1 /* -*- mode: c; c-basic-offset: 8; indent-tabs-mode: nil; -*-
2 * vim:expandtab:shiftwidth=8:tabstop=8:
6 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
8 * This program is free software; you can redistribute it and/or modify
9 * it under the terms of the GNU General Public License version 2 only,
10 * as published by the Free Software Foundation.
12 * This program is distributed in the hope that it will be useful, but
13 * WITHOUT ANY WARRANTY; without even the implied warranty of
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
15 * General Public License version 2 for more details (a copy is included
16 * in the LICENSE file that accompanied this code).
18 * You should have received a copy of the GNU General Public License
19 * version 2 along with this program; If not, see
20 * http://www.sun.com/software/products/lustre/docs/GPLv2.pdf
22 * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara,
23 * CA 95054 USA or visit www.sun.com if you need additional information or
29 * Copyright (c) 2001, 2010, Oracle and/or its affiliates. All rights reserved.
30 * Use is subject to license terms.
32 * Copyright (c) 2011, 2012, Whamcloud, Inc.
35 * This file is part of Lustre, http://www.lustre.org/
36 * Lustre is a trademark of Sun Microsystems, Inc.
38 * lustre/ost/ost_handler.c
40 * Author: Peter J. Braam <braam@clusterfs.com>
41 * Author: Phil Schwan <phil@clusterfs.com>
45 # define EXPORT_SYMTAB
47 #define DEBUG_SUBSYSTEM S_OST
49 #include <linux/module.h>
50 #include <obd_cksum.h>
52 #include <lustre_net.h>
53 #include <lustre_dlm.h>
54 #include <lustre_export.h>
55 #include <lustre_debug.h>
56 #include <linux/init.h>
57 #include <lprocfs_status.h>
58 #include <libcfs/list.h>
59 #include <lustre_quota.h>
60 #include "ost_internal.h"
62 static int oss_num_threads;
63 CFS_MODULE_PARM(oss_num_threads, "i", int, 0444,
64 "number of OSS service threads to start");
66 static int ost_num_threads;
67 CFS_MODULE_PARM(ost_num_threads, "i", int, 0444,
68 "number of OST service threads to start (deprecated)");
70 static int oss_num_create_threads;
71 CFS_MODULE_PARM(oss_num_create_threads, "i", int, 0444,
72 "number of OSS create threads to start");
75 * Do not return server-side uid/gid to remote client
77 static void ost_drop_id(struct obd_export *exp, struct obdo *oa)
79 if (exp_connect_rmtclient(exp)) {
82 oa->o_valid &= ~(OBD_MD_FLUID | OBD_MD_FLGID);
87 * Validate oa from client.
88 * If the request comes from 2.0 clients, currently only RSVD seq and IDIF
90 * a. for single MDS seq = FID_SEQ_OST_MDT0,
91 * b. for CMD, seq = FID_SEQ_OST_MDT0, FID_SEQ_OST_MDT1 - FID_SEQ_OST_MAX
93 static int ost_validate_obdo(struct obd_export *exp, struct obdo *oa,
94 struct obd_ioobj *ioobj)
96 if (oa != NULL && !(oa->o_valid & OBD_MD_FLGROUP)) {
97 oa->o_seq = FID_SEQ_OST_MDT0;
99 ioobj->ioo_seq = FID_SEQ_OST_MDT0;
100 /* remove fid_seq_is_rsvd() after FID-on-OST allows SEQ > 9 */
101 } else if (oa == NULL || !(fid_seq_is_rsvd(oa->o_seq) ||
102 fid_seq_is_mdt0(oa->o_seq))) {
103 CERROR("%s: client %s sent invalid object "POSTID"\n",
104 exp->exp_obd->obd_name, obd_export_nid2str(exp),
105 oa ? oa->o_id : -1, oa ? oa->o_seq : -1);
108 obdo_from_ostid(oa, &oa->o_oi);
110 ioobj_from_obdo(ioobj, oa);
114 void oti_to_request(struct obd_trans_info *oti, struct ptlrpc_request *req)
116 struct oti_req_ack_lock *ack_lock;
122 if (req->rq_repmsg) {
123 __u64 versions[PTLRPC_NUM_VERSIONS] = { 0 };
124 lustre_msg_set_transno(req->rq_repmsg, oti->oti_transno);
125 versions[0] = oti->oti_pre_version;
126 lustre_msg_set_versions(req->rq_repmsg, versions);
128 req->rq_transno = oti->oti_transno;
130 /* XXX 4 == entries in oti_ack_locks??? */
131 for (ack_lock = oti->oti_ack_locks, i = 0; i < 4; i++, ack_lock++) {
134 /* XXX not even calling target_send_reply in some cases... */
135 ptlrpc_save_lock (req, &ack_lock->lock, ack_lock->mode, 0);
139 static int ost_destroy(struct obd_export *exp, struct ptlrpc_request *req,
140 struct obd_trans_info *oti)
142 struct ost_body *body, *repbody;
143 struct lustre_capa *capa = NULL;
147 /* Get the request body */
148 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
152 if (body->oa.o_id == 0)
155 rc = ost_validate_obdo(exp, &body->oa, NULL);
159 /* If there's a DLM request, cancel the locks mentioned in it*/
160 if (req_capsule_field_present(&req->rq_pill, &RMF_DLM_REQ, RCL_CLIENT)) {
161 struct ldlm_request *dlm;
163 dlm = req_capsule_client_get(&req->rq_pill, &RMF_DLM_REQ);
166 ldlm_request_cancel(req, dlm, 0);
169 /* If there's a capability, get it */
170 if (body->oa.o_valid & OBD_MD_FLOSSCAPA) {
171 capa = req_capsule_client_get(&req->rq_pill, &RMF_CAPA1);
173 CERROR("Missing capability for OST DESTROY");
178 /* Prepare the reply */
179 rc = req_capsule_server_pack(&req->rq_pill);
183 /* Get the log cancellation cookie */
184 if (body->oa.o_valid & OBD_MD_FLCOOKIE)
185 oti->oti_logcookies = &body->oa.o_lcookie;
187 /* Finish the reply */
188 repbody = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
189 memcpy(&repbody->oa, &body->oa, sizeof(body->oa));
191 /* Do the destroy and set the reply status accordingly */
192 req->rq_status = obd_destroy(exp, &repbody->oa, NULL, oti, NULL, capa);
197 * Helper function for getting server side [start, start+count] DLM lock
198 * if asked by client.
