1 /* -*- mode: c; c-basic-offset: 8; indent-tabs-mode: nil; -*-
2 * vim:expandtab:shiftwidth=8:tabstop=8:
4 * linux/fs/obdfilter/filter.c
6 * Copyright (c) 2001-2003 Cluster File Systems, Inc.
7 * Author: Peter Braam <braam@clusterfs.com>
8 * Author: Andreas Dilger <adilger@clusterfs.com>
10 * This file is part of Lustre, http://www.lustre.org.
12 * Lustre is free software; you can redistribute it and/or
13 * modify it under the terms of version 2 of the GNU General Public
14 * License as published by the Free Software Foundation.
16 * Lustre is distributed in the hope that it will be useful,
17 * but WITHOUT ANY WARRANTY; without even the implied warranty of
18 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
19 * GNU General Public License for more details.
21 * You should have received a copy of the GNU General Public License
22 * along with Lustre; if not, write to the Free Software
23 * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
27 * Invariant: Get O/R i_sem for lookup, if needed, before any journal ops
28 * (which need to get journal_lock, may block if journal full).
30 * Invariant: Call filter_start_transno() before any journal ops to avoid the
31 * same deadlock problem. We can (and want) to get rid of the
32 * transno sem in favour of the dir/inode i_sem to avoid single
33 * threaded operation on the OST.
36 #define DEBUG_SUBSYSTEM S_FILTER
38 #include <linux/config.h>
39 #include <linux/module.h>
41 #include <linux/dcache.h>
42 #include <linux/init.h>
43 #include <linux/version.h>
44 #if (LINUX_VERSION_CODE > KERNEL_VERSION(2,5,0))
45 # include <linux/mount.h>
46 # include <linux/buffer_head.h>
49 #include <linux/obd_class.h>
50 #include <linux/obd_lov.h>
51 #include <linux/obd_ost.h>
52 #include <linux/lustre_dlm.h>
53 #include <linux/lustre_fsfilt.h>
54 #include <linux/lprocfs_status.h>
55 #include <linux/lustre_log.h>
56 #include <linux/lustre_commit_confd.h>
57 #include <portals/list.h>
59 #include <linux/lustre_smfs.h>
60 #include "filter_internal.h"
62 static struct lvfs_callback_ops filter_lvfs_ops;
64 static int filter_destroy(struct obd_export *exp, struct obdo *oa,
65 struct lov_stripe_md *ea, struct obd_trans_info *);
67 static void filter_commit_cb(struct obd_device *obd, __u64 transno,
68 void *cb_data, int error)
70 obd_transno_commit_cb(obd, transno, error);
73 /* Assumes caller has already pushed us into the kernel context. */
74 int filter_finish_transno(struct obd_export *exp, struct obd_trans_info *oti,
77 struct filter_obd *filter = &exp->exp_obd->u.filter;
78 struct filter_export_data *fed = &exp->exp_filter_data;
79 struct filter_client_data *fcd = fed->fed_fcd;
82 int err, log_pri = D_HA;
84 /* Propagate error code. */
88 if (!exp->exp_obd->obd_replayable || oti == NULL)
91 /* we don't allocate new transnos for replayed requests */
92 if (oti->oti_transno == 0) {
93 spin_lock(&filter->fo_translock);
94 last_rcvd = le64_to_cpu(filter->fo_fsd->fsd_last_transno) + 1;
95 filter->fo_fsd->fsd_last_transno = cpu_to_le64(last_rcvd);
96 spin_unlock(&filter->fo_translock);
97 oti->oti_transno = last_rcvd;
99 spin_lock(&filter->fo_translock);
100 last_rcvd = oti->oti_transno;
101 if (last_rcvd > le64_to_cpu(filter->fo_fsd->fsd_last_transno))
102 filter->fo_fsd->fsd_last_transno =
103 cpu_to_le64(last_rcvd);
104 spin_unlock(&filter->fo_translock);
106 fcd->fcd_last_rcvd = cpu_to_le64(last_rcvd);
108 /* could get xid from oti, if it's ever needed */
109 fcd->fcd_last_xid = 0;
111 off = fed->fed_lr_off;
113 fsfilt_add_journal_cb(exp->exp_obd, filter->fo_sb, last_rcvd,
114 oti->oti_handle, filter_commit_cb, NULL);
116 err = fsfilt_write_record(exp->exp_obd, filter->fo_rcvd_filp, fcd,
117 sizeof(*fcd), &off, 0);
124 CDEBUG(log_pri, "wrote trans "LPU64" for client %s at #%d: err = %d\n",
125 last_rcvd, fcd->fcd_uuid, fed->fed_lr_idx, err);
130 void f_dput(struct dentry *dentry)
132 /* Can't go inside filter_ddelete because it can block */
133 CDEBUG(D_INODE, "putting %s: %p, count = %d\n",
134 dentry->d_name.name, dentry, atomic_read(&dentry->d_count) - 1);
135 LASSERT(atomic_read(&dentry->d_count) > 0);
140 /* Add client data to the FILTER. We use a bitmap to locate a free space
141 * in the last_rcvd file if cl_idx is -1 (i.e. a new client).
142 * Otherwise, we have just read the data from the last_rcvd file and
143 * we know its offset. */
144 static int filter_client_add(struct obd_device *obd, struct filter_obd *filter,
145 struct filter_export_data *fed, int cl_idx)
147 unsigned long *bitmap = filter->fo_last_rcvd_slots;
148 int new_client = (cl_idx == -1);
151 LASSERT(bitmap != NULL);
153 /* XXX if fcd_uuid were a real obd_uuid, I could use obd_uuid_equals */
154 if (!strcmp(fed->fed_fcd->fcd_uuid, obd->obd_uuid.uuid))
157 /* the bitmap operations can handle cl_idx > sizeof(long) * 8, so
158 * there's no need for extra complication here
161 cl_idx = find_first_zero_bit(bitmap, FILTER_LR_MAX_CLIENTS);
163 if (cl_idx >= FILTER_LR_MAX_CLIENTS) {
164 CERROR("no client slots - fix FILTER_LR_MAX_CLIENTS\n");
167 if (test_and_set_bit(cl_idx, bitmap)) {
168 CERROR("FILTER client %d: found bit is set in bitmap\n",
170 cl_idx = find_next_zero_bit(bitmap,
171 FILTER_LR_MAX_CLIENTS,
176 if (test_and_set_bit(cl_idx, bitmap)) {
177 CERROR("FILTER client %d: bit already set in bitmap!\n",
183 fed->fed_lr_idx = cl_idx;
184 fed->fed_lr_off = le32_to_cpu(filter->fo_fsd->fsd_client_start) +
185 cl_idx * le16_to_cpu(filter->fo_fsd->fsd_client_size);
187 CDEBUG(D_INFO, "client at index %d (%llu) with UUID '%s' added\n",
188 fed->fed_lr_idx, fed->fed_lr_off, fed->fed_fcd->fcd_uuid);
191 struct lvfs_run_ctxt saved;
192 loff_t off = fed->fed_lr_off;
196 CDEBUG(D_INFO, "writing client fcd at idx %u (%llu) (len %u)\n",
197 fed->fed_lr_idx,off,(unsigned int)sizeof(*fed->fed_fcd));
199 push_ctxt(&saved, &obd->obd_lvfs_ctxt, NULL);
200 /* Transaction needed to fix bug 1403 */
201 handle = fsfilt_start(obd,
202 filter->fo_rcvd_filp->f_dentry->d_inode,
203 FSFILT_OP_SETATTR, NULL);
204 if (IS_ERR(handle)) {
205 err = PTR_ERR(handle);
206 CERROR("unable to start transaction: rc %d\n", err);
208 err = fsfilt_write_record(obd, filter->fo_rcvd_filp,
210 sizeof(*fed->fed_fcd),
212 fsfilt_commit(obd, filter->fo_sb,
213 filter->fo_rcvd_filp->f_dentry->d_inode,
216 pop_ctxt(&saved, &obd->obd_lvfs_ctxt, NULL);
219 CERROR("error writing %s client idx %u: rc %d\n",
220 LAST_RCVD, fed->fed_lr_idx, err);
227 static int filter_client_free(struct obd_export *exp, int flags)
229 struct filter_export_data *fed = &exp->exp_filter_data;
230 struct filter_obd *filter = &exp->exp_obd->u.filter;
231 struct obd_device *obd = exp->exp_obd;
232 struct filter_client_data zero_fcd;
233 struct lvfs_run_ctxt saved;
238 if (fed->fed_fcd == NULL)
241 if (flags & OBD_OPT_FAILOVER)
244 /* XXX if fcd_uuid were a real obd_uuid, I could use obd_uuid_equals */
245 if (strcmp(fed->fed_fcd->fcd_uuid, obd->obd_uuid.uuid ) == 0)
248 LASSERT(filter->fo_last_rcvd_slots != NULL);
250 off = fed->fed_lr_off;
252 CDEBUG(D_INFO, "freeing client at idx %u (%lld) with UUID '%s'\n",
253 fed->fed_lr_idx, fed->fed_lr_off, fed->fed_fcd->fcd_uuid);
255 /* Clear the bit _after_ zeroing out the client so we don't
256 race with filter_client_add and zero out new clients.*/
257 if (!test_bit(fed->fed_lr_idx, filter->fo_last_rcvd_slots)) {
258 CERROR("FILTER client %u: bit already clear in bitmap!!\n",
263 memset(&zero_fcd, 0, sizeof zero_fcd);
264 push_ctxt(&saved, &obd->obd_lvfs_ctxt, NULL);
265 rc = fsfilt_write_record(obd, filter->fo_rcvd_filp, &zero_fcd,
266 sizeof(zero_fcd), &off, 1);
268 /* update server's transno */
269 filter_update_server_data(obd, filter->fo_rcvd_filp,
271 pop_ctxt(&saved, &obd->obd_lvfs_ctxt, NULL);
273 CDEBUG(rc == 0 ? D_INFO : D_ERROR,
274 "zeroing disconnecting client %s at idx %u (%llu) in %s rc %d\n",
275 fed->fed_fcd->fcd_uuid, fed->fed_lr_idx, fed->fed_lr_off,
278 if (!test_and_clear_bit(fed->fed_lr_idx, filter->fo_last_rcvd_slots)) {
279 CERROR("FILTER client %u: bit already clear in bitmap!!\n",
285 OBD_FREE(fed->fed_fcd, sizeof(*fed->fed_fcd));
290 static int filter_free_server_data(struct filter_obd *filter)
292 OBD_FREE(filter->fo_fsd, sizeof(*filter->fo_fsd));
293 filter->fo_fsd = NULL;
294 OBD_FREE(filter->fo_last_rcvd_slots,
295 FILTER_LR_MAX_CLIENT_WORDS * sizeof(unsigned long));
296 filter->fo_last_rcvd_slots = NULL;
300 /* assumes caller is already in kernel ctxt */
301 int filter_update_server_data(struct obd_device *obd, struct file *filp,
302 struct filter_server_data *fsd, int force_sync)
308 CDEBUG(D_INODE, "server uuid : %s\n", fsd->fsd_uuid);
309 CDEBUG(D_INODE, "server last_rcvd : "LPU64"\n",
310 le64_to_cpu(fsd->fsd_last_transno));
311 CDEBUG(D_INODE, "server last_mount: "LPU64"\n",
312 le64_to_cpu(fsd->fsd_mount_count));
314 rc = fsfilt_write_record(obd, filp, fsd, sizeof(*fsd), &off,force_sync);
316 CERROR("error writing filter_server_data: rc = %d\n", rc);
321 int filter_update_last_objid(struct obd_device *obd, obd_gr group,
324 struct filter_obd *filter = &obd->u.filter;
330 if (filter->fo_last_objid_files[group] == NULL) {
331 CERROR("Object group "LPU64" not fully setup; not updating "
332 "last_objid\n", group);
336 CDEBUG(D_INODE, "server last_objid for group "LPU64": "LPU64"\n",
337 group, filter->fo_last_objids[group]);
339 tmp = cpu_to_le64(filter->fo_last_objids[group]);
340 rc = fsfilt_write_record(obd, filter->fo_last_objid_files[group],
341 &tmp, sizeof(tmp), &off, force_sync);
343 CERROR("error writing group "LPU64" last objid: rc = %d\n",
348 /* assumes caller has already in kernel ctxt */
349 static int filter_init_server_data(struct obd_device *obd, struct file * filp)
351 struct filter_obd *filter = &obd->u.filter;
352 struct filter_server_data *fsd;
353 struct filter_client_data *fcd = NULL;
354 struct inode *inode = filp->f_dentry->d_inode;
355 unsigned long last_rcvd_size = inode->i_size;
361 /* ensure padding in the struct is the correct size */
362 LASSERT (offsetof(struct filter_server_data, fsd_padding) +
363 sizeof(fsd->fsd_padding) == FILTER_LR_SERVER_SIZE);
