm_oi, &lmm->lmm_oi);
1722 /* XXX: We cannot know the stripe size,
1723 * then use the default value (1 MB). */
1724 lmm->lmm_stripe_size = cpu_to_le32(1024 * 1024);
1725 lmm->lmm_layout_gen = cpu_to_le16(0);
1726 objs = &(lmm->lmm_objects[ea_off]);
1728 __u16 count = le16_to_cpu(lmm->lmm_stripe_count);
1729 int gap = ea_off - count;
1730 __u32 magic = le32_to_cpu(lmm->lmm_magic);
1732 /* Currently, we only support LOV_MAGIC_V1/LOV_MAGIC_V3
1733 * which has been verified in lfsck_layout_verify_header()
1734 * already. If some new magic introduced in the future,
1735 * then layout LFSCK needs to be updated also. */
1736 if (magic == LOV_MAGIC_V1) {
1737 objs = &(lmm->lmm_objects[count]);
1739 LASSERT(magic == LOV_MAGIC_V3);
1740 objs = &((struct lov_mds_md_v3 *)lmm)->
1745 memset(objs, 0, gap * sizeof(*objs));
1746 lmm->lmm_layout_gen =
1747 cpu_to_le16(le16_to_cpu(lmm->lmm_layout_gen) + 1);
1750 LASSERT(buf->lb_len == lov_mds_md_size(ea_off + 1, magic));
1753 lmm->lmm_stripe_count = cpu_to_le16(ea_off + 1);
1754 rc = lfsck_layout_refill_lovea(env, handle, parent, cfid, buf, objs,
1761 * \retval +1: repaired
1762 * \retval 0: did nothing
1763 * \retval -ve: on error
1765 static int lfsck_layout_update_pfid(const struct lu_env *env,
1766 struct lfsck_component *com,
1767 struct dt_object *parent,
1768 struct lu_fid *cfid,
1769 struct dt_device *cdev, __u32 ea_off)
1771 struct filter_fid *pfid = &lfsck_env_info(env)->lti_new_pfid;
1772 struct dt_object *child;
1773 struct thandle *handle;
1774 const struct lu_fid *tfid = lu_object_fid(&parent->do_lu);
1779 child = lfsck_object_find_by_dev(env, cdev, cfid);
1781 RETURN(PTR_ERR(child));
1783 handle = dt_trans_create(env, cdev);
1785 GOTO(out, rc = PTR_ERR(handle));
1787 pfid->ff_parent.f_seq = cpu_to_le64(tfid->f_seq);
1788 pfid->ff_parent.f_oid = cpu_to_le32(tfid->f_oid);
1789 /* In fact, the ff_parent::f_ver is not the real parent FID::f_ver,
1790 * instead, it is the OST-object index in its parent MDT-object
1792 pfid->ff_parent.f_ver = cpu_to_le32(ea_off);
1793 buf = lfsck_buf_get(env, pfid, sizeof(struct filter_fid));
1795 rc = dt_declare_xattr_set(env, child, buf, XATTR_NAME_FID, 0, handle);
1799 rc = dt_trans_start(env, cdev, handle);
1803 rc = dt_xattr_set(env, child, buf, XATTR_NAME_FID, 0, handle,
1806 GOTO(stop, rc = (rc == 0 ? 1 : rc));
1809 dt_trans_stop(env, cdev, handle);
1812 lu_object_put(env, &child->do_lu);
1818 * \retval +1: repaired
1819 * \retval 0: did nothing
1820 * \retval -ve: on error
1822 static int lfsck_layout_recreate_parent(const struct lu_env *env,
1823 struct lfsck_component *com,
1824 struct lfsck_tgt_desc *ltd,
1825 struct lu_orphan_rec *rec,
1826 struct lu_fid *cfid,
1828 const char *postfix,
1831 /* XXX: To be extended in other patch. */
1836 * \retval +1: repaired
1837 * \retval 0: did nothing
1838 * \retval -ve: on error
1840 static int lfsck_layout_conflict_create(const struct lu_env *env,
1841 struct lfsck_component *com,
1842 struct lfsck_tgt_desc *ltd,
1843 struct lu_orphan_rec *rec,
1844 struct dt_object *parent,
1845 struct lu_fid *cfid,
1846 struct lu_buf *ea_buf,
1847 struct lov_ost_data_v1 *slot,
1848 __u32 ea_off, __u32 ori_len)
1850 /* XXX: To be extended in other patch. */
1855 * \retval +1: repaired
1856 * \retval 0: did nothing
1857 * \retval -ve: on error
1859 static int lfsck_layout_recreate_lovea(const struct lu_env *env,
1860 struct lfsck_component *com,
1861 struct lfsck_tgt_desc *ltd,
1862 struct lu_orphan_rec *rec,
1863 struct dt_object *parent,
1864 struct lu_fid *cfid,
1865 __u32 ost_idx, __u32 ea_off)
1867 struct lfsck_thread_info *info = lfsck_env_info(env);
1868 struct lu_buf *buf = &info->lti_big_buf;
1869 struct lu_fid *fid = &info->lti_fid2;
1870 struct ost_id *oi = &info->lti_oi;
1871 struct lfsck_instance *lfsck = com->lc_lfsck;
1872 struct dt_device *dt = lfsck->li_bottom;
1873 struct lfsck_bookmark *bk = &lfsck->li_bookmark_ram;
1874 struct thandle *handle = NULL;
1875 size_t buflen = buf->lb_len;
1876 struct lov_mds_md_v1 *lmm;
1877 struct lov_ost_data_v1 *objs;
1878 struct lustre_handle lh = { 0 };
1887 CDEBUG(D_LFSCK, "Re-create the crashed layout EA: parent "
1888 DFID", child "DFID", OST-index %u, stripe-index %u\n",
1889 PFID(lfsck_dto2fid(parent)), PFID(cfid), ost_idx, ea_off);
1891 rc = lfsck_layout_lock(env, com, parent, &lh,
1892 MDS_INODELOCK_LAYOUT | MDS_INODELOCK_XATTR);
1897 if (!(bk->lb_param & LPF_DRYRUN)) {
1898 handle = dt_trans_create(env, dt);
1900 GOTO(unlock_layout, rc = PTR_ERR(handle));
1902 rc = dt_declare_xattr_set(env, parent, buf, XATTR_NAME_LOV,
1907 rc = dt_trans_start_local(env, dt, handle);
1912 dt_write_lock(env, parent, 0);
1913 rc = dt_xattr_get(env, parent, buf, XATTR_NAME_LOV, BYPASS_CAPA);
1914 if (rc == -ERANGE) {
1915 rc = dt_xattr_get(env, parent, &LU_BUF_NULL, XATTR_NAME_LOV,
1919 dt_write_unlock(env, parent);
1920 if (handle != NULL) {
1921 dt_trans_stop(env, dt, handle);
1926 GOTO(unlock_layout, rc);
1928 lu_buf_realloc(buf, rc);
1929 buflen = buf->lb_len;
1930 if (buf->lb_buf == NULL)
1931 GOTO(unlock_layout, rc = -ENOMEM);
1933 fl = LU_XATTR_REPLACE;
1935 } else if (rc == -ENODATA || rc == 0) {
1936 fl = LU_XATTR_CREATE;
1937 } else if (rc < 0) {
1938 GOTO(unlock_parent, rc);
1939 } else if (unlikely(buf->lb_len == 0)) {
1940 dt_write_unlock(env, parent);
1941 if (handle != NULL) {
1942 dt_trans_stop(env, dt, handle);
1946 lu_buf_alloc(buf, rc);
1947 buflen = buf->lb_len;
1948 if (buf->lb_buf == NULL)
1949 GOTO(unlock_layout, rc = -ENOMEM);
1951 fl = LU_XATTR_REPLACE;
1954 fl = LU_XATTR_REPLACE;
1957 if (fl == LU_XATTR_CREATE) {
1958 if (bk->lb_param & LPF_DRYRUN)
1959 GOTO(unlock_parent, rc = 1);
1961 rc = lov_mds_md_size(ea_off + 1, LOV_MAGIC_V1);
1962 /* If the declared is not big enough, re-try. */
1963 if (buf->lb_len < rc) {
1964 dt_write_unlock(env, parent);
1965 if (handle != NULL) {
1966 dt_trans_stop(env, dt, handle);
1970 lu_buf_realloc(buf, rc);
1971 buflen = buf->lb_len;
1972 if (buf->lb_buf == NULL)
1973 GOTO(unlock_layout, rc = -ENOMEM);
1979 rc = lfsck_layout_extend_lovea(env, handle, parent, cfid, buf,
1980 fl, ost_idx, ea_off);
1982 GOTO(unlock_parent, rc);
1986 rc1 = lfsck_layout_verify_header(lmm);
1988 GOTO(unlock_parent, rc = rc1);
1990 /* Currently, we only support LOV_MAGIC_V1/LOV_MAGIC_V3 which has
1991 * been verified in lfsck_layout_verify_header() already. If some
1992 * new magic introduced in the future, then layout LFSCK needs to
1993 * be updated also. */
1994 magic = le32_to_cpu(lmm->lmm_magic);
1995 if (magic == LOV_MAGIC_V1) {
1996 objs = &(lmm->lmm_objects[0]);
1998 LASSERT(magic == LOV_MAGIC_V3);
1999 objs = &((struct lov_mds_md_v3 *)lmm)->lmm_objects[0];
2002 count = le16_to_cpu(lmm->lmm_stripe_count);
2004 GOTO(unlock_parent, rc = -EINVAL);
2007 /* Exceed the current end of MDT-object layout EA. Then extend it. */
2008 if (count <= ea_off) {
2009 if (bk->lb_param & LPF_DRYRUN)
2010 GOTO(unlock_parent, rc = 1);
2012 rc = lov_mds_md_size(ea_off + 1, LOV_MAGIC_V1);
2013 /* If the declared is not big enough, re-try. */
2014 if (buf->lb_len < rc) {
2015 dt_write_unlock(env, parent);
2016 if (handle != NULL) {
2017 dt_trans_stop(env, dt, handle);
2021 lu_buf_realloc(buf, rc);
2022 buflen = buf->lb_len;
2023 if (buf->lb_buf == NULL)
2024 GOTO(unlock_layout, rc = -ENOMEM);
2030 rc = lfsck_layout_extend_lovea(env, handle, parent, cfid, buf,
2031 fl, ost_idx, ea_off);
2032 GOTO(unlock_parent, rc);
2035 LASSERTF(rc > 0, "invalid rc = %d\n", rc);
2038 for (i = 0; i < count; i++, objs++) {
2039 /* The MDT-object was created via lfsck_layout_recover_create()
2040 * by others before, and we fill the dummy layout EA. */
2041 if (is_dummy_lov_ost_data(objs)) {
2045 if (bk->lb_param & LPF_DRYRUN)
2046 GOTO(unlock_parent, rc = 1);
2048 lmm->lmm_layout_gen =
2049 cpu_to_le16(le16_to_cpu(lmm->lmm_layout_gen) + 1);
2050 rc = lfsck_layout_refill_lovea(env, handle, parent,
2051 cfid, buf, objs, fl,
2053 GOTO(unlock_parent, rc);
2056 ostid_le_to_cpu(&objs->l_ost_oi, oi);
2057 ostid_to_fid(fid, oi, le32_to_cpu(objs->l_ost_idx));
2058 /* It should be rare case, the slot is there, but the LFSCK
2059 * does not handle it during the first-phase cycle scanning. */
2060 if (unlikely(lu_fid_eq(fid, cfid))) {
2062 GOTO(unlock_parent, rc = 0);
2064 /* Rare case that the OST-object index
2065 * does not match the parent MDT-object
2066 * layout EA. We trust the later one. */
2067 if (bk->lb_param & LPF_DRYRUN)
2068 GOTO(unlock_parent, rc = 1);
2070 dt_write_unlock(env, parent);
2072 dt_trans_stop(env, dt, handle);
2073 lfsck_layout_unlock(&lh);
2074 buf->lb_len = buflen;
2075 rc = lfsck_layout_update_pfid(env, com, parent,
2076 cfid, ltd->ltd_tgt, i);
2083 /* The MDT-object exists, but related layout EA slot is occupied
2085 if (bk->lb_param & LPF_DRYRUN)
2086 GOTO(unlock_parent, rc = 1);
2088 dt_write_unlock(env, parent);
2090 dt_trans_stop(env, dt, handle);
2091 lfsck_layout_unlock(&lh);
2092 if (le32_to_cpu(lmm->lmm_magic) == LOV_MAGIC_V1)
2093 objs = &(lmm->lmm_objects[ea_off]);
2095 objs = &((struct lov_mds_md_v3 *)lmm)->lmm_objects[ea_off];
2096 rc = lfsck_layout_conflict_create(env, com, ltd, rec, parent, cfid,
2097 buf, objs, ea_off, buflen);
2102 dt_write_unlock(env, parent);
2106 dt_trans_stop(env, dt, handle);
2109 lfsck_layout_unlock(&lh);
2110 buf->lb_len = buflen;
2115 static int lfsck_layout_scan_orphan_one(const struct lu_env *env,
2116 struct lfsck_component *com,
2117 struct lfsck_tgt_desc *ltd,
2118 struct lu_orphan_rec *rec,
2119 struct lu_fid *cfid)
2121 struct lfsck_layout *lo = com->lc_file_ram;
2122 struct lu_fid *pfid = &rec->lor_fid;
2123 struct dt_object *parent = NULL;
2124 __u32 ea_off = pfid->f_ver;
2128 if (!fid_is_sane(cfid))
2129 GOTO(out, rc = -EINVAL);
2131 if (fid_is_zero(pfid)) {
2132 rc = lfsck_layout_recreate_parent(env, com, ltd, rec, cfid,
2138 if (!fid_is_sane(pfid))
2139 GOTO(out, rc = -EINVAL);
2141 parent = lfsck_object_find_by_dev(env, com->lc_lfsck->li_bottom, pfid);
2143 GOTO(out, rc = PTR_ERR(parent));
2145 if (unlikely(dt_object_remote(parent) != 0))
2146 GOTO(put, rc = -EXDEV);
2148 if (dt_object_exists(parent) == 0) {
2149 lu_object_put(env, &parent->do_lu);
2150 rc = lfsck_layout_recreate_parent(env, com, ltd, rec, cfid,
2155 if (!S_ISREG(lu_object_attr(&parent->do_lu)))
2156 GOTO(put, rc = -EISDIR);
2158 rc = lfsck_layout_recreate_lovea(env, com, ltd, rec, parent, cfid,
2159 ltd->ltd_index, ea_off);
2165 lu_object_put(env, &parent->do_lu);
2167 /* The layout EA is changed, need to be reloaded next time. */
2168 lu_object_put_nocache(env, &parent->do_lu);
2171 down_write(&com->lc_sem);
2172 com->lc_new_scanned++;
2173 com->lc_new_checked++;
2175 lo->ll_objs_repaired[LLIT_ORPHAN - 1]++;
2177 } else if (rc < 0) {
2178 lo->ll_objs_failed_phase2++;
2180 up_write(&com->lc_sem);
2185 static int lfsck_layout_scan_orphan(const struct lu_env *env,
2186 struct lfsck_component *com,
2187 struct lfsck_tgt_desc *ltd)
2189 struct lfsck_layout *lo = com->lc_file_ram;
2190 struct lfsck_instance *lfsck = com->lc_lfsck;
2191 struct lfsck_bookmark *bk = &lfsck->li_bookmark_ram;
2192 struct lfsck_thread_info *info = lfsck_env_info(env);
2193 struct ost_id *oi = &info->lti_oi;
2194 struct lu_fid *fid = &info->lti_fid;
2195 struct dt_object *obj;
2196 const struct dt_it_ops *iops;
2201 CDEBUG(D_LFSCK, "%s: start the orphan scanning for OST%04x\n",
2202 lfsck_lfsck2name(lfsck), ltd->ltd_index);
2204 ostid_set_seq(oi, FID_SEQ_IDIF);
2205 ostid_set_id(oi, 0);
2206 ostid_to_fid(fid, oi, ltd->ltd_index);
2207 obj = lfsck_object_find_by_dev(env, ltd->ltd_tgt, fid);
2208 if (unlikely(IS_ERR(obj)))
2209 RETURN(PTR_ERR(obj));
2211 rc = obj->do_ops->do_index_try(env, obj, &dt_lfsck_orphan_features);
2215 iops = &obj->do_index_ops->dio_it;
