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.sun.com/software/products/lustre/docs/GPLv2.pdf
20 * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara,
21 * CA 95054 USA or visit www.sun.com if you need additional information or
27 * Copyright (c) 2003, 2010, Oracle and/or its affiliates. All rights reserved.
28 * Use is subject to license terms.
30 * Copyright (c) 2011, 2015, Intel Corporation.
33 * This file is part of Lustre, http://www.lustre.org/
34 * Lustre is a trademark of Sun Microsystems, Inc.
37 #define DEBUG_SUBSYSTEM S_MDC
39 #include <linux/module.h>
42 #include <obd_class.h>
43 #include <lustre_dlm.h>
44 #include <lustre_fid.h>
45 #include <lustre_intent.h>
46 #include <lustre_mdc.h>
47 #include <lustre_net.h>
48 #include <lustre_req_layout.h>
49 #include <lustre_swab.h>
51 #include "mdc_internal.h"
53 struct mdc_getattr_args {
54 struct obd_export *ga_exp;
55 struct md_enqueue_info *ga_minfo;
58 int it_open_error(int phase, struct lookup_intent *it)
60 if (it_disposition(it, DISP_OPEN_LEASE)) {
61 if (phase >= DISP_OPEN_LEASE)
66 if (it_disposition(it, DISP_OPEN_OPEN)) {
67 if (phase >= DISP_OPEN_OPEN)
73 if (it_disposition(it, DISP_OPEN_CREATE)) {
74 if (phase >= DISP_OPEN_CREATE)
80 if (it_disposition(it, DISP_LOOKUP_EXECD)) {
81 if (phase >= DISP_LOOKUP_EXECD)
87 if (it_disposition(it, DISP_IT_EXECD)) {
88 if (phase >= DISP_IT_EXECD)
94 CERROR("it disp: %X, status: %d\n", it->it_disposition, it->it_status);
99 EXPORT_SYMBOL(it_open_error);
101 /* this must be called on a lockh that is known to have a referenced lock */
102 int mdc_set_lock_data(struct obd_export *exp, __u64 *lockh, void *data,
105 struct ldlm_lock *lock;
106 struct inode *new_inode = data;
115 lock = ldlm_handle2lock((struct lustre_handle *)lockh);
117 LASSERT(lock != NULL);
118 lock_res_and_lock(lock);
119 if (lock->l_resource->lr_lvb_inode &&
120 lock->l_resource->lr_lvb_inode != data) {
121 struct inode *old_inode = lock->l_resource->lr_lvb_inode;
122 LASSERTF(old_inode->i_state & I_FREEING,
123 "Found existing inode %p/%lu/%u state %lu in lock: "
124 "setting data to %p/%lu/%u\n", old_inode,
125 old_inode->i_ino, old_inode->i_generation,
127 new_inode, new_inode->i_ino, new_inode->i_generation);
129 lock->l_resource->lr_lvb_inode = new_inode;
131 *bits = lock->l_policy_data.l_inodebits.bits;
133 unlock_res_and_lock(lock);
139 enum ldlm_mode mdc_lock_match(struct obd_export *exp, __u64 flags,
140 const struct lu_fid *fid, enum ldlm_type type,
141 union ldlm_policy_data *policy,
142 enum ldlm_mode mode, struct lustre_handle *lockh)
144 struct ldlm_res_id res_id;
148 fid_build_reg_res_name(fid, &res_id);
149 /* LU-4405: Clear bits not supported by server */
150 policy->l_inodebits.bits &= exp_connect_ibits(exp);
151 rc = ldlm_lock_match(class_exp2obd(exp)->obd_namespace, flags,
152 &res_id, type, policy, mode, lockh, 0);
156 int mdc_cancel_unused(struct obd_export *exp, const struct lu_fid *fid,
157 union ldlm_policy_data *policy, enum ldlm_mode mode,
158 enum ldlm_cancel_flags flags, void *opaque)
160 struct obd_device *obd = class_exp2obd(exp);
161 struct ldlm_res_id res_id;
166 fid_build_reg_res_name(fid, &res_id);
167 rc = ldlm_cli_cancel_unused_resource(obd->obd_namespace, &res_id,
168 policy, mode, flags, opaque);
172 int mdc_null_inode(struct obd_export *exp,
173 const struct lu_fid *fid)
175 struct ldlm_res_id res_id;
176 struct ldlm_resource *res;
177 struct ldlm_namespace *ns = class_exp2obd(exp)->obd_namespace;
180 LASSERTF(ns != NULL, "no namespace passed\n");
182 fid_build_reg_res_name(fid, &res_id);
184 res = ldlm_resource_get(ns, NULL, &res_id, 0, 0);
189 res->lr_lvb_inode = NULL;
192 ldlm_resource_putref(res);
196 static inline void mdc_clear_replay_flag(struct ptlrpc_request *req, int rc)
198 /* Don't hold error requests for replay. */
199 if (req->rq_replay) {
200 spin_lock(&req->rq_lock);
202 spin_unlock(&req->rq_lock);
204 if (rc && req->rq_transno != 0) {
205 DEBUG_REQ(D_ERROR, req, "transno returned on error rc %d", rc);
210 /* Save a large LOV EA into the request buffer so that it is available
211 * for replay. We don't do this in the initial request because the