364 LASSERT (offsetof(struct filter_client_data, fcd_padding) +
365 sizeof(fcd->fcd_padding) == FILTER_LR_CLIENT_SIZE);
367 OBD_ALLOC(fsd, sizeof(*fsd));
370 filter->fo_fsd = fsd;
372 OBD_ALLOC(filter->fo_last_rcvd_slots,
373 FILTER_LR_MAX_CLIENT_WORDS * sizeof(unsigned long));
374 if (filter->fo_last_rcvd_slots == NULL) {
375 OBD_FREE(fsd, sizeof(*fsd));
379 if (last_rcvd_size == 0) {
380 CWARN("%s: initializing new %s\n", obd->obd_name, LAST_RCVD);
382 memcpy(fsd->fsd_uuid, obd->obd_uuid.uuid,sizeof(fsd->fsd_uuid));
383 fsd->fsd_last_transno = 0;
384 mount_count = fsd->fsd_mount_count = 0;
385 fsd->fsd_server_size = cpu_to_le32(FILTER_LR_SERVER_SIZE);
386 fsd->fsd_client_start = cpu_to_le32(FILTER_LR_CLIENT_START);
387 fsd->fsd_client_size = cpu_to_le16(FILTER_LR_CLIENT_SIZE);
388 fsd->fsd_subdir_count = cpu_to_le16(FILTER_SUBDIR_COUNT);
389 filter->fo_subdir_count = FILTER_SUBDIR_COUNT;
391 rc = fsfilt_read_record(obd, filp, fsd, sizeof(*fsd), &off);
393 CDEBUG(D_INODE,"OBD filter: error reading %s: rc %d\n",
397 if (strcmp(fsd->fsd_uuid, obd->obd_uuid.uuid) != 0) {
398 CERROR("OBD UUID %s does not match last_rcvd UUID %s\n",
399 obd->obd_uuid.uuid, fsd->fsd_uuid);
400 GOTO(err_fsd, rc = -EINVAL);
402 mount_count = le64_to_cpu(fsd->fsd_mount_count);
403 filter->fo_subdir_count = le16_to_cpu(fsd->fsd_subdir_count);
406 if (fsd->fsd_feature_incompat & ~cpu_to_le32(FILTER_INCOMPAT_SUPP)) {
407 CERROR("unsupported feature %x\n",
408 le32_to_cpu(fsd->fsd_feature_incompat) &
409 ~FILTER_INCOMPAT_SUPP);
410 GOTO(err_fsd, rc = -EINVAL);
412 if (fsd->fsd_feature_rocompat & ~cpu_to_le32(FILTER_ROCOMPAT_SUPP)) {
413 CERROR("read-only feature %x\n",
414 le32_to_cpu(fsd->fsd_feature_rocompat) &
415 ~FILTER_ROCOMPAT_SUPP);
416 /* Do something like remount filesystem read-only */
417 GOTO(err_fsd, rc = -EINVAL);
420 CDEBUG(D_INODE, "%s: server last_rcvd : "LPU64"\n",
421 obd->obd_name, le64_to_cpu(fsd->fsd_last_transno));
422 CDEBUG(D_INODE, "%s: server mount_count: "LPU64"\n",
423 obd->obd_name, mount_count + 1);
424 CDEBUG(D_INODE, "%s: server data size: %u\n",
425 obd->obd_name, le32_to_cpu(fsd->fsd_server_size));
426 CDEBUG(D_INODE, "%s: per-client data start: %u\n",
427 obd->obd_name, le32_to_cpu(fsd->fsd_client_start));
428 CDEBUG(D_INODE, "%s: per-client data size: %u\n",
429 obd->obd_name, le32_to_cpu(fsd->fsd_client_size));
430 CDEBUG(D_INODE, "%s: server subdir_count: %u\n",
431 obd->obd_name, le16_to_cpu(fsd->fsd_subdir_count));
432 CDEBUG(D_INODE, "%s: last_rcvd clients: %lu\n", obd->obd_name,
433 last_rcvd_size <= le32_to_cpu(fsd->fsd_client_start) ? 0 :
434 (last_rcvd_size - le32_to_cpu(fsd->fsd_client_start)) /
435 le16_to_cpu(fsd->fsd_client_size));
437 if (!obd->obd_replayable) {
438 CWARN("%s: recovery support OFF\n", obd->obd_name);
442 for (cl_idx = 0, off = le32_to_cpu(fsd->fsd_client_start);
443 off < last_rcvd_size; cl_idx++) {
445 struct obd_export *exp;
446 struct filter_export_data *fed;
449 OBD_ALLOC(fcd, sizeof(*fcd));
451 GOTO(err_client, rc = -ENOMEM);
454 /* Don't assume off is incremented properly by
455 * fsfilt_read_record(), in case sizeof(*fcd)
456 * isn't the same as fsd->fsd_client_size. */
457 off = le32_to_cpu(fsd->fsd_client_start) +
458 cl_idx * le16_to_cpu(fsd->fsd_client_size);
459 rc = fsfilt_read_record(obd, filp, fcd, sizeof(*fcd), &off);
461 CERROR("error reading FILT %s idx %d off %llu: rc %d\n",
462 LAST_RCVD, cl_idx, off, rc);
463 break; /* read error shouldn't cause startup to fail */
466 if (fcd->fcd_uuid[0] == '\0') {
467 CDEBUG(D_INFO, "skipping zeroed client at offset %d\n",
472 last_rcvd = le64_to_cpu(fcd->fcd_last_rcvd);
474 /* These exports are cleaned up by filter_disconnect(), so they
475 * need to be set up like real exports as filter_connect() does.
477 exp = class_new_export(obd);
478 CDEBUG(D_HA, "RCVRNG CLIENT uuid: %s idx: %d lr: "LPU64
479 " srv lr: "LPU64"\n", fcd->fcd_uuid, cl_idx,
480 last_rcvd, le64_to_cpu(fsd->fsd_last_transno));
482 GOTO(err_client, rc = -ENOMEM);
484 memcpy(&exp->exp_client_uuid.uuid, fcd->fcd_uuid,
485 sizeof exp->exp_client_uuid.uuid);
486 fed = &exp->exp_filter_data;
488 filter_client_add(obd, filter, fed, cl_idx);
489 /* create helper if export init gets more complex */
490 spin_lock_init(&fed->fed_lock);
493 exp->exp_replay_needed = 1;
494 obd->obd_recoverable_clients++;
495 obd->obd_max_recoverable_clients++;
496 class_export_put(exp);
498 CDEBUG(D_OTHER, "client at idx %d has last_rcvd = "LPU64"\n",
501 if (last_rcvd > le64_to_cpu(fsd->fsd_last_transno))
502 fsd->fsd_last_transno = cpu_to_le64(last_rcvd);
506 obd->obd_last_committed = le64_to_cpu(fsd->fsd_last_transno);
508 if (obd->obd_recoverable_clients) {
509 CWARN("RECOVERY: %d recoverable clients, last_rcvd "
510 LPU64"\n", obd->obd_recoverable_clients,
511 le64_to_cpu(fsd->fsd_last_transno));
512 obd->obd_next_recovery_transno = obd->obd_last_committed + 1;
513 target_start_recovery_thread(obd, ost_handle);
517 OBD_FREE(fcd, sizeof(*fcd));
520 filter->fo_mount_count = mount_count + 1;
521 fsd->fsd_mount_count = cpu_to_le64(filter->fo_mount_count);
523 /* save it, so mount count and last_transno is current */
524 rc = filter_update_server_data(obd, filp, filter->fo_fsd, 1);
529 class_disconnect_exports(obd, 0);
531 filter_free_server_data(filter);
535 static int filter_cleanup_groups(struct obd_device *obd)
537 struct filter_obd *filter = &obd->u.filter;
538 struct dentry *dentry;
542 for (i = 0; i < filter->fo_group_count; i++) {
543 if (filter->fo_subdirs != NULL) {
544 for (k = 0; k < filter->fo_subdir_count; k++) {
545 dentry = filter->fo_subdirs[i].dentry[k];
549 filter->fo_subdirs[i].dentry[k] = NULL;
552 if (filter->fo_last_objid_files[i] != NULL) {
553 filp_close(filter->fo_last_objid_files[i], 0);
554 filter->fo_last_objid_files[i] = NULL;
556 if (filter->fo_groups[i] != NULL) {
557 dput(filter->fo_groups[i]);
558 filter->fo_groups[i] = NULL;
561 if (filter->fo_subdirs != NULL)
562 OBD_FREE(filter->fo_subdirs,
563 filter->fo_group_count * sizeof(*filter->fo_subdirs));
564 if (filter->fo_groups != NULL)
565 OBD_FREE(filter->fo_groups,
566 filter->fo_group_count * sizeof(*filter->fo_groups));
567 if (filter->fo_last_objids != NULL)
568 OBD_FREE(filter->fo_last_objids,
569 filter->fo_group_count * sizeof(__u64));
570 if (filter->fo_last_objid_files != NULL)
571 OBD_FREE(filter->fo_last_objid_files,
572 filter->fo_group_count * sizeof(struct file *));
573 f_dput(filter->fo_dentry_O);
577 static int filter_read_group_internal(struct obd_device *obd, int group,
580 struct filter_obd *filter = &obd->u.filter;
581 __u64 *new_objids = NULL;
582 struct filter_subdirs *new_subdirs = NULL, *tmp_subdirs = NULL;
583 struct dentry **new_groups = NULL;
584 struct file **new_files = NULL;
585 struct dentry *dentry;
587 int old_count = filter->fo_group_count, rc, stage = 0, i;
592 snprintf(name, 24, "%d", group);
596 dentry = ll_lookup_one_len(name, filter->fo_dentry_O,
598 if (IS_ERR(dentry)) {
599 CERROR("Cannot lookup expected object group %d: %ld\n",
600 group, PTR_ERR(dentry));
601 RETURN(PTR_ERR(dentry));
604 dentry = simple_mkdir(filter->fo_dentry_O, name, 0700, 1);
605 if (IS_ERR(dentry)) {
606 CERROR("cannot lookup/create O/%s: rc = %ld\n", name,
608 RETURN(PTR_ERR(dentry));
613 snprintf(name, 24, "O/%d/LAST_ID", group);
615 filp = filp_open(name, O_CREAT | O_RDWR, 0700);
617 CERROR("cannot create %s: rc = %ld\n", name, PTR_ERR(filp));
618 GOTO(cleanup, rc = PTR_ERR(filp));
622 rc = fsfilt_read_record(obd, filp, &last_objid, sizeof(__u64), &off);
624 CDEBUG(D_INODE, "error reading %s: rc %d\n", name, rc);
628 if (filter->fo_subdir_count) {
629 OBD_ALLOC(tmp_subdirs, sizeof(*tmp_subdirs));
630 if (tmp_subdirs == NULL)
631 GOTO(cleanup, rc = -ENOMEM);
634 for (i = 0; i < filter->fo_subdir_count; i++) {
636 snprintf(dir, sizeof(dir), "d%u", i);
638 tmp_subdirs->dentry[i] =
639 simple_mkdir(dentry, dir, 0700, 1);
640 if (IS_ERR(tmp_subdirs->dentry[i])) {
641 rc = PTR_ERR(tmp_subdirs->dentry[i]);
642 CERROR("can't lookup/create O/%d/%s: rc = %d\n",
646 CDEBUG(D_INODE, "got/created O/%d/%s: %p\n", group, dir,
647 tmp_subdirs->dentry[i]);
651 /* 'group' is an index; we need an array of length 'group + 1' */
652 if (group + 1 > old_count) {
654 OBD_ALLOC(new_objids, len * sizeof(*new_objids));
655 OBD_ALLOC(new_subdirs, len * sizeof(*new_subdirs));
656 OBD_ALLOC(new_groups, len * sizeof(*new_groups));
657 OBD_ALLOC(new_files, len * sizeof(*new_files));
659 if (new_objids == NULL || new_subdirs == NULL ||
660 new_groups == NULL || new_files == NULL)
661 GOTO(cleanup, rc = -ENOMEM);
663 memcpy(new_objids, filter->fo_last_objids,
664 old_count * sizeof(*new_objids));
665 memcpy(new_subdirs, filter->fo_subdirs,
666 old_count * sizeof(*new_subdirs));
667 memcpy(new_groups, filter->fo_groups,
668 old_count * sizeof(*new_groups));
669 memcpy(new_files, filter->fo_last_objid_files,
670 old_count * sizeof(*new_files));
673 OBD_FREE(filter->fo_last_objids,
674 old_count * sizeof(*new_objids));
675 OBD_FREE(filter->fo_subdirs,
676 old_count * sizeof(*new_subdirs));
677 OBD_FREE(filter->fo_groups,
678 old_count * sizeof(*new_groups));
679 OBD_FREE(filter->fo_last_objid_files,
680 old_count * sizeof(*new_files));
682 filter->fo_last_objids = new_objids;
683 filter->fo_subdirs = new_subdirs;
684 filter->fo_groups = new_groups;
685 filter->fo_last_objid_files = new_files;
686 filter->fo_group_count = len;
689 filter->fo_groups[group] = dentry;
690 filter->fo_last_objid_files[group] = filp;
691 if (filter->fo_subdir_count) {
692 filter->fo_subdirs[group] = *tmp_subdirs;
693 OBD_FREE(tmp_subdirs, sizeof(*tmp_subdirs));
696 if (filp->f_dentry->d_inode->i_size == 0) {
697 filter->fo_last_objids[group] = FILTER_INIT_OBJID;
701 filter->fo_last_objids[group] = le64_to_cpu(last_objid);
702 CDEBUG(D_INODE, "%s: server last_objid group %d: "LPU64"\n",
703 obd->obd_name, group, last_objid);
708 if (new_objids != NULL)
709 OBD_FREE(new_objids, group * sizeof(*new_objids));
710 if (new_subdirs != NULL)
711 OBD_FREE(new_subdirs, group * sizeof(*new_subdirs));
712 if (new_groups != NULL)