200 static int ost_lock_get(struct obd_export *exp, struct obdo *oa,
201 __u64 start, __u64 count, struct lustre_handle *lh,
204 struct ldlm_res_id res_id;
205 ldlm_policy_data_t policy;
206 __u64 end = start + count;
210 LASSERT(!lustre_handle_is_used(lh));
211 /* o_id and o_gr are used for localizing resource, if client miss to set
212 * them, do not trigger ASSERTION. */
213 if (unlikely((oa->o_valid & (OBD_MD_FLID | OBD_MD_FLGROUP)) !=
214 (OBD_MD_FLID | OBD_MD_FLGROUP)))
217 if (!(oa->o_valid & OBD_MD_FLFLAGS) ||
218 !(oa->o_flags & OBD_FL_SRVLOCK))
221 osc_build_res_name(oa->o_id, oa->o_seq, &res_id);
222 CDEBUG(D_INODE, "OST-side extent lock.\n");
224 policy.l_extent.start = start & CFS_PAGE_MASK;
226 /* If ->o_blocks is EOF it means "lock till the end of the
227 * file". Otherwise, it's size of a hole being punched (in bytes) */
228 if (count == OBD_OBJECT_EOF || end < start)
229 policy.l_extent.end = OBD_OBJECT_EOF;
231 policy.l_extent.end = end | ~CFS_PAGE_MASK;
233 RETURN(ldlm_cli_enqueue_local(exp->exp_obd->obd_namespace, &res_id,
234 LDLM_EXTENT, &policy, mode, &flags,
235 ldlm_blocking_ast, ldlm_completion_ast,
236 ldlm_glimpse_ast, NULL, 0, NULL, lh));
239 /* Helper function: release lock, if any. */
240 static void ost_lock_put(struct obd_export *exp,
241 struct lustre_handle *lh, int mode)
244 if (lustre_handle_is_used(lh))
245 ldlm_lock_decref(lh, mode);
249 static int ost_getattr(struct obd_export *exp, struct ptlrpc_request *req)
251 struct ost_body *body, *repbody;
252 struct obd_info *oinfo;
253 struct lustre_handle lh = { 0 };
254 struct lustre_capa *capa = NULL;
258 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
262 rc = ost_validate_obdo(exp, &body->oa, NULL);
266 if (body->oa.o_valid & OBD_MD_FLOSSCAPA) {
267 capa = req_capsule_client_get(&req->rq_pill, &RMF_CAPA1);
269 CERROR("Missing capability for OST GETATTR");
274 rc = req_capsule_server_pack(&req->rq_pill);
278 repbody = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
279 repbody->oa = body->oa;
281 rc = ost_lock_get(exp, &repbody->oa, 0, OBD_OBJECT_EOF, &lh, LCK_PR, 0);
285 OBD_ALLOC_PTR(oinfo);
287 GOTO(unlock, rc = -ENOMEM);
288 oinfo->oi_oa = &repbody->oa;
289 oinfo->oi_capa = capa;
291 req->rq_status = obd_getattr(exp, oinfo);
295 ost_drop_id(exp, &repbody->oa);
298 ost_lock_put(exp, &lh, LCK_PR);
302 static int ost_statfs(struct ptlrpc_request *req)
304 struct obd_statfs *osfs;
308 rc = req_capsule_server_pack(&req->rq_pill);
312 osfs = req_capsule_server_get(&req->rq_pill, &RMF_OBD_STATFS);
314 req->rq_status = obd_statfs(req->rq_export->exp_obd, osfs,
315 cfs_time_shift_64(-OBD_STATFS_CACHE_SECONDS),
317 if (req->rq_status != 0)
318 CERROR("ost: statfs failed: rc %d\n", req->rq_status);
323 static int ost_create(struct obd_export *exp, struct ptlrpc_request *req,
324 struct obd_trans_info *oti)
326 struct ost_body *body, *repbody;
330 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
334 rc = ost_validate_obdo(req->rq_export, &body->oa, NULL);
338 rc = req_capsule_server_pack(&req->rq_pill);
342 repbody = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
343 repbody->oa = body->oa;
344 oti->oti_logcookies = &body->oa.o_lcookie;
346 req->rq_status = obd_create(exp, &repbody->oa, NULL, oti);
347 //obd_log_cancel(conn, NULL, 1, oti->oti_logcookies, 0);
351 static int ost_punch(struct obd_export *exp, struct ptlrpc_request *req,
352 struct obd_trans_info *oti)
354 struct ost_body *body, *repbody;
356 struct lustre_handle lh = {0,};
359 /* check that we do support OBD_CONNECT_TRUNCLOCK. */
360 CLASSERT(OST_CONNECT_SUPPORTED & OBD_CONNECT_TRUNCLOCK);
362 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
366 rc = ost_validate_obdo(exp, &body->oa, NULL);
370 if ((body->oa.o_valid & (OBD_MD_FLSIZE | OBD_MD_FLBLOCKS)) !=
371 (OBD_MD_FLSIZE | OBD_MD_FLBLOCKS))
374 rc = req_capsule_server_pack(&req->rq_pill);
378 /* standard truncate optimization: if file body is completely
379 * destroyed, don't send data back to the server. */
380 if (body->oa.o_size == 0)
381 flags |= LDLM_AST_DISCARD_DATA;
383 repbody = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
384 repbody->oa = body->oa;
386 rc = ost_lock_get(exp, &repbody->oa, repbody->oa.o_size,
387 repbody->oa.o_blocks, &lh, LCK_PW, flags);
389 struct obd_info *oinfo;
390 struct lustre_capa *capa = NULL;
392 if (repbody->oa.o_valid & OBD_MD_FLFLAGS &&
393 repbody->oa.o_flags == OBD_FL_SRVLOCK)
395 * If OBD_FL_SRVLOCK is the only bit set in
396 * ->o_flags, clear OBD_MD_FLFLAGS to avoid falling
397 * through filter_setattr() to filter_iocontrol().
399 repbody->oa.o_valid &= ~OBD_MD_FLFLAGS;
401 if (repbody->oa.o_valid & OBD_MD_FLOSSCAPA) {
402 capa = req_capsule_client_get(&req->rq_pill,
405 CERROR("Missing capability for OST PUNCH");
406 GOTO(unlock, rc = -EFAULT);
410 OBD_ALLOC_PTR(oinfo);
412 GOTO(unlock, rc = -ENOMEM);
413 oinfo->oi_oa = &repbody->oa;
414 oinfo->oi_policy.l_extent.start = oinfo->oi_oa->o_size;
415 oinfo->oi_policy.l_extent.end = oinfo->oi_oa->o_blocks;
416 oinfo->oi_capa = capa;
417 oinfo->oi_flags = OBD_FL_PUNCH;
419 req->rq_status = obd_punch(exp, oinfo, oti, NULL);
422 ost_lock_put(exp, &lh, LCK_PW);
425 ost_drop_id(exp, &repbody->oa);
429 static int ost_sync(struct obd_export *exp, struct ptlrpc_request *req)
431 struct ost_body *body, *repbody;
432 struct obd_info *oinfo;
433 struct lustre_capa *capa = NULL;
437 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
441 rc = ost_validate_obdo(exp, &body->oa, NULL);
445 if (body->oa.o_valid & OBD_MD_FLOSSCAPA) {
446 capa = req_capsule_client_get(&req->rq_pill, &RMF_CAPA1);
448 CERROR("Missing capability for OST SYNC");
453 rc = req_capsule_server_pack(&req->rq_pill);
457 repbody = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
458 repbody->oa = body->oa;