2216 di = iops->init(env, obj, 0, BYPASS_CAPA);
2218 GOTO(put, rc = PTR_ERR(di));
2220 rc = iops->load(env, di, 0);
2222 /* -ESRCH means that the orphan OST-objects rbtree has been
2223 * cleanup because of the OSS server restart or other errors. */
2224 lo->ll_flags |= LF_INCOMPLETE;
2229 rc = iops->next(env, di);
2241 struct lu_orphan_rec *rec = &info->lti_rec;
2243 key = iops->key(env, di);
2244 com->lc_fid_latest_scanned_phase2 = *(struct lu_fid *)key;
2245 rc = iops->rec(env, di, (struct dt_rec *)rec, 0);
2247 rc = lfsck_layout_scan_orphan_one(env, com, ltd, rec,
2248 &com->lc_fid_latest_scanned_phase2);
2249 if (rc != 0 && bk->lb_param & LPF_FAILOUT)
2252 lfsck_control_speed_by_self(com);
2254 rc = iops->next(env, di);
2255 } while (rc < 0 && !(bk->lb_param & LPF_FAILOUT));
2262 iops->fini(env, di);
2264 lu_object_put(env, &obj->do_lu);
2266 CDEBUG(D_LFSCK, "%s: finish the orphan scanning for OST%04x, rc = %d\n",
2267 lfsck_lfsck2name(lfsck), ltd->ltd_index, rc);
2269 return rc > 0 ? 0 : rc;
2272 /* For the MDT-object with dangling reference, we need to re-create
2273 * the missed OST-object with the known FID/owner information. */
2274 static int lfsck_layout_recreate_ostobj(const struct lu_env *env,
2275 struct lfsck_component *com,
2276 struct lfsck_layout_req *llr,
2279 struct lfsck_thread_info *info = lfsck_env_info(env);
2280 struct filter_fid *pfid = &info->lti_new_pfid;
2281 struct dt_allocation_hint *hint = &info->lti_hint;
2282 struct dt_object *parent = llr->llr_parent->llo_obj;
2283 struct dt_object *child = llr->llr_child;
2284 struct dt_device *dev = lfsck_obj2dt_dev(child);
2285 const struct lu_fid *tfid = lu_object_fid(&parent->do_lu);
2286 struct thandle *handle;
2288 struct lustre_handle lh = { 0 };
2292 CDEBUG(D_LFSCK, "Repair dangling reference for: parent "DFID
2293 ", child "DFID", OST-index %u, stripe-index %u, owner %u:%u\n",
2294 PFID(lfsck_dto2fid(parent)), PFID(lfsck_dto2fid(child)),
2295 llr->llr_ost_idx, llr->llr_lov_idx, la->la_uid, la->la_gid);
2297 rc = lfsck_layout_lock(env, com, parent, &lh,
2298 MDS_INODELOCK_LAYOUT | MDS_INODELOCK_XATTR);
2302 handle = dt_trans_create(env, dev);
2304 GOTO(unlock1, rc = PTR_ERR(handle));
2306 hint->dah_parent = NULL;
2308 pfid->ff_parent.f_seq = cpu_to_le64(tfid->f_seq);
2309 pfid->ff_parent.f_oid = cpu_to_le32(tfid->f_oid);
2310 pfid->ff_parent.f_ver = cpu_to_le32(llr->llr_lov_idx);
2311 buf = lfsck_buf_get(env, pfid, sizeof(struct filter_fid));
2313 rc = dt_declare_create(env, child, la, hint, NULL, handle);
2317 rc = dt_declare_xattr_set(env, child, buf, XATTR_NAME_FID,
2318 LU_XATTR_CREATE, handle);
2322 rc = dt_trans_start(env, dev, handle);
2326 dt_read_lock(env, parent, 0);
2327 if (unlikely(lu_object_is_dying(parent->do_lu.lo_header)))
2328 GOTO(unlock2, rc = 1);
2330 rc = dt_create(env, child, la, hint, NULL, handle);
2334 rc = dt_xattr_set(env, child, buf, XATTR_NAME_FID, LU_XATTR_CREATE,
2335 handle, BYPASS_CAPA);
2340 dt_read_unlock(env, parent);
2343 rc = lfsck_layout_trans_stop(env, dev, handle, rc);
2346 lfsck_layout_unlock(&lh);
2351 /* If the OST-object does not recognize the MDT-object as its parent, and
2352 * there is no other MDT-object claims as its parent, then just trust the
2353 * given MDT-object as its parent. So update the OST-object filter_fid. */
2354 static int lfsck_layout_repair_unmatched_pair(const struct lu_env *env,
2355 struct lfsck_component *com,
2356 struct lfsck_layout_req *llr,
2357 const struct lu_attr *pla)
2359 struct lfsck_thread_info *info = lfsck_env_info(env);
2360 struct filter_fid *pfid = &info->lti_new_pfid;
2361 struct lu_attr *tla = &info->lti_la3;
2362 struct dt_object *parent = llr->llr_parent->llo_obj;
2363 struct dt_object *child = llr->llr_child;
2364 struct dt_device *dev = lfsck_obj2dt_dev(child);
2365 const struct lu_fid *tfid = lu_object_fid(&parent->do_lu);
2366 struct thandle *handle;
2368 struct lustre_handle lh = { 0 };
2372 CDEBUG(D_LFSCK, "Repair unmatched MDT-OST pair for: parent "DFID
2373 ", child "DFID", OST-index %u, stripe-index %u, owner %u:%u\n",
2374 PFID(lfsck_dto2fid(parent)), PFID(lfsck_dto2fid(child)),
2375 llr->llr_ost_idx, llr->llr_lov_idx, pla->la_uid, pla->la_gid);
2377 rc = lfsck_layout_lock(env, com, parent, &lh,
2378 MDS_INODELOCK_LAYOUT | MDS_INODELOCK_XATTR);
2382 handle = dt_trans_create(env, dev);
2384 GOTO(unlock1, rc = PTR_ERR(handle));
2386 pfid->ff_parent.f_seq = cpu_to_le64(tfid->f_seq);
2387 pfid->ff_parent.f_oid = cpu_to_le32(tfid->f_oid);
2388 /* The ff_parent->f_ver is not the real parent fid->f_ver. Instead,
2389 * it is the OST-object index in the parent MDT-object layout. */
2390 pfid->ff_parent.f_ver = cpu_to_le32(llr->llr_lov_idx);
2391 buf = lfsck_buf_get(env, pfid, sizeof(struct filter_fid));
2393 rc = dt_declare_xattr_set(env, child, buf, XATTR_NAME_FID, 0, handle);
2397 tla->la_valid = LA_UID | LA_GID;
2398 tla->la_uid = pla->la_uid;
2399 tla->la_gid = pla->la_gid;
2400 rc = dt_declare_attr_set(env, child, tla, handle);
2404 rc = dt_trans_start(env, dev, handle);
2408 dt_write_lock(env, parent, 0);
2409 if (unlikely(lu_object_is_dying(parent->do_lu.lo_header)))
2410 GOTO(unlock2, rc = 1);
2412 rc = dt_xattr_set(env, child, buf, XATTR_NAME_FID, 0, handle,
2417 /* Get the latest parent's owner. */
2418 rc = dt_attr_get(env, parent, tla, BYPASS_CAPA);
2422 tla->la_valid = LA_UID | LA_GID;
2423 rc = dt_attr_set(env, child, tla, handle, BYPASS_CAPA);
2428 dt_write_unlock(env, parent);
2431 rc = lfsck_layout_trans_stop(env, dev, handle, rc);
2434 lfsck_layout_unlock(&lh);
2439 /* If there are more than one MDT-objects claim as the OST-object's parent,
2440 * and the OST-object only recognizes one of them, then we need to generate
2441 * new OST-object(s) with new fid(s) for the non-recognized MDT-object(s). */
2442 static int lfsck_layout_repair_multiple_references(const struct lu_env *env,
2443 struct lfsck_component *com,
2444 struct lfsck_layout_req *llr,
2448 struct lfsck_thread_info *info = lfsck_env_info(env);
2449 struct dt_allocation_hint *hint = &info->lti_hint;
2450 struct dt_object_format *dof = &info->lti_dof;
2451 struct dt_device *pdev = com->lc_lfsck->li_next;
2452 struct ost_id *oi = &info->lti_oi;
2453 struct dt_object *parent = llr->llr_parent->llo_obj;
2454 struct dt_device *cdev = lfsck_obj2dt_dev(llr->llr_child);
2455 struct dt_object *child = NULL;
2456 struct lu_device *d = &cdev->dd_lu_dev;
2457 struct lu_object *o = NULL;
2458 struct thandle *handle;
2459 struct lov_mds_md_v1 *lmm;
2460 struct lov_ost_data_v1 *objs;
2461 struct lustre_handle lh = { 0 };
2466 CDEBUG(D_LFSCK, "Repair multiple references for: parent "DFID
2467 ", OST-index %u, stripe-index %u, owner %u:%u\n",
2468 PFID(lfsck_dto2fid(parent)), llr->llr_ost_idx,
2469 llr->llr_lov_idx, la->la_uid, la->la_gid);
2471 rc = lfsck_layout_lock(env, com, parent, &lh,
2472 MDS_INODELOCK_LAYOUT | MDS_INODELOCK_XATTR);
2476 handle = dt_trans_create(env, pdev);
2478 GOTO(unlock1, rc = PTR_ERR(handle));
2480 o = lu_object_anon(env, d, NULL);
2482 GOTO(stop, rc = PTR_ERR(o));
2484 child = container_of(o, struct dt_object, do_lu);
2485 o = lu_object_locate(o->lo_header, d->ld_type);
2486 if (unlikely(o == NULL))
2487 GOTO(stop, rc = -EINVAL);
2489 child = container_of(o, struct dt_object, do_lu);
2490 la->la_valid = LA_UID | LA_GID;
2491 hint->dah_parent = NULL;
2493 dof->dof_type = DFT_REGULAR;
2494 rc = dt_declare_create(env, child, la, NULL, NULL, handle);
2498 rc = dt_declare_xattr_set(env, parent, buf, XATTR_NAME_LOV,
2499 LU_XATTR_REPLACE, handle);
2503 rc = dt_trans_start(env, pdev, handle);
2507 dt_write_lock(env, parent, 0);
2508 if (unlikely(lu_object_is_dying(parent->do_lu.lo_header)))
2509 GOTO(unlock2, rc = 0);
2511 rc = dt_xattr_get(env, parent, buf, XATTR_NAME_LOV, BYPASS_CAPA);
2512 if (unlikely(rc == 0 || rc == -ENODATA || rc == -ERANGE))
2513 GOTO(unlock2, rc = 0);
2516 rc = lfsck_layout_verify_header(lmm);
2520 /* Someone change layout during the LFSCK, no need to repair then. */
2521 if (le16_to_cpu(lmm->lmm_layout_gen) != llr->llr_parent->llo_gen)
2522 GOTO(unlock2, rc = 0);
2524 rc = dt_create(env, child, la, hint, dof, handle);
2528 /* Currently, we only support LOV_MAGIC_V1/LOV_MAGIC_V3 which has
2529 * been verified in lfsck_layout_verify_header() already. If some
2530 * new magic introduced in the future, then layout LFSCK needs to
2531 * be updated also. */
2532 magic = le32_to_cpu(lmm->lmm_magic);
2533 if (magic == LOV_MAGIC_V1) {
2534 objs = &(lmm->lmm_objects[0]);
2536 LASSERT(magic == LOV_MAGIC_V3);
2537 objs = &((struct lov_mds_md_v3 *)lmm)->lmm_objects[0];
2540 lmm->lmm_layout_gen = cpu_to_le16(llr->llr_parent->llo_gen + 1);
2541 fid_to_ostid(lu_object_fid(&child->do_lu), oi);
2542 ostid_cpu_to_le(oi, &objs[llr->llr_lov_idx].l_ost_oi);
2543 objs[llr->llr_lov_idx].l_ost_gen = cpu_to_le32(0);
2544 objs[llr->llr_lov_idx].l_ost_idx = cpu_to_le32(llr->llr_ost_idx);
2545 rc = dt_xattr_set(env, parent, buf, XATTR_NAME_LOV,
2546 LU_XATTR_REPLACE, handle, BYPASS_CAPA);
2548 GOTO(unlock2, rc = (rc == 0 ? 1 : rc));
2551 dt_write_unlock(env, parent);
2555 lu_object_put(env, &child->do_lu);
2557 dt_trans_stop(env, pdev, handle);
2560 lfsck_layout_unlock(&lh);
2565 /* If the MDT-object and the OST-object have different owner information,
2566 * then trust the MDT-object, because the normal chown/chgrp handle order
2567 * is from MDT to OST, and it is possible that some chown/chgrp operation
2568 * is partly done. */
2569 static int lfsck_layout_repair_owner(const struct lu_env *env,
2570 struct lfsck_component *com,
2571 struct lfsck_layout_req *llr,
2572 struct lu_attr *pla)
2574 struct lfsck_thread_info *info = lfsck_env_info(env);
2575 struct lu_attr *tla = &info->lti_la3;
2576 struct dt_object *parent = llr->llr_parent->llo_obj;
2577 struct dt_object *child = llr->llr_child;
2578 struct dt_device *dev = lfsck_obj2dt_dev(child);
2579 struct thandle *handle;
2583 CDEBUG(D_LFSCK, "Repair inconsistent file owner for: parent "DFID
2584 ", child "DFID", OST-index %u, stripe-index %u, owner %u:%u\n",
2585 PFID(lfsck_dto2fid(parent)), PFID(lfsck_dto2fid(child)),
2586 llr->llr_ost_idx, llr->llr_lov_idx, pla->la_uid, pla->la_gid);
2588 handle = dt_trans_create(env, dev);
2590 RETURN(PTR_ERR(handle));
2592 tla->la_uid = pla->la_uid;
2593 tla->la_gid = pla->la_gid;
2594 tla->la_valid = LA_UID | LA_GID;
2595 rc = dt_declare_attr_set(env, child, tla, handle);
2599 rc = dt_trans_start(env, dev, handle);
2603 /* Use the dt_object lock to serialize with destroy and attr_set. */
2604 dt_read_lock(env, parent, 0);
2605 if (unlikely(lu_object_is_dying(parent->do_lu.lo_header)))
2606 GOTO(unlock, rc = 1);
2608 /* Get the latest parent's owner. */
2609 rc = dt_attr_get(env, parent, tla, BYPASS_CAPA);
2611 CWARN("%s: fail to get the latest parent's ("DFID") owner, "
2612 "not sure whether some others chown/chgrp during the "
2613 "LFSCK: rc = %d\n", lfsck_lfsck2name(com->lc_lfsck),
2614 PFID(lfsck_dto2fid(parent)), rc);
2619 /* Some others chown/chgrp during the LFSCK, needs to do nothing. */
2620 if (unlikely(tla->la_uid != pla->la_uid ||
2621 tla->la_gid != pla->la_gid))
2622 GOTO(unlock, rc = 1);
2624 tla->la_valid = LA_UID | LA_GID;
2625 rc = dt_attr_set(env, child, tla, handle, BYPASS_CAPA);
2630 dt_read_unlock(env, parent);
2633 rc = lfsck_layout_trans_stop(env, dev, handle, rc);
2638 /* Check whether the OST-object correctly back points to the
2639 * MDT-object (@parent) via the XATTR_NAME_FID xattr (@pfid). */
2640 static int lfsck_layout_check_parent(const struct lu_env *env,
2641 struct lfsck_component *com,
2642 struct dt_object *parent,
2643 const struct lu_fid *pfid,
2644 const struct lu_fid *cfid,
2645 const struct lu_attr *pla,
