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, but
11 * WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13 * General Public License version 2 for more details (a copy is included
14 * in the LICENSE file that accompanied this code).
16 * You should have received a copy of the GNU General Public License
17 * version 2 along with this program; If not, see
18 * http://www.gnu.org/licenses/gpl-2.0.html
23 * Copyright (c) 2002, 2010, Oracle and/or its affiliates. All rights reserved.
24 * Use is subject to license terms.
26 * Copyright (c) 2011, 2015, Intel Corporation.
29 * This file is part of Lustre, http://www.lustre.org/
30 * Lustre is a trademark of Sun Microsystems, Inc.
32 * lustre/lov/lov_obd.c
34 * Author: Phil Schwan <phil@clusterfs.com>
35 * Author: Peter Braam <braam@clusterfs.com>
36 * Author: Mike Shaver <shaver@clusterfs.com>
37 * Author: Nathan Rutman <nathan@clusterfs.com>
40 #define DEBUG_SUBSYSTEM S_LOV
41 #include "../../include/linux/libcfs/libcfs.h"
43 #include "../include/obd_support.h"
44 #include "../include/lustre_lib.h"
45 #include "../include/lustre_net.h"
46 #include "../include/lustre/lustre_idl.h"
47 #include "../include/lustre_dlm.h"
48 #include "../include/lustre_mds.h"
49 #include "../include/obd_class.h"
50 #include "../include/lprocfs_status.h"
51 #include "../include/lustre_param.h"
52 #include "../include/cl_object.h"
53 #include "../include/lustre/ll_fiemap.h"
54 #include "../include/lustre_fid.h"
56 #include "lov_internal.h"
58 /* Keep a refcount of lov->tgt usage to prevent racing with addition/deletion.
59 * Any function that expects lov_tgts to remain stationary must take a ref.
61 static void lov_getref(struct obd_device *obd)
63 struct lov_obd *lov = &obd->u.lov;
65 /* nobody gets through here until lov_putref is done */
66 mutex_lock(&lov->lov_lock);
67 atomic_inc(&lov->lov_refcount);
68 mutex_unlock(&lov->lov_lock);
72 static void __lov_del_obd(struct obd_device *obd, struct lov_tgt_desc *tgt);
74 static void lov_putref(struct obd_device *obd)
76 struct lov_obd *lov = &obd->u.lov;
78 mutex_lock(&lov->lov_lock);
79 /* ok to dec to 0 more than once -- ltd_exp's will be null */
80 if (atomic_dec_and_test(&lov->lov_refcount) && lov->lov_death_row) {
83 struct lov_tgt_desc *tgt, *n;
85 CDEBUG(D_CONFIG, "destroying %d lov targets\n",
87 for (i = 0; i < lov->desc.ld_tgt_count; i++) {
88 tgt = lov->lov_tgts[i];
90 if (!tgt || !tgt->ltd_reap)
92 list_add(&tgt->ltd_kill, &kill);
93 /* XXX - right now there is a dependency on ld_tgt_count
94 * being the maximum tgt index for computing the
95 * mds_max_easize. So we can't shrink it.
97 lov_ost_pool_remove(&lov->lov_packed, i);
98 lov->lov_tgts[i] = NULL;
101 mutex_unlock(&lov->lov_lock);
103 list_for_each_entry_safe(tgt, n, &kill, ltd_kill) {
104 list_del(&tgt->ltd_kill);
106 __lov_del_obd(obd, tgt);
109 if (lov->lov_tgts_kobj)
110 kobject_put(lov->lov_tgts_kobj);
113 mutex_unlock(&lov->lov_lock);
117 static int lov_set_osc_active(struct obd_device *obd, struct obd_uuid *uuid,
118 enum obd_notify_event ev);
119 static int lov_notify(struct obd_device *obd, struct obd_device *watched,
120 enum obd_notify_event ev, void *data);
122 int lov_connect_obd(struct obd_device *obd, __u32 index, int activate,
123 struct obd_connect_data *data)
125 struct lov_obd *lov = &obd->u.lov;
126 struct obd_uuid *tgt_uuid;
127 struct obd_device *tgt_obd;
128 static struct obd_uuid lov_osc_uuid = { "LOV_OSC_UUID" };
129 struct obd_import *imp;
132 if (!lov->lov_tgts[index])
135 tgt_uuid = &lov->lov_tgts[index]->ltd_uuid;
136 tgt_obd = lov->lov_tgts[index]->ltd_obd;
138 if (!tgt_obd->obd_set_up) {
139 CERROR("Target %s not set up\n", obd_uuid2str(tgt_uuid));
143 /* override the sp_me from lov */
144 tgt_obd->u.cli.cl_sp_me = lov->lov_sp_me;
146 if (data && (data->ocd_connect_flags & OBD_CONNECT_INDEX))
147 data->ocd_index = index;
150 * Divine LOV knows that OBDs under it are OSCs.
152 imp = tgt_obd->u.cli.cl_import;
155 tgt_obd->obd_no_recov = 0;
156 /* FIXME this is probably supposed to be
157 * ptlrpc_set_import_active. Horrible naming.
159 ptlrpc_activate_import(imp);
162 rc = obd_register_observer(tgt_obd, obd);
164 CERROR("Target %s register_observer error %d\n",
165 obd_uuid2str(tgt_uuid), rc);
169 if (imp->imp_invalid) {
170 CDEBUG(D_CONFIG, "not connecting OSC %s; administratively disabled\n",
171 obd_uuid2str(tgt_uuid));
175 rc = obd_connect(NULL, &lov->lov_tgts[index]->ltd_exp, tgt_obd,
176 &lov_osc_uuid, data, NULL);
177 if (rc || !lov->lov_tgts[index]->ltd_exp) {
178 CERROR("Target %s connect error %d\n",
179 obd_uuid2str(tgt_uuid), rc);
183 lov->lov_tgts[index]->ltd_reap = 0;
185 CDEBUG(D_CONFIG, "Connected tgt idx %d %s (%s) %sactive\n", index,
186 obd_uuid2str(tgt_uuid), tgt_obd->obd_name, activate ? "":"in");
188 if (lov->lov_tgts_kobj)
189 /* Even if we failed, that's ok */
190 rc = sysfs_create_link(lov->lov_tgts_kobj, &tgt_obd->obd_kobj,
196 static int lov_connect(const struct lu_env *env,
197 struct obd_export **exp, struct obd_device *obd,
198 struct obd_uuid *cluuid, struct obd_connect_data *data,
201 struct lov_obd *lov = &obd->u.lov;
202 struct lov_tgt_desc *tgt;
203 struct lustre_handle conn;
206 CDEBUG(D_CONFIG, "connect #%d\n", lov->lov_connects);
208 rc = class_connect(&conn, obd, cluuid);
212 *exp = class_conn2export(&conn);
214 /* Why should there ever be more than 1 connect? */
216 LASSERT(lov->lov_connects == 1);
218 memset(&lov->lov_ocd, 0, sizeof(lov->lov_ocd));
220 lov->lov_ocd = *data;
224 lov->lov_tgts_kobj = kobject_create_and_add("target_obds",
227 for (i = 0; i < lov->desc.ld_tgt_count; i++) {
228 tgt = lov->lov_tgts[i];
229 if (!tgt || obd_uuid_empty(&tgt->ltd_uuid))
231 /* Flags will be lowest common denominator */
232 rc = lov_connect_obd(obd, i, tgt->ltd_activate, &lov->lov_ocd);
234 CERROR("%s: lov connect tgt %d failed: %d\n",
235 obd->obd_name, i, rc);
238 /* connect to administrative disabled ost */
239 if (!lov->lov_tgts[i]->ltd_exp)
242 rc = lov_notify(obd, lov->lov_tgts[i]->ltd_exp->exp_obd,
243 OBD_NOTIFY_CONNECT, (void *)&i);
245 CERROR("%s error sending notify %d\n",
254 static int lov_disconnect_obd(struct obd_device *obd, struct lov_tgt_desc *tgt)
256 struct lov_obd *lov = &obd->u.lov;
257 struct obd_device *osc_obd;
260 osc_obd = class_exp2obd(tgt->ltd_exp);
261 CDEBUG(D_CONFIG, "%s: disconnecting target %s\n",
262 obd->obd_name, osc_obd ? osc_obd->obd_name : "NULL");
264 if (tgt->ltd_active) {
266 lov->desc.ld_active_tgt_count--;
267 tgt->ltd_exp->exp_obd->obd_inactive = 1;
271 if (lov->lov_tgts_kobj)
272 sysfs_remove_link(lov->lov_tgts_kobj,
275 /* Pass it on to our clients.
276 * XXX This should be an argument to disconnect,
277 * XXX not a back-door flag on the OBD. Ah well.