212 * original request doesn't need this buffer (at most it sends just the
213 * lov_mds_md) and it is a waste of RAM/bandwidth to send the empty
214 * buffer and may also be difficult to allocate and save a very large
215 * request buffer for each open. (bug 5707)
217 * OOM here may cause recovery failure if lmm is needed (only for the
218 * original open if the MDS crashed just when this client also OOM'd)
219 * but this is incredibly unlikely, and questionable whether the client
220 * could do MDS recovery under OOM anyways... */
221 static void mdc_realloc_openmsg(struct ptlrpc_request *req,
222 struct mdt_body *body)
226 /* FIXME: remove this explicit offset. */
227 rc = sptlrpc_cli_enlarge_reqbuf(req, DLM_INTENT_REC_OFF + 4,
228 body->mbo_eadatasize);
230 CERROR("Can't enlarge segment %d size to %d\n",
231 DLM_INTENT_REC_OFF + 4, body->mbo_eadatasize);
232 body->mbo_valid &= ~OBD_MD_FLEASIZE;
233 body->mbo_eadatasize = 0;
237 static struct ptlrpc_request *
238 mdc_intent_open_pack(struct obd_export *exp, struct lookup_intent *it,
239 struct md_op_data *op_data)
241 struct ptlrpc_request *req;
242 struct obd_device *obddev = class_exp2obd(exp);
243 struct ldlm_intent *lit;
244 const void *lmm = op_data->op_data;
245 __u32 lmmsize = op_data->op_data_size;
246 struct list_head cancels = LIST_HEAD_INIT(cancels);
252 it->it_create_mode = (it->it_create_mode & ~S_IFMT) | S_IFREG;
254 /* XXX: openlock is not cancelled for cross-refs. */
255 /* If inode is known, cancel conflicting OPEN locks. */
256 if (fid_is_sane(&op_data->op_fid2)) {
257 if (it->it_flags & MDS_OPEN_LEASE) { /* try to get lease */
258 if (it->it_flags & FMODE_WRITE)
263 if (it->it_flags & (FMODE_WRITE|MDS_OPEN_TRUNC))
266 else if (it->it_flags & FMODE_EXEC)
272 count = mdc_resource_get_unused(exp, &op_data->op_fid2,
277 /* If CREATE, cancel parent's UPDATE lock. */
278 if (it->it_op & IT_CREAT)
282 count += mdc_resource_get_unused(exp, &op_data->op_fid1,
284 MDS_INODELOCK_UPDATE);
286 req = ptlrpc_request_alloc(class_exp2cliimp(exp),
287 &RQF_LDLM_INTENT_OPEN);
289 ldlm_lock_list_put(&cancels, l_bl_ast, count);
290 RETURN(ERR_PTR(-ENOMEM));
293 req_capsule_set_size(&req->rq_pill, &RMF_NAME, RCL_CLIENT,
294 op_data->op_namelen + 1);
295 req_capsule_set_size(&req->rq_pill, &RMF_EADATA, RCL_CLIENT,
296 max(lmmsize, obddev->u.cli.cl_default_mds_easize));
298 rc = ldlm_prep_enqueue_req(exp, req, &cancels, count);
300 ptlrpc_request_free(req);
304 spin_lock(&req->rq_lock);
305 req->rq_replay = req->rq_import->imp_replayable;
306 spin_unlock(&req->rq_lock);
308 /* pack the intent */
309 lit = req_capsule_client_get(&req->rq_pill, &RMF_LDLM_INTENT);
310 lit->opc = (__u64)it->it_op;
312 /* pack the intended request */
313 mdc_open_pack(req, op_data, it->it_create_mode, 0, it->it_flags, lmm,
316 req_capsule_set_size(&req->rq_pill, &RMF_MDT_MD, RCL_SERVER,
317 obddev->u.cli.cl_max_mds_easize);
319 /* for remote client, fetch remote perm for current user */
320 if (client_is_remote(exp))
321 req_capsule_set_size(&req->rq_pill, &RMF_ACL, RCL_SERVER,
322 sizeof(struct mdt_remote_perm));
323 ptlrpc_request_set_replen(req);
327 static struct ptlrpc_request *
328 mdc_intent_getxattr_pack(struct obd_export *exp,
329 struct lookup_intent *it,
330 struct md_op_data *op_data)
332 struct ptlrpc_request *req;
333 struct ldlm_intent *lit;
336 struct list_head cancels = LIST_HEAD_INIT(cancels);
340 req = ptlrpc_request_alloc(class_exp2cliimp(exp),
341 &RQF_LDLM_INTENT_GETXATTR);
343 RETURN(ERR_PTR(-ENOMEM));
345 rc = ldlm_prep_enqueue_req(exp, req, &cancels, count);
347 ptlrpc_request_free(req);
351 /* pack the intent */
352 lit = req_capsule_client_get(&req->rq_pill, &RMF_LDLM_INTENT);
353 lit->opc = IT_GETXATTR;
355 maxdata = class_exp2cliimp(exp)->imp_connect_data.ocd_max_easize;
357 /* pack the intended request */
358 mdc_pack_body(req, &op_data->op_fid1, op_data->op_valid, maxdata, -1,
361 req_capsule_set_size(&req->rq_pill, &RMF_EADATA,
362 RCL_SERVER, maxdata);
364 req_capsule_set_size(&req->rq_pill, &RMF_EAVALS,
365 RCL_SERVER, maxdata);
367 req_capsule_set_size(&req->rq_pill, &RMF_EAVALS_LENS,