713 OBD_FREE(new_groups, group * sizeof(*new_groups));
714 if (new_files != NULL)
715 OBD_FREE(new_files, group * sizeof(*new_files));
717 if (filter->fo_subdir_count) {
718 for (i = 0; i < filter->fo_subdir_count; i++) {
719 if (tmp_subdirs->dentry[i] != NULL)
720 dput(tmp_subdirs->dentry[i]);
722 OBD_FREE(tmp_subdirs, sizeof(*tmp_subdirs));
732 static int filter_read_groups(struct obd_device *obd, int last_group,
735 struct filter_obd *filter = &obd->u.filter;
736 int old_count = filter->fo_group_count, group = old_count, rc = 0;
738 for (group = old_count; group <= last_group; group++) {
740 continue; /* no group zero */
742 rc = filter_read_group_internal(obd, group, create);
749 static int filter_prep_groups(struct obd_device *obd)
751 struct filter_obd *filter = &obd->u.filter;
752 struct dentry *dentry, *O_dentry;
753 int rc = 0, cleanup_phase = 0;
756 O_dentry = simple_mkdir(current->fs->pwd, "O", 0700, 1);
757 CDEBUG(D_INODE, "got/created O: %p\n", O_dentry);
758 if (IS_ERR(O_dentry)) {
759 rc = PTR_ERR(O_dentry);
760 CERROR("cannot open/create O: rc = %d\n", rc);
763 filter->fo_dentry_O = O_dentry;
764 cleanup_phase = 1; /* O_dentry */
766 /* Lookup "R" to tell if we're on an old OST FS and need to convert
767 * from O/R/<dir>/<objid> to O/0/<dir>/<objid>. This can be removed
768 * some time post 1.0 when all old-style OSTs have converted along
769 * with the init_objid hack. */
770 dentry = ll_lookup_one_len("R", O_dentry, 1);
772 GOTO(cleanup, rc = PTR_ERR(dentry));
773 if (dentry->d_inode && S_ISDIR(dentry->d_inode->i_mode)) {
774 struct dentry *O0_dentry = lookup_one_len("0", O_dentry, 1);
777 CWARN("converting OST to new object layout\n");
778 if (IS_ERR(O0_dentry)) {
779 rc = PTR_ERR(O0_dentry);
780 CERROR("error looking up O/0: rc %d\n", rc);
784 if (O0_dentry->d_inode) {
785 CERROR("Both O/R and O/0 exist. Fix manually.\n");
786 GOTO(cleanup_O0, rc = -EEXIST);
789 down(&O_dentry->d_inode->i_sem);
790 rc = vfs_rename(O_dentry->d_inode, dentry,
791 O_dentry->d_inode, O0_dentry);
792 up(&O_dentry->d_inode->i_sem);
795 CERROR("error renaming O/R to O/0: rc %d\n", rc);
796 GOTO(cleanup_O0, rc);
798 filter->fo_fsd->fsd_feature_incompat |=
799 cpu_to_le32(FILTER_INCOMPAT_GROUPS);
800 rc = filter_update_server_data(obd, filter->fo_rcvd_filp,
802 GOTO(cleanup_O0, rc);
814 cleanup_phase = 2; /* groups */
816 /* Group 0 is no longer a legal group, to catch uninitialized IDs */
817 #define FILTER_MIN_GROUPS 3
818 rc = filter_read_groups(obd, FILTER_MIN_GROUPS, 1);
825 switch (cleanup_phase) {
827 filter_cleanup_groups(obd);
829 f_dput(filter->fo_dentry_O);
830 filter->fo_dentry_O = NULL;
837 /* setup the object store with correct subdirectories */
838 static int filter_prep(struct obd_device *obd)
840 struct lvfs_run_ctxt saved;
841 struct filter_obd *filter = &obd->u.filter;
847 push_ctxt(&saved, &obd->obd_lvfs_ctxt, NULL);
848 file = filp_open(LAST_RCVD, O_RDWR | O_CREAT | O_LARGEFILE, 0700);
849 if (!file || IS_ERR(file)) {
851 CERROR("OBD filter: cannot open/create %s: rc = %d\n",
856 if (!S_ISREG(file->f_dentry->d_inode->i_mode)) {
857 CERROR("%s is not a regular file!: mode = %o\n", LAST_RCVD,
858 file->f_dentry->d_inode->i_mode);
859 GOTO(err_filp, rc = -ENOENT);
862 /* steal operations */
863 inode = file->f_dentry->d_inode;
864 filter->fo_fop = file->f_op;
865 filter->fo_iop = inode->i_op;
866 filter->fo_aops = inode->i_mapping->a_ops;
868 rc = filter_init_server_data(obd, file);
870 CERROR("cannot read %s: rc = %d\n", LAST_RCVD, rc);
873 filter->fo_rcvd_filp = file;
875 rc = filter_prep_groups(obd);
877 GOTO(err_server_data, rc);
880 pop_ctxt(&saved, &obd->obd_lvfs_ctxt, NULL);
885 //class_disconnect_exports(obd, 0);
886 filter_free_server_data(filter);
888 if (filp_close(file, 0))
889 CERROR("can't close %s after error\n", LAST_RCVD);
890 filter->fo_rcvd_filp = NULL;
894 /* cleanup the filter: write last used object id to status file */
895 static void filter_post(struct obd_device *obd)
897 struct lvfs_run_ctxt saved;
898 struct filter_obd *filter = &obd->u.filter;
901 /* XXX: filter_update_lastobjid used to call fsync_dev. It might be
902 * best to start a transaction with h_sync, because we removed this
905 push_ctxt(&saved, &obd->obd_lvfs_ctxt, NULL);
906 rc = filter_update_server_data(obd, filter->fo_rcvd_filp,
909 CERROR("error writing server data: rc = %d\n", rc);
911 for (i = 1; i < filter->fo_group_count; i++) {
912 rc = filter_update_last_objid(obd, i,
913 (i == filter->fo_group_count - 1));
915 CERROR("error writing group %d lastobjid: rc = %d\n",
919 rc = filp_close(filter->fo_rcvd_filp, 0);
920 filter->fo_rcvd_filp = NULL;
922 CERROR("error closing %s: rc = %d\n", LAST_RCVD, rc);
924 filter_cleanup_groups(obd);
925 filter_free_server_data(filter);
926 pop_ctxt(&saved, &obd->obd_lvfs_ctxt, NULL);
929 static void filter_set_last_id(struct filter_obd *filter, int group, obd_id id)
931 LASSERT(filter->fo_fsd != NULL);
933 LASSERT(group < filter->fo_group_count);
935 spin_lock(&filter->fo_objidlock);
936 filter->fo_last_objids[group] = id;
937 spin_unlock(&filter->fo_objidlock);
940 __u64 filter_last_id(struct filter_obd *filter, int group)
943 LASSERT(filter->fo_fsd != NULL);
945 LASSERT(group < filter->fo_group_count);
947 spin_lock(&filter->fo_objidlock);
948 id = filter->fo_last_objids[group];
949 spin_unlock(&filter->fo_objidlock);
954 /* direct cut-n-paste of mds_blocking_ast() */
955 static int filter_blocking_ast(struct ldlm_lock *lock,
956 struct ldlm_lock_desc *desc,
957 void *data, int flag)
962 if (flag == LDLM_CB_CANCELING) {
963 /* Don't need to do anything here. */
967 /* XXX layering violation! -phil */
968 l_lock(&lock->l_resource->lr_namespace->ns_lock);
969 /* Get this: if filter_blocking_ast is racing with ldlm_intent_policy,
970 * such that filter_blocking_ast is called just before l_i_p takes the
971 * ns_lock, then by the time we get the lock, we might not be the
972 * correct blocking function anymore. So check, and return early, if
974 if (lock->l_blocking_ast != filter_blocking_ast) {
975 l_unlock(&lock->l_resource->lr_namespace->ns_lock);
979 lock->l_flags |= LDLM_FL_CBPENDING;
980 do_ast = (!lock->l_readers && !lock->l_writers);
981 l_unlock(&lock->l_resource->lr_namespace->ns_lock);
984 struct lustre_handle lockh;
987 LDLM_DEBUG(lock, "already unused, calling ldlm_cli_cancel");
988 ldlm_lock2handle(lock, &lockh);
989 rc = ldlm_cli_cancel(&lockh);
991 CERROR("ldlm_cli_cancel: %d\n", rc);
993 LDLM_DEBUG(lock, "Lock still has references, will be "
999 extern void *lock_dir(struct inode *dir, struct qstr *name);
1000 extern void unlock_dir(struct inode *dir, void *lock);
1002 static void * filter_lock_dentry(struct obd_device *obd,
1003 struct dentry *dparent, obd_id id)
1009 qstr.len = sprintf(name, LPU64, id);
1010 return lock_dir(dparent->d_inode, &qstr);
1012 down(&dparent->d_inode->i_sem);
1017 /* We never dget the object parent, so DON'T dput it either */
1018 static void filter_parent_unlock(struct dentry *dparent, void *lock)
1021 LASSERT(lock != NULL);
1022 unlock_dir(dparent->d_inode, lock);
1024 up(&dparent->d_inode->i_sem);
1028 /* We never dget the object parent, so DON'T dput it either */
1029 struct dentry *filter_parent(struct obd_device *obd, obd_gr group, obd_id objid)
1031 struct filter_obd *filter = &obd->u.filter;
1032 LASSERT(group < filter->fo_group_count);
1035 if (filter->fo_subdir_count == 0)
1036 return filter->fo_groups[group];
1038 return filter->fo_subdirs[group].dentry[objid & (filter->fo_subdir_count - 1)];
1041 /* We never dget the object parent, so DON'T dput it either */
1042 struct dentry *filter_parent_lock(struct obd_device *obd, obd_gr group,
1043 obd_id objid, void **lock)
1045 unsigned long now = jiffies;
1046 struct dentry *dparent = filter_parent(obd, group, objid);
1048 if (IS_ERR(dparent))
1052 LASSERT(dparent->d_inode);
1054 *lock = filter_lock_dentry(obd, dparent, objid);
1055 if (time_after(jiffies, now + 15 * HZ))
1056 CERROR("slow parent lock %lus\n", (jiffies - now) / HZ);
1060 /* How to get files, dentries, inodes from object id's.
1062 * If dir_dentry is passed, the caller has already locked the parent
1063 * appropriately for this operation (normally a write lock). If
1064 * dir_dentry is NULL, we do a read lock while we do the lookup to
1065 * avoid races with create/destroy and such changing the directory
1066 * internal to the filesystem code. */
1067 struct dentry *filter_fid2dentry(struct obd_device *obd,
1068 struct dentry *dir_dentry,
1069 obd_gr group, obd_id id)
1071 struct dentry *dparent = dir_dentry;
1072 struct dentry *dchild;
1079 CERROR("fatal: invalid object id 0\n");
1080 RETURN(ERR_PTR(-ESTALE));
1083 len = sprintf(name, LPU64, id);
1084 if (dir_dentry == NULL) {
1085 dparent = filter_parent_lock(obd, group, id, &lock);
1086 if (IS_ERR(dparent))
1089 CDEBUG(D_INODE, "looking up object O/%*s/%s\n",
1090 dparent->d_name.len, dparent->d_name.name, name);
1091 dchild = /*ll_*/lookup_one_len(name, dparent, len);
1092 if (dir_dentry == NULL)
1093 filter_parent_unlock(dparent, lock);
1094 if (IS_ERR(dchild)) {
1095 CERROR("child lookup error %ld\n", PTR_ERR(dchild));
1099 CDEBUG(D_INODE, "got child objid %s: %p, count = %d\n",
1100 name, dchild, atomic_read(&dchild->d_count));
1102 LASSERT(atomic_read(&dchild->d_count) > 0);
1107 static int filter_prepare_destroy(struct obd_device *obd, obd_id objid,
1110 struct lustre_handle lockh;
1111 int flags = LDLM_AST_DISCARD_DATA, rc;
1112 struct ldlm_res_id res_id = { .name = { objid, 0, group, 0 } };
1113 ldlm_policy_data_t policy = { .l_extent = { 0, OBD_OBJECT_EOF } };
1116 /* Tell the clients that the object is gone now and that they should
1117 * throw away any cached pages. */
1118 rc = ldlm_cli_enqueue(NULL, NULL, obd->obd_namespace, res_id,
1119 LDLM_EXTENT, &policy, LCK_PW,
1120 &flags, filter_blocking_ast, ldlm_completion_ast,
1121 NULL, NULL, NULL, 0, NULL, &lockh);
1123 /* We only care about the side-effects, just drop the lock. */
1125 ldlm_lock_decref(&lockh, LCK_PW);
1130 /* Caller must hold LCK_PW on parent and push us into kernel context.