2646 const struct lu_attr *cla,
2647 struct lfsck_layout_req *llr,
2648 struct lu_buf *lov_ea, __u32 idx)
2650 struct lfsck_thread_info *info = lfsck_env_info(env);
2651 struct lu_buf *buf = &info->lti_big_buf;
2652 struct dt_object *tobj;
2653 struct lov_mds_md_v1 *lmm;
2654 struct lov_ost_data_v1 *objs;
2661 if (fid_is_zero(pfid)) {
2662 /* client never wrote. */
2663 if (cla->la_size == 0 && cla->la_blocks == 0) {
2664 if (unlikely(cla->la_uid != pla->la_uid ||
2665 cla->la_gid != pla->la_gid))
2666 RETURN (LLIT_INCONSISTENT_OWNER);
2671 RETURN(LLIT_UNMATCHED_PAIR);
2674 if (unlikely(!fid_is_sane(pfid)))
2675 RETURN(LLIT_UNMATCHED_PAIR);
2677 if (lu_fid_eq(pfid, lu_object_fid(&parent->do_lu))) {
2678 if (llr->llr_lov_idx == idx)
2681 RETURN(LLIT_UNMATCHED_PAIR);
2684 tobj = lfsck_object_find(env, com->lc_lfsck, pfid);
2686 RETURN(LLIT_UNMATCHED_PAIR);
2689 RETURN(PTR_ERR(tobj));
2691 if (!dt_object_exists(tobj))
2692 GOTO(out, rc = LLIT_UNMATCHED_PAIR);
2694 /* Load the tobj's layout EA, in spite of it is a local MDT-object or
2695 * remote one on another MDT. Then check whether the given OST-object
2696 * is in such layout. If yes, it is multiple referenced, otherwise it
2697 * is unmatched referenced case. */
2698 rc = lfsck_layout_get_lovea(env, tobj, buf, NULL);
2700 GOTO(out, rc = LLIT_UNMATCHED_PAIR);
2706 rc = lfsck_layout_verify_header(lmm);
2710 /* Currently, we only support LOV_MAGIC_V1/LOV_MAGIC_V3 which has
2711 * been verified in lfsck_layout_verify_header() already. If some
2712 * new magic introduced in the future, then layout LFSCK needs to
2713 * be updated also. */
2714 magic = le32_to_cpu(lmm->lmm_magic);
2715 if (magic == LOV_MAGIC_V1) {
2716 objs = &(lmm->lmm_objects[0]);
2718 LASSERT(magic == LOV_MAGIC_V3);
2719 objs = &((struct lov_mds_md_v3 *)lmm)->lmm_objects[0];
2722 count = le16_to_cpu(lmm->lmm_stripe_count);
2723 for (i = 0; i < count; i++, objs++) {
2724 struct lu_fid *tfid = &info->lti_fid2;
2725 struct ost_id *oi = &info->lti_oi;
2727 if (is_dummy_lov_ost_data(objs))
2730 ostid_le_to_cpu(&objs->l_ost_oi, oi);
2731 ostid_to_fid(tfid, oi, le32_to_cpu(objs->l_ost_idx));
2732 if (lu_fid_eq(cfid, tfid)) {
2735 GOTO(out, rc = LLIT_MULTIPLE_REFERENCED);
2739 GOTO(out, rc = LLIT_UNMATCHED_PAIR);
2742 lfsck_object_put(env, tobj);
2747 static int lfsck_layout_assistant_handle_one(const struct lu_env *env,
2748 struct lfsck_component *com,
2749 struct lfsck_layout_req *llr)
2751 struct lfsck_layout *lo = com->lc_file_ram;
2752 struct lfsck_thread_info *info = lfsck_env_info(env);
2753 struct filter_fid_old *pea = &info->lti_old_pfid;
2754 struct lu_fid *pfid = &info->lti_fid;
2755 struct lu_buf *buf = NULL;
2756 struct dt_object *parent = llr->llr_parent->llo_obj;
2757 struct dt_object *child = llr->llr_child;
2758 struct lu_attr *pla = &info->lti_la;
2759 struct lu_attr *cla = &info->lti_la2;
2760 struct lfsck_instance *lfsck = com->lc_lfsck;
2761 struct lfsck_bookmark *bk = &lfsck->li_bookmark_ram;
2762 enum lfsck_layout_inconsistency_type type = LLIT_NONE;
2767 rc = dt_attr_get(env, parent, pla, BYPASS_CAPA);
2769 if (lu_object_is_dying(parent->do_lu.lo_header))
2775 rc = dt_attr_get(env, child, cla, BYPASS_CAPA);
2776 if (rc == -ENOENT) {
2777 if (lu_object_is_dying(parent->do_lu.lo_header))
2780 type = LLIT_DANGLING;
2787 buf = lfsck_buf_get(env, pea, sizeof(struct filter_fid_old));
2788 rc= dt_xattr_get(env, child, buf, XATTR_NAME_FID, BYPASS_CAPA);
2789 if (unlikely(rc >= 0 && rc != sizeof(struct filter_fid_old) &&
2790 rc != sizeof(struct filter_fid))) {
2791 type = LLIT_UNMATCHED_PAIR;
2795 if (rc < 0 && rc != -ENODATA)
2798 if (rc == -ENODATA) {
2801 fid_le_to_cpu(pfid, &pea->ff_parent);
2802 /* OST-object does not save parent FID::f_ver, instead,
2803 * the OST-object index in the parent MDT-object layout
2804 * EA reuses the pfid->f_ver. */
2809 rc = lfsck_layout_check_parent(env, com, parent, pfid,
2810 lu_object_fid(&child->do_lu),
2811 pla, cla, llr, buf, idx);
2820 if (unlikely(cla->la_uid != pla->la_uid ||
2821 cla->la_gid != pla->la_gid)) {
2822 type = LLIT_INCONSISTENT_OWNER;
2827 if (bk->lb_param & LPF_DRYRUN) {
2828 if (type != LLIT_NONE)
2836 memset(cla, 0, sizeof(*cla));
2837 cla->la_uid = pla->la_uid;
2838 cla->la_gid = pla->la_gid;
2839 cla->la_mode = S_IFREG | 0666;
2840 cla->la_valid = LA_TYPE | LA_MODE | LA_UID | LA_GID |
2841 LA_ATIME | LA_MTIME | LA_CTIME;
2842 rc = lfsck_layout_recreate_ostobj(env, com, llr, cla);
2844 case LLIT_UNMATCHED_PAIR:
2845 rc = lfsck_layout_repair_unmatched_pair(env, com, llr, pla);
2847 case LLIT_MULTIPLE_REFERENCED:
2848 rc = lfsck_layout_repair_multiple_references(env, com, llr,
2851 case LLIT_INCONSISTENT_OWNER:
2852 rc = lfsck_layout_repair_owner(env, com, llr, pla);
2862 down_write(&com->lc_sem);
2864 /* If cannot touch the target server,
2865 * mark the LFSCK as INCOMPLETE. */
2866 if (rc == -ENOTCONN || rc == -ESHUTDOWN || rc == -ETIMEDOUT ||
2867 rc == -EHOSTDOWN || rc == -EHOSTUNREACH) {
2868 CERROR("%s: Fail to talk with OST %x: rc = %d.\n",
2869 lfsck_lfsck2name(lfsck), llr->llr_ost_idx, rc);
2870 lo->ll_flags |= LF_INCOMPLETE;
2871 lo->ll_objs_skipped++;
2874 lo->ll_objs_failed_phase1++;
2876 } else if (rc > 0) {
2877 LASSERTF(type > LLIT_NONE && type <= LLIT_MAX,
2878 "unknown type = %d\n", type);
2880 lo->ll_objs_repaired[type - 1]++;
2882 up_write(&com->lc_sem);
2887 static int lfsck_layout_assistant(void *args)
2889 struct lfsck_thread_args *lta = args;
2890 struct lu_env *env = <a->lta_env;
2891 struct lfsck_component *com = lta->lta_com;
2892 struct lfsck_instance *lfsck = lta->lta_lfsck;
2893 struct lfsck_bookmark *bk = &lfsck->li_bookmark_ram;
2894 struct lfsck_position *pos = &com->lc_pos_start;
2895 struct lfsck_thread_info *info = lfsck_env_info(env);
2896 struct lfsck_request *lr = &info->lti_lr;
2897 struct lfsck_layout_master_data *llmd = com->lc_data;
2898 struct ptlrpc_thread *mthread = &lfsck->li_thread;
2899 struct ptlrpc_thread *athread = &llmd->llmd_thread;
2900 struct lfsck_layout_req *llr;
2901 struct l_wait_info lwi = { 0 };
2906 memset(lr, 0, sizeof(*lr));
2907 lr->lr_event = LE_START;
2908 lr->lr_valid = LSV_SPEED_LIMIT | LSV_ERROR_HANDLE | LSV_DRYRUN |
2910 lr->lr_speed = bk->lb_speed_limit;
2911 lr->lr_version = bk->lb_version;
2912 lr->lr_param = bk->lb_param;
2913 lr->lr_async_windows = bk->lb_async_windows;
2914 lr->lr_flags = LEF_TO_OST;
2915 if (pos->lp_oit_cookie <= 1)
2916 lr->lr_param |= LPF_RESET;
2918 rc = lfsck_layout_master_notify_others(env, com, lr);
2920 CERROR("%s: fail to notify others for layout start: rc = %d\n",
2921 lfsck_lfsck2name(lfsck), rc);
2925 spin_lock(&llmd->llmd_lock);
2926 thread_set_flags(athread, SVC_RUNNING);
2927 spin_unlock(&llmd->llmd_lock);
2928 wake_up_all(&mthread->t_ctl_waitq);
2931 while (!list_empty(&llmd->llmd_req_list)) {
2932 bool wakeup = false;
2934 if (unlikely(llmd->llmd_exit))
2935 GOTO(cleanup1, rc = llmd->llmd_post_result);
2937 llr = list_entry(llmd->llmd_req_list.next,
2938 struct lfsck_layout_req,
2940 /* Only the lfsck_layout_assistant thread itself can
2941 * remove the "llr" from the head of the list, LFSCK
2942 * engine thread only inserts other new "lld" at the
2943 * end of the list. So it is safe to handle current
2944 * "llr" without the spin_lock. */
2945 rc = lfsck_layout_assistant_handle_one(env, com, llr);
2946 spin_lock(&llmd->llmd_lock);
2947 list_del_init(&llr->llr_list);
2948 if (bk->lb_async_windows != 0 &&
2949 llmd->llmd_prefetched >= bk->lb_async_windows)
2952 llmd->llmd_prefetched--;
2953 spin_unlock(&llmd->llmd_lock);
2955 wake_up_all(&mthread->t_ctl_waitq);
2957 lfsck_layout_req_fini(env, llr);
2958 if (rc < 0 && bk->lb_param & LPF_FAILOUT)
2962 /* Wakeup the master engine if it is waiting in checkpoint. */
2963 wake_up_all(&mthread->t_ctl_waitq);
2965 l_wait_event(athread->t_ctl_waitq,
2966 !lfsck_layout_req_empty(llmd) ||
2968 llmd->llmd_to_post ||
2969 llmd->llmd_to_double_scan,
2972 if (unlikely(llmd->llmd_exit))
2973 GOTO(cleanup1, rc = llmd->llmd_post_result);
2975 if (!list_empty(&llmd->llmd_req_list))
2978 if (llmd->llmd_to_post) {
2979 llmd->llmd_to_post = 0;
2980 LASSERT(llmd->llmd_post_result > 0);
2982 memset(lr, 0, sizeof(*lr));
2983 lr->lr_event = LE_PHASE1_DONE;
2984 lr->lr_status = llmd->llmd_post_result;
2985 rc = lfsck_layout_master_notify_others(env, com, lr);
2987 CERROR("%s: failed to notify others "
2988 "for layout post: rc = %d\n",
2989 lfsck_lfsck2name(lfsck), rc);
2991 /* Wakeup the master engine to go ahead. */
2992 wake_up_all(&mthread->t_ctl_waitq);
2995 if (llmd->llmd_to_double_scan) {
2996 llmd->llmd_to_double_scan = 0;
2997 atomic_inc(&lfsck->li_double_scan_count);
2998 llmd->llmd_in_double_scan = 1;
2999 wake_up_all(&mthread->t_ctl_waitq);
3001 com->lc_new_checked = 0;
3002 com->lc_new_scanned = 0;
3003 com->lc_time_last_checkpoint = cfs_time_current();
3004 com->lc_time_next_checkpoint =
3005 com->lc_time_last_checkpoint +
3006 cfs_time_seconds(LFSCK_CHECKPOINT_INTERVAL);
3008 while (llmd->llmd_in_double_scan) {
3009 struct lfsck_tgt_descs *ltds =
3010 &lfsck->li_ost_descs;
3011 struct lfsck_tgt_desc *ltd;
3013 rc = lfsck_layout_master_query_others(env, com);
3014 if (lfsck_layout_master_to_orphan(llmd))
3020 /* Pull LFSCK status on related targets once
3021 * per 30 seconds if we are not notified. */
3022 lwi = LWI_TIMEOUT_INTERVAL(cfs_time_seconds(30),
3023 cfs_time_seconds(1),
3025 rc = l_wait_event(athread->t_ctl_waitq,
3026 lfsck_layout_master_to_orphan(llmd) ||
3028 !thread_is_running(mthread),
3031 if (unlikely(llmd->llmd_exit ||
3032 !thread_is_running(mthread)))
3033 GOTO(cleanup2, rc = 0);
3035 if (rc == -ETIMEDOUT)
3042 spin_lock(<ds->ltd_lock);
3044 &llmd->llmd_ost_phase2_list)) {
3046 llmd->llmd_ost_phase2_list.next,
3047 struct lfsck_tgt_desc,
3048 ltd_layout_phase_list);
3050 <d->ltd_layout_phase_list);
3051 spin_unlock(<ds->ltd_lock);
3053 if (bk->lb_param & LPF_ALL_TGT) {
3054 rc = lfsck_layout_scan_orphan(
3057 bk->lb_param & LPF_FAILOUT)
3061 if (unlikely(llmd->llmd_exit ||
3062 !thread_is_running(mthread)))
3063 GOTO(cleanup2, rc = 0);
3065 spin_lock(<ds->ltd_lock);
3068 if (list_empty(&llmd->llmd_ost_phase1_list)) {
3069 spin_unlock(<ds->ltd_lock);
3070 GOTO(cleanup2, rc = 1);
3072 spin_unlock(<ds->ltd_lock);
3078 /* Cleanup the unfinished requests. */
3079 spin_lock(&llmd->llmd_lock);
3081 llmd->llmd_assistant_status = rc;
3083 while (!list_empty(&llmd->llmd_req_list)) {
3084 llr = list_entry(llmd->llmd_req_list.next,
3085 struct lfsck_layout_req,
3087 list_del_init(&llr->llr_list);
3088 llmd->llmd_prefetched--;
3089 spin_unlock(&llmd->llmd_lock);
3090 lfsck_layout_req_fini(env, llr);
3091 spin_lock(&llmd->llmd_lock);
3093 spin_unlock(&llmd->llmd_lock);
3095 LASSERTF(llmd->llmd_prefetched == 0, "unmatched prefeteched objs %d\n",
3096 llmd->llmd_prefetched);
3099 memset(lr, 0, sizeof(*lr));
3101 lr->lr_event = LE_PHASE2_DONE;
3103 } else if (rc == 0) {
3104 if (lfsck->li_flags & LPF_ALL_TGT) {
3105 lr->lr_event = LE_STOP;
3106 lr->lr_status = LS_STOPPED;
3108 lr->lr_event = LE_PEER_EXIT;
3109 switch (lfsck->li_status) {
3112 lr->lr_status = LS_CO_PAUSED;
3116 lr->lr_status = LS_CO_STOPPED;
3119 CERROR("%s: unknown status: rc = %d\n",
3120 lfsck_lfsck2name(lfsck),
3122 lr->lr_status = LS_CO_FAILED;
3127 if (lfsck->li_flags & LPF_ALL_TGT) {
3128 lr->lr_event = LE_STOP;
3129 lr->lr_status = LS_FAILED;
3131 lr->lr_event = LE_PEER_EXIT;
3132 lr->lr_status = LS_CO_FAILED;
3136 rc1 = lfsck_layout_master_notify_others(env, com, lr);
3138 CERROR("%s: failed to notify others for layout quit: rc = %d\n",
3139 lfsck_lfsck2name(lfsck), rc1);
3143 /* Under force exit case, some requests may be just freed without
3144 * verification, those objects should be re-handled when next run.