368 RCL_SERVER, maxdata);
370 ptlrpc_request_set_replen(req);
375 static struct ptlrpc_request *mdc_intent_unlink_pack(struct obd_export *exp,
376 struct lookup_intent *it,
377 struct md_op_data *op_data)
379 struct ptlrpc_request *req;
380 struct obd_device *obddev = class_exp2obd(exp);
381 struct ldlm_intent *lit;
385 req = ptlrpc_request_alloc(class_exp2cliimp(exp),
386 &RQF_LDLM_INTENT_UNLINK);
388 RETURN(ERR_PTR(-ENOMEM));
390 req_capsule_set_size(&req->rq_pill, &RMF_NAME, RCL_CLIENT,
391 op_data->op_namelen + 1);
393 rc = ldlm_prep_enqueue_req(exp, req, NULL, 0);
395 ptlrpc_request_free(req);
399 /* pack the intent */
400 lit = req_capsule_client_get(&req->rq_pill, &RMF_LDLM_INTENT);
401 lit->opc = (__u64)it->it_op;
403 /* pack the intended request */
404 mdc_unlink_pack(req, op_data);
406 req_capsule_set_size(&req->rq_pill, &RMF_MDT_MD, RCL_SERVER,
407 obddev->u.cli.cl_default_mds_easize);
408 ptlrpc_request_set_replen(req);
412 static struct ptlrpc_request *mdc_intent_getattr_pack(struct obd_export *exp,
413 struct lookup_intent *it,
414 struct md_op_data *op_data)
416 struct ptlrpc_request *req;
417 struct obd_device *obddev = class_exp2obd(exp);
418 u64 valid = OBD_MD_FLGETATTR | OBD_MD_FLEASIZE |
419 OBD_MD_FLMODEASIZE | OBD_MD_FLDIREA |
421 (client_is_remote(exp) ?
422 OBD_MD_FLRMTPERM : OBD_MD_FLACL);
423 struct ldlm_intent *lit;
428 req = ptlrpc_request_alloc(class_exp2cliimp(exp),
429 &RQF_LDLM_INTENT_GETATTR);
431 RETURN(ERR_PTR(-ENOMEM));
433 req_capsule_set_size(&req->rq_pill, &RMF_NAME, RCL_CLIENT,
434 op_data->op_namelen + 1);
436 rc = ldlm_prep_enqueue_req(exp, req, NULL, 0);
438 ptlrpc_request_free(req);
442 /* pack the intent */
443 lit = req_capsule_client_get(&req->rq_pill, &RMF_LDLM_INTENT);
444 lit->opc = (__u64)it->it_op;
446 if (obddev->u.cli.cl_default_mds_easize > 0)
447 easize = obddev->u.cli.cl_default_mds_easize;
449 easize = obddev->u.cli.cl_max_mds_easize;
451 /* pack the intended request */
452 mdc_getattr_pack(req, valid, it->it_flags, op_data, easize);
454 req_capsule_set_size(&req->rq_pill, &RMF_MDT_MD, RCL_SERVER, easize);
455 if (client_is_remote(exp))
456 req_capsule_set_size(&req->rq_pill, &RMF_ACL, RCL_SERVER,
457 sizeof(struct mdt_remote_perm));
458 ptlrpc_request_set_replen(req);
462 static struct ptlrpc_request *mdc_intent_layout_pack(struct obd_export *exp,
463 struct lookup_intent *it,
464 struct md_op_data *unused)
466 struct obd_device *obd = class_exp2obd(exp);
467 struct ptlrpc_request *req;
468 struct ldlm_intent *lit;
469 struct layout_intent *layout;
473 req = ptlrpc_request_alloc(class_exp2cliimp(exp),
474 &RQF_LDLM_INTENT_LAYOUT);
476 RETURN(ERR_PTR(-ENOMEM));
478 req_capsule_set_size(&req->rq_pill, &RMF_EADATA, RCL_CLIENT, 0);
479 rc = ldlm_prep_enqueue_req(exp, req, NULL, 0);
481 ptlrpc_request_free(req);
485 /* pack the intent */
486 lit = req_capsule_client_get(&req->rq_pill, &RMF_LDLM_INTENT);
487 lit->opc = (__u64)it->it_op;
489 /* pack the layout intent request */
490 layout = req_capsule_client_get(&req->rq_pill, &RMF_LAYOUT_INTENT);
491 /* LAYOUT_INTENT_ACCESS is generic, specific operation will be
492 * set for replication */
493 layout->li_opc = LAYOUT_INTENT_ACCESS;
495 req_capsule_set_size(&req->rq_pill, &RMF_DLM_LVB, RCL_SERVER,
496 obd->u.cli.cl_default_mds_easize);
497 ptlrpc_request_set_replen(req);
501 static struct ptlrpc_request *
502 mdc_enqueue_pack(struct obd_export *exp, int lvb_len)
504 struct ptlrpc_request *req;
508 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_LDLM_ENQUEUE);
510 RETURN(ERR_PTR(-ENOMEM));
512 rc = ldlm_prep_enqueue_req(exp, req, NULL, 0);
514 ptlrpc_request_free(req);
518 req_capsule_set_size(&req->rq_pill, &RMF_DLM_LVB, RCL_SERVER, lvb_len);
519 ptlrpc_request_set_replen(req);
523 static int mdc_finish_enqueue(struct obd_export *exp,
524 struct ptlrpc_request *req,
525 struct ldlm_enqueue_info *einfo,
526 struct lookup_intent *it,
527 struct lustre_handle *lockh,
530 struct req_capsule *pill = &req->rq_pill;
531 struct ldlm_request *lockreq;
532 struct ldlm_reply *lockrep;
533 struct ldlm_lock *lock;
534 void *lvb_data = NULL;