279 osc_obd->obd_force = obd->obd_force;
280 osc_obd->obd_fail = obd->obd_fail;
281 osc_obd->obd_no_recov = obd->obd_no_recov;
284 obd_register_observer(osc_obd, NULL);
286 rc = obd_disconnect(tgt->ltd_exp);
288 CERROR("Target %s disconnect error %d\n",
289 tgt->ltd_uuid.uuid, rc);
297 static int lov_disconnect(struct obd_export *exp)
299 struct obd_device *obd = class_exp2obd(exp);
300 struct lov_obd *lov = &obd->u.lov;
306 /* Only disconnect the underlying layers on the final disconnect. */
308 if (lov->lov_connects != 0) {
309 /* why should there be more than 1 connect? */
310 CERROR("disconnect #%d\n", lov->lov_connects);
314 /* Let's hold another reference so lov_del_obd doesn't spin through
319 for (i = 0; i < lov->desc.ld_tgt_count; i++) {
320 if (lov->lov_tgts[i] && lov->lov_tgts[i]->ltd_exp) {
321 /* Disconnection is the last we know about an obd */
322 lov_del_target(obd, i, NULL, lov->lov_tgts[i]->ltd_gen);
329 rc = class_disconnect(exp); /* bz 9811 */
335 * -EINVAL : UUID can't be found in the LOV's target list
336 * -ENOTCONN: The UUID is found, but the target connection is bad (!)
337 * -EBADF : The UUID is found, but the OBD is the wrong type (!)
338 * any >= 0 : is log target index
340 static int lov_set_osc_active(struct obd_device *obd, struct obd_uuid *uuid,
341 enum obd_notify_event ev)
343 struct lov_obd *lov = &obd->u.lov;
344 struct lov_tgt_desc *tgt;
345 int index, activate, active;
347 CDEBUG(D_INFO, "Searching in lov %p for uuid %s event(%d)\n",
348 lov, uuid->uuid, ev);
351 for (index = 0; index < lov->desc.ld_tgt_count; index++) {
352 tgt = lov->lov_tgts[index];
356 * LU-642, initially inactive OSC could miss the obd_connect,
357 * we make up for it here.
359 if (ev == OBD_NOTIFY_ACTIVATE && !tgt->ltd_exp &&
360 obd_uuid_equals(uuid, &tgt->ltd_uuid)) {
361 struct obd_uuid lov_osc_uuid = {"LOV_OSC_UUID"};
363 obd_connect(NULL, &tgt->ltd_exp, tgt->ltd_obd,
364 &lov_osc_uuid, &lov->lov_ocd, NULL);
369 CDEBUG(D_INFO, "lov idx %d is %s conn %#llx\n",
370 index, obd_uuid2str(&tgt->ltd_uuid),
371 tgt->ltd_exp->exp_handle.h_cookie);
372 if (obd_uuid_equals(uuid, &tgt->ltd_uuid))
376 if (index == lov->desc.ld_tgt_count) {
381 if (ev == OBD_NOTIFY_DEACTIVATE || ev == OBD_NOTIFY_ACTIVATE) {
382 activate = (ev == OBD_NOTIFY_ACTIVATE) ? 1 : 0;
384 if (lov->lov_tgts[index]->ltd_activate == activate) {
385 CDEBUG(D_INFO, "OSC %s already %sactivate!\n",
386 uuid->uuid, activate ? "" : "de");
388 lov->lov_tgts[index]->ltd_activate = activate;
389 CDEBUG(D_CONFIG, "%sactivate OSC %s\n",
390 activate ? "" : "de", obd_uuid2str(uuid));
393 } else if (ev == OBD_NOTIFY_INACTIVE || ev == OBD_NOTIFY_ACTIVE) {
394 active = (ev == OBD_NOTIFY_ACTIVE) ? 1 : 0;
396 if (lov->lov_tgts[index]->ltd_active == active) {
397 CDEBUG(D_INFO, "OSC %s already %sactive!\n",
398 uuid->uuid, active ? "" : "in");
401 CDEBUG(D_CONFIG, "Marking OSC %s %sactive\n",
402 obd_uuid2str(uuid), active ? "" : "in");
404 lov->lov_tgts[index]->ltd_active = active;
406 lov->desc.ld_active_tgt_count++;
407 lov->lov_tgts[index]->ltd_exp->exp_obd->obd_inactive = 0;
409 lov->desc.ld_active_tgt_count--;
410 lov->lov_tgts[index]->ltd_exp->exp_obd->obd_inactive = 1;
413 CERROR("Unknown event(%d) for uuid %s", ev, uuid->uuid);
421 static int lov_notify(struct obd_device *obd, struct obd_device *watched,
422 enum obd_notify_event ev, void *data)
425 struct lov_obd *lov = &obd->u.lov;
427 down_read(&lov->lov_notify_lock);
428 if (!lov->lov_connects) {
429 up_read(&lov->lov_notify_lock);
433 if (ev == OBD_NOTIFY_ACTIVE || ev == OBD_NOTIFY_INACTIVE ||
434 ev == OBD_NOTIFY_ACTIVATE || ev == OBD_NOTIFY_DEACTIVATE) {
435 struct obd_uuid *uuid;
439 if (strcmp(watched->obd_type->typ_name, LUSTRE_OSC_NAME)) {
440 up_read(&lov->lov_notify_lock);
441 CERROR("unexpected notification of %s %s!\n",
442 watched->obd_type->typ_name,
446 uuid = &watched->u.cli.cl_target_uuid;
448 /* Set OSC as active before notifying the observer, so the
449 * observer can use the OSC normally.
451 rc = lov_set_osc_active(obd, uuid, ev);
453 up_read(&lov->lov_notify_lock);
454 CERROR("event(%d) of %s failed: %d\n", ev,
455 obd_uuid2str(uuid), rc);
458 /* active event should be pass lov target index as data */
462 /* Pass the notification up the chain. */
464 rc = obd_notify_observer(obd, watched, ev, data);
466 /* NULL watched means all osc's in the lov (only for syncs) */
467 /* sync event should be send lov idx as data */
468 struct lov_obd *lov = &obd->u.lov;
472 is_sync = (ev == OBD_NOTIFY_SYNC) ||
473 (ev == OBD_NOTIFY_SYNC_NONBLOCK);
476 for (i = 0; i < lov->desc.ld_tgt_count; i++) {
477 if (!lov->lov_tgts[i])
480 /* don't send sync event if target not
481 * connected/activated
483 if (is_sync && !lov->lov_tgts[i]->ltd_active)
486 rc = obd_notify_observer(obd, lov->lov_tgts[i]->ltd_obd,
489 CERROR("%s: notify %s of %s failed %d\n",
491 obd->obd_observer->obd_name,
492 lov->lov_tgts[i]->ltd_obd->obd_name,
499 up_read(&lov->lov_notify_lock);
503 static int lov_add_target(struct obd_device *obd, struct obd_uuid *uuidp,
504 __u32 index, int gen, int active)
506 struct lov_obd *lov = &obd->u.lov;
507 struct lov_tgt_desc *tgt;
508 struct obd_device *tgt_obd;
511 CDEBUG(D_CONFIG, "uuid:%s idx:%d gen:%d active:%d\n",
512 uuidp->uuid, index, gen, active);
515 CERROR("request to add OBD %s with invalid generation: %d\n",
520 tgt_obd = class_find_client_obd(uuidp, LUSTRE_OSC_NAME,
525 mutex_lock(&lov->lov_lock);
527 if ((index < lov->lov_tgt_size) && lov->lov_tgts[index]) {
528 tgt = lov->lov_tgts[index];
529 CERROR("UUID %s already assigned at LOV target index %d\n",
530 obd_uuid2str(&tgt->ltd_uuid), index);
531 mutex_unlock(&lov->lov_lock);
535 if (index >= lov->lov_tgt_size) {
536 /* We need to reallocate the lov target array. */
537 struct lov_tgt_desc **newtgts, **old = NULL;
538 __u32 newsize, oldsize = 0;
540 newsize = max_t(__u32, lov->lov_tgt_size, 2);
541 while (newsize < index + 1)
543 newtgts = kcalloc(newsize, sizeof(*newtgts), GFP_NOFS);
545 mutex_unlock(&lov->lov_lock);
549 if (lov->lov_tgt_size) {
550 memcpy(newtgts, lov->lov_tgts, sizeof(*newtgts) *
553 oldsize = lov->lov_tgt_size;
556 lov->lov_tgts = newtgts;
557 lov->lov_tgt_size = newsize;
561 CDEBUG(D_CONFIG, "tgts: %p size: %d\n",
562 lov->lov_tgts, lov->lov_tgt_size);
565 tgt = kzalloc(sizeof(*tgt), GFP_NOFS);
567 mutex_unlock(&lov->lov_lock);
571 rc = lov_ost_pool_add(&lov->lov_packed, index, lov->lov_tgt_size);
573 mutex_unlock(&lov->lov_lock);
578 tgt->ltd_uuid = *uuidp;
579 tgt->ltd_obd = tgt_obd;
580 /* XXX - add a sanity check on the generation number. */
582 tgt->ltd_index = index;
583 tgt->ltd_activate = active;
584 lov->lov_tgts[index] = tgt;
585 if (index >= lov->desc.ld_tgt_count)
586 lov->desc.ld_tgt_count = index + 1;
588 mutex_unlock(&lov->lov_lock);
590 CDEBUG(D_CONFIG, "idx=%d ltd_gen=%d ld_tgt_count=%d\n",
591 index, tgt->ltd_gen, lov->desc.ld_tgt_count);
593 rc = obd_notify(obd, tgt_obd, OBD_NOTIFY_CREATE, &index);
595 if (lov->lov_connects == 0) {
596 /* lov_connect hasn't been called yet. We'll do the
597 * lov_connect_obd on this target when that fn first runs,
598 * because we don't know the connect flags yet.