1131 * Caller is also required to ensure that dchild->d_inode exists. */
1132 static int filter_destroy_internal(struct obd_device *obd, obd_id objid,
1133 struct dentry *dparent,
1134 struct dentry *dchild)
1136 struct inode *inode = dchild->d_inode;
1140 if (inode->i_nlink != 1 || atomic_read(&inode->i_count) != 1) {
1141 CERROR("destroying objid %*s nlink = %lu, count = %d\n",
1142 dchild->d_name.len, dchild->d_name.name,
1143 (unsigned long)inode->i_nlink,
1144 atomic_read(&inode->i_count));
1147 rc = vfs_unlink(dparent->d_inode, dchild);
1150 CERROR("error unlinking objid %*s: rc %d\n",
1151 dchild->d_name.len, dchild->d_name.name, rc);
1156 static int filter_intent_policy(struct ldlm_namespace *ns,
1157 struct ldlm_lock **lockp, void *req_cookie,
1158 ldlm_mode_t mode, int flags, void *data)
1160 struct list_head rpc_list = LIST_HEAD_INIT(rpc_list);
1161 struct ptlrpc_request *req = req_cookie;
1162 struct ldlm_lock *lock = *lockp, *l = NULL;
1163 struct ldlm_resource *res = lock->l_resource;
1164 ldlm_processing_policy policy;
1165 struct ost_lvb *res_lvb, *reply_lvb;
1166 struct list_head *tmp;
1168 int tmpflags = 0, rc, repsize[2] = {sizeof(struct ldlm_reply),
1169 sizeof(struct ost_lvb) };
1172 policy = ldlm_get_processing_policy(res);
1173 LASSERT(policy != NULL);
1174 LASSERT(req != NULL);
1176 rc = lustre_pack_reply(req, 2, repsize, NULL);
1178 RETURN(req->rq_status = rc);
1180 reply_lvb = lustre_msg_buf(req->rq_repmsg, 1, sizeof(*reply_lvb));
1181 LASSERT(reply_lvb != NULL);
1183 //fixup_handle_for_resent_req(req, lock, &lockh);
1185 /* If we grant any lock at all, it will be a whole-file read lock.
1186 * Call the extent policy function to see if our request can be
1187 * granted, or is blocked. */
1188 lock->l_policy_data.l_extent.start = 0;
1189 lock->l_policy_data.l_extent.end = OBD_OBJECT_EOF;
1190 lock->l_req_mode = LCK_PR;
1192 l_lock(&res->lr_namespace->ns_lock);
1194 res->lr_tmp = &rpc_list;
1195 rc = policy(lock, &tmpflags, 0, &err);
1198 /* FIXME: we should change the policy function slightly, to not make
1199 * this list at all, since we just turn around and free it */
1200 while (!list_empty(&rpc_list)) {
1201 struct ldlm_ast_work *w =
1202 list_entry(rpc_list.next, struct ldlm_ast_work, w_list);
1203 list_del(&w->w_list);
1204 LDLM_LOCK_PUT(w->w_lock);
1205 OBD_FREE(w, sizeof(*w));
1208 if (rc == LDLM_ITER_CONTINUE) {
1209 /* The lock met with no resistance; we're finished. */
1210 l_unlock(&res->lr_namespace->ns_lock);
1211 RETURN(ELDLM_LOCK_REPLACED);
1214 /* Do not grant any lock, but instead send GL callbacks. The extent
1215 * policy nicely created a list of all PW locks for us. We will choose
1216 * the highest of those which are larger than the size in the LVB, if
1217 * any, and perform a glimpse callback. */
1218 down(&res->lr_lvb_sem);
1219 res_lvb = res->lr_lvb_data;
1220 LASSERT(res_lvb != NULL);
1221 reply_lvb->lvb_size = res_lvb->lvb_size;
1222 reply_lvb->lvb_blocks = res_lvb->lvb_blocks;
1223 up(&res->lr_lvb_sem);
1225 list_for_each(tmp, &res->lr_granted) {
1226 struct ldlm_lock *tmplock =
1227 list_entry(tmp, struct ldlm_lock, l_res_link);
1229 if (tmplock->l_granted_mode == LCK_PR)
1232 if (tmplock->l_policy_data.l_extent.end <= reply_lvb->lvb_size)
1236 l = LDLM_LOCK_GET(tmplock);
1240 if (l->l_policy_data.l_extent.start >
1241 tmplock->l_policy_data.l_extent.start)
1245 l = LDLM_LOCK_GET(tmplock);
1247 l_unlock(&res->lr_namespace->ns_lock);
1249 /* There were no PW locks beyond the size in the LVB; finished. */
1251 RETURN(ELDLM_LOCK_ABORTED);
1253 LASSERT(l->l_glimpse_ast != NULL);
1254 rc = l->l_glimpse_ast(l, NULL); /* this will update the LVB */
1255 if (rc != 0 && res->lr_namespace->ns_lvbo &&
1256 res->lr_namespace->ns_lvbo->lvbo_update) {
1257 res->lr_namespace->ns_lvbo->lvbo_update(res, NULL, 0, 1);
1260 down(&res->lr_lvb_sem);
1261 reply_lvb->lvb_size = res_lvb->lvb_size;
1262 reply_lvb->lvb_blocks = res_lvb->lvb_blocks;
1263 up(&res->lr_lvb_sem);
1267 RETURN(ELDLM_LOCK_ABORTED);
1270 static int filter_post_fs_cleanup(struct obd_device *obd)
1274 rc = fsfilt_post_cleanup(obd);
1279 static int filter_group_set_fs_flags(struct obd_device *obd, int group)
1281 struct filter_obd *filter = &obd->u.filter;
1285 /* zero group is not longer valid. */
1288 for (i = 0; i < filter->fo_subdir_count; i++) {
1289 struct dentry *dentry;
1290 dentry = (filter->fo_subdirs + group)->dentry[i];
1291 rc = fsfilt_set_fs_flags(obd, dentry->d_inode,
1292 SM_DO_REC | SM_DO_COW);
1298 static int filter_post_fs_setup(struct obd_device *obd)
1300 struct filter_obd *filter = &obd->u.filter;
1302 struct llog_ctxt *ctxt = NULL;
1304 rc = fsfilt_post_setup(obd);
1308 for (j = 0; j < filter->fo_group_count; j++) {
1309 rc = filter_group_set_fs_flags(obd, j);
1314 fsfilt_get_reint_log_ctxt(obd, filter->fo_sb, &ctxt);
1316 ctxt->loc_obd = obd;
1317 ctxt->loc_idx = LLOG_REINT_ORIG_CTXT;
1318 obd->obd_llog_ctxt[LLOG_REINT_ORIG_CTXT] = ctxt;
1320 fsfilt_set_ost_flags(obd, filter->fo_sb);
1324 /* mount the file system (secretly) */
1325 int filter_common_setup(struct obd_device *obd, obd_count len,
1326 void *buf, char *option)
1328 struct lustre_cfg* lcfg = buf;
1329 struct filter_obd *filter = &obd->u.filter;
1330 struct vfsmount *mnt;
1331 char name[32] = "CATLIST";
1335 dev_clear_rdonly(2);
1337 if (!lcfg->lcfg_inlbuf1 || !lcfg->lcfg_inlbuf2)
1340 obd->obd_fsops = fsfilt_get_ops(lcfg->lcfg_inlbuf2);
1341 if (IS_ERR(obd->obd_fsops))
1342 RETURN(PTR_ERR(obd->obd_fsops));
1344 mnt = do_kern_mount(lcfg->lcfg_inlbuf2, MS_NOATIME | MS_NODIRATIME,
1345 lcfg->lcfg_inlbuf1, option);
1350 if (lcfg->lcfg_inllen3 > 0 && lcfg->lcfg_inlbuf3) {
1351 if (*lcfg->lcfg_inlbuf3 == 'f') {
1352 obd->obd_replayable = 1;
1353 obd_sync_filter = 1;
1354 CWARN("%s: recovery enabled\n", obd->obd_name);
1356 if (*lcfg->lcfg_inlbuf3 != 'n') {
1357 CERROR("unrecognised flag '%c'\n",
1358 *lcfg->lcfg_inlbuf3);
1360 // XXX Robert? Why do we get errors here
1361 // GOTO(err_mntput, rc = -EINVAL);
1365 filter->fo_vfsmnt = mnt;
1366 filter->fo_sb = mnt->mnt_sb;
1367 filter->fo_fstype = mnt->mnt_sb->s_type->name;
1368 CDEBUG(D_SUPER, "%s: mnt = %p\n", filter->fo_fstype, mnt);
1370 OBD_SET_CTXT_MAGIC(&obd->obd_lvfs_ctxt);
1371 obd->obd_lvfs_ctxt.pwdmnt = mnt;
1372 obd->obd_lvfs_ctxt.pwd = mnt->mnt_root;
1373 obd->obd_lvfs_ctxt.fs = get_ds();
1374 obd->obd_lvfs_ctxt.cb_ops = filter_lvfs_ops;
1376 rc = fsfilt_setup(obd, mnt->mnt_sb);
1378 GOTO(err_mntput, rc);
1380 rc = filter_prep(obd);
1382 GOTO(err_mntput, rc);
1385 filter->fo_destroy_in_progress = 0;
1386 sema_init(&filter->fo_create_lock, 1);
1388 spin_lock_init(&filter->fo_translock);
1389 spin_lock_init(&filter->fo_objidlock);
1390 INIT_LIST_HEAD(&filter->fo_export_list);
1391 sema_init(&filter->fo_alloc_lock, 1);
1392 spin_lock_init(&filter->fo_r_pages.oh_lock);
1393 spin_lock_init(&filter->fo_w_pages.oh_lock);
1394 spin_lock_init(&filter->fo_r_discont_pages.oh_lock);
1395 spin_lock_init(&filter->fo_w_discont_pages.oh_lock);
1396 spin_lock_init(&filter->fo_r_discont_blocks.oh_lock);
1397 spin_lock_init(&filter->fo_w_discont_blocks.oh_lock);
1398 filter->fo_readcache_max_filesize = FILTER_MAX_CACHE_SIZE;
1400 INIT_LIST_HEAD(&filter->fo_llog_list);
1401 spin_lock_init(&filter->fo_llog_list_lock);
1403 obd->obd_namespace = ldlm_namespace_new("filter-tgt",
1404 LDLM_NAMESPACE_SERVER);
1405 if (obd->obd_namespace == NULL)
1406 GOTO(err_post, rc = -ENOMEM);
1407 obd->obd_namespace->ns_lvbp = obd;
1408 obd->obd_namespace->ns_lvbo = &filter_lvbo;
1409 ldlm_register_intent(obd->obd_namespace, filter_intent_policy);
1411 ptlrpc_init_client(LDLM_CB_REQUEST_PORTAL, LDLM_CB_REPLY_PORTAL,
1412 "filter_ldlm_cb_client", &obd->obd_ldlm_client);
1414 rc = obd_llog_cat_initialize(obd, &obd->obd_llogs, 1, name);
1416 CERROR("failed to setup llogging subsystems\n");
1429 fsfilt_put_ops(obd->obd_fsops);
1433 static int filter_attach(struct obd_device *obd, obd_count len, void *data)
1435 struct lprocfs_static_vars lvars;
1438 lprocfs_init_vars(filter, &lvars);
1439 rc = lprocfs_obd_attach(obd, lvars.obd_vars);
1443 rc = lprocfs_alloc_obd_stats(obd, LPROC_FILTER_LAST);
1447 /* Init obdfilter private stats here */
1448 lprocfs_counter_init(obd->obd_stats, LPROC_FILTER_READ_BYTES,