539 /* Similarly, if we're going to replay this request, we don't want to
540 * actually get a lock, just perform the intent. */
541 if (req->rq_transno || req->rq_replay) {
542 lockreq = req_capsule_client_get(pill, &RMF_DLM_REQ);
543 lockreq->lock_flags |= ldlm_flags_to_wire(LDLM_FL_INTENT_ONLY);
546 if (rc == ELDLM_LOCK_ABORTED) {
548 memset(lockh, 0, sizeof(*lockh));
550 } else { /* rc = 0 */
551 lock = ldlm_handle2lock(lockh);
552 LASSERT(lock != NULL);
554 /* If the server gave us back a different lock mode, we should
555 * fix up our variables. */
556 if (lock->l_req_mode != einfo->ei_mode) {
557 ldlm_lock_addref(lockh, lock->l_req_mode);
558 ldlm_lock_decref(lockh, einfo->ei_mode);
559 einfo->ei_mode = lock->l_req_mode;
564 lockrep = req_capsule_server_get(pill, &RMF_DLM_REP);
565 LASSERT(lockrep != NULL); /* checked by ldlm_cli_enqueue() */
567 it->it_disposition = (int)lockrep->lock_policy_res1;
568 it->it_status = (int)lockrep->lock_policy_res2;
569 it->it_lock_mode = einfo->ei_mode;
570 it->it_lock_handle = lockh->cookie;
573 /* Technically speaking rq_transno must already be zero if
574 * it_status is in error, so the check is a bit redundant */
575 if ((!req->rq_transno || it->it_status < 0) && req->rq_replay)
576 mdc_clear_replay_flag(req, it->it_status);
578 /* If we're doing an IT_OPEN which did not result in an actual
579 * successful open, then we need to remove the bit which saves
580 * this request for unconditional replay.
582 * It's important that we do this first! Otherwise we might exit the
583 * function without doing so, and try to replay a failed create
585 if (it->it_op & IT_OPEN && req->rq_replay &&
586 (!it_disposition(it, DISP_OPEN_OPEN) || it->it_status != 0))
587 mdc_clear_replay_flag(req, it->it_status);
589 DEBUG_REQ(D_RPCTRACE, req, "op: %d disposition: %x, status: %d",
590 it->it_op, it->it_disposition, it->it_status);
592 /* We know what to expect, so we do any byte flipping required here */
593 if (it->it_op & (IT_OPEN | IT_UNLINK | IT_LOOKUP | IT_GETATTR)) {
594 struct mdt_body *body;
596 body = req_capsule_server_get(pill, &RMF_MDT_BODY);
598 CERROR ("Can't swab mdt_body\n");
602 if (it_disposition(it, DISP_OPEN_OPEN) &&
603 !it_open_error(DISP_OPEN_OPEN, it)) {
605 * If this is a successful OPEN request, we need to set
606 * replay handler and data early, so that if replay
607 * happens immediately after swabbing below, new reply
608 * is swabbed by that handler correctly.
610 mdc_set_open_replay_data(NULL, NULL, it);
613 if (body->mbo_valid & (OBD_MD_FLDIREA | OBD_MD_FLEASIZE)) {
616 mdc_update_max_ea_from_body(exp, body);
619 * The eadata is opaque; just check that it is there.
620 * Eventually, obd_unpackmd() will check the contents.
622 eadata = req_capsule_server_sized_get(pill, &RMF_MDT_MD,
623 body->mbo_eadatasize);
627 /* save lvb data and length in case this is for layout
630 lvb_len = body->mbo_eadatasize;
633 * We save the reply LOV EA in case we have to replay a
634 * create for recovery. If we didn't allocate a large
635 * enough request buffer above we need to reallocate it
636 * here to hold the actual LOV EA.
638 * To not save LOV EA if request is not going to replay
639 * (for example error one).
641 if ((it->it_op & IT_OPEN) && req->rq_replay) {
643 if (req_capsule_get_size(pill, &RMF_EADATA,
645 body->mbo_eadatasize)
646 mdc_realloc_openmsg(req, body);
648 req_capsule_shrink(pill, &RMF_EADATA,
649 body->mbo_eadatasize,
652 req_capsule_set_size(pill, &RMF_EADATA,
654 body->mbo_eadatasize);
656 lmm = req_capsule_client_get(pill, &RMF_EADATA);
659 body->mbo_eadatasize);
663 if (body->mbo_valid & OBD_MD_FLRMTPERM) {
664 struct mdt_remote_perm *perm;
666 LASSERT(client_is_remote(exp));
667 perm = req_capsule_server_swab_get(pill, &RMF_ACL,
668 lustre_swab_mdt_remote_perm);
672 } else if (it->it_op & IT_LAYOUT) {
673 /* maybe the lock was granted right away and layout
674 * is packed into RMF_DLM_LVB of req */
675 lvb_len = req_capsule_get_size(pill, &RMF_DLM_LVB, RCL_SERVER);
677 lvb_data = req_capsule_server_sized_get(pill,
678 &RMF_DLM_LVB, lvb_len);
679 if (lvb_data == NULL)
684 /* fill in stripe data for layout lock.