605 rc = lov_connect_obd(obd, index, active, &lov->lov_ocd);
609 /* connect to administrative disabled ost */
615 if (lov->lov_cache) {
616 rc = obd_set_info_async(NULL, tgt->ltd_exp,
617 sizeof(KEY_CACHE_SET), KEY_CACHE_SET,
618 sizeof(struct cl_client_cache),
619 lov->lov_cache, NULL);
624 rc = lov_notify(obd, tgt->ltd_exp->exp_obd,
625 active ? OBD_NOTIFY_CONNECT : OBD_NOTIFY_INACTIVE,
630 CERROR("add failed (%d), deleting %s\n", rc,
631 obd_uuid2str(&tgt->ltd_uuid));
632 lov_del_target(obd, index, NULL, 0);
638 /* Schedule a target for deletion */
639 int lov_del_target(struct obd_device *obd, __u32 index,
640 struct obd_uuid *uuidp, int gen)
642 struct lov_obd *lov = &obd->u.lov;
643 int count = lov->desc.ld_tgt_count;
646 if (index >= count) {
647 CERROR("LOV target index %d >= number of LOV OBDs %d.\n",
652 /* to make sure there's no ongoing lov_notify() now */
653 down_write(&lov->lov_notify_lock);
656 if (!lov->lov_tgts[index]) {
657 CERROR("LOV target at index %d is not setup.\n", index);
662 if (uuidp && !obd_uuid_equals(uuidp, &lov->lov_tgts[index]->ltd_uuid)) {
663 CERROR("LOV target UUID %s at index %d doesn't match %s.\n",
664 lov_uuid2str(lov, index), index,
665 obd_uuid2str(uuidp));
670 CDEBUG(D_CONFIG, "uuid: %s idx: %d gen: %d exp: %p active: %d\n",
671 lov_uuid2str(lov, index), index,
672 lov->lov_tgts[index]->ltd_gen, lov->lov_tgts[index]->ltd_exp,
673 lov->lov_tgts[index]->ltd_active);
675 lov->lov_tgts[index]->ltd_reap = 1;
676 lov->lov_death_row++;
677 /* we really delete it from obd_putref */
680 up_write(&lov->lov_notify_lock);
685 static void __lov_del_obd(struct obd_device *obd, struct lov_tgt_desc *tgt)
687 struct obd_device *osc_obd;
690 LASSERT(tgt->ltd_reap);
692 osc_obd = class_exp2obd(tgt->ltd_exp);
694 CDEBUG(D_CONFIG, "Removing tgt %s : %s\n",
696 osc_obd ? osc_obd->obd_name : "<no obd>");
699 lov_disconnect_obd(obd, tgt);
703 /* Manual cleanup - no cleanup logs to clean up the osc's. We must
704 * do it ourselves. And we can't do it from lov_cleanup,
705 * because we just lost our only reference to it.
708 class_manual_cleanup(osc_obd);
711 void lov_fix_desc_stripe_size(__u64 *val)
713 if (*val < LOV_MIN_STRIPE_SIZE) {
715 LCONSOLE_INFO("Increasing default stripe size to minimum %u\n",
716 LOV_DESC_STRIPE_SIZE_DEFAULT);
717 *val = LOV_DESC_STRIPE_SIZE_DEFAULT;
718 } else if (*val & (LOV_MIN_STRIPE_SIZE - 1)) {
719 *val &= ~(LOV_MIN_STRIPE_SIZE - 1);
720 LCONSOLE_WARN("Changing default stripe size to %llu (a multiple of %u)\n",
721 *val, LOV_MIN_STRIPE_SIZE);
725 void lov_fix_desc_stripe_count(__u32 *val)
731 void lov_fix_desc_pattern(__u32 *val)
733 /* from lov_setstripe */
734 if ((*val != 0) && (*val != LOV_PATTERN_RAID0)) {
735 LCONSOLE_WARN("Unknown stripe pattern: %#x\n", *val);
740 void lov_fix_desc_qos_maxage(__u32 *val)
743 *val = LOV_DESC_QOS_MAXAGE_DEFAULT;
746 void lov_fix_desc(struct lov_desc *desc)
748 lov_fix_desc_stripe_size(&desc->ld_default_stripe_size);
749 lov_fix_desc_stripe_count(&desc->ld_default_stripe_count);
750 lov_fix_desc_pattern(&desc->ld_pattern);
751 lov_fix_desc_qos_maxage(&desc->ld_qos_maxage);
754 int lov_setup(struct obd_device *obd, struct lustre_cfg *lcfg)
756 struct lprocfs_static_vars lvars = { NULL };
757 struct lov_desc *desc;
758 struct lov_obd *lov = &obd->u.lov;
761 if (LUSTRE_CFG_BUFLEN(lcfg, 1) < 1) {
762 CERROR("LOV setup requires a descriptor\n");
766 desc = (struct lov_desc *)lustre_cfg_buf(lcfg, 1);
768 if (sizeof(*desc) > LUSTRE_CFG_BUFLEN(lcfg, 1)) {
769 CERROR("descriptor size wrong: %d > %d\n",
770 (int)sizeof(*desc), LUSTRE_CFG_BUFLEN(lcfg, 1));
774 if (desc->ld_magic != LOV_DESC_MAGIC) {
775 if (desc->ld_magic == __swab32(LOV_DESC_MAGIC)) {
776 CDEBUG(D_OTHER, "%s: Swabbing lov desc %p\n",
777 obd->obd_name, desc);
778 lustre_swab_lov_desc(desc);
780 CERROR("%s: Bad lov desc magic: %#x\n",
781 obd->obd_name, desc->ld_magic);
788 desc->ld_active_tgt_count = 0;
790 lov->lov_tgt_size = 0;
792 mutex_init(&lov->lov_lock);
793 atomic_set(&lov->lov_refcount, 0);
794 lov->lov_sp_me = LUSTRE_SP_CLI;
796 init_rwsem(&lov->lov_notify_lock);
798 lov->lov_pools_hash_body = cfs_hash_create("POOLS", HASH_POOLS_CUR_BITS,
800 HASH_POOLS_BKT_BITS, 0,
803 &pool_hash_operations,
805 INIT_LIST_HEAD(&lov->lov_pool_list);
806 lov->lov_pool_count = 0;
807 rc = lov_ost_pool_init(&lov->lov_packed, 0);
811 lprocfs_lov_init_vars(&lvars);
812 lprocfs_obd_setup(obd, lvars.obd_vars, lvars.sysfs_vars);
814 rc = ldebugfs_seq_create(obd->obd_debugfs_entry, "target_obd",
815 0444, &lov_proc_target_fops, obd);
817 CWARN("Error adding the target_obd file\n");
819 lov->lov_pool_debugfs_entry = ldebugfs_register("pools",
820 obd->obd_debugfs_entry,
828 static int lov_precleanup(struct obd_device *obd, enum obd_cleanup_stage stage)
830 struct lov_obd *lov = &obd->u.lov;
833 case OBD_CLEANUP_EARLY: {
836 for (i = 0; i < lov->desc.ld_tgt_count; i++) {
837 if (!lov->lov_tgts[i] || !lov->lov_tgts[i]->ltd_active)
839 obd_precleanup(class_exp2obd(lov->lov_tgts[i]->ltd_exp),
851 static int lov_cleanup(struct obd_device *obd)
853 struct lov_obd *lov = &obd->u.lov;
854 struct list_head *pos, *tmp;
855 struct pool_desc *pool;
857 list_for_each_safe(pos, tmp, &lov->lov_pool_list) {
858 pool = list_entry(pos, struct pool_desc, pool_list);
859 /* free pool structs */
860 CDEBUG(D_INFO, "delete pool %p\n", pool);
861 /* In the function below, .hs_keycmp resolves to
862 * pool_hashkey_keycmp()
864 /* coverity[overrun-buffer-val] */
865 lov_pool_del(obd, pool->pool_name);
867 cfs_hash_putref(lov->lov_pools_hash_body);
868 lov_ost_pool_free(&lov->lov_packed);
870 lprocfs_obd_cleanup(obd);
875 for (i = 0; i < lov->desc.ld_tgt_count; i++) {
876 if (!lov->lov_tgts[i])
879 /* Inactive targets may never have connected */
880 if (lov->lov_tgts[i]->ltd_active ||
881 atomic_read(&lov->lov_refcount))
882 /* We should never get here - these
883 * should have been removed in the
886 CERROR("lov tgt %d not cleaned! deathrow=%d, lovrc=%d\n",
887 i, lov->lov_death_row,
888 atomic_read(&lov->lov_refcount));
889 lov_del_target(obd, i, NULL, 0);
892 kfree(lov->lov_tgts);
893 lov->lov_tgt_size = 0;
896 if (lov->lov_cache) {