1449 LPROCFS_CNTR_AVGMINMAX, "read_bytes", "bytes");
1450 lprocfs_counter_init(obd->obd_stats, LPROC_FILTER_WRITE_BYTES,
1451 LPROCFS_CNTR_AVGMINMAX, "write_bytes", "bytes");
1453 return lproc_filter_attach_seqstat(obd);
1456 static int filter_detach(struct obd_device *dev)
1458 lprocfs_free_obd_stats(dev);
1459 return lprocfs_obd_detach(dev);
1462 static int filter_setup(struct obd_device *obd, obd_count len, void *buf)
1464 struct lustre_cfg* lcfg = buf;
1467 /* all mount options including errors=remount-ro and asyncdel are passed
1468 * using 4th lcfg param. And it is good, finally we have got rid of
1469 * hardcoded fs types in the code. */
1470 rc = filter_common_setup(obd, len, buf, lcfg->lcfg_inlbuf4);
1473 rc = filter_post_fs_setup(obd);
1477 static int filter_cleanup(struct obd_device *obd, int flags)
1479 struct filter_obd *filter = &obd->u.filter;
1482 if (flags & OBD_OPT_FAILOVER)
1483 CERROR("%s: shutting down for failover; client state will"
1484 " be preserved.\n", obd->obd_name);
1486 if (!list_empty(&obd->obd_exports)) {
1487 CERROR("%s: still has clients!\n", obd->obd_name);
1488 class_disconnect_exports(obd, flags);
1489 if (!list_empty(&obd->obd_exports)) {
1490 CERROR("still has exports after forced cleanup?\n");
1495 ldlm_namespace_free(obd->obd_namespace, flags & OBD_OPT_FORCE);
1497 if (filter->fo_sb == NULL)
1500 filter_post_fs_cleanup(obd);
1503 shrink_dcache_parent(filter->fo_sb->s_root);
1506 if (atomic_read(&filter->fo_vfsmnt->mnt_count) > 1)
1507 CERROR("%s: mount point %p busy, mnt_count: %d\n",
1508 obd->obd_name, filter->fo_vfsmnt,
1509 atomic_read(&filter->fo_vfsmnt->mnt_count));
1512 mntput(filter->fo_vfsmnt);
1513 //destroy_buffers(filter->fo_sb->s_dev);
1514 filter->fo_sb = NULL;
1515 fsfilt_put_ops(obd->obd_fsops);
1518 dev_clear_rdonly(2);
1523 /* nearly identical to mds_connect */
1524 static int filter_connect(struct lustre_handle *conn, struct obd_device *obd,
1525 struct obd_uuid *cluuid)
1527 struct obd_export *exp;
1528 struct filter_export_data *fed;
1529 struct filter_client_data *fcd = NULL;
1530 struct filter_obd *filter = &obd->u.filter;
1534 if (conn == NULL || obd == NULL || cluuid == NULL)
1537 rc = class_connect(conn, obd, cluuid);
1540 exp = class_conn2export(conn);
1541 LASSERT(exp != NULL);
1543 fed = &exp->exp_filter_data;
1545 spin_lock_init(&fed->fed_lock);
1547 if (!obd->obd_replayable)
1548 GOTO(cleanup, rc = 0);
1550 OBD_ALLOC(fcd, sizeof(*fcd));
1552 CERROR("filter: out of memory for client data\n");
1553 GOTO(cleanup, rc = -ENOMEM);
1556 memcpy(fcd->fcd_uuid, cluuid, sizeof(fcd->fcd_uuid));
1559 rc = filter_client_add(obd, filter, fed, -1);
1564 OBD_FREE(fcd, sizeof(*fcd));
1565 class_disconnect(exp, 0);
1567 class_export_put(exp);
1572 static int filter_precleanup(struct obd_device *obd, int flags)
1574 struct filter_group_llog *log;
1575 struct filter_obd *filter;
1579 filter = &obd->u.filter;
1581 spin_lock(&filter->fo_llog_list_lock);
1582 while (!list_empty(&filter->fo_llog_list)) {
1583 log = list_entry(filter->fo_llog_list.next,
1584 struct filter_group_llog, list);
1585 list_del(&log->list);
1586 spin_unlock(&filter->fo_llog_list_lock);
1588 rc = obd_llog_finish(obd, log->llogs, 0);
1590 CERROR("failed to cleanup llogging subsystem for %u\n",
1592 OBD_FREE(log->llogs, sizeof(*(log->llogs)));
1593 OBD_FREE(log, sizeof(*log));
1594 spin_lock(&filter->fo_llog_list_lock);
1596 spin_unlock(&filter->fo_llog_list_lock);
1598 rc = obd_llog_finish(obd, &obd->obd_llogs, 0);
1600 CERROR("failed to cleanup llogging subsystem\n");
1605 /* Do extra sanity checks for grant accounting. We do this at connect,
1606 * disconnect, and statfs RPC time, so it shouldn't be too bad. We can
1607 * always get rid of it or turn it off when we know accounting is good. */
1608 static void filter_grant_sanity_check(struct obd_device *obd, const char *func)
1610 struct filter_export_data *fed;
1611 struct obd_export *exp;
1612 obd_size maxsize = obd->obd_osfs.os_blocks * obd->obd_osfs.os_bsize;
1613 obd_size tot_dirty = 0, tot_pending = 0, tot_granted = 0;
1614 obd_size fo_tot_dirty, fo_tot_pending, fo_tot_granted;
1616 if (list_empty(&obd->obd_exports))
1619 spin_lock(&obd->obd_osfs_lock);
1620 spin_lock(&obd->obd_dev_lock);
1621 list_for_each_entry(exp, &obd->obd_exports, exp_obd_chain) {
1622 fed = &exp->exp_filter_data;
1623 LASSERTF(fed->fed_grant + fed->fed_pending <= maxsize,
1624 "cli %s/%p %lu+%lu > "LPU64"\n",
1625 exp->exp_client_uuid.uuid, exp,
1626 fed->fed_grant, fed->fed_pending, maxsize);
1627 LASSERTF(fed->fed_dirty <= maxsize, "cli %s/%p %lu > "LPU64"\n",
1628 exp->exp_client_uuid.uuid, exp,fed->fed_dirty,maxsize);
1629 CDEBUG(D_CACHE,"%s: cli %s/%p dirty %lu pend %lu grant %lu\n",
1630 obd->obd_name, exp->exp_client_uuid.uuid, exp,
1631 fed->fed_dirty, fed->fed_pending, fed->fed_grant);
1632 tot_granted += fed->fed_grant + fed->fed_pending;
1633 tot_pending += fed->fed_pending;
1634 tot_dirty += fed->fed_dirty;
1636 fo_tot_granted = obd->u.filter.fo_tot_granted;
1637 fo_tot_pending = obd->u.filter.fo_tot_pending;
1638 fo_tot_dirty = obd->u.filter.fo_tot_dirty;
1639 spin_unlock(&obd->obd_dev_lock);
1640 spin_unlock(&obd->obd_osfs_lock);
1642 /* Do these assertions outside the spinlocks so we don't kill system */
1643 if (tot_granted != fo_tot_granted)
1644 CERROR("%s: tot_granted "LPU64" != fo_tot_granted "LPU64"\n",
1645 func, tot_granted, fo_tot_granted);
1646 if (tot_pending != fo_tot_pending)
1647 CERROR("%s: tot_pending "LPU64" != fo_tot_pending "LPU64"\n",
1648 func, tot_pending, fo_tot_pending);
1649 if (tot_dirty != fo_tot_dirty)
1650 CERROR("%s: tot_dirty "LPU64" != fo_tot_dirty "LPU64"\n",
1651 func, tot_dirty, fo_tot_dirty);
1652 if (tot_pending > tot_granted)
1653 CERROR("%s: tot_pending "LPU64" > tot_granted "LPU64"\n",
1654 func, tot_pending, tot_granted);
1655 if (tot_granted > maxsize)
1656 CERROR("%s: tot_granted "LPU64" > maxsize "LPU64"\n",
1657 func, tot_granted, maxsize);
1658 if (tot_dirty > maxsize)
1659 CERROR("%s: tot_dirty "LPU64" > maxsize "LPU64"\n",
1660 func, tot_dirty, maxsize);
1663 /* Remove this client from the grant accounting totals. We also remove
1664 * the export from the obd device under the osfs and dev locks to ensure
1665 * that the filter_grant_sanity_check() calculations are always valid.
1666 * The client should do something similar when it invalidates its import. */
1667 static void filter_grant_discard(struct obd_export *exp)
1669 struct obd_device *obd = exp->exp_obd;
1670 struct filter_obd *filter = &obd->u.filter;
1671 struct filter_export_data *fed = &exp->exp_filter_data;
1673 spin_lock(&obd->obd_osfs_lock);
1674 spin_lock(&exp->exp_obd->obd_dev_lock);
1675 list_del_init(&exp->exp_obd_chain);
1676 spin_unlock(&exp->exp_obd->obd_dev_lock);
1678 CDEBUG(D_CACHE, "%s: cli %s/%p dirty %lu pend %lu grant %lu\n",
1679 obd->obd_name, exp->exp_client_uuid.uuid, exp,
1680 fed->fed_dirty, fed->fed_pending, fed->fed_grant);
1682 LASSERTF(filter->fo_tot_granted >= fed->fed_grant,
1683 "%s: tot_granted "LPU64" cli %s/%p fed_grant %lu\n",
1684 obd->obd_name, filter->fo_tot_granted,
1685 exp->exp_client_uuid.uuid, exp, fed->fed_grant);
1686 filter->fo_tot_granted -= fed->fed_grant;
1687 LASSERTF(exp->exp_obd->u.filter.fo_tot_pending >= fed->fed_pending,
1688 "%s: tot_pending "LPU64" cli %s/%p fed_pending %lu\n",
1689 obd->obd_name, filter->fo_tot_pending,
1690 exp->exp_client_uuid.uuid, exp, fed->fed_pending);
1691 LASSERTF(filter->fo_tot_dirty >= fed->fed_dirty,
1692 "%s: tot_dirty "LPU64" cli %s/%p fed_dirty %lu\n",
1693 obd->obd_name, filter->fo_tot_dirty,
1694 exp->exp_client_uuid.uuid, exp, fed->fed_dirty);
1695 filter->fo_tot_dirty -= fed->fed_dirty;
1699 spin_unlock(&obd->obd_osfs_lock);
1702 static int filter_destroy_export(struct obd_export *exp)
1706 if (exp->exp_filter_data.fed_pending)
1707 CERROR("%s: cli %s/%p has %lu pending on destroyed export\n",
1708 exp->exp_obd->obd_name, exp->exp_client_uuid.uuid,
1709 exp, exp->exp_filter_data.fed_pending);
1711 target_destroy_export(exp);
1713 if (exp->exp_obd->obd_replayable)
1714 filter_client_free(exp, exp->exp_flags);
1716 filter_grant_discard(exp);
1717 if (!(exp->exp_flags & OBD_OPT_FORCE))
1718 filter_grant_sanity_check(exp->exp_obd, __FUNCTION__);