685 * LU-6581: trust layout data only if layout lock is granted. The MDT
686 * has stopped sending layout unless the layout lock is granted. The
687 * client still does this checking in case it's talking with an old
688 * server. - Jinshan */
689 lock = ldlm_handle2lock(lockh);
690 if (lock != NULL && ldlm_has_layout(lock) && lvb_data != NULL &&
691 !(lockrep->lock_flags & LDLM_FL_BLOCKED_MASK)) {
694 LDLM_DEBUG(lock, "layout lock returned by: %s, lvb_len: %d",
695 ldlm_it2str(it->it_op), lvb_len);
697 OBD_ALLOC_LARGE(lmm, lvb_len);
702 memcpy(lmm, lvb_data, lvb_len);
704 /* install lvb_data */
705 lock_res_and_lock(lock);
706 if (lock->l_lvb_data == NULL) {
707 lock->l_lvb_type = LVB_T_LAYOUT;
708 lock->l_lvb_data = lmm;
709 lock->l_lvb_len = lvb_len;
712 unlock_res_and_lock(lock);
714 OBD_FREE_LARGE(lmm, lvb_len);
722 /* We always reserve enough space in the reply packet for a stripe MD, because
723 * we don't know in advance the file type. */
724 int mdc_enqueue(struct obd_export *exp,
725 struct ldlm_enqueue_info *einfo,
726 const union ldlm_policy_data *policy,
727 struct lookup_intent *it, struct md_op_data *op_data,
728 struct lustre_handle *lockh, __u64 extra_lock_flags)
730 struct obd_device *obddev = class_exp2obd(exp);
731 struct ptlrpc_request *req = NULL;
732 __u64 flags, saved_flags = extra_lock_flags;
733 struct ldlm_res_id res_id;
734 static const union ldlm_policy_data lookup_policy = {
735 .l_inodebits = { MDS_INODELOCK_LOOKUP } };
736 static const union ldlm_policy_data update_policy = {
737 .l_inodebits = { MDS_INODELOCK_UPDATE } };
738 static const union ldlm_policy_data layout_policy = {
739 .l_inodebits = { MDS_INODELOCK_LAYOUT } };
740 static const union ldlm_policy_data getxattr_policy = {
741 .l_inodebits = { MDS_INODELOCK_XATTR } };
742 int generation, resends = 0;
743 struct ldlm_reply *lockrep;
744 enum lvb_type lvb_type = 0;
748 LASSERTF(!it || einfo->ei_type == LDLM_IBITS, "lock type %d\n",
750 fid_build_reg_res_name(&op_data->op_fid1, &res_id);
753 LASSERT(policy == NULL);
755 saved_flags |= LDLM_FL_HAS_INTENT;
756 if (it->it_op & (IT_OPEN | IT_UNLINK | IT_GETATTR | IT_READDIR))
757 policy = &update_policy;
758 else if (it->it_op & IT_LAYOUT)
759 policy = &layout_policy;
760 else if (it->it_op & (IT_GETXATTR | IT_SETXATTR))
761 policy = &getxattr_policy;
763 policy = &lookup_policy;
766 generation = obddev->u.cli.cl_import->imp_generation;
770 /* The only way right now is FLOCK. */
771 LASSERTF(einfo->ei_type == LDLM_FLOCK, "lock type %d\n",
773 res_id.name[3] = LDLM_FLOCK;
774 } else if (it->it_op & IT_OPEN) {
775 req = mdc_intent_open_pack(exp, it, op_data);
776 } else if (it->it_op & IT_UNLINK) {
777 req = mdc_intent_unlink_pack(exp, it, op_data);
778 } else if (it->it_op & (IT_GETATTR | IT_LOOKUP)) {
779 req = mdc_intent_getattr_pack(exp, it, op_data);
780 } else if (it->it_op & IT_READDIR) {
781 req = mdc_enqueue_pack(exp, 0);
782 } else if (it->it_op & IT_LAYOUT) {
783 if (!imp_connect_lvb_type(class_exp2cliimp(exp)))
785 req = mdc_intent_layout_pack(exp, it, op_data);
786 lvb_type = LVB_T_LAYOUT;
787 } else if (it->it_op & IT_GETXATTR) {
788 req = mdc_intent_getxattr_pack(exp, it, op_data);
795 RETURN(PTR_ERR(req));
798 req->rq_generation_set = 1;
799 req->rq_import_generation = generation;
800 req->rq_sent = cfs_time_current_sec() + resends;
803 /* It is important to obtain modify RPC slot first (if applicable), so
804 * that threads that are waiting for a modify RPC slot are not polluting
805 * our rpcs in flight counter.