897 cl_cache_decref(lov->lov_cache);
898 lov->lov_cache = NULL;
904 int lov_process_config_base(struct obd_device *obd, struct lustre_cfg *lcfg,
905 __u32 *indexp, int *genp)
907 struct obd_uuid obd_uuid;
911 switch (cmd = lcfg->lcfg_command) {
912 case LCFG_LOV_ADD_OBD:
913 case LCFG_LOV_ADD_INA:
914 case LCFG_LOV_DEL_OBD: {
917 /* lov_modify_tgts add 0:lov_mdsA 1:ost1_UUID 2:0 3:1 */
918 if (LUSTRE_CFG_BUFLEN(lcfg, 1) > sizeof(obd_uuid.uuid)) {
923 obd_str2uuid(&obd_uuid, lustre_cfg_buf(lcfg, 1));
925 rc = kstrtoint(lustre_cfg_buf(lcfg, 2), 10, indexp);
928 rc = kstrtoint(lustre_cfg_buf(lcfg, 3), 10, genp);
933 if (cmd == LCFG_LOV_ADD_OBD)
934 rc = lov_add_target(obd, &obd_uuid, index, gen, 1);
935 else if (cmd == LCFG_LOV_ADD_INA)
936 rc = lov_add_target(obd, &obd_uuid, index, gen, 0);
938 rc = lov_del_target(obd, index, &obd_uuid, gen);
942 struct lprocfs_static_vars lvars = { NULL };
943 struct lov_desc *desc = &(obd->u.lov.desc);
950 lprocfs_lov_init_vars(&lvars);
952 rc = class_process_proc_param(PARAM_LOV, lvars.obd_vars,
965 CERROR("Unknown command: %d\n", lcfg->lcfg_command);
974 static int lov_recreate(struct obd_export *exp, struct obdo *src_oa,
975 struct lov_stripe_md **ea, struct obd_trans_info *oti)
977 struct lov_stripe_md *obj_mdp, *lsm;
978 struct lov_obd *lov = &exp->exp_obd->u.lov;
982 LASSERT(src_oa->o_valid & OBD_MD_FLFLAGS &&
983 src_oa->o_flags & OBD_FL_RECREATE_OBJS);
985 obj_mdp = kzalloc(sizeof(*obj_mdp), GFP_NOFS);
989 ost_idx = src_oa->o_nlink;
995 if (ost_idx >= lov->desc.ld_tgt_count ||
996 !lov->lov_tgts[ost_idx]) {
1001 for (i = 0; i < lsm->lsm_stripe_count; i++) {
1002 struct lov_oinfo *loi = lsm->lsm_oinfo[i];
1004 if (lov_oinfo_is_dummy(loi))
1007 if (loi->loi_ost_idx == ost_idx) {
1008 if (ostid_id(&loi->loi_oi) != ostid_id(&src_oa->o_oi)) {
1015 if (i == lsm->lsm_stripe_count) {
1020 rc = obd_create(NULL, lov->lov_tgts[ost_idx]->ltd_exp,
1021 src_oa, &obj_mdp, oti);
1027 /* the LOV expects oa->o_id to be set to the LOV object id */
1028 static int lov_create(const struct lu_env *env, struct obd_export *exp,
1029 struct obdo *src_oa, struct lov_stripe_md **ea,
1030 struct obd_trans_info *oti)
1032 struct lov_obd *lov;
1039 if ((src_oa->o_valid & OBD_MD_FLFLAGS) &&
1040 src_oa->o_flags == OBD_FL_DELORPHAN) {
1041 /* should be used with LOV anymore */
1045 lov = &exp->exp_obd->u.lov;
1046 if (!lov->desc.ld_active_tgt_count)
1049 obd_getref(exp->exp_obd);
1050 /* Recreate a specific object id at the given OST index */
1051 if ((src_oa->o_valid & OBD_MD_FLFLAGS) &&
1052 (src_oa->o_flags & OBD_FL_RECREATE_OBJS)) {
1053 rc = lov_recreate(exp, src_oa, ea, oti);
1056 obd_putref(exp->exp_obd);
1060 #define ASSERT_LSM_MAGIC(lsmp) \
1063 LASSERTF(((lsmp)->lsm_magic == LOV_MAGIC_V1 || \
1064 (lsmp)->lsm_magic == LOV_MAGIC_V3), \
1065 "%p->lsm_magic=%x\n", (lsmp), (lsmp)->lsm_magic); \
1068 static int lov_destroy(const struct lu_env *env, struct obd_export *exp,
1069 struct obdo *oa, struct lov_stripe_md *lsm,
1070 struct obd_trans_info *oti, struct obd_export *md_exp)
1072 struct lov_request_set *set;
1073 struct obd_info oinfo;
1074 struct lov_request *req;
1075 struct lov_obd *lov;
1076 int rc = 0, err = 0;
1078 ASSERT_LSM_MAGIC(lsm);
1080 if (!exp || !exp->exp_obd)
1083 if (oa->o_valid & OBD_MD_FLCOOKIE) {
1085 LASSERT(oti->oti_logcookies);
1088 lov = &exp->exp_obd->u.lov;
1089 obd_getref(exp->exp_obd);
1090 rc = lov_prep_destroy_set(exp, &oinfo, oa, lsm, oti, &set);
1094 list_for_each_entry(req, &set->set_list, rq_link) {
1095 if (oa->o_valid & OBD_MD_FLCOOKIE)
1096 oti->oti_logcookies = set->set_cookies + req->rq_stripe;
1098 err = obd_destroy(env, lov->lov_tgts[req->rq_idx]->ltd_exp,
1099 req->rq_oi.oi_oa, NULL, oti, NULL);
1100 err = lov_update_common_set(set, req, err);
1102 CERROR("%s: destroying objid "DOSTID" subobj "
1103 DOSTID" on OST idx %d: rc = %d\n",
1104 exp->exp_obd->obd_name, POSTID(&oa->o_oi),
1105 POSTID(&req->rq_oi.oi_oa->o_oi),
1113 rc = lsm_op_find(lsm->lsm_magic)->lsm_destroy(lsm, oa, md_exp);
1115 err = lov_fini_destroy_set(set);
1117 obd_putref(exp->exp_obd);
1118 return rc ? rc : err;
1121 static int lov_getattr_interpret(struct ptlrpc_request_set *rqset,
1124 struct lov_request_set *lovset = (struct lov_request_set *)data;
1127 /* don't do attribute merge if this async op failed */
1129 atomic_set(&lovset->set_completes, 0);
1130 err = lov_fini_getattr_set(lovset);
1131 return rc ? rc : err;
1134 static int lov_getattr_async(struct obd_export *exp, struct obd_info *oinfo,
1135 struct ptlrpc_request_set *rqset)
1137 struct lov_request_set *lovset;
1138 struct lov_obd *lov;
1139 struct lov_request *req;
1143 ASSERT_LSM_MAGIC(oinfo->oi_md);
1145 if (!exp || !exp->exp_obd)
1148 lov = &exp->exp_obd->u.lov;
1150 rc = lov_prep_getattr_set(exp, oinfo, &lovset);
1154 CDEBUG(D_INFO, "objid "DOSTID": %ux%u byte stripes\n",
1155 POSTID(&oinfo->oi_md->lsm_oi), oinfo->oi_md->lsm_stripe_count,
1156 oinfo->oi_md->lsm_stripe_size);
1158 list_for_each_entry(req, &lovset->set_list, rq_link) {
1159 CDEBUG(D_INFO, "objid " DOSTID "[%d] has subobj " DOSTID " at idx%u\n",
1160 POSTID(&oinfo->oi_oa->o_oi), req->rq_stripe,
1161 POSTID(&req->rq_oi.oi_oa->o_oi), req->rq_idx);
1162 rc = obd_getattr_async(lov->lov_tgts[req->rq_idx]->ltd_exp,
1163 &req->rq_oi, rqset);
1165 CERROR("%s: getattr objid "DOSTID" subobj"
1166 DOSTID" on OST idx %d: rc = %d\n",
1167 exp->exp_obd->obd_name,
1168 POSTID(&oinfo->oi_oa->o_oi),
1169 POSTID(&req->rq_oi.oi_oa->o_oi),
1175 if (!list_empty(&rqset->set_requests)) {
1177 LASSERT(!rqset->set_interpret);
1178 rqset->set_interpret = lov_getattr_interpret;
1179 rqset->set_arg = (void *)lovset;
1184 atomic_set(&lovset->set_completes, 0);
1185 err = lov_fini_getattr_set(lovset);
1186 return rc ? rc : err;
1189 static int lov_setattr_interpret(struct ptlrpc_request_set *rqset,
1192 struct lov_request_set *lovset = (struct lov_request_set *)data;
1196 atomic_set(&lovset->set_completes, 0);
1197 err = lov_fini_setattr_set(lovset);
1198 return rc ? rc : err;
1201 /* If @oti is given, the request goes from MDS and responses from OSTs are not
1202 * needed. Otherwise, a client is waiting for responses.