1723 static void filter_sync_llogs(struct obd_export *dexp)
1725 struct filter_group_llog *fglog, *nlog;
1726 struct obd_device *obd = dexp->exp_obd;
1727 struct filter_obd *filter;
1728 int worked = 0, group;
1729 struct llog_ctxt *ctxt;
1731 filter = &obd->u.filter;
1733 /* we can't sync log holding spinlock. also, we do not want to get
1734 * into livelock. so we do following: loop over MDS's exports in
1735 * group order and skip already synced llogs -bzzz */
1737 /* look for group with min. number, but > worked */
1740 spin_lock(&filter->fo_llog_list_lock);
1741 list_for_each_entry(nlog, &filter->fo_llog_list, list) {
1743 if (nlog->group <= worked) {
1744 /* this group is already synced */
1748 if (group < nlog->group) {
1749 /* we have group with smaller number to sync */
1753 /* store current minimal group */
1755 group = nlog->group;
1757 spin_unlock(&filter->fo_llog_list_lock);
1760 worked = fglog->group;
1761 ctxt = llog_get_context(fglog->llogs,
1762 LLOG_UNLINK_REPL_CTXT);
1763 llog_sync(ctxt, dexp);
1765 } while (fglog != NULL);
1768 /* also incredibly similar to mds_disconnect */
1769 static int filter_disconnect(struct obd_export *exp, int flags)
1771 struct obd_device *obd = exp->exp_obd;
1772 unsigned long irqflags;
1777 class_export_get(exp);
1779 spin_lock_irqsave(&exp->exp_lock, irqflags);
1780 exp->exp_flags = flags;
1781 spin_unlock_irqrestore(&exp->exp_lock, irqflags);
1783 if (!(flags & OBD_OPT_FORCE))
1784 filter_grant_sanity_check(obd, __FUNCTION__);
1785 filter_grant_discard(exp);
1787 /* Disconnect early so that clients can't keep using export */
1788 rc = class_disconnect(exp, flags);
1790 ldlm_cancel_locks_for_export(exp);
1792 fsfilt_sync(obd, obd->u.filter.fo_sb);
1794 /* flush any remaining cancel messages out to the target */
1795 filter_sync_llogs(exp);
1797 class_export_put(exp);
1801 struct dentry *__filter_oa2dentry(struct obd_device *obd,
1802 struct obdo *oa, const char *what)
1804 struct dentry *dchild = NULL;
1807 if (oa->o_valid & OBD_MD_FLGROUP)
1810 dchild = filter_fid2dentry(obd, NULL, group, oa->o_id);
1812 if (IS_ERR(dchild)) {
1813 CERROR("%s error looking up object: "LPU64"\n", what, oa->o_id);
1817 if (dchild->d_inode == NULL) {
1818 CERROR("%s: %s on non-existent object: "LPU64"\n",
1819 obd->obd_name, what, oa->o_id);
1821 RETURN(ERR_PTR(-ENOENT));
1827 static int filter_getattr(struct obd_export *exp, struct obdo *oa,
1828 struct lov_stripe_md *md)
1830 struct dentry *dentry = NULL;
1831 struct obd_device *obd;
1835 obd = class_exp2obd(exp);
1837 CDEBUG(D_IOCTL, "invalid client cookie "LPX64"\n",
1838 exp->exp_handle.h_cookie);
1842 dentry = filter_oa2dentry(obd, oa);
1844 RETURN(PTR_ERR(dentry));
1846 /* Limit the valid bits in the return data to what we actually use */
1847 oa->o_valid = OBD_MD_FLID;
1848 obdo_from_inode(oa, dentry->d_inode, FILTER_VALID_FLAGS);
1854 /* this is called from filter_truncate() until we have filter_punch() */
1855 static int filter_setattr(struct obd_export *exp, struct obdo *oa,
1856 struct lov_stripe_md *md, struct obd_trans_info *oti)
1858 struct lvfs_run_ctxt saved;
1859 struct filter_obd *filter;
1860 struct dentry *dentry;
1862 struct ldlm_res_id res_id = { .name = { oa->o_id, 0, oa->o_gr, 0 } };
1863 struct ldlm_resource *res;
1868 LASSERT(oti != NULL);
1870 dentry = filter_oa2dentry(exp->exp_obd, oa);
1872 RETURN(PTR_ERR(dentry));
1874 filter = &exp->exp_obd->u.filter;
1876 iattr_from_obdo(&iattr, oa, oa->o_valid);
1878 push_ctxt(&saved, &exp->exp_obd->obd_lvfs_ctxt, NULL);
1881 if (iattr.ia_valid & ATTR_SIZE)
1882 down(&dentry->d_inode->i_sem);
1883 handle = fsfilt_start(exp->exp_obd, dentry->d_inode, FSFILT_OP_SETATTR,
1886 GOTO(out_unlock, rc = PTR_ERR(handle));
1888 /* XXX this could be a rwsem instead, if filter_preprw played along */
1889 if (iattr.ia_valid & ATTR_ATTR_FLAG)
1890 rc = fsfilt_iocontrol(exp->exp_obd, dentry->d_inode, NULL,
1892 (long)&iattr.ia_attr_flags);
1894 rc = fsfilt_setattr(exp->exp_obd, dentry, handle, &iattr, 1);
1895 rc = filter_finish_transno(exp, oti, rc);
1896 rc2 = fsfilt_commit(exp->exp_obd, filter->fo_sb, dentry->d_inode, handle, 0);
1898 CERROR("error on commit, err = %d\n", rc2);
1903 if (iattr.ia_valid & ATTR_SIZE) {
1904 res = ldlm_resource_get(exp->exp_obd->obd_namespace, NULL,
1905 res_id, LDLM_EXTENT, 0);
1907 CERROR("!!! resource_get failed for object "LPU64" -- "
1908 "filter_setattr with no lock?\n", oa->o_id);
1910 if (res->lr_namespace->ns_lvbo &&
1911 res->lr_namespace->ns_lvbo->lvbo_update) {
1912 rc = res->lr_namespace->ns_lvbo->lvbo_update
1915 ldlm_resource_putref(res);
1919 oa->o_valid = OBD_MD_FLID;
1920 obdo_from_inode(oa, dentry->d_inode, FILTER_VALID_FLAGS);
1923 if (iattr.ia_valid & ATTR_SIZE)
1924 up(&dentry->d_inode->i_sem);
1926 pop_ctxt(&saved, &exp->exp_obd->obd_lvfs_ctxt, NULL);
1932 /* XXX identical to osc_unpackmd */
1933 static int filter_unpackmd(struct obd_export *exp, struct lov_stripe_md **lsmp,
1934 struct lov_mds_md *lmm, int lmm_bytes)
1940 if (lmm_bytes < sizeof (*lmm)) {
1941 CERROR("lov_mds_md too small: %d, need %d\n",
1942 lmm_bytes, (int)sizeof(*lmm));
1945 /* XXX LOV_MAGIC etc check? */
1947 if (lmm->lmm_object_id == cpu_to_le64(0)) {
1948 CERROR("lov_mds_md: zero lmm_object_id\n");
1953 lsm_size = lov_stripe_md_size(1);
1957 if (*lsmp != NULL && lmm == NULL) {
1958 OBD_FREE(*lsmp, lsm_size);
1963 if (*lsmp == NULL) {
1964 OBD_ALLOC(*lsmp, lsm_size);
1968 loi_init((*lsmp)->lsm_oinfo);
1972 /* XXX zero *lsmp? */
1973 (*lsmp)->lsm_object_id = le64_to_cpu (lmm->lmm_object_id);
1974 LASSERT((*lsmp)->lsm_object_id);
1977 (*lsmp)->lsm_maxbytes = LUSTRE_STRIPE_MAXBYTES;
1982 static void filter_destroy_precreated(struct obd_export *exp, struct obdo *oa,
1983 struct filter_obd *filter)
1985 struct obdo doa; /* XXX obdo on stack */
1990 memset(&doa, 0, sizeof(doa));
1991 if (oa->o_valid & OBD_MD_FLGROUP) {
1992 doa.o_valid |= OBD_MD_FLGROUP;
1993 doa.o_gr = oa->o_gr;
1997 doa.o_mode = S_IFREG;
1998 doa.o_gr = oa->o_gr;
1999 doa.o_valid = oa->o_valid & OBD_MD_FLGROUP;
2001 filter->fo_destroy_in_progress = 1;
2002 down(&filter->fo_create_lock);
2003 if (!filter->fo_destroy_in_progress) {
2004 CERROR("%s: destroy_in_progress already cleared\n",
2005 exp->exp_obd->obd_name);
2006 up(&filter->fo_create_lock);
2011 last = filter_last_id(filter, doa.o_gr);
2012 CWARN("%s: deleting orphan objects from "LPU64" to "LPU64"\n",
2013 exp->exp_obd->obd_name, oa->o_id + 1, last);
2014 for (id = oa->o_id + 1; id <= last; id++) {
2016 filter_destroy(exp, &doa, NULL, NULL);
2019 CDEBUG(D_HA, "%s: after destroy: set last_objids["LPU64"] = "LPU64"\n",
2020 exp->exp_obd->obd_name, doa.o_gr, oa->o_id);
2022 spin_lock(&filter->fo_objidlock);
2023 filter->fo_last_objids[doa.o_gr] = oa->o_id;
2024 spin_unlock(&filter->fo_objidlock);
2026 filter->fo_destroy_in_progress = 0;
2027 up(&filter->fo_create_lock);
2032 /* returns a negative error or a nonnegative number of files to create */
2033 static int filter_should_precreate(struct obd_export *exp, struct obdo *oa,
2036 struct obd_device *obd = exp->exp_obd;
2037 struct filter_obd *filter = &obd->u.filter;
2041 diff = oa->o_id - filter_last_id(filter, oa->o_gr);
2042 CDEBUG(D_INFO, "filter_last_id() = "LPU64" -> diff = %d\n",
2043 filter_last_id(filter, oa->o_gr), diff);
2045 /* delete orphans request */
2046 if ((oa->o_valid & OBD_MD_FLFLAGS) &&
2047 (oa->o_flags & OBD_FL_DELORPHAN)) {
2050 if (-diff > 10000) { /* XXX make this smarter */
2051 CERROR("ignoring bogus orphan destroy request: obdid "
2052 LPU64" last_id "LPU64"\n",
2053 oa->o_id, filter_last_id(filter, oa->o_gr));
2056 filter_destroy_precreated(exp, oa, filter);
2057 rc = filter_update_last_objid(obd, group, 0);
2059 CERROR("unable to write lastobjid, but orphans"
2063 /* only precreate if group == 0 and o_id is specfied */
2064 if (!(oa->o_valid & OBD_FL_DELORPHAN) &&
2065 (/*group != 0 ||*/ oa->o_id == 0))
2072 static int filter_precreate_rec(struct obd_device *obd, struct dentry *dentry,
2073 int *number, struct obdo *oa)
2078 rc = fsfilt_precreate_rec(obd, dentry, number, oa);
2083 /* We rely on the fact that only one thread will be creating files in a given
2084 * group at a time, which is why we don't need an atomic filter_get_new_id.