460 OBD_ALLOC_PTR(oinfo);
464 oinfo->oi_oa = &repbody->oa;
465 oinfo->oi_capa = capa;
466 req->rq_status = obd_sync(exp, oinfo, repbody->oa.o_size,
467 repbody->oa.o_blocks, NULL);
470 ost_drop_id(exp, &repbody->oa);
474 static int ost_setattr(struct obd_export *exp, struct ptlrpc_request *req,
475 struct obd_trans_info *oti)
477 struct ost_body *body, *repbody;
478 struct obd_info *oinfo;
479 struct lustre_capa *capa = NULL;
483 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
487 rc = ost_validate_obdo(req->rq_export, &body->oa, NULL);
491 rc = req_capsule_server_pack(&req->rq_pill);
495 if (body->oa.o_valid & OBD_MD_FLOSSCAPA) {
496 capa = req_capsule_client_get(&req->rq_pill, &RMF_CAPA1);
498 CERROR("Missing capability for OST SETATTR");
503 repbody = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
504 repbody->oa = body->oa;
506 OBD_ALLOC_PTR(oinfo);
509 oinfo->oi_oa = &repbody->oa;
510 oinfo->oi_capa = capa;
512 req->rq_status = obd_setattr(exp, oinfo, oti);
516 ost_drop_id(exp, &repbody->oa);
520 static __u32 ost_checksum_bulk(struct ptlrpc_bulk_desc *desc, int opc,
521 cksum_type_t cksum_type)
526 cksum = init_checksum(cksum_type);
527 for (i = 0; i < desc->bd_iov_count; i++) {
528 struct page *page = desc->bd_iov[i].kiov_page;
529 int off = desc->bd_iov[i].kiov_offset & ~CFS_PAGE_MASK;
530 char *ptr = kmap(page) + off;
531 int len = desc->bd_iov[i].kiov_len;
533 /* corrupt the data before we compute the checksum, to
534 * simulate a client->OST data error */
535 if (i == 0 && opc == OST_WRITE &&
536 OBD_FAIL_CHECK(OBD_FAIL_OST_CHECKSUM_RECEIVE))
537 memcpy(ptr, "bad3", min(4, len));
538 cksum = compute_checksum(cksum, ptr, len, cksum_type);
539 /* corrupt the data after we compute the checksum, to
540 * simulate an OST->client data error */
541 if (i == 0 && opc == OST_READ &&
542 OBD_FAIL_CHECK(OBD_FAIL_OST_CHECKSUM_SEND)) {
543 memcpy(ptr, "bad4", min(4, len));
544 /* nobody should use corrupted page again */
545 ClearPageUptodate(page);
550 return fini_checksum(cksum, cksum_type);
553 static int ost_brw_lock_get(int mode, struct obd_export *exp,
554 struct obd_ioobj *obj, struct niobuf_remote *nb,
555 struct lustre_handle *lh)
558 int nrbufs = obj->ioo_bufcnt;
559 struct ldlm_res_id res_id;
560 ldlm_policy_data_t policy;
564 osc_build_res_name(obj->ioo_id, obj->ioo_seq, &res_id);
565 LASSERT(mode == LCK_PR || mode == LCK_PW);
566 LASSERT(!lustre_handle_is_used(lh));
568 if (nrbufs == 0 || !(nb[0].flags & OBD_BRW_SRVLOCK))
571 for (i = 1; i < nrbufs; i ++)
572 if ((nb[0].flags & OBD_BRW_SRVLOCK) !=
573 (nb[i].flags & OBD_BRW_SRVLOCK))
576 policy.l_extent.start = nb[0].offset & CFS_PAGE_MASK;
577 policy.l_extent.end = (nb[nrbufs - 1].offset +
578 nb[nrbufs - 1].len - 1) | ~CFS_PAGE_MASK;
580 RETURN(ldlm_cli_enqueue_local(exp->exp_obd->obd_namespace, &res_id,
581 LDLM_EXTENT, &policy, mode, &flags,
582 ldlm_blocking_ast, ldlm_completion_ast,
583 ldlm_glimpse_ast, NULL, 0, NULL, lh));
586 static void ost_brw_lock_put(int mode,
587 struct obd_ioobj *obj, struct niobuf_remote *niob,
588 struct lustre_handle *lh)
591 LASSERT(mode == LCK_PR || mode == LCK_PW);
592 LASSERT((obj->ioo_bufcnt > 0 && (niob[0].flags & OBD_BRW_SRVLOCK)) ==
593 lustre_handle_is_used(lh));
594 if (lustre_handle_is_used(lh))
595 ldlm_lock_decref(lh, mode);
599 /* Allocate thread local buffers if needed */
600 static struct ost_thread_local_cache *ost_tls_get(struct ptlrpc_request *r)
602 struct ost_thread_local_cache *tls =
603 (struct ost_thread_local_cache *)(r->rq_svc_thread->t_data);
605 /* In normal mode of operation an I/O request is serviced only
606 * by ll_ost_io threads each of them has own tls buffers allocated by
608 * During recovery, an I/O request may be queued until any of the ost
609 * service threads process it. Not necessary it should be one of
610 * ll_ost_io threads. In that case we dynamically allocating tls
611 * buffers for the request service time. */
612 if (unlikely(tls == NULL)) {
613 LASSERT(r->rq_export->exp_in_recovery);
617 r->rq_svc_thread->t_data = tls;
623 /* Free thread local buffers if they were allocated only for servicing
624 * this one request */
625 static void ost_tls_put(struct ptlrpc_request *r)
627 struct ost_thread_local_cache *tls =
628 (struct ost_thread_local_cache *)(r->rq_svc_thread->t_data);
630 if (unlikely(tls->temporary)) {
632 r->rq_svc_thread->t_data = NULL;
636 static int ost_brw_read(struct ptlrpc_request *req, struct obd_trans_info *oti)
638 struct ptlrpc_bulk_desc *desc = NULL;
639 struct obd_export *exp = req->rq_export;
640 struct niobuf_remote *remote_nb;
641 struct niobuf_local *local_nb;
642 struct obd_ioobj *ioo;
643 struct ost_body *body, *repbody;
644 struct lustre_capa *capa = NULL;
645 struct l_wait_info lwi;
646 struct lustre_handle lockh = { 0 };
647 int niocount, npages, nob = 0, rc, i;
649 struct ost_thread_local_cache *tls;
652 req->rq_bulk_read = 1;
654 if (OBD_FAIL_CHECK(OBD_FAIL_OST_BRW_READ_BULK))
655 GOTO(out, rc = -EIO);
657 OBD_FAIL_TIMEOUT(OBD_FAIL_OST_BRW_PAUSE_BULK, (obd_timeout + 1) / 4);
659 /* Check if there is eviction in progress, and if so, wait for it to
661 if (unlikely(cfs_atomic_read(&exp->exp_obd->obd_evict_inprogress))) {
662 lwi = LWI_INTR(NULL, NULL); // We do not care how long it takes
663 rc = l_wait_event(exp->exp_obd->obd_evict_inprogress_waitq,
664 !cfs_atomic_read(&exp->exp_obd->obd_evict_inprogress),
668 GOTO(out, rc = -ENOTCONN);
670 /* ost_body, ioobj & noibuf_remote are verified and swabbed in
671 * ost_rw_hpreq_check(). */
672 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
674 GOTO(out, rc = -EFAULT);
677 * A req_capsule_X_get_array(pill, field, ptr_to_element_count) function
678 * would be useful here and wherever we get &RMF_OBD_IOOBJ and
679 * &RMF_NIOBUF_REMOTE.