1204 static int lov_setattr_async(struct obd_export *exp, struct obd_info *oinfo,
1205 struct obd_trans_info *oti,
1206 struct ptlrpc_request_set *rqset)
1208 struct lov_request_set *set;
1209 struct lov_request *req;
1210 struct lov_obd *lov;
1214 ASSERT_LSM_MAGIC(oinfo->oi_md);
1215 if (oinfo->oi_oa->o_valid & OBD_MD_FLCOOKIE) {
1217 LASSERT(oti->oti_logcookies);
1220 if (!exp || !exp->exp_obd)
1223 lov = &exp->exp_obd->u.lov;
1224 rc = lov_prep_setattr_set(exp, oinfo, oti, &set);
1228 CDEBUG(D_INFO, "objid "DOSTID": %ux%u byte stripes\n",
1229 POSTID(&oinfo->oi_md->lsm_oi),
1230 oinfo->oi_md->lsm_stripe_count,
1231 oinfo->oi_md->lsm_stripe_size);
1233 list_for_each_entry(req, &set->set_list, rq_link) {
1234 if (oinfo->oi_oa->o_valid & OBD_MD_FLCOOKIE)
1235 oti->oti_logcookies = set->set_cookies + req->rq_stripe;
1237 CDEBUG(D_INFO, "objid " DOSTID "[%d] has subobj " DOSTID " at idx%u\n",
1238 POSTID(&oinfo->oi_oa->o_oi), req->rq_stripe,
1239 POSTID(&req->rq_oi.oi_oa->o_oi), req->rq_idx);
1241 rc = obd_setattr_async(lov->lov_tgts[req->rq_idx]->ltd_exp,
1242 &req->rq_oi, oti, rqset);
1244 CERROR("error: setattr objid "DOSTID" subobj"
1245 DOSTID" on OST idx %d: rc = %d\n",
1246 POSTID(&set->set_oi->oi_oa->o_oi),
1247 POSTID(&req->rq_oi.oi_oa->o_oi),
1253 /* If we are not waiting for responses on async requests, return. */
1254 if (rc || !rqset || list_empty(&rqset->set_requests)) {
1258 atomic_set(&set->set_completes, 0);
1259 err = lov_fini_setattr_set(set);
1260 return rc ? rc : err;
1263 LASSERT(!rqset->set_interpret);
1264 rqset->set_interpret = lov_setattr_interpret;
1265 rqset->set_arg = (void *)set;
1270 /* find any ldlm lock of the inode in lov
1275 static int lov_find_cbdata(struct obd_export *exp,
1276 struct lov_stripe_md *lsm, ldlm_iterator_t it,
1279 struct lov_obd *lov;
1282 ASSERT_LSM_MAGIC(lsm);
1284 if (!exp || !exp->exp_obd)
1287 lov = &exp->exp_obd->u.lov;
1288 for (i = 0; i < lsm->lsm_stripe_count; i++) {
1289 struct lov_stripe_md submd;
1290 struct lov_oinfo *loi = lsm->lsm_oinfo[i];
1292 if (lov_oinfo_is_dummy(loi))
1295 if (!lov->lov_tgts[loi->loi_ost_idx]) {
1296 CDEBUG(D_HA, "lov idx %d NULL\n", loi->loi_ost_idx);
1300 submd.lsm_oi = loi->loi_oi;
1301 submd.lsm_stripe_count = 0;
1302 rc = obd_find_cbdata(lov->lov_tgts[loi->loi_ost_idx]->ltd_exp,
1310 int lov_statfs_interpret(struct ptlrpc_request_set *rqset, void *data, int rc)
1312 struct lov_request_set *lovset = (struct lov_request_set *)data;
1316 atomic_set(&lovset->set_completes, 0);
1318 err = lov_fini_statfs_set(lovset);
1319 return rc ? rc : err;
1322 static int lov_statfs_async(struct obd_export *exp, struct obd_info *oinfo,
1323 __u64 max_age, struct ptlrpc_request_set *rqset)
1325 struct obd_device *obd = class_exp2obd(exp);
1326 struct lov_request_set *set;
1327 struct lov_request *req;
1328 struct lov_obd *lov;
1331 LASSERT(oinfo->oi_osfs);
1334 rc = lov_prep_statfs_set(obd, oinfo, &set);
1338 list_for_each_entry(req, &set->set_list, rq_link) {
1339 rc = obd_statfs_async(lov->lov_tgts[req->rq_idx]->ltd_exp,
1340 &req->rq_oi, max_age, rqset);
1345 if (rc || list_empty(&rqset->set_requests)) {
1349 atomic_set(&set->set_completes, 0);
1350 err = lov_fini_statfs_set(set);
1351 return rc ? rc : err;
1354 LASSERT(!rqset->set_interpret);
1355 rqset->set_interpret = lov_statfs_interpret;
1356 rqset->set_arg = (void *)set;
1360 static int lov_statfs(const struct lu_env *env, struct obd_export *exp,
1361 struct obd_statfs *osfs, __u64 max_age, __u32 flags)
1363 struct ptlrpc_request_set *set = NULL;
1364 struct obd_info oinfo = { };
1367 /* for obdclass we forbid using obd_statfs_rqset, but prefer using async
1370 set = ptlrpc_prep_set();
1374 oinfo.oi_osfs = osfs;
1375 oinfo.oi_flags = flags;
1376 rc = lov_statfs_async(exp, &oinfo, max_age, set);
1378 rc = ptlrpc_set_wait(set);
1379 ptlrpc_set_destroy(set);
1384 static int lov_iocontrol(unsigned int cmd, struct obd_export *exp, int len,
1385 void *karg, void __user *uarg)
1387 struct obd_device *obddev = class_exp2obd(exp);
1388 struct lov_obd *lov = &obddev->u.lov;
1389 int i = 0, rc = 0, count = lov->desc.ld_tgt_count;
1390 struct obd_uuid *uuidp;
1393 case IOC_OBD_STATFS: {
1394 struct obd_ioctl_data *data = karg;
1395 struct obd_device *osc_obd;
1396 struct obd_statfs stat_buf = {0};
1400 memcpy(&index, data->ioc_inlbuf2, sizeof(__u32));
1404 if (!lov->lov_tgts[index])
1405 /* Try again with the next index */
1407 if (!lov->lov_tgts[index]->ltd_active)
1410 osc_obd = class_exp2obd(lov->lov_tgts[index]->ltd_exp);
1415 if (copy_to_user(data->ioc_pbuf2, obd2cli_tgt(osc_obd),
1416 min((int)data->ioc_plen2,
1417 (int)sizeof(struct obd_uuid))))
1420 memcpy(&flags, data->ioc_inlbuf1, sizeof(__u32));
1421 flags = flags & LL_STATFS_NODELAY ? OBD_STATFS_NODELAY : 0;
1423 /* got statfs data */
1424 rc = obd_statfs(NULL, lov->lov_tgts[index]->ltd_exp, &stat_buf,
1425 cfs_time_shift_64(-OBD_STATFS_CACHE_SECONDS),
1429 if (copy_to_user(data->ioc_pbuf1, &stat_buf,
1430 min((int)data->ioc_plen1,
1431 (int)sizeof(stat_buf))))
1435 case OBD_IOC_LOV_GET_CONFIG: {
1436 struct obd_ioctl_data *data;
1437 struct lov_desc *desc;
1442 if (obd_ioctl_getdata(&buf, &len, uarg))
1445 data = (struct obd_ioctl_data *)buf;
1447 if (sizeof(*desc) > data->ioc_inllen1) {
1448 obd_ioctl_freedata(buf, len);
1452 if (sizeof(uuidp->uuid) * count > data->ioc_inllen2) {
1453 obd_ioctl_freedata(buf, len);
1457 if (sizeof(__u32) * count > data->ioc_inllen3) {
1458 obd_ioctl_freedata(buf, len);
1462 desc = (struct lov_desc *)data->ioc_inlbuf1;
1463 memcpy(desc, &(lov->desc), sizeof(*desc));
1465 uuidp = (struct obd_uuid *)data->ioc_inlbuf2;
1466 genp = (__u32 *)data->ioc_inlbuf3;
1467 /* the uuid will be empty for deleted OSTs */
1468 for (i = 0; i < count; i++, uuidp++, genp++) {
1469 if (!lov->lov_tgts[i])
1471 *uuidp = lov->lov_tgts[i]->ltd_uuid;
1472 *genp = lov->lov_tgts[i]->ltd_gen;
1475 if (copy_to_user(uarg, buf, len))
1477 obd_ioctl_freedata(buf, len);
1480 case LL_IOC_LOV_GETSTRIPE:
1481 rc = lov_getstripe(exp, karg, uarg);
1483 case OBD_IOC_QUOTACTL: {
1484 struct if_quotactl *qctl = karg;
1485 struct lov_tgt_desc *tgt = NULL;
1486 struct obd_quotactl *oqctl;
1488 if (qctl->qc_valid == QC_OSTIDX) {
1489 if (count <= qctl->qc_idx)
1492 tgt = lov->lov_tgts[qctl->qc_idx];
1493 if (!tgt || !tgt->ltd_exp)
1495 } else if (qctl->qc_valid == QC_UUID) {
1496 for (i = 0; i < count; i++) {
1497 tgt = lov->lov_tgts[i];
1499 !obd_uuid_equals(&tgt->ltd_uuid,
1515 LASSERT(tgt && tgt->ltd_exp);
1516 oqctl = kzalloc(sizeof(*oqctl), GFP_NOFS);
1520 QCTL_COPY(oqctl, qctl);
1521 rc = obd_quotactl(tgt->ltd_exp, oqctl);
1523 QCTL_COPY(qctl, oqctl);
1524 qctl->qc_valid = QC_OSTIDX;
1525 qctl->obd_uuid = tgt->ltd_uuid;
1536 for (i = 0; i < count; i++) {
1538 struct obd_device *osc_obd;
1540 /* OST was disconnected */
1541 if (!lov->lov_tgts[i] || !lov->lov_tgts[i]->ltd_exp)
1544 /* ll_umount_begin() sets force flag but for lov, not
1545 * osc. Let's pass it through
1547 osc_obd = class_exp2obd(lov->lov_tgts[i]->ltd_exp);
1548 osc_obd->obd_force = obddev->obd_force;
1549 err = obd_iocontrol(cmd, lov->lov_tgts[i]->ltd_exp,
1551 if (err == -ENODATA && cmd == OBD_IOC_POLL_QUOTACHECK)
1554 if (lov->lov_tgts[i]->ltd_active) {
1555 CDEBUG(err == -ENOTTY ?
1556 D_IOCTL : D_WARNING,
1557 "iocontrol OSC %s on OST idx %d cmd %x: err = %d\n",
1558 lov_uuid2str(lov, i),
1575 #define FIEMAP_BUFFER_SIZE 4096
1578 * Non-zero fe_logical indicates that this is a continuation FIEMAP
1579 * call. The local end offset and the device are sent in the first
1580 * fm_extent. This function calculates the stripe number from the index.