806 * We do not do flock request limiting, though */
808 mdc_get_mod_rpc_slot(req, it);
809 rc = obd_get_request_slot(&obddev->u.cli);
811 mdc_put_mod_rpc_slot(req, it);
812 mdc_clear_replay_flag(req, 0);
813 ptlrpc_req_finished(req);
818 rc = ldlm_cli_enqueue(exp, &req, einfo, &res_id, policy, &flags, NULL,
819 0, lvb_type, lockh, 0);
821 /* For flock requests we immediatelly return without further
822 delay and let caller deal with the rest, since rest of
823 this function metadata processing makes no sense for flock
824 requests anyway. But in case of problem during comms with
825 Server (ETIMEDOUT) or any signal/kill attempt (EINTR), we
826 can not rely on caller and this mainly for F_UNLCKs
827 (explicits or automatically generated by Kernel to clean
828 current FLocks upon exit) that can't be trashed */
829 if (((rc == -EINTR) || (rc == -ETIMEDOUT)) &&
830 (einfo->ei_type == LDLM_FLOCK) &&
831 (einfo->ei_mode == LCK_NL))
836 obd_put_request_slot(&obddev->u.cli);
837 mdc_put_mod_rpc_slot(req, it);
840 CDEBUG(D_INFO, "%s: ldlm_cli_enqueue failed: rc = %d\n",
841 obddev->obd_name, rc);
843 mdc_clear_replay_flag(req, rc);
844 ptlrpc_req_finished(req);
848 lockrep = req_capsule_server_get(&req->rq_pill, &RMF_DLM_REP);
849 LASSERT(lockrep != NULL);
851 lockrep->lock_policy_res2 =
852 ptlrpc_status_ntoh(lockrep->lock_policy_res2);
854 /* Retry infinitely when the server returns -EINPROGRESS for the
855 * intent operation, when server returns -EINPROGRESS for acquiring
856 * intent lock, we'll retry in after_reply(). */
857 if (it && (int)lockrep->lock_policy_res2 == -EINPROGRESS) {
858 mdc_clear_replay_flag(req, rc);
859 ptlrpc_req_finished(req);
862 CDEBUG(D_HA, "%s: resend:%d op:%d "DFID"/"DFID"\n",
863 obddev->obd_name, resends, it->it_op,
864 PFID(&op_data->op_fid1), PFID(&op_data->op_fid2));
866 if (generation == obddev->u.cli.cl_import->imp_generation) {
869 CDEBUG(D_HA, "resend cross eviction\n");
874 rc = mdc_finish_enqueue(exp, req, einfo, it, lockh, rc);
876 if (lustre_handle_is_used(lockh)) {
877 ldlm_lock_decref(lockh, einfo->ei_mode);
878 memset(lockh, 0, sizeof(*lockh));
880 ptlrpc_req_finished(req);
882 it->it_lock_handle = 0;
883 it->it_lock_mode = 0;
890 static int mdc_finish_intent_lock(struct obd_export *exp,
891 struct ptlrpc_request *request,
892 struct md_op_data *op_data,
893 struct lookup_intent *it,
894 struct lustre_handle *lockh)
896 struct lustre_handle old_lock;
897 struct mdt_body *mdt_body;
898 struct ldlm_lock *lock;
902 LASSERT(request != NULL);
903 LASSERT(request != LP_POISON);
904 LASSERT(request->rq_repmsg != LP_POISON);
906 if (it->it_op & IT_READDIR)
909 if (!it_disposition(it, DISP_IT_EXECD)) {
910 /* The server failed before it even started executing the
911 * intent, i.e. because it couldn't unpack the request. */
912 LASSERT(it->it_status != 0);
913 RETURN(it->it_status);
915 rc = it_open_error(DISP_IT_EXECD, it);
919 mdt_body = req_capsule_server_get(&request->rq_pill, &RMF_MDT_BODY);
920 LASSERT(mdt_body != NULL); /* mdc_enqueue checked */
922 rc = it_open_error(DISP_LOOKUP_EXECD, it);
926 /* keep requests around for the multiple phases of the call
927 * this shows the DISP_XX must guarantee we make it into the call
929 if (!it_disposition(it, DISP_ENQ_CREATE_REF) &&
930 it_disposition(it, DISP_OPEN_CREATE) &&
931 !it_open_error(DISP_OPEN_CREATE, it)) {
932 it_set_disposition(it, DISP_ENQ_CREATE_REF);
933 ptlrpc_request_addref(request); /* balanced in ll_create_node */
935 if (!it_disposition(it, DISP_ENQ_OPEN_REF) &&
936 it_disposition(it, DISP_OPEN_OPEN) &&
937 !it_open_error(DISP_OPEN_OPEN, it)) {
938 it_set_disposition(it, DISP_ENQ_OPEN_REF);
939 ptlrpc_request_addref(request); /* balanced in ll_file_open */
940 /* BUG 11546 - eviction in the middle of open rpc processing */
941 OBD_FAIL_TIMEOUT(OBD_FAIL_MDC_ENQUEUE_PAUSE, obd_timeout);
944 if (it->it_op & IT_CREAT) {