3145 * So not update the on-disk tracing file under such case. */
3146 if (!llmd->llmd_exit)
3147 rc1 = lfsck_layout_double_scan_result(env, com, rc);
3150 if (llmd->llmd_in_double_scan)
3151 atomic_dec(&lfsck->li_double_scan_count);
3153 spin_lock(&llmd->llmd_lock);
3154 llmd->llmd_assistant_status = (rc1 != 0 ? rc1 : rc);
3155 thread_set_flags(athread, SVC_STOPPED);
3156 wake_up_all(&mthread->t_ctl_waitq);
3157 spin_unlock(&llmd->llmd_lock);
3158 lfsck_thread_args_fini(lta);
3164 lfsck_layout_slave_async_interpret(const struct lu_env *env,
3165 struct ptlrpc_request *req,
3168 struct lfsck_layout_slave_async_args *llsaa = args;
3169 struct obd_export *exp = llsaa->llsaa_exp;
3170 struct lfsck_component *com = llsaa->llsaa_com;
3171 struct lfsck_layout_slave_target *llst = llsaa->llsaa_llst;
3172 struct lfsck_layout_slave_data *llsd = com->lc_data;
3176 /* It is quite probably caused by target crash,
3177 * to make the LFSCK can go ahead, assume that
3178 * the target finished the LFSCK prcoessing. */
3181 struct lfsck_reply *lr;
3183 lr = req_capsule_server_get(&req->rq_pill, &RMF_LFSCK_REPLY);
3184 if (lr->lr_status != LS_SCANNING_PHASE1 &&
3185 lr->lr_status != LS_SCANNING_PHASE2)
3189 lfsck_layout_llst_del(llsd, llst);
3190 lfsck_layout_llst_put(llst);
3191 lfsck_component_put(env, com);
3192 class_export_put(exp);
3197 static int lfsck_layout_async_query(const struct lu_env *env,
3198 struct lfsck_component *com,
3199 struct obd_export *exp,
3200 struct lfsck_layout_slave_target *llst,
3201 struct lfsck_request *lr,
3202 struct ptlrpc_request_set *set)
3204 struct lfsck_layout_slave_async_args *llsaa;
3205 struct ptlrpc_request *req;
3206 struct lfsck_request *tmp;
3210 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_LFSCK_QUERY);
3214 rc = ptlrpc_request_pack(req, LUSTRE_OBD_VERSION, LFSCK_QUERY);
3216 ptlrpc_request_free(req);
3220 tmp = req_capsule_client_get(&req->rq_pill, &RMF_LFSCK_REQUEST);
3222 ptlrpc_request_set_replen(req);
3224 llsaa = ptlrpc_req_async_args(req);
3225 llsaa->llsaa_exp = exp;
3226 llsaa->llsaa_com = lfsck_component_get(com);
3227 llsaa->llsaa_llst = llst;
3228 req->rq_interpret_reply = lfsck_layout_slave_async_interpret;
3229 ptlrpc_set_add_req(set, req);
3234 static int lfsck_layout_async_notify(const struct lu_env *env,
3235 struct obd_export *exp,
3236 struct lfsck_request *lr,
3237 struct ptlrpc_request_set *set)
3239 struct ptlrpc_request *req;
3240 struct lfsck_request *tmp;
3244 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_LFSCK_NOTIFY);
3248 rc = ptlrpc_request_pack(req, LUSTRE_OBD_VERSION, LFSCK_NOTIFY);
3250 ptlrpc_request_free(req);
3254 tmp = req_capsule_client_get(&req->rq_pill, &RMF_LFSCK_REQUEST);
3256 ptlrpc_request_set_replen(req);
3257 ptlrpc_set_add_req(set, req);
3263 lfsck_layout_slave_query_master(const struct lu_env *env,
3264 struct lfsck_component *com)
3266 struct lfsck_request *lr = &lfsck_env_info(env)->lti_lr;
3267 struct lfsck_instance *lfsck = com->lc_lfsck;
3268 struct lfsck_layout_slave_data *llsd = com->lc_data;
3269 struct lfsck_layout_slave_target *llst;
3270 struct obd_export *exp;
3271 struct ptlrpc_request_set *set;
3276 set = ptlrpc_prep_set();
3280 memset(lr, 0, sizeof(*lr));
3281 lr->lr_index = lfsck_dev_idx(lfsck->li_bottom);
3282 lr->lr_event = LE_QUERY;
3283 lr->lr_active = LT_LAYOUT;
3285 llsd->llsd_touch_gen++;
3286 spin_lock(&llsd->llsd_lock);
3287 while (!list_empty(&llsd->llsd_master_list)) {
3288 llst = list_entry(llsd->llsd_master_list.next,
3289 struct lfsck_layout_slave_target,
3291 if (llst->llst_gen == llsd->llsd_touch_gen)
3294 llst->llst_gen = llsd->llsd_touch_gen;
3295 list_del(&llst->llst_list);
3296 list_add_tail(&llst->llst_list,
3297 &llsd->llsd_master_list);
3298 atomic_inc(&llst->llst_ref);
3299 spin_unlock(&llsd->llsd_lock);
3301 exp = lustre_find_lwp_by_index(lfsck->li_obd->obd_name,
3304 lfsck_layout_llst_del(llsd, llst);
3305 lfsck_layout_llst_put(llst);
3306 spin_lock(&llsd->llsd_lock);
3310 rc = lfsck_layout_async_query(env, com, exp, llst, lr, set);
3312 CERROR("%s: slave fail to query %s for layout: "
3313 "rc = %d\n", lfsck_lfsck2name(lfsck),
3314 exp->exp_obd->obd_name, rc);
3316 lfsck_layout_llst_put(llst);
3317 class_export_put(exp);
3319 spin_lock(&llsd->llsd_lock);
3321 spin_unlock(&llsd->llsd_lock);
3323 rc = ptlrpc_set_wait(set);
3324 ptlrpc_set_destroy(set);
3326 RETURN(rc1 != 0 ? rc1 : rc);
3330 lfsck_layout_slave_notify_master(const struct lu_env *env,
3331 struct lfsck_component *com,
3332 enum lfsck_events event, int result)
3334 struct lfsck_instance *lfsck = com->lc_lfsck;
3335 struct lfsck_layout_slave_data *llsd = com->lc_data;
3336 struct lfsck_request *lr = &lfsck_env_info(env)->lti_lr;
3337 struct lfsck_layout_slave_target *llst;
3338 struct obd_export *exp;
3339 struct ptlrpc_request_set *set;
3343 set = ptlrpc_prep_set();
3347 memset(lr, 0, sizeof(*lr));
3348 lr->lr_event = event;
3349 lr->lr_flags = LEF_FROM_OST;
3350 lr->lr_status = result;
3351 lr->lr_index = lfsck_dev_idx(lfsck->li_bottom);
3352 lr->lr_active = LT_LAYOUT;
3353 llsd->llsd_touch_gen++;
3354 spin_lock(&llsd->llsd_lock);
3355 while (!list_empty(&llsd->llsd_master_list)) {
3356 llst = list_entry(llsd->llsd_master_list.next,
3357 struct lfsck_layout_slave_target,
3359 if (llst->llst_gen == llsd->llsd_touch_gen)
3362 llst->llst_gen = llsd->llsd_touch_gen;
3363 list_del(&llst->llst_list);
3364 list_add_tail(&llst->llst_list,
3365 &llsd->llsd_master_list);
3366 atomic_inc(&llst->llst_ref);
3367 spin_unlock(&llsd->llsd_lock);
3369 exp = lustre_find_lwp_by_index(lfsck->li_obd->obd_name,
3372 lfsck_layout_llst_del(llsd, llst);
3373 lfsck_layout_llst_put(llst);
3374 spin_lock(&llsd->llsd_lock);
3378 rc = lfsck_layout_async_notify(env, exp, lr, set);
3380 CERROR("%s: slave fail to notify %s for layout: "
3381 "rc = %d\n", lfsck_lfsck2name(lfsck),
3382 exp->exp_obd->obd_name, rc);
3383 lfsck_layout_llst_put(llst);
3384 class_export_put(exp);
3385 spin_lock(&llsd->llsd_lock);
3387 spin_unlock(&llsd->llsd_lock);
3389 ptlrpc_set_wait(set);
3390 ptlrpc_set_destroy(set);
3397 static int lfsck_layout_reset(const struct lu_env *env,
3398 struct lfsck_component *com, bool init)
3400 struct lfsck_layout *lo = com->lc_file_ram;
3403 down_write(&com->lc_sem);
3405 memset(lo, 0, com->lc_file_size);
3407 __u32 count = lo->ll_success_count;
3408 __u64 last_time = lo->ll_time_last_complete;
3410 memset(lo, 0, com->lc_file_size);
3411 lo->ll_success_count = count;
3412 lo->ll_time_last_complete = last_time;
3415 lo->ll_magic = LFSCK_LAYOUT_MAGIC;
3416 lo->ll_status = LS_INIT;
3418 rc = lfsck_layout_store(env, com);
3419 up_write(&com->lc_sem);
3424 static void lfsck_layout_fail(const struct lu_env *env,
3425 struct lfsck_component *com, bool new_checked)
3427 struct lfsck_layout *lo = com->lc_file_ram;
3429 down_write(&com->lc_sem);
3431 com->lc_new_checked++;
3432 lo->ll_objs_failed_phase1++;
3433 if (lo->ll_pos_first_inconsistent == 0) {
3434 struct lfsck_instance *lfsck = com->lc_lfsck;
3436 lo->ll_pos_first_inconsistent =
3437 lfsck->li_obj_oit->do_index_ops->dio_it.store(env,
3440 up_write(&com->lc_sem);
3443 static int lfsck_layout_master_checkpoint(const struct lu_env *env,
3444 struct lfsck_component *com, bool init)
3446 struct lfsck_instance *lfsck = com->lc_lfsck;
3447 struct lfsck_layout *lo = com->lc_file_ram;
3448 struct lfsck_layout_master_data *llmd = com->lc_data;
3449 struct ptlrpc_thread *mthread = &lfsck->li_thread;
3450 struct ptlrpc_thread *athread = &llmd->llmd_thread;
3451 struct l_wait_info lwi = { 0 };
3454 if (com->lc_new_checked == 0 && !init)
3457 l_wait_event(mthread->t_ctl_waitq,
3458 list_empty(&llmd->llmd_req_list) ||
3459 !thread_is_running(mthread) ||
3460 thread_is_stopped(athread),
3463 if (!thread_is_running(mthread) || thread_is_stopped(athread))
3466 down_write(&com->lc_sem);
3468 lo->ll_pos_latest_start = lfsck->li_pos_current.lp_oit_cookie;
3470 lo->ll_pos_last_checkpoint =
3471 lfsck->li_pos_current.lp_oit_cookie;
3472 lo->ll_run_time_phase1 += cfs_duration_sec(cfs_time_current() +
3473 HALF_SEC - lfsck->li_time_last_checkpoint);
3474 lo->ll_time_last_checkpoint = cfs_time_current_sec();
3475 lo->ll_objs_checked_phase1 += com->lc_new_checked;
3476 com->lc_new_checked = 0;
3479 rc = lfsck_layout_store(env, com);
3480 up_write(&com->lc_sem);
3485 static int lfsck_layout_slave_checkpoint(const struct lu_env *env,
3486 struct lfsck_component *com, bool init)
3488 struct lfsck_instance *lfsck = com->lc_lfsck;
3489 struct lfsck_layout *lo = com->lc_file_ram;
3492 if (com->lc_new_checked == 0 && !init)
3495 down_write(&com->lc_sem);
3498 lo->ll_pos_latest_start = lfsck->li_pos_current.lp_oit_cookie;
3500 lo->ll_pos_last_checkpoint =
3501 lfsck->li_pos_current.lp_oit_cookie;
3502 lo->ll_run_time_phase1 += cfs_duration_sec(cfs_time_current() +
3503 HALF_SEC - lfsck->li_time_last_checkpoint);
3504 lo->ll_time_last_checkpoint = cfs_time_current_sec();
3505 lo->ll_objs_checked_phase1 += com->lc_new_checked;
3506 com->lc_new_checked = 0;
3509 rc = lfsck_layout_store(env, com);
3511 up_write(&com->lc_sem);
3516 static int lfsck_layout_prep(const struct lu_env *env,
3517 struct lfsck_component *com,
3518 struct lfsck_start *start)
3520 struct lfsck_instance *lfsck = com->lc_lfsck;
3521 struct lfsck_layout *lo = com->lc_file_ram;
3522 struct lfsck_position *pos = &com->lc_pos_start;
3524 fid_zero(&pos->lp_dir_parent);
3525 pos->lp_dir_cookie = 0;
3526 if (lo->ll_status == LS_COMPLETED ||
3527 lo->ll_status == LS_PARTIAL ||
3528 /* To handle orphan, must scan from the beginning. */
3529 (start != NULL && start->ls_flags & LPF_ORPHAN)) {
3532 rc = lfsck_layout_reset(env, com, false);
3537 down_write(&com->lc_sem);
3538 lo->ll_time_latest_start = cfs_time_current_sec();
3539 spin_lock(&lfsck->li_lock);
3540 if (lo->ll_flags & LF_SCANNED_ONCE) {
3541 if (!lfsck->li_drop_dryrun ||
3542 lo->ll_pos_first_inconsistent == 0) {
3543 lo->ll_status = LS_SCANNING_PHASE2;
3544 list_del_init(&com->lc_link);
3545 list_add_tail(&com->lc_link,
3546 &lfsck->li_list_double_scan);
3547 pos->lp_oit_cookie = 0;
3551 lo->ll_status = LS_SCANNING_PHASE1;
3552 lo->ll_run_time_phase1 = 0;
3553 lo->ll_run_time_phase2 = 0;
3554 lo->ll_objs_checked_phase1 = 0;
3555 lo->ll_objs_checked_phase2 = 0;
3556 lo->ll_objs_failed_phase1 = 0;
3557 lo->ll_objs_failed_phase2 = 0;
3558 for (i = 0; i < LLIT_MAX; i++)
3559 lo->ll_objs_repaired[i] = 0;
3561 pos->lp_oit_cookie = lo->ll_pos_first_inconsistent;
3562 fid_zero(&com->lc_fid_latest_scanned_phase2);
3565 lo->ll_status = LS_SCANNING_PHASE1;
3566 if (!lfsck->li_drop_dryrun ||
3567 lo->ll_pos_first_inconsistent == 0)
3568 pos->lp_oit_cookie = lo->ll_pos_last_checkpoint + 1;
3570 pos->lp_oit_cookie = lo->ll_pos_first_inconsistent;
3572 spin_unlock(&lfsck->li_lock);
3573 up_write(&com->lc_sem);
3578 static int lfsck_layout_slave_prep(const struct lu_env *env,
3579 struct lfsck_component *com,
3580 struct lfsck_start_param *lsp)
3582 struct lfsck_layout_slave_data *llsd = com->lc_data;
3583 struct lfsck_start *start = lsp->lsp_start;
3586 rc = lfsck_layout_prep(env, com, start);
3587 if (rc != 0 || !lsp->lsp_index_valid)