681 ioo = req_capsule_client_get(&req->rq_pill, &RMF_OBD_IOOBJ);
683 GOTO(out, rc = -EFAULT);
685 rc = ost_validate_obdo(exp, &body->oa, ioo);
689 niocount = ioo->ioo_bufcnt;
690 remote_nb = req_capsule_client_get(&req->rq_pill, &RMF_NIOBUF_REMOTE);
691 if (remote_nb == NULL)
692 GOTO(out, rc = -EFAULT);
694 if (body->oa.o_valid & OBD_MD_FLOSSCAPA) {
695 capa = req_capsule_client_get(&req->rq_pill, &RMF_CAPA1);
697 CERROR("Missing capability for OST BRW READ");
698 GOTO(out, rc = -EFAULT);
702 rc = req_capsule_server_pack(&req->rq_pill);
706 tls = ost_tls_get(req);
708 GOTO(out_bulk, rc = -ENOMEM);
709 local_nb = tls->local;
711 rc = ost_brw_lock_get(LCK_PR, exp, ioo, remote_nb, &lockh);
716 * If getting the lock took more time than
717 * client was willing to wait, drop it. b=11330
719 if (cfs_time_current_sec() > req->rq_deadline ||
720 OBD_FAIL_CHECK(OBD_FAIL_OST_DROP_REQ)) {
722 CERROR("Dropping timed-out read from %s because locking"
723 "object "LPX64" took %ld seconds (limit was %ld).\n",
724 libcfs_id2str(req->rq_peer), ioo->ioo_id,
725 cfs_time_current_sec() - req->rq_arrival_time.tv_sec,
726 req->rq_deadline - req->rq_arrival_time.tv_sec);
727 GOTO(out_lock, rc = -ETIMEDOUT);
730 repbody = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
731 memcpy(&repbody->oa, &body->oa, sizeof(repbody->oa));
733 npages = OST_THREAD_POOL_SIZE;
734 rc = obd_preprw(OBD_BRW_READ, exp, &repbody->oa, 1, ioo,
735 remote_nb, &npages, local_nb, oti, capa);
739 desc = ptlrpc_prep_bulk_exp(req, npages,
740 BULK_PUT_SOURCE, OST_BULK_PORTAL);
742 GOTO(out_commitrw, rc = -ENOMEM);
745 for (i = 0; i < npages; i++) {
746 int page_rc = local_nb[i].rc;
748 if (page_rc < 0) { /* error */
754 if (page_rc != 0) { /* some data! */
755 LASSERT (local_nb[i].page != NULL);
756 ptlrpc_prep_bulk_page(desc, local_nb[i].page,
757 local_nb[i].offset & ~CFS_PAGE_MASK,
761 if (page_rc != local_nb[i].len) { /* short read */
762 /* All subsequent pages should be 0 */
764 LASSERT(local_nb[i].rc == 0);
769 if (body->oa.o_valid & OBD_MD_FLCKSUM) {
770 cksum_type_t cksum_type =
771 cksum_type_unpack(repbody->oa.o_valid & OBD_MD_FLFLAGS ?
772 repbody->oa.o_flags : 0);
773 repbody->oa.o_flags = cksum_type_pack(cksum_type);
774 repbody->oa.o_valid = OBD_MD_FLCKSUM | OBD_MD_FLFLAGS;
775 repbody->oa.o_cksum = ost_checksum_bulk(desc, OST_READ,cksum_type);
776 CDEBUG(D_PAGE, "checksum at read origin: %x\n",
777 repbody->oa.o_cksum);
779 repbody->oa.o_valid = 0;
781 /* We're finishing using body->oa as an input variable */
783 /* Check if client was evicted while we were doing i/o before touching
786 if (likely(!CFS_FAIL_PRECHECK(OBD_FAIL_PTLRPC_CLIENT_BULK_CB2)))
787 rc = target_bulk_io(exp, desc, &lwi);
792 /* Must commit after prep above in all cases */
793 rc = obd_commitrw(OBD_BRW_READ, exp, &repbody->oa, 1, ioo,
794 remote_nb, npages, local_nb, oti, rc);
797 ost_drop_id(exp, &repbody->oa);
800 ost_brw_lock_put(LCK_PR, ioo, remote_nb, &lockh);
804 if (desc && !CFS_FAIL_PRECHECK(OBD_FAIL_PTLRPC_CLIENT_BULK_CB2))
805 ptlrpc_free_bulk(desc);
809 req->rq_status = nob;
810 ptlrpc_lprocfs_brw(req, nob);
811 target_committed_to_req(req);
813 } else if (!no_reply) {
814 /* Only reply if there was no comms problem with bulk */
815 target_committed_to_req(req);
819 /* reply out callback would free */
820 ptlrpc_req_drop_rs(req);
821 CWARN("%s: ignoring bulk IO comm error with %s@%s id %s - "
822 "client will retry\n",
823 exp->exp_obd->obd_name,
824 exp->exp_client_uuid.uuid,
825 exp->exp_connection->c_remote_uuid.uuid,
826 libcfs_id2str(req->rq_peer));
828 /* send a bulk after reply to simulate a network delay or reordering
830 if (unlikely(CFS_FAIL_PRECHECK(OBD_FAIL_PTLRPC_CLIENT_BULK_CB2))) {
832 struct l_wait_info lwi1;
834 CDEBUG(D_INFO, "reorder BULK\n");
835 cfs_waitq_init(&waitq);
837 lwi1 = LWI_TIMEOUT_INTR(cfs_time_seconds(3), NULL, NULL, NULL);
838 l_wait_event(waitq, 0, &lwi1);
839 rc = target_bulk_io(exp, desc, &lwi);
840 ptlrpc_free_bulk(desc);
846 static int ost_brw_write(struct ptlrpc_request *req, struct obd_trans_info *oti)
848 struct ptlrpc_bulk_desc *desc = NULL;
849 struct obd_export *exp = req->rq_export;
850 struct niobuf_remote *remote_nb;
851 struct niobuf_local *local_nb;
852 struct obd_ioobj *ioo;
853 struct ost_body *body, *repbody;
854 struct l_wait_info lwi;
855 struct lustre_handle lockh = {0};
856 struct lustre_capa *capa = NULL;
858 int objcount, niocount, npages;
860 obd_count client_cksum = 0, server_cksum = 0;
861 cksum_type_t cksum_type = OBD_CKSUM_CRC32;
862 int no_reply = 0, mmap = 0;
863 __u32 o_uid = 0, o_gid = 0;
864 struct ost_thread_local_cache *tls;
867 req->rq_bulk_write = 1;
869 if (OBD_FAIL_CHECK(OBD_FAIL_OST_BRW_WRITE_BULK))
870 GOTO(out, rc = -EIO);
871 if (OBD_FAIL_CHECK(OBD_FAIL_OST_BRW_WRITE_BULK2))
872 GOTO(out, rc = -EFAULT);
874 /* pause before transaction has been started */
875 OBD_FAIL_TIMEOUT(OBD_FAIL_OST_BRW_PAUSE_BULK, (obd_timeout + 1) / 4);
877 /* ost_body, ioobj & noibuf_remote are verified and swabbed in
878 * ost_rw_hpreq_check(). */
879 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
881 GOTO(out, rc = -EFAULT);
883 objcount = req_capsule_get_size(&req->rq_pill, &RMF_OBD_IOOBJ,
884 RCL_CLIENT) / sizeof(*ioo);
885 ioo = req_capsule_client_get(&req->rq_pill, &RMF_OBD_IOOBJ);
887 GOTO(out, rc = -EFAULT);
889 rc = ost_validate_obdo(exp, &body->oa, ioo);
893 for (niocount = i = 0; i < objcount; i++)
894 niocount += ioo[i].ioo_bufcnt;
897 * It'd be nice to have a capsule function to indicate how many elements
898 * there were in a buffer for an RMF that's declared to be an array.
899 * It's easy enough to compute the number of elements here though.