945 /* XXX this belongs in ll_create_it */
946 } else if (it->it_op == IT_OPEN) {
947 LASSERT(!it_disposition(it, DISP_OPEN_CREATE));
949 LASSERT(it->it_op & (IT_GETATTR | IT_LOOKUP | IT_LAYOUT));
952 /* If we already have a matching lock, then cancel the new
953 * one. We have to set the data here instead of in
954 * mdc_enqueue, because we need to use the child's inode as
955 * the l_ast_data to match, and that's not available until
956 * intent_finish has performed the iget().) */
957 lock = ldlm_handle2lock(lockh);
959 union ldlm_policy_data policy = lock->l_policy_data;
960 LDLM_DEBUG(lock, "matching against this");
962 LASSERTF(fid_res_name_eq(&mdt_body->mbo_fid1,
963 &lock->l_resource->lr_name),
964 "Lock res_id: "DLDLMRES", fid: "DFID"\n",
965 PLDLMRES(lock->l_resource), PFID(&mdt_body->mbo_fid1));
968 memcpy(&old_lock, lockh, sizeof(*lockh));
969 if (ldlm_lock_match(NULL, LDLM_FL_BLOCK_GRANTED, NULL,
970 LDLM_IBITS, &policy, LCK_NL, &old_lock, 0)) {
971 ldlm_lock_decref_and_cancel(lockh, it->it_lock_mode);
972 memcpy(lockh, &old_lock, sizeof(old_lock));
973 it->it_lock_handle = lockh->cookie;
977 CDEBUG(D_DENTRY,"D_IT dentry %.*s intent: %s status %d disp %x rc %d\n",
978 (int)op_data->op_namelen, op_data->op_name,
979 ldlm_it2str(it->it_op), it->it_status,
980 it->it_disposition, rc);
985 int mdc_revalidate_lock(struct obd_export *exp, struct lookup_intent *it,
986 struct lu_fid *fid, __u64 *bits)
988 /* We could just return 1 immediately, but since we should only
989 * be called in revalidate_it if we already have a lock, let's
991 struct ldlm_res_id res_id;
992 struct lustre_handle lockh;
993 union ldlm_policy_data policy;
997 if (it->it_lock_handle) {
998 lockh.cookie = it->it_lock_handle;
999 mode = ldlm_revalidate_lock_handle(&lockh, bits);
1001 fid_build_reg_res_name(fid, &res_id);
1002 switch (it->it_op) {
1004 /* File attributes are held under multiple bits:
1005 * nlink is under lookup lock, size and times are
1006 * under UPDATE lock and recently we've also got
1007 * a separate permissions lock for owner/group/acl that
1008 * were protected by lookup lock before.
1009 * Getattr must provide all of that information,
1010 * so we need to ensure we have all of those locks.
1011 * Unfortunately, if the bits are split across multiple
1012 * locks, there's no easy way to match all of them here,
1013 * so an extra RPC would be performed to fetch all
1014 * of those bits at once for now. */
1015 /* For new MDTs(> 2.4), UPDATE|PERM should be enough,
1016 * but for old MDTs (< 2.4), permission is covered
1017 * by LOOKUP lock, so it needs to match all bits here.*/
1018 policy.l_inodebits.bits = MDS_INODELOCK_UPDATE |
1019 MDS_INODELOCK_LOOKUP |
1023 policy.l_inodebits.bits = MDS_INODELOCK_UPDATE;
1026 policy.l_inodebits.bits = MDS_INODELOCK_LAYOUT;
1029 policy.l_inodebits.bits = MDS_INODELOCK_LOOKUP;
1033 mode = mdc_lock_match(exp, LDLM_FL_BLOCK_GRANTED, fid,
1034 LDLM_IBITS, &policy,
1035 LCK_CR | LCK_CW | LCK_PR | LCK_PW,
1040 it->it_lock_handle = lockh.cookie;
1041 it->it_lock_mode = mode;
1043 it->it_lock_handle = 0;
1044 it->it_lock_mode = 0;
1051 * This long block is all about fixing up the lock and request state
1052 * so that it is correct as of the moment _before_ the operation was
1053 * applied; that way, the VFS will think that everything is normal and
1054 * call Lustre's regular VFS methods.
1056 * If we're performing a creation, that means that unless the creation
1057 * failed with EEXIST, we should fake up a negative dentry.
1059 * For everything else, we want to lookup to succeed.
1061 * One additional note: if CREATE or OPEN succeeded, we add an extra
1062 * reference to the request because we need to keep it around until
1063 * ll_create/ll_open gets called.
1065 * The server will return to us, in it_disposition, an indication of
1066 * exactly what it_status refers to.