3590 rc = lfsck_layout_llst_add(llsd, lsp->lsp_index);
3591 if (rc == 0 && start != NULL && start->ls_flags & LPF_ORPHAN) {
3592 LASSERT(!llsd->llsd_rbtree_valid);
3594 write_lock(&llsd->llsd_rb_lock);
3595 rc = lfsck_rbtree_setup(env, com);
3596 write_unlock(&llsd->llsd_rb_lock);
3602 static int lfsck_layout_master_prep(const struct lu_env *env,
3603 struct lfsck_component *com,
3604 struct lfsck_start_param *lsp)
3606 struct lfsck_instance *lfsck = com->lc_lfsck;
3607 struct lfsck_layout_master_data *llmd = com->lc_data;
3608 struct ptlrpc_thread *mthread = &lfsck->li_thread;
3609 struct ptlrpc_thread *athread = &llmd->llmd_thread;
3610 struct lfsck_thread_args *lta;
3614 rc = lfsck_layout_prep(env, com, lsp->lsp_start);
3618 llmd->llmd_assistant_status = 0;
3619 llmd->llmd_post_result = 0;
3620 llmd->llmd_to_post = 0;
3621 llmd->llmd_to_double_scan = 0;
3622 llmd->llmd_in_double_scan = 0;
3623 llmd->llmd_exit = 0;
3624 thread_set_flags(athread, 0);
3626 lta = lfsck_thread_args_init(lfsck, com, lsp);
3628 RETURN(PTR_ERR(lta));
3630 rc = PTR_ERR(kthread_run(lfsck_layout_assistant, lta, "lfsck_layout"));
3631 if (IS_ERR_VALUE(rc)) {
3632 CERROR("%s: Cannot start LFSCK layout assistant thread: "
3633 "rc = %ld\n", lfsck_lfsck2name(lfsck), rc);
3634 lfsck_thread_args_fini(lta);
3636 struct l_wait_info lwi = { 0 };
3638 l_wait_event(mthread->t_ctl_waitq,
3639 thread_is_running(athread) ||
3640 thread_is_stopped(athread),
3642 if (unlikely(!thread_is_running(athread)))
3643 rc = llmd->llmd_assistant_status;
3651 /* Pre-fetch the attribute for each stripe in the given layout EA. */
3652 static int lfsck_layout_scan_stripes(const struct lu_env *env,
3653 struct lfsck_component *com,
3654 struct dt_object *parent,
3655 struct lov_mds_md_v1 *lmm)
3657 struct lfsck_thread_info *info = lfsck_env_info(env);
3658 struct lfsck_instance *lfsck = com->lc_lfsck;
3659 struct lfsck_bookmark *bk = &lfsck->li_bookmark_ram;
3660 struct lfsck_layout *lo = com->lc_file_ram;
3661 struct lfsck_layout_master_data *llmd = com->lc_data;
3662 struct lfsck_layout_object *llo = NULL;
3663 struct lov_ost_data_v1 *objs;
3664 struct lfsck_tgt_descs *ltds = &lfsck->li_ost_descs;
3665 struct ptlrpc_thread *mthread = &lfsck->li_thread;
3666 struct ptlrpc_thread *athread = &llmd->llmd_thread;
3667 struct l_wait_info lwi = { 0 };
3676 buf = lfsck_buf_get(env, &info->lti_old_pfid,
3677 sizeof(struct filter_fid_old));
3678 count = le16_to_cpu(lmm->lmm_stripe_count);
3679 gen = le16_to_cpu(lmm->lmm_layout_gen);
3680 /* Currently, we only support LOV_MAGIC_V1/LOV_MAGIC_V3 which has
3681 * been verified in lfsck_layout_verify_header() already. If some
3682 * new magic introduced in the future, then layout LFSCK needs to
3683 * be updated also. */
3684 magic = le32_to_cpu(lmm->lmm_magic);
3685 if (magic == LOV_MAGIC_V1) {
3686 objs = &(lmm->lmm_objects[0]);
3688 LASSERT(magic == LOV_MAGIC_V3);
3689 objs = &((struct lov_mds_md_v3 *)lmm)->lmm_objects[0];
3692 for (i = 0; i < count; i++, objs++) {
3693 struct lu_fid *fid = &info->lti_fid;
3694 struct ost_id *oi = &info->lti_oi;
3695 struct lfsck_layout_req *llr;
3696 struct lfsck_tgt_desc *tgt = NULL;
3697 struct dt_object *cobj = NULL;
3699 le32_to_cpu(objs->l_ost_idx);
3700 bool wakeup = false;
3702 if (is_dummy_lov_ost_data(objs))
3705 l_wait_event(mthread->t_ctl_waitq,
3706 bk->lb_async_windows == 0 ||
3707 llmd->llmd_prefetched < bk->lb_async_windows ||
3708 !thread_is_running(mthread) ||
3709 thread_is_stopped(athread),
3712 if (unlikely(!thread_is_running(mthread)) ||
3713 thread_is_stopped(athread))
3716 ostid_le_to_cpu(&objs->l_ost_oi, oi);
3717 ostid_to_fid(fid, oi, index);
3718 tgt = lfsck_tgt_get(ltds, index);
3719 if (unlikely(tgt == NULL)) {
3720 CERROR("%s: Cannot talk with OST %x which did not join "
3721 "the layout LFSCK.\n",
3722 lfsck_lfsck2name(lfsck), index);
3723 lo->ll_flags |= LF_INCOMPLETE;
3727 cobj = lfsck_object_find_by_dev(env, tgt->ltd_tgt, fid);
3733 rc = dt_declare_attr_get(env, cobj, BYPASS_CAPA);
3737 rc = dt_declare_xattr_get(env, cobj, buf, XATTR_NAME_FID,
3743 llo = lfsck_layout_object_init(env, parent, gen);
3750 llr = lfsck_layout_req_init(llo, cobj, index, i);
3757 spin_lock(&llmd->llmd_lock);
3758 if (llmd->llmd_assistant_status < 0) {
3759 spin_unlock(&llmd->llmd_lock);
3760 lfsck_layout_req_fini(env, llr);
3762 RETURN(llmd->llmd_assistant_status);
3765 list_add_tail(&llr->llr_list, &llmd->llmd_req_list);
3766 if (llmd->llmd_prefetched == 0)
3769 llmd->llmd_prefetched++;
3770 spin_unlock(&llmd->llmd_lock);
3772 wake_up_all(&athread->t_ctl_waitq);
3775 down_write(&com->lc_sem);
3776 com->lc_new_checked++;
3778 lo->ll_objs_failed_phase1++;
3779 up_write(&com->lc_sem);
3781 if (cobj != NULL && !IS_ERR(cobj))
3782 lu_object_put(env, &cobj->do_lu);
3784 if (likely(tgt != NULL))
3787 if (rc < 0 && bk->lb_param & LPF_FAILOUT)
3794 if (llo != NULL && !IS_ERR(llo))
3795 lfsck_layout_object_put(env, llo);
3800 /* For the given object, read its layout EA locally. For each stripe, pre-fetch
3801 * the OST-object's attribute and generate an structure lfsck_layout_req on the
3802 * list ::llmd_req_list.
3804 * For each request on above list, the lfsck_layout_assistant thread compares
3805 * the OST side attribute with local attribute, if inconsistent, then repair it.
3807 * All above processing is async mode with pipeline. */
3808 static int lfsck_layout_master_exec_oit(const struct lu_env *env,
3809 struct lfsck_component *com,
3810 struct dt_object *obj)
3812 struct lfsck_thread_info *info = lfsck_env_info(env);
3813 struct ost_id *oi = &info->lti_oi;
3814 struct lfsck_layout *lo = com->lc_file_ram;
3815 struct lfsck_layout_master_data *llmd = com->lc_data;
3816 struct lfsck_instance *lfsck = com->lc_lfsck;
3817 struct lfsck_bookmark *bk = &lfsck->li_bookmark_ram;
3818 struct thandle *handle = NULL;
3819 struct lu_buf *buf = &info->lti_big_buf;
3820 struct lov_mds_md_v1 *lmm = NULL;
3821 struct dt_device *dev = lfsck->li_bottom;
3822 struct lustre_handle lh = { 0 };
3823 ssize_t buflen = buf->lb_len;
3825 bool locked = false;
3826 bool stripe = false;
3829 if (!S_ISREG(lfsck_object_type(obj)))
3832 if (llmd->llmd_assistant_status < 0)
3833 GOTO(out, rc = -ESRCH);
3835 fid_to_lmm_oi(lfsck_dto2fid(obj), oi);
3836 lmm_oi_cpu_to_le(oi, oi);
3837 dt_read_lock(env, obj, 0);
3841 rc = lfsck_layout_get_lovea(env, obj, buf, &buflen);
3847 rc = lfsck_layout_verify_header(lmm);
3851 if (memcmp(oi, &lmm->lmm_oi, sizeof(*oi)) == 0)
3852 GOTO(out, stripe = true);
3854 /* Inconsistent lmm_oi, should be repaired. */
3855 CDEBUG(D_LFSCK, "Repair bad lmm_oi for "DFID"\n",
3856 PFID(lfsck_dto2fid(obj)));
3858 if (bk->lb_param & LPF_DRYRUN) {
3859 down_write(&com->lc_sem);
3860 lo->ll_objs_repaired[LLIT_OTHERS - 1]++;
3861 up_write(&com->lc_sem);
3863 GOTO(out, stripe = true);
3866 if (!lustre_handle_is_used(&lh)) {
3867 dt_read_unlock(env, obj);
3869 buf->lb_len = buflen;
3870 rc = lfsck_layout_lock(env, com, obj, &lh,
3871 MDS_INODELOCK_LAYOUT |
3872 MDS_INODELOCK_XATTR);
3876 handle = dt_trans_create(env, dev);
3878 GOTO(out, rc = PTR_ERR(handle));
3880 rc = dt_declare_xattr_set(env, obj, buf, XATTR_NAME_LOV,
3881 LU_XATTR_REPLACE, handle);
3885 rc = dt_trans_start_local(env, dev, handle);
3889 dt_write_lock(env, obj, 0);
3896 rc = dt_xattr_set(env, obj, buf, XATTR_NAME_LOV,
3897 LU_XATTR_REPLACE, handle, BYPASS_CAPA);
3901 down_write(&com->lc_sem);
3902 lo->ll_objs_repaired[LLIT_OTHERS - 1]++;
3903 up_write(&com->lc_sem);
3905 GOTO(out, stripe = true);
3909 if (lustre_handle_is_used(&lh))
3910 dt_write_unlock(env, obj);
3912 dt_read_unlock(env, obj);
3915 if (handle != NULL && !IS_ERR(handle))
3916 dt_trans_stop(env, dev, handle);
3918 lfsck_layout_unlock(&lh);
3920 rc = lfsck_layout_scan_stripes(env, com, obj, lmm);
3922 down_write(&com->lc_sem);
3923 com->lc_new_checked++;
3925 lo->ll_objs_failed_phase1++;
3926 up_write(&com->lc_sem);
3928 buf->lb_len = buflen;
3933 static int lfsck_layout_slave_exec_oit(const struct lu_env *env,
3934 struct lfsck_component *com,
3935 struct dt_object *obj)
3937 struct lfsck_instance *lfsck = com->lc_lfsck;
3938 struct lfsck_layout *lo = com->lc_file_ram;
3939 const struct lu_fid *fid = lfsck_dto2fid(obj);
3940 struct lfsck_layout_slave_data *llsd = com->lc_data;
3941 struct lfsck_layout_seq *lls;
3947 LASSERT(llsd != NULL);
3949 lfsck_rbtree_update_bitmap(env, com, fid, false);
3951 down_write(&com->lc_sem);
3952 if (fid_is_idif(fid))
3954 else if (!fid_is_norm(fid) ||
3955 !fid_is_for_ostobj(env, lfsck->li_next, obj, fid))
3956 GOTO(unlock, rc = 0);
3959 com->lc_new_checked++;
3961 lls = lfsck_layout_seq_lookup(llsd, seq);
3964 if (unlikely(lls == NULL))
3965 GOTO(unlock, rc = -ENOMEM);
3967 INIT_LIST_HEAD(&lls->lls_list);
3969 rc = lfsck_layout_lastid_load(env, com, lls);
3971 lo->ll_objs_failed_phase1++;
3976 lfsck_layout_seq_insert(llsd, lls);
3979 if (unlikely(fid_is_last_id(fid)))
3980 GOTO(unlock, rc = 0);
3983 if (oid > lls->lls_lastid_known)
3984 lls->lls_lastid_known = oid;
3986 if (oid > lls->lls_lastid) {
3987 if (!(lo->ll_flags & LF_CRASHED_LASTID)) {
3988 /* OFD may create new objects during LFSCK scanning. */
3989 rc = lfsck_layout_lastid_reload(env, com, lls);
3990 if (unlikely(rc != 0))
3991 CWARN("%s: failed to reload LAST_ID for "LPX64
3993 lfsck_lfsck2name(com->lc_lfsck),
3995 if (oid <= lls->lls_lastid)
3996 GOTO(unlock, rc = 0);
3998 LASSERT(lfsck->li_out_notify != NULL);
4000 lfsck->li_out_notify(env, lfsck->li_out_notify_data,
4001 LE_LASTID_REBUILDING);
4002 lo->ll_flags |= LF_CRASHED_LASTID;
4005 lls->lls_lastid = oid;
4009 GOTO(unlock, rc = 0);
4012 up_write(&com->lc_sem);
4017 static int lfsck_layout_exec_dir(const struct lu_env *env,
4018 struct lfsck_component *com,
4019 struct dt_object *obj,
4020 struct lu_dirent *ent)
4025 static int lfsck_layout_master_post(const struct lu_env *env,
4026 struct lfsck_component *com,
4027 int result, bool init)
4029 struct lfsck_instance *lfsck = com->lc_lfsck;
4030 struct lfsck_layout *lo = com->lc_file_ram;
4031 struct lfsck_layout_master_data *llmd = com->lc_data;
4032 struct ptlrpc_thread *mthread = &lfsck->li_thread;
4033 struct ptlrpc_thread *athread = &llmd->llmd_thread;
4034 struct l_wait_info lwi = { 0 };
4039 llmd->llmd_post_result = result;
4040 llmd->llmd_to_post = 1;
4041 if (llmd->llmd_post_result <= 0)
4042 llmd->llmd_exit = 1;
4044 wake_up_all(&athread->t_ctl_waitq);
4045 l_wait_event(mthread->t_ctl_waitq,
4046 (result > 0 && list_empty(&llmd->llmd_req_list)) ||
4047 thread_is_stopped(athread),
4050 if (llmd->llmd_assistant_status < 0)
4051 result = llmd->llmd_assistant_status;
4053 down_write(&com->lc_sem);
4054 spin_lock(&lfsck->li_lock);
4055 /* When LFSCK failed, there may be some prefetched objects those are
4056 * not been processed yet, we do not know the exactly position, then
4057 * just restart from last check-point next time. */
4058 if (!init && !llmd->llmd_exit)
4059 lo->ll_pos_last_checkpoint =
4060 lfsck->li_pos_current.lp_oit_cookie;
4063 lo->ll_status = LS_SCANNING_PHASE2;
4064 lo->ll_flags |= LF_SCANNED_ONCE;
4065 lo->ll_flags &= ~LF_UPGRADE;
4066 list_del_init(&com->lc_link);