2085 * Even if we had that atomic function, the following race would exist:
2087 * thread 1: gets id x from filter_next_id
2088 * thread 2: gets id (x + 1) from filter_next_id
2089 * thread 2: creates object (x + 1)
2090 * thread 1: tries to create object x, gets -ENOSPC
2092 static int filter_precreate(struct obd_device *obd, struct obdo *oa,
2093 obd_gr group, int *num)
2095 struct dentry *dchild = NULL, *dparent = NULL;
2096 struct filter_obd *filter;
2097 int err = 0, rc = 0, recreate_obj = 0, i;
2099 void *handle = NULL;
2103 filter = &obd->u.filter;
2105 if ((oa->o_valid & OBD_MD_FLFLAGS) &&
2106 (oa->o_flags & OBD_FL_RECREATE_OBJS)) {
2110 CDEBUG(D_HA, "%s: precreating %d objects\n", obd->obd_name, *num);
2112 down(&filter->fo_create_lock);
2114 for (i = 0; i < *num && err == 0; i++) {
2115 int cleanup_phase = 0;
2117 if (filter->fo_destroy_in_progress) {
2118 CWARN("%s: precreate aborted by destroy\n",
2126 last_id = filter_last_id(filter, group);
2127 if (next_id > last_id) {
2128 CERROR("Error: Trying to recreate obj greater"
2129 "than last id "LPD64" > "LPD64"\n",
2131 GOTO(cleanup, rc = -EINVAL);
2134 next_id = filter_last_id(filter, group) + 1;
2137 CDEBUG(D_INFO, "precreate objid "LPU64"\n", next_id);
2139 dparent = filter_parent_lock(obd, group, next_id, &lock);
2140 if (IS_ERR(dparent))
2141 GOTO(cleanup, rc = PTR_ERR(dparent));
2144 /*only do precreate rec record. so clean kml flags here*/
2145 fsfilt_clear_fs_flags(obd, dparent->d_inode,
2148 dchild = filter_fid2dentry(obd, dparent, group, next_id);
2150 GOTO(cleanup, rc = PTR_ERR(dchild));
2153 if (dchild->d_inode != NULL) {
2154 /* This would only happen if lastobjid was bad on disk*/
2155 /* Could also happen if recreating missing obj but
2159 CERROR("%s: Serious error: recreating obj %*s "
2160 "but obj already exists \n",
2161 obd->obd_name, dchild->d_name.len,
2162 dchild->d_name.name);
2165 CERROR("%s: Serious error: objid %*s already "
2166 "exists; is this filesystem corrupt?\n",
2167 obd->obd_name, dchild->d_name.len,
2168 dchild->d_name.name);
2171 GOTO(cleanup, rc = -EEXIST);
2174 handle = fsfilt_start_log(obd, dparent->d_inode,
2175 FSFILT_OP_CREATE, NULL, 1);
2177 GOTO(cleanup, rc = PTR_ERR(handle));
2180 rc = ll_vfs_create(dparent->d_inode, dchild, S_IFREG, NULL);
2182 CERROR("create failed rc = %d\n", rc);
2186 if (!recreate_obj) {
2187 filter_set_last_id(filter, group, next_id);
2188 err = filter_update_last_objid(obd, group, 0);
2190 CERROR("unable to write lastobjid "
2191 "but file created\n");
2193 fsfilt_set_fs_flags(obd, dparent->d_inode, SM_DO_REC);
2196 switch(cleanup_phase) {
2198 err = fsfilt_commit(obd, filter->fo_sb, dparent->d_inode, handle, 0);
2200 CERROR("error on commit, err = %d\n", err);
2207 filter_parent_unlock(dparent, lock);
2218 /* check if we have an error after ll_vfs_create(). It is possible that
2219 * there will be say -ENOSPC and we will leak it. */
2221 rc = filter_precreate_rec(obd, dparent, num, oa);
2223 up(&filter->fo_create_lock);
2225 CDEBUG(D_HA, "%s: server last_objid for group "LPU64": "LPU64"\n",
2226 obd->obd_name, group, filter->fo_last_objids[group]);
2228 CDEBUG(D_HA, "%s: filter_precreate() created %d objects\n",
2234 static int filter_create(struct obd_export *exp, struct obdo *oa,
2235 struct lov_stripe_md **ea, struct obd_trans_info *oti)
2237 struct obd_device *obd = NULL;
2238 struct filter_obd *filter;
2239 struct lvfs_run_ctxt saved;
2240 struct lov_stripe_md *lsm = NULL;
2241 struct filter_export_data *fed;
2242 char str[PTL_NALFMT_SIZE];
2243 int group = oa->o_gr, rc = 0, diff, recreate_objs = 0;
2246 if (!(oa->o_valid & OBD_MD_FLGROUP) || group == 0) {
2247 portals_nid2str(exp->exp_connection->c_peer.peer_ni->pni_number,
2248 exp->exp_connection->c_peer.peer_nid, str);
2249 CERROR("!!! nid "LPX64"/%s sent invalid object group %d\n",
2250 exp->exp_connection->c_peer.peer_nid, str, group);
2254 if ((oa->o_valid & OBD_MD_FLFLAGS) &&
2255 (oa->o_flags & OBD_FL_RECREATE_OBJS)) {
2260 fed = &exp->exp_filter_data;
2261 filter = &obd->u.filter;
2263 if (fed->fed_group != group && !recreate_objs &&
2264 !(oa->o_valid & OBD_MD_REINT)) {
2265 portals_nid2str(exp->exp_connection->c_peer.peer_ni->pni_number,
2266 exp->exp_connection->c_peer.peer_nid, str);
2267 CERROR("!!! This export (nid "LPX64"/%s) used object group %d "
2268 "earlier; now it's trying to use group %d! This could "
2269 "be a bug in the MDS. Tell CFS.\n",
2270 exp->exp_connection->c_peer.peer_nid, str,
2271 fed->fed_group, group);
2275 CDEBUG(D_INFO, "filter_create(od->o_gr=%d,od->o_id="LPU64")\n",
2280 rc = obd_alloc_memmd(exp, &lsm);
2287 push_ctxt(&saved, &obd->obd_lvfs_ctxt, NULL);
2289 if (oa->o_valid & OBD_MD_REINT) {
2290 int num = *((int*)oa->o_inline);
2291 rc = filter_precreate(obd, oa, oa->o_gr, &num);
2292 } else if (recreate_objs) {
2293 if (oa->o_id > filter_last_id(&obd->u.filter, group)) {
2294 CERROR("recreate objid "LPU64" > last id "LPU64"\n",
2295 oa->o_id, filter_last_id(&obd->u.filter, group));
2299 rc = filter_precreate(obd, oa, group, &diff);
2302 diff = filter_should_precreate(exp, oa, group);
2304 oa->o_id = filter_last_id(&obd->u.filter, group);
2305 rc = filter_precreate(obd, oa, group, &diff);
2307 oa->o_valid = OBD_MD_FLID;
2311 pop_ctxt(&saved, &obd->obd_lvfs_ctxt, NULL);
2312 if (rc && ea != NULL && *ea != lsm) {
2313 obd_free_memmd(exp, &lsm);
2314 } else if (rc == 0 && ea != NULL) {
2315 /* XXX LOV STACKING: the lsm that is passed to us from
2316 * LOV does not have valid lsm_oinfo data structs, so
2317 * don't go touching that. This needs to be fixed in a
2319 lsm->lsm_object_id = oa->o_id;
2320 lsm->lsm_object_gr = oa->o_gr;
2327 static int filter_destroy(struct obd_export *exp, struct obdo *oa,
2328 struct lov_stripe_md *ea, struct obd_trans_info *oti)
2330 struct obd_device *obd;
2331 struct filter_obd *filter;
2332 struct dentry *dchild = NULL, *dparent = NULL;
2333 struct lvfs_run_ctxt saved;
2334 void *handle = NULL;
2335 struct llog_cookie *fcc = NULL;
2336 int rc, rc2, cleanup_phase = 0, have_prepared = 0;
2340 LASSERT(oa->o_valid & OBD_MD_FLGROUP);
2343 filter = &obd->u.filter;
2345 push_ctxt(&saved, &obd->obd_lvfs_ctxt, NULL);
2348 dparent = filter_parent_lock(obd, oa->o_gr, oa->o_id, &lock);
2349 if (IS_ERR(dparent))
2350 GOTO(cleanup, rc = PTR_ERR(dparent));
2353 dchild = filter_fid2dentry(obd, dparent, oa->o_gr, oa->o_id);
2355 GOTO(cleanup, rc = -ENOENT);
2358 if (dchild->d_inode == NULL) {
2359 CDEBUG(D_INODE, "destroying non-existent object "LPU64"\n",
2361 GOTO(cleanup, rc = -ENOENT);
2364 if (!have_prepared) {
2365 /* If we're really going to destroy the object, get ready
2366 * by getting the clients to discard their cached data.
2368 * We have to drop the parent lock, because
2369 * filter_prepare_destroy will acquire a PW on the object, and
2370 * we don't want to deadlock with an incoming write to the
2371 * object, which has the extent PW and then wants to get the
2372 * parent dentry to do the lookup.
2374 * We dput the child because it's not worth the extra
2375 * complication of condition the above code to skip it on the
2376 * second time through. */
2378 filter_parent_unlock(dparent, lock);
2380 filter_prepare_destroy(obd, oa->o_id, oa->o_gr);
2385 handle = fsfilt_start_log(obd, dparent->d_inode, FSFILT_OP_UNLINK, oti, 1);
2387 GOTO(cleanup, rc = PTR_ERR(handle));
2391 /* Our MDC connection is established by the MDS to us */
2392 if (oa->o_valid & OBD_MD_FLCOOKIE) {
2393 OBD_ALLOC(fcc, sizeof(*fcc));
2395 memcpy(fcc, obdo_logcookie(oa), sizeof(*fcc));
2398 rc = filter_destroy_internal(obd, oa->o_id, dparent, dchild);
2401 switch(cleanup_phase) {
2405 fsfilt_add_journal_cb(obd, filter->fo_sb, 0,
2407 filter_cancel_cookies_cb,
2410 fsfilt_add_journal_cb(obd, filter->fo_sb, 0,
2412 filter_cancel_cookies_cb,
2415 rc = filter_finish_transno(exp, oti, rc);
2416 rc2 = fsfilt_commit(obd, filter->fo_sb, dparent->d_inode,
2419 CERROR("error on commit, err = %d\n", rc2);
2426 filter_parent_unlock(dparent, lock);
2428 pop_ctxt(&saved, &obd->obd_lvfs_ctxt, NULL);
2431 CERROR("invalid cleanup_phase %d\n", cleanup_phase);
2438 /* NB start and end are used for punch, but not truncate */
2439 static int filter_truncate(struct obd_export *exp, struct obdo *oa,
2440 struct lov_stripe_md *lsm,
2441 obd_off start, obd_off end,
2442 struct obd_trans_info *oti)
2447 if (end != OBD_OBJECT_EOF)
2448 CERROR("PUNCH not supported, only truncate: end = "LPX64"\n",
2451 CDEBUG(D_INODE, "calling truncate for object "LPU64", valid = %x, "
2452 "o_size = "LPD64"\n", oa->o_id, oa->o_valid, start);
2454 error = filter_setattr(exp, oa, NULL, oti);
2458 static int filter_sync(struct obd_export *exp, struct obdo *oa,
2459 struct lov_stripe_md *lsm, obd_off start, obd_off end)
2461 struct obd_device *obd = exp->exp_obd;
2462 struct lvfs_run_ctxt saved;
2463 struct filter_obd *filter;
2464 struct dentry *dentry;
2465 struct llog_ctxt *ctxt;
2469 filter = &obd->u.filter;
2471 /* an objid of zero is taken to mean "sync whole filesystem" */
2472 if (!oa || !(oa->o_valid & OBD_MD_FLID)) {
2473 rc = fsfilt_sync(obd, filter->fo_sb);
2474 /* flush any remaining cancel messages out to the target */
2475 ctxt = llog_get_context(&obd->obd_llogs, LLOG_UNLINK_REPL_CTXT);
2476 llog_sync(ctxt, exp);
2480 dentry = filter_oa2dentry(obd, oa);
2482 RETURN(PTR_ERR(dentry));
2484 push_ctxt(&saved, &exp->exp_obd->obd_lvfs_ctxt, NULL);
2486 down(&dentry->d_inode->i_sem);
2487 rc = filemap_fdatasync(dentry->d_inode->i_mapping);
2489 /* just any file to grab fsync method - "file" arg unused */
2490 struct file *file = filter->fo_rcvd_filp;
2492 if (file->f_op && file->f_op->fsync)
2493 rc = file->f_op->fsync(NULL, dentry, 1);
2495 rc2 = filemap_fdatawait(dentry->d_inode->i_mapping);
2499 up(&dentry->d_inode->i_sem);
2501 oa->o_valid = OBD_MD_FLID;
2502 obdo_from_inode(oa, dentry->d_inode, FILTER_VALID_FLAGS);
2504 pop_ctxt(&saved, &exp->exp_obd->obd_lvfs_ctxt, NULL);
2510 static int filter_statfs(struct obd_device *obd, struct obd_statfs *osfs,
2511 unsigned long max_age)
2513 struct filter_obd *filter = &obd->u.filter;
2514 int blockbits = filter->fo_sb->s_blocksize_bits;
2518 /* at least try to account for cached pages. its still racey and
2519 * might be under-reporting if clients haven't announced their
2520 * caches with brw recently */
2521 spin_lock(&obd->obd_osfs_lock);
2522 rc = fsfilt_statfs(obd, filter->fo_sb, max_age);
2523 memcpy(osfs, &obd->obd_osfs, sizeof(*osfs));
2524 spin_unlock(&obd->obd_osfs_lock);
2526 CDEBUG(D_SUPER | D_CACHE, "blocks cached "LPU64" granted "LPU64
2527 " pending "LPU64" free "LPU64" avail "LPU64"\n",
2528 filter->fo_tot_dirty, filter->fo_tot_granted,
2529 filter->fo_tot_pending,
2530 osfs->os_bfree << blockbits, osfs->os_bavail << blockbits);
2532 filter_grant_sanity_check(obd, __FUNCTION__);
2534 osfs->os_bavail -= min(osfs->os_bavail,