1581 * This function returns a stripe_no on which mapping is to be restarted.
1583 * This function returns fm_end_offset which is the in-OST offset at which
1584 * mapping should be restarted. If fm_end_offset=0 is returned then caller
1585 * will re-calculate proper offset in next stripe.
1586 * Note that the first extent is passed to lov_get_info via the value field.
1588 * \param fiemap fiemap request header
1589 * \param lsm striping information for the file
1590 * \param fm_start logical start of mapping
1591 * \param fm_end logical end of mapping
1592 * \param start_stripe starting stripe will be returned in this
1594 static u64 fiemap_calc_fm_end_offset(struct ll_user_fiemap *fiemap,
1595 struct lov_stripe_md *lsm, u64 fm_start,
1596 u64 fm_end, int *start_stripe)
1598 u64 local_end = fiemap->fm_extents[0].fe_logical;
1599 u64 lun_start, lun_end;
1601 int stripe_no = -1, i;
1603 if (fiemap->fm_extent_count == 0 ||
1604 fiemap->fm_extents[0].fe_logical == 0)
1607 /* Find out stripe_no from ost_index saved in the fe_device */
1608 for (i = 0; i < lsm->lsm_stripe_count; i++) {
1609 struct lov_oinfo *oinfo = lsm->lsm_oinfo[i];
1611 if (lov_oinfo_is_dummy(oinfo))
1614 if (oinfo->loi_ost_idx == fiemap->fm_extents[0].fe_device) {
1619 if (stripe_no == -1)
1622 /* If we have finished mapping on previous device, shift logical
1623 * offset to start of next device
1625 if ((lov_stripe_intersects(lsm, stripe_no, fm_start, fm_end,
1626 &lun_start, &lun_end)) != 0 &&
1627 local_end < lun_end) {
1628 fm_end_offset = local_end;
1629 *start_stripe = stripe_no;
1631 /* This is a special value to indicate that caller should
1632 * calculate offset in next stripe.
1635 *start_stripe = (stripe_no + 1) % lsm->lsm_stripe_count;
1638 return fm_end_offset;
1642 * We calculate on which OST the mapping will end. If the length of mapping
1643 * is greater than (stripe_size * stripe_count) then the last_stripe will
1644 * will be one just before start_stripe. Else we check if the mapping
1645 * intersects each OST and find last_stripe.
1646 * This function returns the last_stripe and also sets the stripe_count
1647 * over which the mapping is spread
1649 * \param lsm striping information for the file
1650 * \param fm_start logical start of mapping
1651 * \param fm_end logical end of mapping
1652 * \param start_stripe starting stripe of the mapping
1653 * \param stripe_count the number of stripes across which to map is returned
1655 * \retval last_stripe return the last stripe of the mapping
1657 static int fiemap_calc_last_stripe(struct lov_stripe_md *lsm, u64 fm_start,
1658 u64 fm_end, int start_stripe,
1662 u64 obd_start, obd_end;
1665 if (fm_end - fm_start > lsm->lsm_stripe_size * lsm->lsm_stripe_count) {
1666 last_stripe = start_stripe < 1 ? lsm->lsm_stripe_count - 1 :
1668 *stripe_count = lsm->lsm_stripe_count;
1670 for (j = 0, i = start_stripe; j < lsm->lsm_stripe_count;
1671 i = (i + 1) % lsm->lsm_stripe_count, j++) {
1672 if ((lov_stripe_intersects(lsm, i, fm_start, fm_end,
1673 &obd_start, &obd_end)) == 0)
1677 last_stripe = (start_stripe + j - 1) % lsm->lsm_stripe_count;
1684 * Set fe_device and copy extents from local buffer into main return buffer.
1686 * \param fiemap fiemap request header
1687 * \param lcl_fm_ext array of local fiemap extents to be copied
1688 * \param ost_index OST index to be written into the fm_device field for each
1690 * \param ext_count number of extents to be copied
1691 * \param current_extent where to start copying in main extent array
1693 static void fiemap_prepare_and_copy_exts(struct ll_user_fiemap *fiemap,
1694 struct ll_fiemap_extent *lcl_fm_ext,
1695 int ost_index, unsigned int ext_count,
1701 for (ext = 0; ext < ext_count; ext++) {
1702 lcl_fm_ext[ext].fe_device = ost_index;
1703 lcl_fm_ext[ext].fe_flags |= FIEMAP_EXTENT_NET;
1706 /* Copy fm_extent's from fm_local to return buffer */
1707 to = (char *)fiemap + fiemap_count_to_size(current_extent);
1708 memcpy(to, lcl_fm_ext, ext_count * sizeof(struct ll_fiemap_extent));
1712 * Break down the FIEMAP request and send appropriate calls to individual OSTs.
1713 * This also handles the restarting of FIEMAP calls in case mapping overflows
1714 * the available number of extents in single call.
1716 static int lov_fiemap(struct lov_obd *lov, __u32 keylen, void *key,
1717 __u32 *vallen, void *val, struct lov_stripe_md *lsm)
1719 struct ll_fiemap_info_key *fm_key = key;
1720 struct ll_user_fiemap *fiemap = val;
1721 struct ll_user_fiemap *fm_local = NULL;
1722 struct ll_fiemap_extent *lcl_fm_ext;
1724 unsigned int get_num_extents = 0;
1725 int ost_index = 0, actual_start_stripe, start_stripe;
1726 u64 fm_start, fm_end, fm_length, fm_end_offset;
1728 int current_extent = 0, rc = 0, i;
1729 int ost_eof = 0; /* EOF for object */
1730 int ost_done = 0; /* done with required mapping for this OST? */
1732 int cur_stripe = 0, cur_stripe_wrap = 0, stripe_count;
1733 unsigned int buffer_size = FIEMAP_BUFFER_SIZE;
1735 if (!lsm_has_objects(lsm)) {
1736 if (lsm && lsm_is_released(lsm) && (fm_key->fiemap.fm_start <
1737 fm_key->oa.o_size)) {
1739 * released file, return a minimal FIEMAP if
1740 * request fits in file-size.