4067 list_add_tail(&com->lc_link, &lfsck->li_list_double_scan);
4068 } else if (result == 0) {
4069 lo->ll_status = lfsck->li_status;
4070 if (lo->ll_status == 0)
4071 lo->ll_status = LS_STOPPED;
4072 if (lo->ll_status != LS_PAUSED) {
4073 list_del_init(&com->lc_link);
4074 list_add_tail(&com->lc_link, &lfsck->li_list_idle);
4077 lo->ll_status = LS_FAILED;
4078 list_del_init(&com->lc_link);
4079 list_add_tail(&com->lc_link, &lfsck->li_list_idle);
4081 spin_unlock(&lfsck->li_lock);
4084 lo->ll_run_time_phase1 += cfs_duration_sec(cfs_time_current() +
4085 HALF_SEC - lfsck->li_time_last_checkpoint);
4086 lo->ll_time_last_checkpoint = cfs_time_current_sec();
4087 lo->ll_objs_checked_phase1 += com->lc_new_checked;
4088 com->lc_new_checked = 0;
4091 rc = lfsck_layout_store(env, com);
4092 up_write(&com->lc_sem);
4097 static int lfsck_layout_slave_post(const struct lu_env *env,
4098 struct lfsck_component *com,
4099 int result, bool init)
4101 struct lfsck_instance *lfsck = com->lc_lfsck;
4102 struct lfsck_layout *lo = com->lc_file_ram;
4106 rc = lfsck_layout_lastid_store(env, com);
4110 LASSERT(lfsck->li_out_notify != NULL);
4112 down_write(&com->lc_sem);
4114 spin_lock(&lfsck->li_lock);
4116 lo->ll_pos_last_checkpoint =
4117 lfsck->li_pos_current.lp_oit_cookie;
4119 lo->ll_status = LS_SCANNING_PHASE2;
4120 lo->ll_flags |= LF_SCANNED_ONCE;
4121 if (lo->ll_flags & LF_CRASHED_LASTID) {
4123 lo->ll_flags &= ~LF_CRASHED_LASTID;
4125 lo->ll_flags &= ~LF_UPGRADE;
4126 list_del_init(&com->lc_link);
4127 list_add_tail(&com->lc_link, &lfsck->li_list_double_scan);
4128 } else if (result == 0) {
4129 lo->ll_status = lfsck->li_status;
4130 if (lo->ll_status == 0)
4131 lo->ll_status = LS_STOPPED;
4132 if (lo->ll_status != LS_PAUSED) {
4133 list_del_init(&com->lc_link);
4134 list_add_tail(&com->lc_link, &lfsck->li_list_idle);
4137 lo->ll_status = LS_FAILED;
4138 list_del_init(&com->lc_link);
4139 list_add_tail(&com->lc_link, &lfsck->li_list_idle);
4141 spin_unlock(&lfsck->li_lock);
4144 lfsck->li_out_notify(env, lfsck->li_out_notify_data,
4148 lo->ll_run_time_phase1 += cfs_duration_sec(cfs_time_current() +
4149 HALF_SEC - lfsck->li_time_last_checkpoint);
4150 lo->ll_time_last_checkpoint = cfs_time_current_sec();
4151 lo->ll_objs_checked_phase1 += com->lc_new_checked;
4152 com->lc_new_checked = 0;
4155 rc = lfsck_layout_store(env, com);
4157 up_write(&com->lc_sem);
4159 lfsck_layout_slave_notify_master(env, com, LE_PHASE1_DONE, result);
4162 lfsck_rbtree_cleanup(env, com);
4167 static int lfsck_layout_dump(const struct lu_env *env,
4168 struct lfsck_component *com, char *buf, int len)
4170 struct lfsck_instance *lfsck = com->lc_lfsck;
4171 struct lfsck_bookmark *bk = &lfsck->li_bookmark_ram;
4172 struct lfsck_layout *lo = com->lc_file_ram;
4177 down_read(&com->lc_sem);
4178 rc = snprintf(buf, len,
4179 "name: lfsck_layout\n"
4185 lfsck_status2names(lo->ll_status));
4191 rc = lfsck_bits_dump(&buf, &len, lo->ll_flags, lfsck_flags_names,
4196 rc = lfsck_bits_dump(&buf, &len, bk->lb_param, lfsck_param_names,
4201 rc = lfsck_time_dump(&buf, &len, lo->ll_time_last_complete,
4202 "time_since_last_completed");
4206 rc = lfsck_time_dump(&buf, &len, lo->ll_time_latest_start,
4207 "time_since_latest_start");
4211 rc = lfsck_time_dump(&buf, &len, lo->ll_time_last_checkpoint,
4212 "time_since_last_checkpoint");
4216 rc = snprintf(buf, len,
4217 "latest_start_position: "LPU64"\n"
4218 "last_checkpoint_position: "LPU64"\n"
4219 "first_failure_position: "LPU64"\n",
4220 lo->ll_pos_latest_start,
4221 lo->ll_pos_last_checkpoint,
4222 lo->ll_pos_first_inconsistent);
4229 rc = snprintf(buf, len,
4230 "success_count: %u\n"
4231 "repaired_dangling: "LPU64"\n"
4232 "repaired_unmatched_pair: "LPU64"\n"
4233 "repaired_multiple_referenced: "LPU64"\n"
4234 "repaired_orphan: "LPU64"\n"
4235 "repaired_inconsistent_owner: "LPU64"\n"
4236 "repaired_others: "LPU64"\n"
4237 "skipped: "LPU64"\n"
4238 "failed_phase1: "LPU64"\n"
4239 "failed_phase2: "LPU64"\n",
4240 lo->ll_success_count,
4241 lo->ll_objs_repaired[LLIT_DANGLING - 1],
4242 lo->ll_objs_repaired[LLIT_UNMATCHED_PAIR - 1],
4243 lo->ll_objs_repaired[LLIT_MULTIPLE_REFERENCED - 1],
4244 lo->ll_objs_repaired[LLIT_ORPHAN - 1],
4245 lo->ll_objs_repaired[LLIT_INCONSISTENT_OWNER - 1],
4246 lo->ll_objs_repaired[LLIT_OTHERS - 1],
4247 lo->ll_objs_skipped,
4248 lo->ll_objs_failed_phase1,
4249 lo->ll_objs_failed_phase2);
4256 if (lo->ll_status == LS_SCANNING_PHASE1) {
4258 const struct dt_it_ops *iops;
4259 cfs_duration_t duration = cfs_time_current() -
4260 lfsck->li_time_last_checkpoint;
4261 __u64 checked = lo->ll_objs_checked_phase1 + com->lc_new_checked;
4262 __u64 speed = checked;
4263 __u64 new_checked = com->lc_new_checked * HZ;
4264 __u32 rtime = lo->ll_run_time_phase1 +
4265 cfs_duration_sec(duration + HALF_SEC);
4268 do_div(new_checked, duration);
4270 do_div(speed, rtime);
4271 rc = snprintf(buf, len,
4272 "checked_phase1: "LPU64"\n"
4273 "checked_phase2: "LPU64"\n"
4274 "run_time_phase1: %u seconds\n"
4275 "run_time_phase2: %u seconds\n"
4276 "average_speed_phase1: "LPU64" items/sec\n"
4277 "average_speed_phase2: N/A\n"
4278 "real-time_speed_phase1: "LPU64" items/sec\n"
4279 "real-time_speed_phase2: N/A\n",
4281 lo->ll_objs_checked_phase2,
4283 lo->ll_run_time_phase2,
4292 LASSERT(lfsck->li_di_oit != NULL);
4294 iops = &lfsck->li_obj_oit->do_index_ops->dio_it;
4296 /* The low layer otable-based iteration position may NOT
4297 * exactly match the layout-based directory traversal
4298 * cookie. Generally, it is not a serious issue. But the
4299 * caller should NOT make assumption on that. */
4300 pos = iops->store(env, lfsck->li_di_oit);
4301 if (!lfsck->li_current_oit_processed)
4303 rc = snprintf(buf, len, "current_position: "LPU64"\n", pos);
4309 } else if (lo->ll_status == LS_SCANNING_PHASE2) {
4310 cfs_duration_t duration = cfs_time_current() -
4311 lfsck->li_time_last_checkpoint;
4312 __u64 checked = lo->ll_objs_checked_phase1 + com->lc_new_checked;
4313 __u64 speed = checked;
4314 __u64 new_checked = com->lc_new_checked * HZ;
4315 __u32 rtime = lo->ll_run_time_phase1 +
4316 cfs_duration_sec(duration + HALF_SEC);
4319 do_div(new_checked, duration);
4321 do_div(speed, rtime);
4322 rc = snprintf(buf, len,
4323 "checked_phase1: "LPU64"\n"
4324 "checked_phase2: "LPU64"\n"
4325 "run_time_phase1: %u seconds\n"
4326 "run_time_phase2: %u seconds\n"
4327 "average_speed_phase1: "LPU64" items/sec\n"
4328 "average_speed_phase2: N/A\n"
4329 "real-time_speed_phase1: "LPU64" items/sec\n"
4330 "real-time_speed_phase2: N/A\n"
4331 "current_position: "DFID"\n",
4333 lo->ll_objs_checked_phase2,
4335 lo->ll_run_time_phase2,
4338 PFID(&com->lc_fid_latest_scanned_phase2));
4345 __u64 speed1 = lo->ll_objs_checked_phase1;
4346 __u64 speed2 = lo->ll_objs_checked_phase2;
4348 if (lo->ll_run_time_phase1 != 0)
4349 do_div(speed1, lo->ll_run_time_phase1);
4350 if (lo->ll_run_time_phase2 != 0)
4351 do_div(speed2, lo->ll_run_time_phase2);
4352 rc = snprintf(buf, len,
4353 "checked_phase1: "LPU64"\n"
4354 "checked_phase2: "LPU64"\n"
4355 "run_time_phase1: %u seconds\n"
4356 "run_time_phase2: %u seconds\n"
4357 "average_speed_phase1: "LPU64" items/sec\n"
4358 "average_speed_phase2: "LPU64" objs/sec\n"
4359 "real-time_speed_phase1: N/A\n"
4360 "real-time_speed_phase2: N/A\n"
4361 "current_position: N/A\n",
4362 lo->ll_objs_checked_phase1,
4363 lo->ll_objs_checked_phase2,
4364 lo->ll_run_time_phase1,
4365 lo->ll_run_time_phase2,
4377 up_read(&com->lc_sem);
4382 static int lfsck_layout_master_double_scan(const struct lu_env *env,
4383 struct lfsck_component *com)
4385 struct lfsck_layout_master_data *llmd = com->lc_data;
4386 struct ptlrpc_thread *mthread = &com->lc_lfsck->li_thread;
4387 struct ptlrpc_thread *athread = &llmd->llmd_thread;
4388 struct lfsck_layout *lo = com->lc_file_ram;
4389 struct l_wait_info lwi = { 0 };
4391 if (unlikely(lo->ll_status != LS_SCANNING_PHASE2))
4394 llmd->llmd_to_double_scan = 1;
4395 wake_up_all(&athread->t_ctl_waitq);
4396 l_wait_event(mthread->t_ctl_waitq,
4397 llmd->llmd_in_double_scan ||
4398 thread_is_stopped(athread),
4400 if (llmd->llmd_assistant_status < 0)
4401 return llmd->llmd_assistant_status;
4406 static int lfsck_layout_slave_double_scan(const struct lu_env *env,
4407 struct lfsck_component *com)
4409 struct lfsck_instance *lfsck = com->lc_lfsck;
4410 struct lfsck_layout_slave_data *llsd = com->lc_data;
4411 struct lfsck_layout *lo = com->lc_file_ram;
4412 struct ptlrpc_thread *thread = &lfsck->li_thread;
4416 if (unlikely(lo->ll_status != LS_SCANNING_PHASE2)) {
4417 lfsck_rbtree_cleanup(env, com);
4418 lfsck_layout_slave_notify_master(env, com, LE_PHASE2_DONE, 0);
4422 atomic_inc(&lfsck->li_double_scan_count);
4424 com->lc_new_checked = 0;
4425 com->lc_new_scanned = 0;
4426 com->lc_time_last_checkpoint = cfs_time_current();
4427 com->lc_time_next_checkpoint = com->lc_time_last_checkpoint +
4428 cfs_time_seconds(LFSCK_CHECKPOINT_INTERVAL);
4431 struct l_wait_info lwi = LWI_TIMEOUT(cfs_time_seconds(30),
4434 rc = lfsck_layout_slave_query_master(env, com);
4435 if (list_empty(&llsd->llsd_master_list)) {
4436 if (unlikely(!thread_is_running(thread)))
4447 rc = l_wait_event(thread->t_ctl_waitq,
4448 !thread_is_running(thread) ||
4449 list_empty(&llsd->llsd_master_list),
4451 if (unlikely(!thread_is_running(thread)))
4454 if (rc == -ETIMEDOUT)
4457 GOTO(done, rc = (rc < 0 ? rc : 1));
4461 rc = lfsck_layout_double_scan_result(env, com, rc);
4463 lfsck_rbtree_cleanup(env, com);
4464 lfsck_layout_slave_notify_master(env, com, LE_PHASE2_DONE, rc);
4465 if (atomic_dec_and_test(&lfsck->li_double_scan_count))
4466 wake_up_all(&lfsck->li_thread.t_ctl_waitq);
4471 static void lfsck_layout_master_data_release(const struct lu_env *env,
4472 struct lfsck_component *com)
4474 struct lfsck_layout_master_data *llmd = com->lc_data;
4475 struct lfsck_instance *lfsck = com->lc_lfsck;
4476 struct lfsck_tgt_descs *ltds;
4477 struct lfsck_tgt_desc *ltd;
4478 struct lfsck_tgt_desc *next;
4480 LASSERT(llmd != NULL);
4481 LASSERT(thread_is_init(&llmd->llmd_thread) ||
4482 thread_is_stopped(&llmd->llmd_thread));
4483 LASSERT(list_empty(&llmd->llmd_req_list));
4485 com->lc_data = NULL;
4487 ltds = &lfsck->li_ost_descs;
4488 spin_lock(<ds->ltd_lock);
4489 list_for_each_entry_safe(ltd, next, &llmd->llmd_ost_phase1_list,
4490 ltd_layout_phase_list) {
4491 list_del_init(<d->ltd_layout_phase_list);
4493 list_for_each_entry_safe(ltd, next, &llmd->llmd_ost_phase2_list,
4494 ltd_layout_phase_list) {
4495 list_del_init(<d->ltd_layout_phase_list);
4497 list_for_each_entry_safe(ltd, next, &llmd->llmd_ost_list,
4499 list_del_init(<d->ltd_layout_list);
4501 list_for_each_entry_safe(ltd, next, &llmd->llmd_mdt_phase1_list,
4502 ltd_layout_phase_list) {
4503 list_del_init(<d->ltd_layout_phase_list);
4505 list_for_each_entry_safe(ltd, next, &llmd->llmd_mdt_phase2_list,
4506 ltd_layout_phase_list) {
4507 list_del_init(<d->ltd_layout_phase_list);
4509 list_for_each_entry_safe(ltd, next, &llmd->llmd_mdt_list,
4511 list_del_init(<d->ltd_layout_list);
4513 spin_unlock(<ds->ltd_lock);
4518 static void lfsck_layout_slave_data_release(const struct lu_env *env,