2535 (filter->fo_tot_dirty + filter->fo_tot_pending +
2536 osfs->os_bsize -1) >> blockbits);
2541 static int filter_get_info(struct obd_export *exp, __u32 keylen,
2542 void *key, __u32 *vallen, void *val)
2544 struct filter_export_data *fed = &exp->exp_filter_data;
2545 struct obd_device *obd;
2548 obd = class_exp2obd(exp);
2550 CDEBUG(D_IOCTL, "invalid client cookie "LPX64"\n",
2551 exp->exp_handle.h_cookie);
2555 if (keylen == strlen("blocksize") &&
2556 memcmp(key, "blocksize", keylen) == 0) {
2557 __u32 *blocksize = val;
2558 *vallen = sizeof(*blocksize);
2559 *blocksize = obd->u.filter.fo_sb->s_blocksize;
2563 if (keylen == strlen("blocksize_bits") &&
2564 memcmp(key, "blocksize_bits", keylen) == 0) {
2565 __u32 *blocksize_bits = val;
2566 *vallen = sizeof(*blocksize_bits);
2567 *blocksize_bits = obd->u.filter.fo_sb->s_blocksize_bits;
2571 if (keylen >= strlen("last_id") && memcmp(key, "last_id", 7) == 0) {
2572 obd_id *last_id = val;
2573 *last_id = filter_last_id(&obd->u.filter, fed->fed_group);
2576 if (keylen >= strlen("reint_log") && memcmp(key, "reint_log", 9) == 0) {
2577 /*Get log_context handle*/
2578 unsigned long *llh_handle = val;
2579 *vallen = sizeof(unsigned long);
2580 *llh_handle = (unsigned long)obd->obd_llog_ctxt[LLOG_REINT_ORIG_CTXT];
2583 if (keylen >= strlen("cache_sb") && memcmp(key, "cache_sb", 8) == 0) {
2584 /*Get log_context handle*/
2585 unsigned long *sb = val;
2586 *vallen = sizeof(unsigned long);
2587 *sb = (unsigned long)obd->u.filter.fo_sb;
2591 CDEBUG(D_IOCTL, "invalid key\n");
2595 struct obd_llogs *filter_grab_llog_for_group(struct obd_device *obd, int group)
2597 struct filter_group_llog *fglog, *nlog;
2598 char name[32] = "CATLIST";
2599 struct filter_obd *filter;
2600 struct list_head *cur;
2603 filter = &obd->u.filter;
2605 spin_lock(&filter->fo_llog_list_lock);
2606 list_for_each(cur, &filter->fo_llog_list) {
2607 fglog = list_entry(cur, struct filter_group_llog, list);
2608 if (fglog->group == group) {
2609 spin_unlock(&filter->fo_llog_list_lock);
2610 RETURN(fglog->llogs);
2613 spin_unlock(&filter->fo_llog_list_lock);
2615 OBD_ALLOC(fglog, sizeof(*fglog));
2618 fglog->group = group;
2620 OBD_ALLOC(fglog->llogs, sizeof(struct obd_llogs));
2621 if (fglog->llogs == NULL) {
2622 OBD_FREE(fglog, sizeof(*fglog));
2626 spin_lock(&filter->fo_llog_list_lock);
2627 list_for_each(cur, &filter->fo_llog_list) {
2628 nlog = list_entry(cur, struct filter_group_llog, list);
2629 LASSERT(nlog->group != group);
2631 list_add(&fglog->list, &filter->fo_llog_list);
2632 spin_unlock(&filter->fo_llog_list_lock);
2634 rc = obd_llog_cat_initialize(obd, fglog->llogs, 1, name);
2636 OBD_FREE(fglog->llogs, sizeof(*(fglog->llogs)));
2637 OBD_FREE(fglog, sizeof(*fglog));
2641 CDEBUG(D_OTHER, "%s: new llog 0x%p for group %u\n", obd->obd_name,
2642 fglog->llogs, group);
2644 RETURN(fglog->llogs);
2647 static int filter_set_info(struct obd_export *exp, __u32 keylen,
2648 void *key, __u32 vallen, void *val)
2650 struct lvfs_run_ctxt saved;
2651 struct filter_export_data *fed = &exp->exp_filter_data;
2652 struct obd_device *obd;
2653 struct lustre_handle conn;
2654 struct obd_llogs *llog;
2655 struct llog_ctxt *ctxt;
2660 conn.cookie = exp->exp_handle.h_cookie;
2664 CDEBUG(D_IOCTL, "invalid exp %p cookie "LPX64"\n",
2669 if (keylen < strlen("mds_conn") ||
2670 memcmp(key, "mds_conn", keylen) != 0)
2673 group = *((__u32 *)val);
2674 if (fed->fed_group != 0 && fed->fed_group != group) {
2675 char str[PTL_NALFMT_SIZE];
2676 portals_nid2str(exp->exp_connection->c_peer.peer_ni->pni_number,
2677 exp->exp_connection->c_peer.peer_nid, str);
2678 CERROR("!!! This export (nid "LPX64"/%s) used object group %d "
2679 "earlier; now it's trying to use group %d! This could "
2680 "be a bug in the MDS. Tell CFS.\n",
2681 exp->exp_connection->c_peer.peer_nid, str,
2682 fed->fed_group, group);
2685 fed->fed_group = group;
2686 CWARN("Received MDS connection ("LPX64"); group %d\n", conn.cookie,
2691 push_ctxt(&saved, &obd->obd_lvfs_ctxt, NULL);
2692 rc = filter_read_groups(obd, group, 1);
2693 pop_ctxt(&saved, &obd->obd_lvfs_ctxt, NULL);
2695 CERROR("can't read group %u\n", group);
2698 rc = filter_group_set_fs_flags(obd, group);
2700 CERROR("can't set kml flags %u\n", group);
2703 llog = filter_grab_llog_for_group(obd, group);
2704 LASSERT(llog != NULL);
2706 ctxt = llog_get_context(llog, LLOG_UNLINK_REPL_CTXT);
2707 LASSERT(ctxt != NULL);
2708 rc = llog_receptor_accept(ctxt, exp->exp_imp_reverse);
2712 int filter_iocontrol(unsigned int cmd, struct obd_export *exp,
2713 int len, void *karg, void *uarg)
2715 struct obd_device *obd = exp->exp_obd;
2716 struct obd_ioctl_data *data = karg;
2720 case OBD_IOC_ABORT_RECOVERY:
2721 target_stop_recovery_thread(obd);
2724 case OBD_IOC_SET_READONLY: {
2726 struct super_block *sb = obd->u.filter.fo_sb;
2727 struct inode *inode = sb->s_root->d_inode;
2728 BDEVNAME_DECLARE_STORAGE(tmp);
2729 CERROR("setting device %s read-only\n",
2730 ll_bdevname(sb, tmp));
2732 handle = fsfilt_start(obd, inode, FSFILT_OP_MKNOD, NULL);
2734 (void)fsfilt_commit(obd, sb, inode, handle, 1);
2736 dev_set_rdonly(ll_sbdev(obd->u.filter.fo_sb), 2);
2740 case OBD_IOC_CATLOGLIST: {
2741 rc = llog_catalog_list(obd, 1, data);
2745 case OBD_IOC_SNAP_ADD: {
2746 char *name = data->ioc_inlbuf1;
2748 rc = fsfilt_set_snap_item(obd, obd->u.filter.fo_sb, name);
2752 case OBD_IOC_LLOG_CANCEL:
2753 case OBD_IOC_LLOG_REMOVE:
2754 case OBD_IOC_LLOG_INFO:
2755 case OBD_IOC_LLOG_PRINT: {
2756 /* FIXME to be finished */
2757 RETURN(-EOPNOTSUPP);
2759 struct llog_ctxt *ctxt = NULL;
2761 push_ctxt(&saved, &ctxt->loc_ctxt, NULL);
2762 rc = llog_ioctl(ctxt, cmd, data);
2763 pop_ctxt(&saved, &ctxt->loc_ctxt, NULL);
2776 static struct llog_operations filter_unlink_repl_logops;
2777 static struct llog_operations filter_size_orig_logops = {
2778 lop_setup: llog_obd_origin_setup,
2779 lop_cleanup: llog_catalog_cleanup,
2780 lop_add: llog_catalog_add,
2783 static int filter_llog_init(struct obd_device *obd, struct obd_llogs *llogs,
2784 struct obd_device *tgt, int count,
2785 struct llog_catid *catid)
2787 struct llog_ctxt *ctxt;
2791 filter_unlink_repl_logops = llog_client_ops;
2792 filter_unlink_repl_logops.lop_cancel = llog_obd_repl_cancel;
2793 filter_unlink_repl_logops.lop_connect = llog_repl_connect;
2794 filter_unlink_repl_logops.lop_sync = llog_obd_repl_sync;
2796 rc = obd_llog_setup(obd, llogs, LLOG_UNLINK_REPL_CTXT, tgt, 0, NULL,
2797 &filter_unlink_repl_logops);
2800 /* FIXME - assign unlink_cb for filter's recovery */
2801 ctxt = llog_get_context(llogs, LLOG_UNLINK_REPL_CTXT);
2802 ctxt->llog_proc_cb = filter_recov_log_unlink_cb;
2804 /* FIXME - count should be 1 to setup size log */
2805 rc = obd_llog_setup(obd, llogs, LLOG_SIZE_ORIG_CTXT, tgt, 0,
2806 &catid->lci_logid, &filter_size_orig_logops);
2810 static int filter_llog_finish(struct obd_device *obd,
2811 struct obd_llogs *llogs, int count)
2816 rc = obd_llog_cleanup(llog_get_context(llogs, LLOG_UNLINK_REPL_CTXT));
2820 rc = obd_llog_cleanup(llog_get_context(llogs, LLOG_SIZE_ORIG_CTXT));
2824 static int filter_llog_connect(struct obd_device *obd,
2825 struct llogd_conn_body *body)
2827 struct llog_ctxt *ctxt;
2828 struct obd_llogs *llog;
2832 CDEBUG(D_OTHER, "handle connect for %s: %u/%u/%u\n", obd->obd_name,
2833 (unsigned) body->lgdc_logid.lgl_ogr,
2834 (unsigned) body->lgdc_logid.lgl_oid,
2835 (unsigned) body->lgdc_logid.lgl_ogen);
2836 llog = filter_grab_llog_for_group(obd, body->lgdc_logid.lgl_ogr);
2837 LASSERT(llog != NULL);
2838 ctxt = llog_get_context(llog, body->lgdc_ctxt_idx);
2839 rc = llog_connect(ctxt, 1, &body->lgdc_logid,
2840 &body->lgdc_gen, NULL);
2842 CERROR("failed to connect\n");
2847 static struct dentry *filter_lvfs_fid2dentry(__u64 id, __u32 gen, __u64 gr,
2850 return filter_fid2dentry(data, NULL, gr, id);
2853 static struct lvfs_callback_ops filter_lvfs_ops = {
2854 l_fid2dentry: filter_lvfs_fid2dentry,
2857 static struct obd_ops filter_obd_ops = {
2858 .o_owner = THIS_MODULE,
2859 .o_attach = filter_attach,
2860 .o_detach = filter_detach,
2861 .o_get_info = filter_get_info,
2862 .o_set_info = filter_set_info,
2863 .o_setup = filter_setup,
2864 .o_precleanup = filter_precleanup,
2865 .o_cleanup = filter_cleanup,
2866 .o_connect = filter_connect,
2867 .o_disconnect = filter_disconnect,
2868 .o_statfs = filter_statfs,
2869 .o_getattr = filter_getattr,
2870 .o_unpackmd = filter_unpackmd,
2871 .o_create = filter_create,
2872 .o_setattr = filter_setattr,
2873 .o_destroy = filter_destroy,
2874 .o_brw = filter_brw,
2875 .o_punch = filter_truncate,
2876 .o_sync = filter_sync,
2877 .o_preprw = filter_preprw,
2878 .o_commitrw = filter_commitrw,
2879 .o_do_cow = filter_do_cow,
2880 .o_write_extents = filter_write_extents,
2881 .o_destroy_export = filter_destroy_export,
2882 .o_llog_init = filter_llog_init,
2883 .o_llog_finish = filter_llog_finish,
2884 .o_llog_connect = filter_llog_connect,
2885 .o_iocontrol = filter_iocontrol,
2888 static struct obd_ops filter_sanobd_ops = {
2889 .o_owner = THIS_MODULE,
2890 .o_attach = filter_attach,
2891 .o_detach = filter_detach,
2892 .o_get_info = filter_get_info,
2893 .o_set_info = filter_set_info,
2894 .o_setup = filter_san_setup,
2895 .o_precleanup = filter_precleanup,
2896 .o_cleanup = filter_cleanup,
2897 .o_connect = filter_connect,
2898 .o_disconnect = filter_disconnect,
2899 .o_statfs = filter_statfs,
2900 .o_getattr = filter_getattr,
2901 .o_unpackmd = filter_unpackmd,
2902 .o_create = filter_create,
2903 .o_setattr = filter_setattr,
2904 .o_destroy = filter_destroy,
2905 .o_brw = filter_brw,
2906 .o_punch = filter_truncate,
2907 .o_sync = filter_sync,
2908 .o_preprw = filter_preprw,
2909 .o_commitrw = filter_commitrw,
2910 .o_do_cow = filter_do_cow,
2911 .o_write_extents = filter_write_extents,
2912 .o_san_preprw = filter_san_preprw,
2913 .o_destroy_export = filter_destroy_export,
2914 .o_llog_init = filter_llog_init,
2915 .o_llog_finish = filter_llog_finish,
2916 .o_llog_connect = filter_llog_connect,
2917 .o_iocontrol = filter_iocontrol,
2920 static int __init obdfilter_init(void)
2922 struct lprocfs_static_vars lvars;
2925 printk(KERN_INFO "Lustre: Filtering OBD driver; info@clusterfs.com\n");
2927 lprocfs_init_vars(filter, &lvars);
2929 rc = class_register_type(&filter_obd_ops, NULL, lvars.module_vars,
2930 OBD_FILTER_DEVICENAME);
2934 rc = class_register_type(&filter_sanobd_ops, NULL, lvars.module_vars,
2935 OBD_FILTER_SAN_DEVICENAME);
2937 class_unregister_type(OBD_FILTER_DEVICENAME);
2941 static void __exit obdfilter_exit(void)
2943 class_unregister_type(OBD_FILTER_SAN_DEVICENAME);
2944 class_unregister_type(OBD_FILTER_DEVICENAME);
2947 MODULE_AUTHOR("Cluster File Systems, Inc. <info@clusterfs.com>");
2948 MODULE_DESCRIPTION("Lustre Filtering OBD driver");
2949 MODULE_LICENSE("GPL");
2951 module_init(obdfilter_init);
2952 module_exit(obdfilter_exit);