901 remote_nb = req_capsule_client_get(&req->rq_pill, &RMF_NIOBUF_REMOTE);
902 if (remote_nb == NULL || niocount != (req_capsule_get_size(&req->rq_pill,
903 &RMF_NIOBUF_REMOTE, RCL_CLIENT) / sizeof(*remote_nb)))
904 GOTO(out, rc = -EFAULT);
906 if ((remote_nb[0].flags & OBD_BRW_MEMALLOC) &&
907 (exp->exp_connection->c_peer.nid == exp->exp_connection->c_self))
908 cfs_memory_pressure_set();
910 if (body->oa.o_valid & OBD_MD_FLOSSCAPA) {
911 capa = req_capsule_client_get(&req->rq_pill, &RMF_CAPA1);
913 CERROR("Missing capability for OST BRW WRITE");
914 GOTO(out, rc = -EFAULT);
918 req_capsule_set_size(&req->rq_pill, &RMF_RCS, RCL_SERVER,
919 niocount * sizeof(*rcs));
920 rc = req_capsule_server_pack(&req->rq_pill);
923 CFS_FAIL_TIMEOUT(OBD_FAIL_OST_BRW_PAUSE_PACK, cfs_fail_val);
924 rcs = req_capsule_server_get(&req->rq_pill, &RMF_RCS);
926 tls = ost_tls_get(req);
928 GOTO(out_bulk, rc = -ENOMEM);
929 local_nb = tls->local;
931 rc = ost_brw_lock_get(LCK_PW, exp, ioo, remote_nb, &lockh);
936 * If getting the lock took more time than
937 * client was willing to wait, drop it. b=11330
939 if (cfs_time_current_sec() > req->rq_deadline ||
940 OBD_FAIL_CHECK(OBD_FAIL_OST_DROP_REQ)) {
942 CERROR("Dropping timed-out write from %s because locking "
943 "object "LPX64" took %ld seconds (limit was %ld).\n",
944 libcfs_id2str(req->rq_peer), ioo->ioo_id,
945 cfs_time_current_sec() - req->rq_arrival_time.tv_sec,
946 req->rq_deadline - req->rq_arrival_time.tv_sec);
947 GOTO(out_lock, rc = -ETIMEDOUT);
950 /* obd_preprw clobbers oa->valid, so save what we need */
951 if (body->oa.o_valid & OBD_MD_FLCKSUM) {
952 client_cksum = body->oa.o_cksum;
953 if (body->oa.o_valid & OBD_MD_FLFLAGS)
954 cksum_type = cksum_type_unpack(body->oa.o_flags);
956 if (body->oa.o_valid & OBD_MD_FLFLAGS && body->oa.o_flags & OBD_FL_MMAP)
959 /* Because we already sync grant info with client when reconnect,
960 * grant info will be cleared for resent req, then fed_grant and
961 * total_grant will not be modified in following preprw_write */
962 if (lustre_msg_get_flags(req->rq_reqmsg) & (MSG_RESENT | MSG_REPLAY)) {
963 DEBUG_REQ(D_CACHE, req, "clear resent/replay req grant info");
964 body->oa.o_valid &= ~OBD_MD_FLGRANT;
967 if (exp_connect_rmtclient(exp)) {
968 o_uid = body->oa.o_uid;
969 o_gid = body->oa.o_gid;
972 repbody = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
973 memcpy(&repbody->oa, &body->oa, sizeof(repbody->oa));
975 npages = OST_THREAD_POOL_SIZE;
976 rc = obd_preprw(OBD_BRW_WRITE, exp, &repbody->oa, objcount,
977 ioo, remote_nb, &npages, local_nb, oti, capa);
981 desc = ptlrpc_prep_bulk_exp(req, npages,
982 BULK_GET_SINK, OST_BULK_PORTAL);
984 GOTO(skip_transfer, rc = -ENOMEM);
986 /* NB Having prepped, we must commit... */
988 for (i = 0; i < npages; i++)
989 ptlrpc_prep_bulk_page(desc, local_nb[i].page,
990 local_nb[i].offset & ~CFS_PAGE_MASK,
993 rc = sptlrpc_svc_prep_bulk(req, desc);
997 rc = target_bulk_io(exp, desc, &lwi);
1001 if (client_cksum != 0 && rc == 0) {
1002 static int cksum_counter;
1003 repbody->oa.o_valid |= OBD_MD_FLCKSUM | OBD_MD_FLFLAGS;
1004 repbody->oa.o_flags &= ~OBD_FL_CKSUM_ALL;
1005 repbody->oa.o_flags |= cksum_type_pack(cksum_type);
1006 server_cksum = ost_checksum_bulk(desc, OST_WRITE, cksum_type);
1007 repbody->oa.o_cksum = server_cksum;
1009 if (unlikely(client_cksum != server_cksum)) {
1010 CDEBUG_LIMIT(mmap ? D_INFO : D_ERROR,
1011 "client csum %x, server csum %x\n",
1012 client_cksum, server_cksum);
1014 } else if ((cksum_counter & (-cksum_counter)) == cksum_counter){
1015 CDEBUG(D_INFO, "Checksum %u from %s OK: %x\n",
1016 cksum_counter, libcfs_id2str(req->rq_peer),
1021 /* Must commit after prep above in all cases */
1022 rc = obd_commitrw(OBD_BRW_WRITE, exp, &repbody->oa, objcount, ioo,
1023 remote_nb, npages, local_nb, oti, rc);
1024 if (rc == -ENOTCONN)
1025 /* quota acquire process has been given up because
1026 * either the client has been evicted or the client
1027 * has timed out the request already */
1030 if (exp_connect_rmtclient(exp)) {
1031 repbody->oa.o_uid = o_uid;
1032 repbody->oa.o_gid = o_gid;
1036 * Disable sending mtime back to the client. If the client locked the
1037 * whole object, then it has already updated the mtime on its side,
1038 * otherwise it will have to glimpse anyway (see bug 21489, comment 32)
1040 repbody->oa.o_valid &= ~(OBD_MD_FLMTIME | OBD_MD_FLATIME);
1042 if (unlikely(client_cksum != server_cksum && rc == 0 && !mmap)) {
1043 int new_cksum = ost_checksum_bulk(desc, OST_WRITE, cksum_type);
1048 if (new_cksum == server_cksum)
1049 msg = "changed in transit before arrival at OST";
1050 else if (new_cksum == client_cksum)
1051 msg = "initial checksum before message complete";
1053 msg = "changed in transit AND after initial checksum";
1055 if (req->rq_peer.nid == desc->bd_sender) {
1059 router = libcfs_nid2str(desc->bd_sender);
1062 LCONSOLE_ERROR_MSG(0x168, "%s: BAD WRITE CHECKSUM: %s from "
1063 "%s%s%s inode "DFID" object "
1064 LPU64"/"LPU64" extent ["LPU64"-"LPU64"]\n",
1065 exp->exp_obd->obd_name, msg,
1066 libcfs_id2str(req->rq_peer),
1068 body->oa.o_valid & OBD_MD_FLFID ?
1069 body->oa.o_parent_seq : (__u64)0,
1070 body->oa.o_valid & OBD_MD_FLFID ?
1071 body->oa.o_parent_oid : 0,
1072 body->oa.o_valid & OBD_MD_FLFID ?
1073 body->oa.o_parent_ver : 0,
1075 body->oa.o_valid & OBD_MD_FLGROUP ?
1076 body->oa.o_seq : (__u64)0,
1078 local_nb[npages-1].offset +
1079 local_nb[npages-1].len - 1 );
1080 CERROR("client csum %x, original server csum %x, "
1081 "server csum now %x\n",
1082 client_cksum, server_cksum, new_cksum);
1088 /* set per-requested niobuf return codes */
1089 for (i = j = 0; i < niocount; i++) {
1090 int len = remote_nb[i].len;
1095 LASSERT(j < npages);
1096 if (local_nb[j].rc < 0)
1097 rcs[i] = local_nb[j].rc;
1098 len -= local_nb[j].len;
1103 LASSERT(j == npages);
1104 ptlrpc_lprocfs_brw(req, nob);
1108 ost_brw_lock_put(LCK_PW, ioo, remote_nb, &lockh);
1113 ptlrpc_free_bulk(desc);
1116 oti_to_request(oti, req);
1117 target_committed_to_req(req);
1118 rc = ptlrpc_reply(req);
1119 } else if (!no_reply) {
1120 /* Only reply if there was no comms problem with bulk */
1121 target_committed_to_req(req);
1122 req->rq_status = rc;
1125 /* reply out callback would free */
1126 ptlrpc_req_drop_rs(req);
1127 CWARN("%s: ignoring bulk IO comm error with %s@%s id %s - "
1128 "client will retry\n",
1129 exp->exp_obd->obd_name,
1130 exp->exp_client_uuid.uuid,
1131 exp->exp_connection->c_remote_uuid.uuid,
1132 libcfs_id2str(req->rq_peer));
1134 cfs_memory_pressure_clr();
1139 * Implementation of OST_SET_INFO.