4519 struct lfsck_component *com)
4521 struct lfsck_layout_slave_data *llsd = com->lc_data;
4522 struct lfsck_layout_seq *lls;
4523 struct lfsck_layout_seq *next;
4524 struct lfsck_layout_slave_target *llst;
4525 struct lfsck_layout_slave_target *tmp;
4527 LASSERT(llsd != NULL);
4529 list_for_each_entry_safe(lls, next, &llsd->llsd_seq_list,
4531 list_del_init(&lls->lls_list);
4532 lfsck_object_put(env, lls->lls_lastid_obj);
4536 list_for_each_entry_safe(llst, tmp, &llsd->llsd_master_list,
4538 list_del_init(&llst->llst_list);
4542 lfsck_rbtree_cleanup(env, com);
4543 com->lc_data = NULL;
4547 static void lfsck_layout_master_quit(const struct lu_env *env,
4548 struct lfsck_component *com)
4550 struct lfsck_layout_master_data *llmd = com->lc_data;
4551 struct ptlrpc_thread *mthread = &com->lc_lfsck->li_thread;
4552 struct ptlrpc_thread *athread = &llmd->llmd_thread;
4553 struct l_wait_info lwi = { 0 };
4555 llmd->llmd_exit = 1;
4556 wake_up_all(&athread->t_ctl_waitq);
4557 l_wait_event(mthread->t_ctl_waitq,
4558 thread_is_init(athread) ||
4559 thread_is_stopped(athread),
4563 static void lfsck_layout_slave_quit(const struct lu_env *env,
4564 struct lfsck_component *com)
4566 lfsck_rbtree_cleanup(env, com);
4569 static int lfsck_layout_master_in_notify(const struct lu_env *env,
4570 struct lfsck_component *com,
4571 struct lfsck_request *lr)
4573 struct lfsck_instance *lfsck = com->lc_lfsck;
4574 struct lfsck_layout *lo = com->lc_file_ram;
4575 struct lfsck_layout_master_data *llmd = com->lc_data;
4576 struct lfsck_tgt_descs *ltds;
4577 struct lfsck_tgt_desc *ltd;
4581 if (lr->lr_event != LE_PHASE1_DONE &&
4582 lr->lr_event != LE_PHASE2_DONE &&
4583 lr->lr_event != LE_PEER_EXIT)
4586 if (lr->lr_flags & LEF_FROM_OST)
4587 ltds = &lfsck->li_ost_descs;
4589 ltds = &lfsck->li_mdt_descs;
4590 spin_lock(<ds->ltd_lock);
4591 ltd = LTD_TGT(ltds, lr->lr_index);
4593 spin_unlock(<ds->ltd_lock);
4598 list_del_init(<d->ltd_layout_phase_list);
4599 switch (lr->lr_event) {
4600 case LE_PHASE1_DONE:
4601 if (lr->lr_status <= 0) {
4602 ltd->ltd_layout_done = 1;
4603 list_del_init(<d->ltd_layout_list);
4604 CWARN("%s: %s %x failed/stopped at phase1: rc = %d.\n",
4605 lfsck_lfsck2name(lfsck),
4606 (lr->lr_flags & LEF_TO_OST) ? "OST" : "MDT",
4607 ltd->ltd_index, lr->lr_status);
4608 lo->ll_flags |= LF_INCOMPLETE;
4613 if (lr->lr_flags & LEF_FROM_OST) {
4614 if (list_empty(<d->ltd_layout_list))
4615 list_add_tail(<d->ltd_layout_list,
4616 &llmd->llmd_ost_list);
4617 list_add_tail(<d->ltd_layout_phase_list,
4618 &llmd->llmd_ost_phase2_list);
4620 if (list_empty(<d->ltd_layout_list))
4621 list_add_tail(<d->ltd_layout_list,
4622 &llmd->llmd_mdt_list);
4623 list_add_tail(<d->ltd_layout_phase_list,
4624 &llmd->llmd_mdt_phase2_list);
4627 case LE_PHASE2_DONE:
4628 ltd->ltd_layout_done = 1;
4629 list_del_init(<d->ltd_layout_list);
4633 ltd->ltd_layout_done = 1;
4634 list_del_init(<d->ltd_layout_list);
4635 if (!(lfsck->li_bookmark_ram.lb_param & LPF_FAILOUT)) {
4636 CWARN("%s: the peer %s %x exit layout LFSCK.\n",
4637 lfsck_lfsck2name(lfsck),
4638 (lr->lr_flags & LEF_TO_OST) ? "OST" : "MDT",
4640 lo->ll_flags |= LF_INCOMPLETE;
4646 spin_unlock(<ds->ltd_lock);
4648 if (fail && lfsck->li_bookmark_ram.lb_param & LPF_FAILOUT) {
4649 struct lfsck_stop *stop = &lfsck_env_info(env)->lti_stop;
4651 memset(stop, 0, sizeof(*stop));
4652 stop->ls_status = lr->lr_status;
4653 stop->ls_flags = lr->lr_param & ~LPF_BROADCAST;
4654 lfsck_stop(env, lfsck->li_bottom, stop);
4655 } else if (lfsck_layout_master_to_orphan(llmd)) {
4656 wake_up_all(&llmd->llmd_thread.t_ctl_waitq);
4662 static int lfsck_layout_slave_in_notify(const struct lu_env *env,
4663 struct lfsck_component *com,
4664 struct lfsck_request *lr)
4666 struct lfsck_instance *lfsck = com->lc_lfsck;
4667 struct lfsck_layout_slave_data *llsd = com->lc_data;
4668 struct lfsck_layout_slave_target *llst;
4671 if (lr->lr_event == LE_FID_ACCESSED) {
4672 lfsck_rbtree_update_bitmap(env, com, &lr->lr_fid, true);
4677 if (lr->lr_event != LE_PHASE2_DONE && lr->lr_event != LE_PEER_EXIT)
4680 llst = lfsck_layout_llst_find_and_del(llsd, lr->lr_index, true);
4684 lfsck_layout_llst_put(llst);
4685 if (list_empty(&llsd->llsd_master_list))
4686 wake_up_all(&lfsck->li_thread.t_ctl_waitq);
4688 if (lr->lr_event == LE_PEER_EXIT &&
4689 lfsck->li_bookmark_ram.lb_param & LPF_FAILOUT) {
4690 struct lfsck_stop *stop = &lfsck_env_info(env)->lti_stop;
4692 memset(stop, 0, sizeof(*stop));
4693 stop->ls_status = lr->lr_status;
4694 stop->ls_flags = lr->lr_param & ~LPF_BROADCAST;
4695 lfsck_stop(env, lfsck->li_bottom, stop);
4701 static int lfsck_layout_query(const struct lu_env *env,
4702 struct lfsck_component *com)
4704 struct lfsck_layout *lo = com->lc_file_ram;
4706 return lo->ll_status;
4709 static int lfsck_layout_master_stop_notify(const struct lu_env *env,
4710 struct lfsck_component *com,
4711 struct lfsck_tgt_descs *ltds,
4712 struct lfsck_tgt_desc *ltd,
4713 struct ptlrpc_request_set *set)
4715 struct lfsck_thread_info *info = lfsck_env_info(env);
4716 struct lfsck_async_interpret_args *laia = &info->lti_laia;
4717 struct lfsck_request *lr = &info->lti_lr;
4718 struct lfsck_instance *lfsck = com->lc_lfsck;
4721 spin_lock(<ds->ltd_lock);
4722 if (list_empty(<d->ltd_layout_list)) {
4723 LASSERT(list_empty(<d->ltd_layout_phase_list));
4724 spin_unlock(<ds->ltd_lock);
4729 list_del_init(<d->ltd_layout_phase_list);
4730 list_del_init(<d->ltd_layout_list);
4731 spin_unlock(<ds->ltd_lock);
4733 memset(lr, 0, sizeof(*lr));
4734 lr->lr_index = lfsck_dev_idx(lfsck->li_bottom);
4735 lr->lr_event = LE_PEER_EXIT;
4736 lr->lr_active = LT_LAYOUT;
4737 lr->lr_status = LS_CO_PAUSED;
4738 if (ltds == &lfsck->li_ost_descs)
4739 lr->lr_flags = LEF_TO_OST;
4741 laia->laia_com = com;
4742 laia->laia_ltds = ltds;
4743 atomic_inc(<d->ltd_ref);
4744 laia->laia_ltd = ltd;
4746 laia->laia_shared = 0;
4748 rc = lfsck_async_request(env, ltd->ltd_exp, lr, set,
4749 lfsck_layout_master_async_interpret,
4750 laia, LFSCK_NOTIFY);
4752 CERROR("%s: Fail to notify %s %x for co-stop: rc = %d\n",
4753 lfsck_lfsck2name(lfsck),
4754 (lr->lr_flags & LEF_TO_OST) ? "OST" : "MDT",
4755 ltd->ltd_index, rc);
4762 /* with lfsck::li_lock held */
4763 static int lfsck_layout_slave_join(const struct lu_env *env,
4764 struct lfsck_component *com,
4765 struct lfsck_start_param *lsp)
4767 struct lfsck_instance *lfsck = com->lc_lfsck;
4768 struct lfsck_layout_slave_data *llsd = com->lc_data;
4769 struct lfsck_layout_slave_target *llst;
4770 struct lfsck_start *start = lsp->lsp_start;
4774 if (!lsp->lsp_index_valid || start == NULL ||
4775 !(start->ls_flags & LPF_ALL_TGT) ||
4776 !(lfsck->li_bookmark_ram.lb_param & LPF_ALL_TGT))
4779 spin_unlock(&lfsck->li_lock);
4780 rc = lfsck_layout_llst_add(llsd, lsp->lsp_index);
4781 spin_lock(&lfsck->li_lock);
4782 if (rc == 0 && !thread_is_running(&lfsck->li_thread)) {
4783 spin_unlock(&lfsck->li_lock);
4784 llst = lfsck_layout_llst_find_and_del(llsd, lsp->lsp_index,
4787 lfsck_layout_llst_put(llst);
4788 spin_lock(&lfsck->li_lock);
4795 static struct lfsck_operations lfsck_layout_master_ops = {
4796 .lfsck_reset = lfsck_layout_reset,
4797 .lfsck_fail = lfsck_layout_fail,
4798 .lfsck_checkpoint = lfsck_layout_master_checkpoint,
4799 .lfsck_prep = lfsck_layout_master_prep,
4800 .lfsck_exec_oit = lfsck_layout_master_exec_oit,
4801 .lfsck_exec_dir = lfsck_layout_exec_dir,
4802 .lfsck_post = lfsck_layout_master_post,
4803 .lfsck_interpret = lfsck_layout_master_async_interpret,
4804 .lfsck_dump = lfsck_layout_dump,
4805 .lfsck_double_scan = lfsck_layout_master_double_scan,
4806 .lfsck_data_release = lfsck_layout_master_data_release,
4807 .lfsck_quit = lfsck_layout_master_quit,
4808 .lfsck_in_notify = lfsck_layout_master_in_notify,
4809 .lfsck_query = lfsck_layout_query,
4810 .lfsck_stop_notify = lfsck_layout_master_stop_notify,
4813 static struct lfsck_operations lfsck_layout_slave_ops = {
4814 .lfsck_reset = lfsck_layout_reset,
4815 .lfsck_fail = lfsck_layout_fail,
4816 .lfsck_checkpoint = lfsck_layout_slave_checkpoint,
4817 .lfsck_prep = lfsck_layout_slave_prep,
4818 .lfsck_exec_oit = lfsck_layout_slave_exec_oit,
4819 .lfsck_exec_dir = lfsck_layout_exec_dir,
4820 .lfsck_post = lfsck_layout_slave_post,
4821 .lfsck_dump = lfsck_layout_dump,
4822 .lfsck_double_scan = lfsck_layout_slave_double_scan,
4823 .lfsck_data_release = lfsck_layout_slave_data_release,
4824 .lfsck_quit = lfsck_layout_slave_quit,
4825 .lfsck_in_notify = lfsck_layout_slave_in_notify,
4826 .lfsck_query = lfsck_layout_query,
4827 .lfsck_join = lfsck_layout_slave_join,
4830 int lfsck_layout_setup(const struct lu_env *env, struct lfsck_instance *lfsck)
4832 struct lfsck_component *com;
4833 struct lfsck_layout *lo;
4834 struct dt_object *root = NULL;
4835 struct dt_object *obj;
4843 INIT_LIST_HEAD(&com->lc_link);
4844 INIT_LIST_HEAD(&com->lc_link_dir);
4845 init_rwsem(&com->lc_sem);
4846 atomic_set(&com->lc_ref, 1);
4847 com->lc_lfsck = lfsck;
4848 com->lc_type = LT_LAYOUT;
4849 if (lfsck->li_master) {
4850 struct lfsck_layout_master_data *llmd;
4852 com->lc_ops = &lfsck_layout_master_ops;
4853 OBD_ALLOC_PTR(llmd);
4855 GOTO(out, rc = -ENOMEM);
4857 INIT_LIST_HEAD(&llmd->llmd_req_list);
4858 spin_lock_init(&llmd->llmd_lock);
4859 INIT_LIST_HEAD(&llmd->llmd_ost_list);
4860 INIT_LIST_HEAD(&llmd->llmd_ost_phase1_list);
4861 INIT_LIST_HEAD(&llmd->llmd_ost_phase2_list);
4862 INIT_LIST_HEAD(&llmd->llmd_mdt_list);
4863 INIT_LIST_HEAD(&llmd->llmd_mdt_phase1_list);
4864 INIT_LIST_HEAD(&llmd->llmd_mdt_phase2_list);
4865 init_waitqueue_head(&llmd->llmd_thread.t_ctl_waitq);
4866 com->lc_data = llmd;
4868 struct lfsck_layout_slave_data *llsd;
4870 com->lc_ops = &lfsck_layout_slave_ops;
4871 OBD_ALLOC_PTR(llsd);
4873 GOTO(out, rc = -ENOMEM);
4875 INIT_LIST_HEAD(&llsd->llsd_seq_list);
4876 INIT_LIST_HEAD(&llsd->llsd_master_list);
4877 spin_lock_init(&llsd->llsd_lock);
4878 llsd->llsd_rb_root = RB_ROOT;
4879 rwlock_init(&llsd->llsd_rb_lock);
4880 com->lc_data = llsd;
4882 com->lc_file_size = sizeof(*lo);
4883 OBD_ALLOC(com->lc_file_ram, com->lc_file_size);
4884 if (com->lc_file_ram == NULL)
4885 GOTO(out, rc = -ENOMEM);
4887 OBD_ALLOC(com->lc_file_disk, com->lc_file_size);
4888 if (com->lc_file_disk == NULL)
4889 GOTO(out, rc = -ENOMEM);
4891 root = dt_locate(env, lfsck->li_bottom, &lfsck->li_local_root_fid);
4893 GOTO(out, rc = PTR_ERR(root));
4895 if (unlikely(!dt_try_as_dir(env, root)))
4896 GOTO(out, rc = -ENOTDIR);
4898 obj = local_file_find_or_create(env, lfsck->li_los, root,
4900 S_IFREG | S_IRUGO | S_IWUSR);
4902 GOTO(out, rc = PTR_ERR(obj));
4905 rc = lfsck_layout_load(env, com);
4907 rc = lfsck_layout_reset(env, com, true);
4908 else if (rc == -ENOENT)
4909 rc = lfsck_layout_init(env, com);
4914 lo = com->lc_file_ram;
4915 switch (lo->ll_status) {
4921 spin_lock(&lfsck->li_lock);
4922 list_add_tail(&com->lc_link, &lfsck->li_list_idle);
4923 spin_unlock(&lfsck->li_lock);
4926 CERROR("%s: unknown lfsck_layout status: rc = %u\n",
4927 lfsck_lfsck2name(lfsck), lo->ll_status);
4929 case LS_SCANNING_PHASE1:
4930 case LS_SCANNING_PHASE2:
4931 /* No need to store the status to disk right now.