1742 fiemap->fm_mapped_extents = 1;
1743 fiemap->fm_extents[0].fe_logical =
1744 fm_key->fiemap.fm_start;
1745 if (fm_key->fiemap.fm_start + fm_key->fiemap.fm_length <
1746 fm_key->oa.o_size) {
1747 fiemap->fm_extents[0].fe_length =
1748 fm_key->fiemap.fm_length;
1750 fiemap->fm_extents[0].fe_length =
1751 fm_key->oa.o_size - fm_key->fiemap.fm_start;
1752 fiemap->fm_extents[0].fe_flags |=
1753 (FIEMAP_EXTENT_UNKNOWN |
1754 FIEMAP_EXTENT_LAST);
1761 if (fiemap_count_to_size(fm_key->fiemap.fm_extent_count) < buffer_size)
1762 buffer_size = fiemap_count_to_size(fm_key->fiemap.fm_extent_count);
1764 fm_local = libcfs_kvzalloc(buffer_size, GFP_NOFS);
1769 lcl_fm_ext = &fm_local->fm_extents[0];
1771 count_local = fiemap_size_to_count(buffer_size);
1773 memcpy(fiemap, &fm_key->fiemap, sizeof(*fiemap));
1774 fm_start = fiemap->fm_start;
1775 fm_length = fiemap->fm_length;
1776 /* Calculate start stripe, last stripe and length of mapping */
1777 start_stripe = lov_stripe_number(lsm, fm_start);
1778 actual_start_stripe = start_stripe;
1779 fm_end = (fm_length == ~0ULL ? fm_key->oa.o_size :
1780 fm_start + fm_length - 1);
1781 /* If fm_length != ~0ULL but fm_start+fm_length-1 exceeds file size */
1782 if (fm_end > fm_key->oa.o_size)
1783 fm_end = fm_key->oa.o_size;
1785 last_stripe = fiemap_calc_last_stripe(lsm, fm_start, fm_end,
1786 actual_start_stripe,
1789 fm_end_offset = fiemap_calc_fm_end_offset(fiemap, lsm, fm_start,
1790 fm_end, &start_stripe);
1791 if (fm_end_offset == -EINVAL) {
1796 if (fiemap_count_to_size(fiemap->fm_extent_count) > *vallen)
1797 fiemap->fm_extent_count = fiemap_size_to_count(*vallen);
1798 if (fiemap->fm_extent_count == 0) {
1799 get_num_extents = 1;
1802 /* Check each stripe */
1803 for (cur_stripe = start_stripe, i = 0; i < stripe_count;
1804 i++, cur_stripe = (cur_stripe + 1) % lsm->lsm_stripe_count) {
1805 u64 req_fm_len; /* Stores length of required mapping */
1806 u64 len_mapped_single_call;
1807 u64 lun_start, lun_end, obd_object_end;
1808 unsigned int ext_count;
1810 cur_stripe_wrap = cur_stripe;
1812 /* Find out range of mapping on this stripe */
1813 if ((lov_stripe_intersects(lsm, cur_stripe, fm_start, fm_end,
1814 &lun_start, &obd_object_end)) == 0)
1817 if (lov_oinfo_is_dummy(lsm->lsm_oinfo[cur_stripe])) {
1822 /* If this is a continuation FIEMAP call and we are on
1823 * starting stripe then lun_start needs to be set to
1826 if (fm_end_offset != 0 && cur_stripe == start_stripe)
1827 lun_start = fm_end_offset;
1829 if (fm_length != ~0ULL) {
1830 /* Handle fm_start + fm_length overflow */
1831 if (fm_start + fm_length < fm_start)
1832 fm_length = ~0ULL - fm_start;
1833 lun_end = lov_size_to_stripe(lsm, fm_start + fm_length,
1839 if (lun_start == lun_end)
1842 req_fm_len = obd_object_end - lun_start;
1843 fm_local->fm_length = 0;
1844 len_mapped_single_call = 0;
1846 /* If the output buffer is very large and the objects have many
1847 * extents we may need to loop on a single OST repeatedly
1852 if (get_num_extents == 0) {
1853 /* Don't get too many extents. */
1854 if (current_extent + count_local >
1855 fiemap->fm_extent_count)
1856 count_local = fiemap->fm_extent_count -
1860 lun_start += len_mapped_single_call;
1861 fm_local->fm_length = req_fm_len - len_mapped_single_call;
1862 req_fm_len = fm_local->fm_length;
1863 fm_local->fm_extent_count = count_local;
1864 fm_local->fm_mapped_extents = 0;
1865 fm_local->fm_flags = fiemap->fm_flags;
1867 fm_key->oa.o_oi = lsm->lsm_oinfo[cur_stripe]->loi_oi;
1868 ost_index = lsm->lsm_oinfo[cur_stripe]->loi_ost_idx;
1870 if (ost_index < 0 ||
1871 ost_index >= lov->desc.ld_tgt_count) {
1876 /* If OST is inactive, return extent with UNKNOWN flag */
1877 if (!lov->lov_tgts[ost_index]->ltd_active) {
1878 fm_local->fm_flags |= FIEMAP_EXTENT_LAST;
1879 fm_local->fm_mapped_extents = 1;
1881 lcl_fm_ext[0].fe_logical = lun_start;
1882 lcl_fm_ext[0].fe_length = obd_object_end -
1884 lcl_fm_ext[0].fe_flags |= FIEMAP_EXTENT_UNKNOWN;
1889 fm_local->fm_start = lun_start;
1890 fm_local->fm_flags &= ~FIEMAP_FLAG_DEVICE_ORDER;
1891 memcpy(&fm_key->fiemap, fm_local, sizeof(*fm_local));
1892 *vallen = fiemap_count_to_size(fm_local->fm_extent_count);
1893 rc = obd_get_info(NULL,
1894 lov->lov_tgts[ost_index]->ltd_exp,
1895 keylen, key, vallen, fm_local, lsm);
1900 ext_count = fm_local->fm_mapped_extents;
1901 if (ext_count == 0) {
1903 /* If last stripe has hole at the end,
1904 * then we need to return
1906 if (cur_stripe_wrap == last_stripe) {
1907 fiemap->fm_mapped_extents = 0;
1913 /* If we just need num of extents then go to next device */
1914 if (get_num_extents) {
1915 current_extent += ext_count;
1919 len_mapped_single_call = lcl_fm_ext[ext_count-1].fe_logical -
1920 lun_start + lcl_fm_ext[ext_count - 1].fe_length;
1922 /* Have we finished mapping on this device? */
1923 if (req_fm_len <= len_mapped_single_call)
1926 /* Clear the EXTENT_LAST flag which can be present on
1929 if (lcl_fm_ext[ext_count-1].fe_flags & FIEMAP_EXTENT_LAST)
1930 lcl_fm_ext[ext_count - 1].fe_flags &=
1931 ~FIEMAP_EXTENT_LAST;
1933 curr_loc = lov_stripe_size(lsm,
1934 lcl_fm_ext[ext_count - 1].fe_logical+
1935 lcl_fm_ext[ext_count - 1].fe_length,
1937 if (curr_loc >= fm_key->oa.o_size)
1940 fiemap_prepare_and_copy_exts(fiemap, lcl_fm_ext,
1941 ost_index, ext_count,
1944 current_extent += ext_count;
1946 /* Ran out of available extents? */
1947 if (current_extent >= fiemap->fm_extent_count)
1949 } while (ost_done == 0 && ost_eof == 0);
1951 if (cur_stripe_wrap == last_stripe)
1956 /* Indicate that we are returning device offsets unless file just has
1959 if (lsm->lsm_stripe_count > 1)
1960 fiemap->fm_flags |= FIEMAP_FLAG_DEVICE_ORDER;
1962 if (get_num_extents)
1963 goto skip_last_device_calc;
1965 /* Check if we have reached the last stripe and whether mapping for that
1968 if (cur_stripe_wrap == last_stripe) {
1969 if (ost_done || ost_eof)
1970 fiemap->fm_extents[current_extent - 1].fe_flags |=
1974 skip_last_device_calc:
1975 fiemap->fm_mapped_extents = current_extent;
1982 static int lov_get_info(const struct lu_env *env, struct obd_export *exp,
1983 __u32 keylen, void *key, __u32 *vallen, void *val,
1984 struct lov_stripe_md *lsm)
1986 struct obd_device *obddev = class_exp2obd(exp);
1987 struct lov_obd *lov = &obddev->u.lov;
1990 if (!vallen || !val)
1995 if (KEY_IS(KEY_LOCK_TO_STRIPE)) {
1998 struct ldlm_lock *lock;
2000 struct ldlm_res_id *res_id = &data->lock->l_resource->lr_name;
2001 struct lov_oinfo *loi;
2002 __u32 *stripe = val;
2004 if (*vallen < sizeof(*stripe)) {
2008 *vallen = sizeof(*stripe);
2010 /* XXX This is another one of those bits that will need to
2011 * change if we ever actually support nested LOVs. It uses
2012 * the lock's export to find out which stripe it is.
2014 /* XXX - it's assumed all the locks for deleted OSTs have
2015 * been cancelled. Also, the export for deleted OSTs will
2016 * be NULL and won't match the lock's export.