1068 * If DISP_OPEN_OPEN is set, then it_status refers to the open() call,
1069 * otherwise if DISP_OPEN_CREATE is set, then it status is the
1070 * creation failure mode. In either case, one of DISP_LOOKUP_NEG or
1071 * DISP_LOOKUP_POS will be set, indicating whether the child lookup
1074 * Else, if DISP_LOOKUP_EXECD then it_status is the rc of the
1077 int mdc_intent_lock(struct obd_export *exp, struct md_op_data *op_data,
1078 struct lookup_intent *it, struct ptlrpc_request **reqp,
1079 ldlm_blocking_callback cb_blocking, __u64 extra_lock_flags)
1081 struct ldlm_enqueue_info einfo = {
1082 .ei_type = LDLM_IBITS,
1083 .ei_mode = it_to_lock_mode(it),
1084 .ei_cb_bl = cb_blocking,
1085 .ei_cb_cp = ldlm_completion_ast,
1087 struct lustre_handle lockh;
1092 CDEBUG(D_DLMTRACE, "(name: %.*s,"DFID") in obj "DFID
1093 ", intent: %s flags %#"LPF64"o\n", (int)op_data->op_namelen,
1094 op_data->op_name, PFID(&op_data->op_fid2),
1095 PFID(&op_data->op_fid1), ldlm_it2str(it->it_op),
1099 if (fid_is_sane(&op_data->op_fid2) &&
1100 (it->it_op & (IT_LOOKUP | IT_GETATTR | IT_READDIR))) {
1101 /* We could just return 1 immediately, but since we should only
1102 * be called in revalidate_it if we already have a lock, let's
1104 it->it_lock_handle = 0;
1105 rc = mdc_revalidate_lock(exp, it, &op_data->op_fid2, NULL);
1106 /* Only return failure if it was not GETATTR by cfid
1107 (from inode_revalidate) */
1108 if (rc || op_data->op_namelen != 0)
1112 /* For case if upper layer did not alloc fid, do it now. */
1113 if (!fid_is_sane(&op_data->op_fid2) && it->it_op & IT_CREAT) {
1114 rc = mdc_fid_alloc(NULL, exp, &op_data->op_fid2, op_data);
1116 CERROR("Can't alloc new fid, rc %d\n", rc);
1121 rc = mdc_enqueue(exp, &einfo, NULL, it, op_data, &lockh,
1126 *reqp = it->it_data;
1127 rc = mdc_finish_intent_lock(exp, *reqp, op_data, it, &lockh);
1131 static int mdc_intent_getattr_async_interpret(const struct lu_env *env,
1132 struct ptlrpc_request *req,
1135 struct mdc_getattr_args *ga = args;
1136 struct obd_export *exp = ga->ga_exp;
1137 struct md_enqueue_info *minfo = ga->ga_minfo;
1138 struct ldlm_enqueue_info *einfo = &minfo->mi_einfo;
1139 struct lookup_intent *it;
1140 struct lustre_handle *lockh;
1141 struct obd_device *obddev;
1142 struct ldlm_reply *lockrep;
1143 __u64 flags = LDLM_FL_HAS_INTENT;
1147 lockh = &minfo->mi_lockh;
1149 obddev = class_exp2obd(exp);
1151 obd_put_request_slot(&obddev->u.cli);
1152 if (OBD_FAIL_CHECK(OBD_FAIL_MDC_GETATTR_ENQUEUE))
1155 rc = ldlm_cli_enqueue_fini(exp, req, einfo->ei_type, 1, einfo->ei_mode,
1156 &flags, NULL, 0, lockh, rc);
1158 CERROR("ldlm_cli_enqueue_fini: %d\n", rc);
1159 mdc_clear_replay_flag(req, rc);
1163 lockrep = req_capsule_server_get(&req->rq_pill, &RMF_DLM_REP);
1164 LASSERT(lockrep != NULL);
1166 lockrep->lock_policy_res2 =
1167 ptlrpc_status_ntoh(lockrep->lock_policy_res2);
1169 rc = mdc_finish_enqueue(exp, req, einfo, it, lockh, rc);
1173 rc = mdc_finish_intent_lock(exp, req, &minfo->mi_data, it, lockh);
1177 minfo->mi_cb(req, minfo, rc);
1181 int mdc_intent_getattr_async(struct obd_export *exp,
1182 struct md_enqueue_info *minfo)
1184 struct md_op_data *op_data = &minfo->mi_data;
1185 struct lookup_intent *it = &minfo->mi_it;
1186 struct ptlrpc_request *req;
1187 struct mdc_getattr_args *ga;
1188 struct obd_device *obddev = class_exp2obd(exp);
1189 struct ldlm_res_id res_id;
1190 union ldlm_policy_data policy = {
1191 .l_inodebits = { MDS_INODELOCK_LOOKUP |
1192 MDS_INODELOCK_UPDATE } };
1194 __u64 flags = LDLM_FL_HAS_INTENT;
1197 CDEBUG(D_DLMTRACE, "name: %.*s in inode "DFID", intent: %s flags %#"
1199 (int)op_data->op_namelen, op_data->op_name,
1200 PFID(&op_data->op_fid1), ldlm_it2str(it->it_op), it->it_flags);
1202 fid_build_reg_res_name(&op_data->op_fid1, &res_id);
1203 req = mdc_intent_getattr_pack(exp, it, op_data);
1205 RETURN(PTR_ERR(req));
1207 rc = obd_get_request_slot(&obddev->u.cli);
1209 ptlrpc_req_finished(req);
1213 rc = ldlm_cli_enqueue(exp, &req, &minfo->mi_einfo, &res_id, &policy,
1214 &flags, NULL, 0, LVB_T_NONE, &minfo->mi_lockh, 1);
1216 obd_put_request_slot(&obddev->u.cli);
1217 ptlrpc_req_finished(req);
1221 CLASSERT(sizeof(*ga) <= sizeof(req->rq_async_args));
1222 ga = ptlrpc_req_async_args(req);
1224 ga->ga_minfo = minfo;
1226 req->rq_interpret_reply = mdc_intent_getattr_async_interpret;
1227 ptlrpcd_add_req(req);