4932 * If the system crashed before the status stored,
4933 * it will be loaded back when next time. */
4934 lo->ll_status = LS_CRASHED;
4935 lo->ll_flags |= LF_INCOMPLETE;
4942 spin_lock(&lfsck->li_lock);
4943 list_add_tail(&com->lc_link, &lfsck->li_list_scan);
4944 spin_unlock(&lfsck->li_lock);
4948 if (lo->ll_flags & LF_CRASHED_LASTID) {
4949 LASSERT(lfsck->li_out_notify != NULL);
4951 lfsck->li_out_notify(env, lfsck->li_out_notify_data,
4952 LE_LASTID_REBUILDING);
4958 if (root != NULL && !IS_ERR(root))
4959 lu_object_put(env, &root->do_lu);
4962 lfsck_component_cleanup(env, com);
4967 struct lfsck_orphan_it {
4968 struct lfsck_component *loi_com;
4969 struct lfsck_rbtree_node *loi_lrn;
4970 struct lfsck_layout_slave_target *loi_llst;
4971 struct lu_fid loi_key;
4972 struct lu_orphan_rec loi_rec;
4974 unsigned int loi_over:1;
4977 static int lfsck_fid_match_idx(const struct lu_env *env,
4978 struct lfsck_instance *lfsck,
4979 const struct lu_fid *fid, int idx)
4981 struct seq_server_site *ss;
4982 struct lu_server_fld *sf;
4983 struct lu_seq_range range = { 0 };
4986 /* All abnormal cases will be returned to MDT0. */
4987 if (!fid_is_norm(fid)) {
4994 ss = lu_site2seq(lfsck->li_bottom->dd_lu_dev.ld_site);
4995 if (unlikely(ss == NULL))
4998 sf = ss->ss_server_fld;
4999 LASSERT(sf != NULL);
5001 fld_range_set_any(&range);
5002 rc = fld_server_lookup(env, sf, fid_seq(fid), &range);
5006 if (!fld_range_is_mdt(&range))
5009 if (range.lsr_index == idx)
5015 static void lfsck_layout_destroy_orphan(const struct lu_env *env,
5016 struct dt_device *dev,
5017 struct dt_object *obj)
5019 struct thandle *handle;
5023 handle = dt_trans_create(env, dev);
5027 rc = dt_declare_ref_del(env, obj, handle);
5031 rc = dt_declare_destroy(env, obj, handle);
5035 rc = dt_trans_start_local(env, dev, handle);
5039 dt_write_lock(env, obj, 0);
5040 rc = dt_ref_del(env, obj, handle);
5042 rc = dt_destroy(env, obj, handle);
5043 dt_write_unlock(env, obj);
5048 dt_trans_stop(env, dev, handle);
5053 static int lfsck_orphan_index_lookup(const struct lu_env *env,
5054 struct dt_object *dt,
5056 const struct dt_key *key,
5057 struct lustre_capa *capa)
5062 static int lfsck_orphan_index_declare_insert(const struct lu_env *env,
5063 struct dt_object *dt,
5064 const struct dt_rec *rec,
5065 const struct dt_key *key,
5066 struct thandle *handle)
5071 static int lfsck_orphan_index_insert(const struct lu_env *env,
5072 struct dt_object *dt,
5073 const struct dt_rec *rec,
5074 const struct dt_key *key,
5075 struct thandle *handle,
5076 struct lustre_capa *capa,
5082 static int lfsck_orphan_index_declare_delete(const struct lu_env *env,
5083 struct dt_object *dt,
5084 const struct dt_key *key,
5085 struct thandle *handle)
5090 static int lfsck_orphan_index_delete(const struct lu_env *env,
5091 struct dt_object *dt,
5092 const struct dt_key *key,
5093 struct thandle *handle,
5094 struct lustre_capa *capa)
5099 static struct dt_it *lfsck_orphan_it_init(const struct lu_env *env,
5100 struct dt_object *dt,
5102 struct lustre_capa *capa)
5104 struct dt_device *dev = lu2dt_dev(dt->do_lu.lo_dev);
5105 struct lfsck_instance *lfsck;
5106 struct lfsck_component *com = NULL;
5107 struct lfsck_layout_slave_data *llsd;
5108 struct lfsck_orphan_it *it = NULL;
5112 lfsck = lfsck_instance_find(dev, true, false);
5113 if (unlikely(lfsck == NULL))
5114 RETURN(ERR_PTR(-ENODEV));
5116 com = lfsck_component_find(lfsck, LT_LAYOUT);
5117 if (unlikely(com == NULL))
5118 GOTO(out, rc = -ENOENT);
5120 llsd = com->lc_data;
5121 if (!llsd->llsd_rbtree_valid)
5122 GOTO(out, rc = -ESRCH);
5126 GOTO(out, rc = -ENOMEM);
5128 it->loi_llst = lfsck_layout_llst_find_and_del(llsd, attr, false);
5129 if (it->loi_llst == NULL)
5130 GOTO(out, rc = -ENODEV);
5132 if (dev->dd_record_fid_accessed) {
5133 /* The first iteration against the rbtree, scan the whole rbtree
5134 * to remove the nodes which do NOT need to be handled. */
5135 write_lock(&llsd->llsd_rb_lock);
5136 if (dev->dd_record_fid_accessed) {
5137 struct rb_node *node;
5138 struct rb_node *next;
5139 struct lfsck_rbtree_node *lrn;
5141 /* No need to record the fid accessing anymore. */
5142 dev->dd_record_fid_accessed = 0;
5144 node = rb_first(&llsd->llsd_rb_root);
5145 while (node != NULL) {
5146 next = rb_next(node);
5147 lrn = rb_entry(node, struct lfsck_rbtree_node,
5149 if (atomic_read(&lrn->lrn_known_count) <=
5150 atomic_read(&lrn->lrn_accessed_count)) {
5151 rb_erase(node, &llsd->llsd_rb_root);
5152 lfsck_rbtree_free(lrn);
5157 write_unlock(&llsd->llsd_rb_lock);
5160 /* read lock the rbtree when init, and unlock when fini */
5161 read_lock(&llsd->llsd_rb_lock);
5169 lfsck_component_put(env, com);
5170 lfsck_instance_put(env, lfsck);
5175 it = (struct lfsck_orphan_it *)ERR_PTR(rc);
5178 return (struct dt_it *)it;
5181 static void lfsck_orphan_it_fini(const struct lu_env *env,
5184 struct lfsck_orphan_it *it = (struct lfsck_orphan_it *)di;
5185 struct lfsck_component *com = it->loi_com;
5186 struct lfsck_layout_slave_data *llsd;
5187 struct lfsck_layout_slave_target *llst;
5190 llsd = com->lc_data;
5191 read_unlock(&llsd->llsd_rb_lock);
5192 llst = it->loi_llst;
5193 LASSERT(llst != NULL);
5195 /* Save the key and hash for iterate next. */
5196 llst->llst_fid = it->loi_key;
5197 llst->llst_hash = it->loi_hash;
5198 lfsck_layout_llst_put(llst);
5199 lfsck_component_put(env, com);
5205 * \retval +1: the iteration finished
5206 * \retval 0: on success, not finished
5207 * \retval -ve: on error
5209 static int lfsck_orphan_it_next(const struct lu_env *env,
5212 struct lfsck_thread_info *info = lfsck_env_info(env);
5213 struct filter_fid_old *pfid = &info->lti_old_pfid;
5214 struct lu_attr *la = &info->lti_la;
5215 struct lfsck_orphan_it *it = (struct lfsck_orphan_it *)di;
5216 struct lu_fid *key = &it->loi_key;
5217 struct lu_orphan_rec *rec = &it->loi_rec;
5218 struct lfsck_component *com = it->loi_com;
5219 struct lfsck_instance *lfsck = com->lc_lfsck;
5220 struct lfsck_layout_slave_data *llsd = com->lc_data;
5221 struct dt_object *obj;
5222 struct lfsck_rbtree_node *lrn;
5226 __u32 idx = it->loi_llst->llst_index;
5236 lrn = lfsck_rbtree_search(llsd, key, &exact);
5244 key->f_seq = lrn->lrn_seq;
5245 key->f_oid = lrn->lrn_first_oid;
5250 if (unlikely(key->f_oid == 0)) {
5257 lrn->lrn_first_oid + LFSCK_RBTREE_BITMAP_WIDTH) {
5263 if (unlikely(atomic_read(&lrn->lrn_known_count) <=
5264 atomic_read(&lrn->lrn_accessed_count))) {
5265 struct rb_node *next = rb_next(&lrn->lrn_node);
5267 while (next != NULL) {
5268 lrn = rb_entry(next, struct lfsck_rbtree_node,
5270 if (atomic_read(&lrn->lrn_known_count) >
5271 atomic_read(&lrn->lrn_accessed_count))
5273 next = rb_next(next);
5282 key->f_seq = lrn->lrn_seq;
5283 key->f_oid = lrn->lrn_first_oid;
5287 pos = key->f_oid - lrn->lrn_first_oid;
5290 pos = find_next_bit(lrn->lrn_known_bitmap,
5291 LFSCK_RBTREE_BITMAP_WIDTH, pos);
5292 if (pos >= LFSCK_RBTREE_BITMAP_WIDTH) {
5293 key->f_oid = lrn->lrn_first_oid + pos;
5294 if (unlikely(key->f_oid < lrn->lrn_first_oid)) {
5302 if (test_bit(pos, lrn->lrn_accessed_bitmap)) {
5307 key->f_oid = lrn->lrn_first_oid + pos;
5308 obj = lfsck_object_find(env, lfsck, key);
5311 if (rc == -ENOENT) {
5318 dt_read_lock(env, obj, 0);
5319 if (!dt_object_exists(obj)) {
5320 dt_read_unlock(env, obj);
5321 lfsck_object_put(env, obj);
5326 rc = dt_attr_get(env, obj, la, BYPASS_CAPA);
5330 rc = dt_xattr_get(env, obj, lfsck_buf_get(env, pfid, sizeof(*pfid)),
5331 XATTR_NAME_FID, BYPASS_CAPA);
5332 if (rc == -ENODATA) {
5333 /* For the pre-created OST-object, update the bitmap to avoid
5334 * others LFSCK (second phase) iteration to touch it again. */
5335 if (la->la_ctime == 0) {
5336 if (!test_and_set_bit(pos, lrn->lrn_accessed_bitmap))
5337 atomic_inc(&lrn->lrn_accessed_count);
5339 /* For the race between repairing dangling referenced
5340 * MDT-object and unlink the file, it may left orphan
5341 * OST-object there. Destroy it now! */
5342 if (unlikely(!(la->la_mode & S_ISUID))) {
5343 dt_read_unlock(env, obj);
5344 lfsck_layout_destroy_orphan(env,
5347 lfsck_object_put(env, obj);
5351 } else if (idx == 0) {
5352 /* If the orphan OST-object has no parent information,
5353 * regard it as referenced by the MDT-object on MDT0. */
5354 fid_zero(&rec->lor_fid);
5355 rec->lor_uid = la->la_uid;
5356 rec->lor_gid = la->la_gid;
5360 dt_read_unlock(env, obj);
5361 lfsck_object_put(env, obj);
5369 if (rc != sizeof(struct filter_fid) &&
5370 rc != sizeof(struct filter_fid_old))
5371 GOTO(out, rc = -EINVAL);
5373 fid_le_to_cpu(&rec->lor_fid, &pfid->ff_parent);
5374 /* In fact, the ff_parent::f_ver is not the real parent FID::f_ver,
5375 * instead, it is the OST-object index in its parent MDT-object
5377 save = rec->lor_fid.f_ver;
5378 rec->lor_fid.f_ver = 0;
5379 rc = lfsck_fid_match_idx(env, lfsck, &rec->lor_fid, idx);
5380 /* If the orphan OST-object does not claim the MDT, then next.
5382 * If we do not know whether it matches or not, then return it
5383 * to the MDT for further check. */
5385 dt_read_unlock(env, obj);
5386 lfsck_object_put(env, obj);
5391 rec->lor_fid.f_ver = save;
5392 rec->lor_uid = la->la_uid;
5393 rec->lor_gid = la->la_gid;
5395 CDEBUG(D_LFSCK, "%s: return orphan "DFID", PFID "DFID", owner %u:%u\n",
5396 lfsck_lfsck2name(com->lc_lfsck), PFID(key), PFID(&rec->lor_fid),
5397 rec->lor_uid, rec->lor_gid);
5402 dt_read_unlock(env, obj);
5403 lfsck_object_put(env, obj);
5411 * \retval +1: locate to the exactly position
5412 * \retval 0: cannot locate to the exactly position,
5413 * call next() to move to a valid position.
5414 * \retval -ve: on error
5416 static int lfsck_orphan_it_get(const struct lu_env *env,
5418 const struct dt_key *key)
5420 struct lfsck_orphan_it *it = (struct lfsck_orphan_it *)di;
5423 it->loi_key = *(struct lu_fid *)key;
5424 rc = lfsck_orphan_it_next(env, di);
5434 static void lfsck_orphan_it_put(const struct lu_env *env,
5439 static struct dt_key *lfsck_orphan_it_key(const struct lu_env *env,
5440 const struct dt_it *di)
5442 struct lfsck_orphan_it *it = (struct lfsck_orphan_it *)di;
5444 return (struct dt_key *)&it->loi_key;
5447 static int lfsck_orphan_it_key_size(const struct lu_env *env,
5448 const struct dt_it *di)
5450 return sizeof(struct lu_fid);
5453 static int lfsck_orphan_it_rec(const struct lu_env *env,
5454 const struct dt_it *di,
5458 struct lfsck_orphan_it *it = (struct lfsck_orphan_it *)di;
5460 *(struct lu_orphan_rec *)rec = it->loi_rec;
5465 static __u64 lfsck_orphan_it_store(const struct lu_env *env,
5466 const struct dt_it *di)
5468 struct lfsck_orphan_it *it = (struct lfsck_orphan_it *)di;
5470 return it->loi_hash;
5474 * \retval +1: locate to the exactly position
5475 * \retval 0: cannot locate to the exactly position,
5476 * call next() to move to a valid position.
5477 * \retval -ve: on error
5479 static int lfsck_orphan_it_load(const struct lu_env *env,
5480 const struct dt_it *di,
5483 struct lfsck_orphan_it *it = (struct lfsck_orphan_it *)di;
5484 struct lfsck_layout_slave_target *llst = it->loi_llst;
5487 LASSERT(llst != NULL);
5489 if (hash != llst->llst_hash) {
5490 CWARN("%s: the given hash "LPU64" for orphan iteration does "
5491 "not match the one when fini "LPU64", to be reset.\n",
5492 lfsck_lfsck2name(it->loi_com->lc_lfsck), hash,
5494 fid_zero(&llst->llst_fid);
5495 llst->llst_hash = 0;
5498 it->loi_key = llst->llst_fid;
5499 it->loi_hash = llst->llst_hash;
5500 rc = lfsck_orphan_it_next(env, (struct dt_it *)di);
5510 static int lfsck_orphan_it_key_rec(const struct lu_env *env,
5511 const struct dt_it *di,
5517 const struct dt_index_operations lfsck_orphan_index_ops = {
5518 .dio_lookup = lfsck_orphan_index_lookup,
5519 .dio_declare_insert = lfsck_orphan_index_declare_insert,
5520 .dio_insert = lfsck_orphan_index_insert,
5521 .dio_declare_delete = lfsck_orphan_index_declare_delete,
5522 .dio_delete = lfsck_orphan_index_delete,
5524 .init = lfsck_orphan_it_init,
5525 .fini = lfsck_orphan_it_fini,
5526 .get = lfsck_orphan_it_get,
5527 .put = lfsck_orphan_it_put,
5528 .next = lfsck_orphan_it_next,
5529 .key = lfsck_orphan_it_key,
5530 .key_size = lfsck_orphan_it_key_size,
5531 .rec = lfsck_orphan_it_rec,
5532 .store = lfsck_orphan_it_store,
5533 .load = lfsck_orphan_it_load,
5534 .key_rec = lfsck_orphan_it_key_rec,