1141 * OST_SET_INFO is like ioctl(): heavily overloaded. Specifically, it takes a
1142 * "key" and a value RPC buffers as arguments, with the value's contents
1143 * interpreted according to the key.
1145 * Value types that need swabbing have swabbing done explicitly, either here or
1146 * in functions called from here. This should be corrected: all swabbing should
1147 * be done in the capsule abstraction, as that will then allow us to move
1148 * swabbing exclusively to the client without having to modify server code
1149 * outside the capsule abstraction's implementation itself. To correct this
1150 * will require minor changes to the capsule abstraction; see the comments for
1151 * req_capsule_extend() in layout.c.
1153 static int ost_set_info(struct obd_export *exp, struct ptlrpc_request *req)
1155 struct ost_body *body = NULL, *repbody;
1156 char *key, *val = NULL;
1157 int keylen, vallen, rc = 0;
1158 int is_grant_shrink = 0;
1161 key = req_capsule_client_get(&req->rq_pill, &RMF_SETINFO_KEY);
1163 DEBUG_REQ(D_HA, req, "no set_info key");
1166 keylen = req_capsule_get_size(&req->rq_pill, &RMF_SETINFO_KEY,
1169 vallen = req_capsule_get_size(&req->rq_pill, &RMF_SETINFO_VAL,
1172 if ((is_grant_shrink = KEY_IS(KEY_GRANT_SHRINK)))
1173 /* In this case the value is actually an RMF_OST_BODY, so we
1174 * transmutate the type of this PTLRPC */
1175 req_capsule_extend(&req->rq_pill, &RQF_OST_SET_GRANT_INFO);
1177 rc = req_capsule_server_pack(&req->rq_pill);
1182 if (is_grant_shrink) {
1183 body = req_capsule_client_get(&req->rq_pill,
1188 repbody = req_capsule_server_get(&req->rq_pill,
1190 memcpy(repbody, body, sizeof(*body));
1191 val = (char*)repbody;
1193 val = req_capsule_client_get(&req->rq_pill,
1198 if (KEY_IS(KEY_EVICT_BY_NID)) {
1200 obd_export_evict_by_nid(exp->exp_obd, val);
1202 } else if (KEY_IS(KEY_MDS_CONN) && ptlrpc_req_need_swab(req)) {
1203 if (vallen < sizeof(__u32))
1205 __swab32s((__u32 *)val);
1208 /* OBD will also check if KEY_IS(KEY_GRANT_SHRINK), and will cast val to
1209 * a struct ost_body * value */
1210 rc = obd_set_info_async(exp, keylen, key, vallen, val, NULL);
1212 lustre_msg_set_status(req->rq_repmsg, 0);
1216 static int ost_get_info(struct obd_export *exp, struct ptlrpc_request *req)
1219 int keylen, replylen, rc = 0;
1220 struct req_capsule *pill = &req->rq_pill;
1223 /* this common part for get_info rpc */
1224 key = req_capsule_client_get(pill, &RMF_SETINFO_KEY);
1226 DEBUG_REQ(D_HA, req, "no get_info key");
1229 keylen = req_capsule_get_size(pill, &RMF_SETINFO_KEY, RCL_CLIENT);
1231 if (KEY_IS(KEY_FIEMAP)) {
1232 struct ll_fiemap_info_key *fm_key = key;
1235 rc = ost_validate_obdo(exp, &fm_key->oa, NULL);
1240 rc = obd_get_info(exp, keylen, key, &replylen, NULL, NULL);
1244 req_capsule_set_size(pill, &RMF_GENERIC_DATA,
1245 RCL_SERVER, replylen);
1247 rc = req_capsule_server_pack(pill);
1251 reply = req_capsule_server_get(pill, &RMF_GENERIC_DATA);
1255 /* call again to fill in the reply buffer */
1256 rc = obd_get_info(exp, keylen, key, &replylen, reply, NULL);
1258 lustre_msg_set_status(req->rq_repmsg, 0);
1262 #ifdef HAVE_QUOTA_SUPPORT
1263 static int ost_handle_quotactl(struct ptlrpc_request *req)
1265 struct obd_quotactl *oqctl, *repoqc;
1269 oqctl = req_capsule_client_get(&req->rq_pill, &RMF_OBD_QUOTACTL);
1271 GOTO(out, rc = -EPROTO);
1273 rc = req_capsule_server_pack(&req->rq_pill);
1277 repoqc = req_capsule_server_get(&req->rq_pill, &RMF_OBD_QUOTACTL);
1278 req->rq_status = obd_quotactl(req->rq_export, oqctl);
1285 static int ost_handle_quotacheck(struct ptlrpc_request *req)
1287 struct obd_quotactl *oqctl;
1291 oqctl = req_capsule_client_get(&req->rq_pill, &RMF_OBD_QUOTACTL);
1295 rc = req_capsule_server_pack(&req->rq_pill);
1299 req->rq_status = obd_quotacheck(req->rq_export, oqctl);
1303 static int ost_handle_quota_adjust_qunit(struct ptlrpc_request *req)
1305 struct quota_adjust_qunit *oqaq, *repoqa;
1306 struct lustre_quota_ctxt *qctxt;
1310 qctxt = &req->rq_export->exp_obd->u.obt.obt_qctxt;
1311 oqaq = req_capsule_client_get(&req->rq_pill, &RMF_QUOTA_ADJUST_QUNIT);
1313 GOTO(out, rc = -EPROTO);
1315 rc = req_capsule_server_pack(&req->rq_pill);
1319 repoqa = req_capsule_server_get(&req->rq_pill, &RMF_QUOTA_ADJUST_QUNIT);
1320 req->rq_status = obd_quota_adjust_qunit(req->rq_export, oqaq, qctxt, NULL);
1328 static int ost_llog_handle_connect(struct obd_export *exp,
1329 struct ptlrpc_request *req)
1331 struct llogd_conn_body *body;
1335 body = req_capsule_client_get(&req->rq_pill, &RMF_LLOGD_CONN_BODY);
1336 rc = obd_llog_connect(exp, body);
1340 #define ost_init_sec_none(reply, exp) \
1342 reply->ocd_connect_flags &= ~(OBD_CONNECT_RMT_CLIENT | \
1343 OBD_CONNECT_RMT_CLIENT_FORCE | \
1344 OBD_CONNECT_OSS_CAPA); \
1345 cfs_spin_lock(&exp->exp_lock); \
1346 exp->exp_connect_flags = reply->ocd_connect_flags; \
1347 cfs_spin_unlock(&exp->exp_lock); \
1350 static int ost_init_sec_level(struct ptlrpc_request *req)
1352 struct obd_export *exp = req->rq_export;
1353 struct req_capsule *pill = &req->rq_pill;
1354 struct obd_device *obd = exp->exp_obd;
1355 struct filter_obd *filter = &obd->u.filter;
1356 char *client = libcfs_nid2str(req->rq_peer.nid);
1357 struct obd_connect_data *data, *reply;
1361 data = req_capsule_client_get(pill, &RMF_CONNECT_DATA);
1362 reply = req_capsule_server_get(pill, &RMF_CONNECT_DATA);
1363 if (data == NULL || reply == NULL)
1366 /* connection from MDT is always trusted */
1367 if (req->rq_auth_usr_mdt) {
1368 ost_init_sec_none(reply, exp);
1372 /* no GSS support case */
1373 if (!req->rq_auth_gss) {