2018 for (i = 0; i < lsm->lsm_stripe_count; i++) {
2019 loi = lsm->lsm_oinfo[i];
2020 if (lov_oinfo_is_dummy(loi))
2023 if (!lov->lov_tgts[loi->loi_ost_idx])
2025 if (lov->lov_tgts[loi->loi_ost_idx]->ltd_exp ==
2026 data->lock->l_conn_export &&
2027 ostid_res_name_eq(&loi->loi_oi, res_id)) {
2033 LDLM_ERROR(data->lock, "lock on inode without such object");
2034 dump_lsm(D_ERROR, lsm);
2037 } else if (KEY_IS(KEY_LAST_ID)) {
2038 struct obd_id_info *info = val;
2039 __u32 size = sizeof(u64);
2040 struct lov_tgt_desc *tgt;
2042 LASSERT(*vallen == sizeof(struct obd_id_info));
2043 tgt = lov->lov_tgts[info->idx];
2045 if (!tgt || !tgt->ltd_active) {
2050 rc = obd_get_info(env, tgt->ltd_exp, keylen, key,
2051 &size, info->data, NULL);
2054 } else if (KEY_IS(KEY_LOVDESC)) {
2055 struct lov_desc *desc_ret = val;
2056 *desc_ret = lov->desc;
2060 } else if (KEY_IS(KEY_FIEMAP)) {
2061 rc = lov_fiemap(lov, keylen, key, vallen, val, lsm);
2063 } else if (KEY_IS(KEY_CONNECT_FLAG)) {
2064 struct lov_tgt_desc *tgt;
2065 __u64 ost_idx = *((__u64 *)val);
2067 LASSERT(*vallen == sizeof(__u64));
2068 LASSERT(ost_idx < lov->desc.ld_tgt_count);
2069 tgt = lov->lov_tgts[ost_idx];
2071 if (!tgt || !tgt->ltd_exp) {
2076 *((__u64 *)val) = exp_connect_flags(tgt->ltd_exp);
2079 } else if (KEY_IS(KEY_TGT_COUNT)) {
2080 *((int *)val) = lov->desc.ld_tgt_count;
2092 static int lov_set_info_async(const struct lu_env *env, struct obd_export *exp,
2093 u32 keylen, void *key, u32 vallen,
2094 void *val, struct ptlrpc_request_set *set)
2096 struct obd_device *obddev = class_exp2obd(exp);
2097 struct lov_obd *lov = &obddev->u.lov;
2100 struct lov_tgt_desc *tgt;
2101 unsigned int incr = 0, check_uuid = 0, do_inactive = 0, no_set = 0;
2102 unsigned int next_id = 0, mds_con = 0;
2106 set = ptlrpc_prep_set();
2112 count = lov->desc.ld_tgt_count;
2114 if (KEY_IS(KEY_NEXT_ID)) {
2115 count = vallen / sizeof(struct obd_id_info);
2116 vallen = sizeof(u64);
2117 incr = sizeof(struct obd_id_info);
2120 } else if (KEY_IS(KEY_CHECKSUM)) {
2122 } else if (KEY_IS(KEY_EVICT_BY_NID)) {
2123 /* use defaults: do_inactive = incr = 0; */
2124 } else if (KEY_IS(KEY_MDS_CONN)) {
2126 } else if (KEY_IS(KEY_CACHE_SET)) {
2127 LASSERT(!lov->lov_cache);
2128 lov->lov_cache = val;
2130 cl_cache_incref(lov->lov_cache);
2133 for (i = 0; i < count; i++, val = (char *)val + incr) {
2135 tgt = lov->lov_tgts[((struct obd_id_info *)val)->idx];
2137 tgt = lov->lov_tgts[i];
2138 /* OST was disconnected */
2139 if (!tgt || !tgt->ltd_exp)
2142 /* OST is inactive and we don't want inactive OSCs */
2143 if (!tgt->ltd_active && !do_inactive)
2147 struct mds_group_info *mgi;
2149 LASSERT(vallen == sizeof(*mgi));
2150 mgi = (struct mds_group_info *)val;
2152 /* Only want a specific OSC */
2153 if (mgi->uuid && !obd_uuid_equals(mgi->uuid,
2157 err = obd_set_info_async(env, tgt->ltd_exp,
2158 keylen, key, sizeof(int),
2160 } else if (next_id) {
2161 err = obd_set_info_async(env, tgt->ltd_exp,
2162 keylen, key, vallen,
2163 ((struct obd_id_info *)val)->data, set);
2165 /* Only want a specific OSC */
2167 !obd_uuid_equals(val, &tgt->ltd_uuid))
2170 err = obd_set_info_async(env, tgt->ltd_exp,
2171 keylen, key, vallen, val, set);
2180 err = ptlrpc_set_wait(set);
2183 ptlrpc_set_destroy(set);
2188 void lov_stripe_lock(struct lov_stripe_md *md)
2189 __acquires(&md->lsm_lock)
2191 LASSERT(md->lsm_lock_owner != current_pid());
2192 spin_lock(&md->lsm_lock);
2193 LASSERT(md->lsm_lock_owner == 0);
2194 md->lsm_lock_owner = current_pid();
2197 void lov_stripe_unlock(struct lov_stripe_md *md)
2198 __releases(&md->lsm_lock)
2200 LASSERT(md->lsm_lock_owner == current_pid());
2201 md->lsm_lock_owner = 0;
2202 spin_unlock(&md->lsm_lock);
2205 static int lov_quotactl(struct obd_device *obd, struct obd_export *exp,
2206 struct obd_quotactl *oqctl)
2208 struct lov_obd *lov = &obd->u.lov;
2209 struct lov_tgt_desc *tgt;
2211 __u64 bhardlimit = 0;
2214 if (oqctl->qc_cmd != LUSTRE_Q_QUOTAON &&
2215 oqctl->qc_cmd != LUSTRE_Q_QUOTAOFF &&
2216 oqctl->qc_cmd != Q_GETOQUOTA &&
2217 oqctl->qc_cmd != Q_INITQUOTA &&
2218 oqctl->qc_cmd != LUSTRE_Q_SETQUOTA &&
2219 oqctl->qc_cmd != Q_FINVALIDATE) {
2220 CERROR("bad quota opc %x for lov obd\n", oqctl->qc_cmd);
2226 for (i = 0; i < lov->desc.ld_tgt_count; i++) {
2229 tgt = lov->lov_tgts[i];
2234 if (!tgt->ltd_active || tgt->ltd_reap) {
2235 if (oqctl->qc_cmd == Q_GETOQUOTA &&
2236 lov->lov_tgts[i]->ltd_activate) {
2238 CERROR("ost %d is inactive\n", i);
2240 CDEBUG(D_HA, "ost %d is inactive\n", i);
2245 err = obd_quotactl(tgt->ltd_exp, oqctl);
2247 if (tgt->ltd_active && !rc)
2252 if (oqctl->qc_cmd == Q_GETOQUOTA) {
2253 curspace += oqctl->qc_dqblk.dqb_curspace;
2254 bhardlimit += oqctl->qc_dqblk.dqb_bhardlimit;
2259 if (oqctl->qc_cmd == Q_GETOQUOTA) {
2260 oqctl->qc_dqblk.dqb_curspace = curspace;
2261 oqctl->qc_dqblk.dqb_bhardlimit = bhardlimit;
2266 static int lov_quotacheck(struct obd_device *obd, struct obd_export *exp,
2267 struct obd_quotactl *oqctl)
2269 struct lov_obd *lov = &obd->u.lov;
2274 for (i = 0; i < lov->desc.ld_tgt_count; i++) {
2275 if (!lov->lov_tgts[i])
2278 /* Skip quota check on the administratively disabled OSTs. */
2279 if (!lov->lov_tgts[i]->ltd_activate) {
2280 CWARN("lov idx %d was administratively disabled, skip quotacheck on it.\n",
2285 if (!lov->lov_tgts[i]->ltd_active) {
2286 CERROR("lov idx %d inactive\n", i);
2292 for (i = 0; i < lov->desc.ld_tgt_count; i++) {
2295 if (!lov->lov_tgts[i] || !lov->lov_tgts[i]->ltd_activate)
2298 err = obd_quotacheck(lov->lov_tgts[i]->ltd_exp, oqctl);
2309 static struct obd_ops lov_obd_ops = {
2310 .owner = THIS_MODULE,
2312 .precleanup = lov_precleanup,
2313 .cleanup = lov_cleanup,
2314 /*.process_config = lov_process_config,*/
2315 .connect = lov_connect,
2316 .disconnect = lov_disconnect,
2317 .statfs = lov_statfs,
2318 .statfs_async = lov_statfs_async,
2319 .packmd = lov_packmd,
2320 .unpackmd = lov_unpackmd,
2321 .create = lov_create,
2322 .destroy = lov_destroy,
2323 .getattr_async = lov_getattr_async,
2324 .setattr_async = lov_setattr_async,
2325 .adjust_kms = lov_adjust_kms,
2326 .find_cbdata = lov_find_cbdata,
2327 .iocontrol = lov_iocontrol,
2328 .get_info = lov_get_info,
2329 .set_info_async = lov_set_info_async,
2330 .notify = lov_notify,
2331 .pool_new = lov_pool_new,
2332 .pool_rem = lov_pool_remove,
2333 .pool_add = lov_pool_add,
2334 .pool_del = lov_pool_del,
2335 .getref = lov_getref,
2336 .putref = lov_putref,
2337 .quotactl = lov_quotactl,
2338 .quotacheck = lov_quotacheck,
2341 struct kmem_cache *lov_oinfo_slab;
2343 static int __init lov_init(void)
2345 struct lprocfs_static_vars lvars = { NULL };
2348 /* print an address of _any_ initialized kernel symbol from this
2349 * module, to allow debugging with gdb that doesn't support data
2350 * symbols from modules.
2352 CDEBUG(D_INFO, "Lustre LOV module (%p).\n", &lov_caches);
2354 rc = lu_kmem_init(lov_caches);
2358 lov_oinfo_slab = kmem_cache_create("lov_oinfo",
2359 sizeof(struct lov_oinfo),
2360 0, SLAB_HWCACHE_ALIGN, NULL);
2361 if (!lov_oinfo_slab) {
2362 lu_kmem_fini(lov_caches);
2365 lprocfs_lov_init_vars(&lvars);
2367 rc = class_register_type(&lov_obd_ops, NULL,
2368 LUSTRE_LOV_NAME, &lov_device_type);
2371 kmem_cache_destroy(lov_oinfo_slab);
2372 lu_kmem_fini(lov_caches);
2378 static void /*__exit*/ lov_exit(void)
2380 class_unregister_type(LUSTRE_LOV_NAME);
2381 kmem_cache_destroy(lov_oinfo_slab);
2383 lu_kmem_fini(lov_caches);
2386 MODULE_AUTHOR("OpenSFS, Inc. <http://www.lustre.org/>");
2387 MODULE_DESCRIPTION("Lustre Logical Object Volume");
2388 MODULE_LICENSE("GPL");
2389 MODULE_VERSION(LUSTRE_VERSION_STRING);
2391 module_init(lov_init);
2392 module_exit(lov_exit);