1374 if (filter->fo_sec_level > LUSTRE_SEC_NONE) {
1375 CWARN("client %s -> target %s does not user GSS, "
1376 "can not run under security level %d.\n",
1377 client, obd->obd_name, filter->fo_sec_level);
1380 ost_init_sec_none(reply, exp);
1385 /* old version case */
1386 if (unlikely(!(data->ocd_connect_flags & OBD_CONNECT_RMT_CLIENT) ||
1387 !(data->ocd_connect_flags & OBD_CONNECT_OSS_CAPA))) {
1388 if (filter->fo_sec_level > LUSTRE_SEC_NONE) {
1389 CWARN("client %s -> target %s uses old version, "
1390 "can not run under security level %d.\n",
1391 client, obd->obd_name, filter->fo_sec_level);
1394 CWARN("client %s -> target %s uses old version, "
1395 "run under security level %d.\n",
1396 client, obd->obd_name, filter->fo_sec_level);
1397 ost_init_sec_none(reply, exp);
1402 remote = data->ocd_connect_flags & OBD_CONNECT_RMT_CLIENT_FORCE;
1404 if (!req->rq_auth_remote)
1405 CDEBUG(D_SEC, "client (local realm) %s -> target %s "
1406 "asked to be remote.\n", client, obd->obd_name);
1407 } else if (req->rq_auth_remote) {
1409 CDEBUG(D_SEC, "client (remote realm) %s -> target %s is set "
1410 "as remote by default.\n", client, obd->obd_name);
1414 if (!filter->fo_fl_oss_capa) {
1415 CDEBUG(D_SEC, "client %s -> target %s is set as remote,"
1416 " but OSS capabilities are not enabled: %d.\n",
1417 client, obd->obd_name, filter->fo_fl_oss_capa);
1422 switch (filter->fo_sec_level) {
1423 case LUSTRE_SEC_NONE:
1425 ost_init_sec_none(reply, exp);
1428 CDEBUG(D_SEC, "client %s -> target %s is set as remote, "
1429 "can not run under security level %d.\n",
1430 client, obd->obd_name, filter->fo_sec_level);
1433 case LUSTRE_SEC_REMOTE:
1435 ost_init_sec_none(reply, exp);
1437 case LUSTRE_SEC_ALL:
1439 reply->ocd_connect_flags &= ~(OBD_CONNECT_RMT_CLIENT |
1440 OBD_CONNECT_RMT_CLIENT_FORCE);
1441 if (!filter->fo_fl_oss_capa)
1442 reply->ocd_connect_flags &= ~OBD_CONNECT_OSS_CAPA;
1444 cfs_spin_lock(&exp->exp_lock);
1445 exp->exp_connect_flags = reply->ocd_connect_flags;
1446 cfs_spin_unlock(&exp->exp_lock);
1458 * this should be done in filter_connect()/filter_reconnect(), but
1459 * we can't obtain information like NID, which stored in incoming
1460 * request, thus can't decide what flavor to use. so we do it here.
1462 * This hack should be removed after the OST stack be rewritten, just
1463 * like what we are doing in mdt_obd_connect()/mdt_obd_reconnect().
1465 static int ost_connect_check_sptlrpc(struct ptlrpc_request *req)
1467 struct obd_export *exp = req->rq_export;
1468 struct filter_obd *filter = &exp->exp_obd->u.filter;
1469 struct sptlrpc_flavor flvr;
1472 if (unlikely(strcmp(exp->exp_obd->obd_type->typ_name,
1473 LUSTRE_ECHO_NAME) == 0)) {
1474 exp->exp_flvr.sf_rpc = SPTLRPC_FLVR_ANY;
1478 if (exp->exp_flvr.sf_rpc == SPTLRPC_FLVR_INVALID) {
1479 cfs_read_lock(&filter->fo_sptlrpc_lock);
1480 sptlrpc_target_choose_flavor(&filter->fo_sptlrpc_rset,
1484 cfs_read_unlock(&filter->fo_sptlrpc_lock);
1486 cfs_spin_lock(&exp->exp_lock);
1488 exp->exp_sp_peer = req->rq_sp_from;
1489 exp->exp_flvr = flvr;
1491 if (exp->exp_flvr.sf_rpc != SPTLRPC_FLVR_ANY &&
1492 exp->exp_flvr.sf_rpc != req->rq_flvr.sf_rpc) {
1493 CERROR("unauthorized rpc flavor %x from %s, "
1494 "expect %x\n", req->rq_flvr.sf_rpc,
1495 libcfs_nid2str(req->rq_peer.nid),
1496 exp->exp_flvr.sf_rpc);
1500 cfs_spin_unlock(&exp->exp_lock);
1502 if (exp->exp_sp_peer != req->rq_sp_from) {
1503 CERROR("RPC source %s doesn't match %s\n",
1504 sptlrpc_part2name(req->rq_sp_from),
1505 sptlrpc_part2name(exp->exp_sp_peer));
1508 rc = sptlrpc_target_export_check(exp, req);
1515 /* Ensure that data and metadata are synced to the disk when lock is cancelled
1517 int ost_blocking_ast(struct ldlm_lock *lock,
1518 struct ldlm_lock_desc *desc,
1519 void *data, int flag)
1521 __u32 sync_lock_cancel = 0;
1522 __u32 len = sizeof(sync_lock_cancel);
1526 rc = obd_get_info(lock->l_export, sizeof(KEY_SYNC_LOCK_CANCEL),
1527 KEY_SYNC_LOCK_CANCEL, &len, &sync_lock_cancel, NULL);
1529 if (!rc && flag == LDLM_CB_CANCELING &&
1530 (lock->l_granted_mode & (LCK_PW|LCK_GROUP)) &&
1531 (sync_lock_cancel == ALWAYS_SYNC_ON_CANCEL ||
1532 (sync_lock_cancel == BLOCKING_SYNC_ON_CANCEL &&
1533 lock->l_flags & LDLM_FL_CBPENDING))) {
1534 struct obd_info *oinfo;
1538 OBD_ALLOC_PTR(oinfo);
1543 OBD_FREE_PTR(oinfo);
1546 oa->o_id = lock->l_resource->lr_name.name[0];
1547 oa->o_seq = lock->l_resource->lr_name.name[1];
1548 oa->o_valid = OBD_MD_FLID|OBD_MD_FLGROUP;
1551 rc = obd_sync(lock->l_export, oinfo,
1552 lock->l_policy_data.l_extent.start,
1553 lock->l_policy_data.l_extent.end, NULL);
1555 CERROR("Error %d syncing data on lock cancel\n", rc);
1558 OBD_FREE_PTR(oinfo);
1561 rc = ldlm_server_blocking_ast(lock, desc, data, flag);
1565 static int ost_filter_recovery_request(struct ptlrpc_request *req,
1566 struct obd_device *obd, int *process)
1568 switch (lustre_msg_get_opc(req->rq_reqmsg)) {
1569 case OST_CONNECT: /* This will never get here, but for completeness. */
1570 case OST_DISCONNECT:
1581 case OBD_LOG_CANCEL:
1583 *process = target_queue_recovery_request(req, obd);
1587 DEBUG_REQ(D_ERROR, req, "not permitted during recovery");
1593 int ost_msg_check_version(struct lustre_msg *msg)
1597 switch(lustre_msg_get_opc(msg)) {
1599 case OST_DISCONNECT:
1602 case SEC_CTX_INIT_CONT:
1604 rc = lustre_msg_check_version(msg, LUSTRE_OBD_VERSION);
1606 CERROR("bad opc %u version %08x, expecting %08x\n",
1607 lustre_msg_get_opc(msg),
1608 lustre_msg_get_version(msg),
1609 LUSTRE_OBD_VERSION);
1622 #ifdef HAVE_QUOTA_SUPPORT
1623 case OST_QUOTACHECK:
1625 case OST_QUOTA_ADJUST_QUNIT:
1627 rc = lustre_msg_check_version(msg, LUSTRE_OST_VERSION);
1629 CERROR("bad opc %u version %08x, expecting %08x\n",
1630 lustre_msg_get_opc(msg),
1631 lustre_msg_get_version(msg),
1632 LUSTRE_OST_VERSION);
1637 case LDLM_BL_CALLBACK:
1638 case LDLM_CP_CALLBACK:
1639 rc = lustre_msg_check_version(msg, LUSTRE_DLM_VERSION);
1641 CERROR("bad opc %u version %08x, expecting %08x\n",
1642 lustre_msg_get_opc(msg),
1643 lustre_msg_get_version(msg),
1644 LUSTRE_DLM_VERSION);
1646 case LLOG_ORIGIN_CONNECT:
1647 case OBD_LOG_CANCEL:
1648 rc = lustre_msg_check_version(msg, LUSTRE_LOG_VERSION);
1650 CERROR("bad opc %u version %08x, expecting %08x\n",
1651 lustre_msg_get_opc(msg),
1652 lustre_msg_get_version(msg),
1653 LUSTRE_LOG_VERSION);
1656 CERROR("Unexpected opcode %d\n", lustre_msg_get_opc(msg));
1662 struct ost_prolong_data {
1663 struct ptlrpc_request *opd_req;
1664 struct obd_export *opd_exp;
1665 struct obdo *opd_oa;
1666 struct ldlm_res_id opd_resid;
1667 struct ldlm_extent opd_extent;
1668 ldlm_mode_t opd_mode;
1669 unsigned int opd_locks;
1673 /* prolong locks for the current service time of the corresponding
1674 * portal (= OST_IO_PORTAL)
1676 static inline int prolong_timeout(struct ptlrpc_request *req)
1678 struct ptlrpc_service *svc = req->rq_rqbd->rqbd_service;
1681 return obd_timeout / 2;
1683 return max(at_est2timeout(at_get(&svc->srv_at_estimate)), ldlm_timeout);
1686 static void ost_prolong_lock_one(struct ost_prolong_data *opd,
1687 struct ldlm_lock *lock)
1689 LASSERT(lock->l_req_mode == lock->l_granted_mode);
1690 LASSERT(lock->l_export == opd->opd_exp);
1692 /* XXX: never try to grab resource lock here because we're inside
1693 * exp_bl_list_lock; in ldlm_lockd.c to handle waiting list we take
1694 * res lock and then exp_bl_list_lock. */
1696 if (!(lock->l_flags & LDLM_FL_AST_SENT))
1697 /* ignore locks not being cancelled */
1701 "refreshed for req x"LPU64" ext("LPU64"->"LPU64") to %ds.\n",
1702 opd->opd_req->rq_xid, opd->opd_extent.start,
1703 opd->opd_extent.end, opd->opd_timeout);
1705 /* OK. this is a possible lock the user holds doing I/O
1706 * let's refresh eviction timer for it */
1707 ldlm_refresh_waiting_lock(lock, opd->opd_timeout);
1711 static void ost_prolong_locks(struct ost_prolong_data *data)
1713 struct obd_export *exp = data->opd_exp;
1714 struct obdo *oa = data->opd_oa;
1715 struct ldlm_lock *lock;
1718 if (oa->o_valid & OBD_MD_FLHANDLE) {
1719 /* mostly a request should be covered by only one lock, try
1721 lock = ldlm_handle2lock(&oa->o_handle);
1723 /* Fast path to check if the lock covers the whole IO
1724 * region exclusively. */
1725 if (lock->l_granted_mode == LCK_PW &&
1726 ldlm_extent_contain(&lock->l_policy_data.l_extent,
1727 &data->opd_extent)) {
1729 ost_prolong_lock_one(data, lock);
1730 LDLM_LOCK_PUT(lock);
1733 LDLM_LOCK_PUT(lock);
1738 cfs_spin_lock_bh(&exp->exp_bl_list_lock);
1739 cfs_list_for_each_entry(lock, &exp->exp_bl_list, l_exp_list) {
1740 LASSERT(lock->l_flags & LDLM_FL_AST_SENT);
1741 LASSERT(lock->l_resource->lr_type == LDLM_EXTENT);
1743 if (!ldlm_res_eq(&data->opd_resid, &lock->l_resource->lr_name))
1746 if (!ldlm_extent_overlap(&lock->l_policy_data.l_extent,
1750 ost_prolong_lock_one(data, lock);
1752 cfs_spin_unlock_bh(&exp->exp_bl_list_lock);
1758 * Returns 1 if the given PTLRPC matches the given LDLM locks, or 0 if it does
1761 static int ost_rw_hpreq_lock_match(struct ptlrpc_request *req,
1762 struct ldlm_lock *lock)
1764 struct niobuf_remote *nb;
1765 struct obd_ioobj *ioo;
1767 struct ldlm_extent ext;
1770 opc = lustre_msg_get_opc(req->rq_reqmsg);
1771 LASSERT(opc == OST_READ || opc == OST_WRITE);
1773 ioo = req_capsule_client_get(&req->rq_pill, &RMF_OBD_IOOBJ);
1774 LASSERT(ioo != NULL);
1776 nb = req_capsule_client_get(&req->rq_pill, &RMF_NIOBUF_REMOTE);
1777 LASSERT(nb != NULL);
1779 ext.start = nb->offset;
1780 nb += ioo->ioo_bufcnt - 1;
1781 ext.end = nb->offset + nb->len - 1;
1783 LASSERT(lock->l_resource != NULL);
1784 if (!osc_res_name_eq(ioo->ioo_id, ioo->ioo_seq,
1785 &lock->l_resource->lr_name))
1789 if (opc == OST_READ)
1791 if (!(lock->l_granted_mode & mode))
1794 RETURN(ldlm_extent_overlap(&lock->l_policy_data.l_extent, &ext));
1798 * High-priority queue request check for whether the given PTLRPC request (\a
1799 * req) is blocking an LDLM lock cancel.
1801 * Returns 1 if the given given PTLRPC request (\a req) is blocking an LDLM lock
1802 * cancel, 0 if it is not, and -EFAULT if the request is malformed.
1804 * Only OST_READs, OST_WRITEs and OST_PUNCHes go on the h-p RPC queue. This
1805 * function looks only at OST_READs and OST_WRITEs.
1807 static int ost_rw_hpreq_check(struct ptlrpc_request *req)
1809 struct obd_device *obd = req->rq_export->exp_obd;
1810 struct ost_body *body;
1811 struct obd_ioobj *ioo;
1812 struct niobuf_remote *nb;
1813 struct ost_prolong_data opd = { 0 };
1818 * Use LASSERT to do sanity check because malformed RPCs should have
1819 * been filtered out in ost_hpreq_handler().
1821 opc = lustre_msg_get_opc(req->rq_reqmsg);
1822 LASSERT(opc == OST_READ || opc == OST_WRITE);
1824 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
1825 LASSERT(body != NULL);
1827 ioo = req_capsule_client_get(&req->rq_pill, &RMF_OBD_IOOBJ);
1828 LASSERT(ioo != NULL);
1830 nb = req_capsule_client_get(&req->rq_pill, &RMF_NIOBUF_REMOTE);
1831 LASSERT(nb != NULL);
1832 LASSERT(!(nb->flags & OBD_BRW_SRVLOCK));
1834 osc_build_res_name(ioo->ioo_id, ioo->ioo_seq, &opd.opd_resid);
1838 if (opc == OST_READ)
1840 opd.opd_mode = mode;
1841 opd.opd_exp = req->rq_export;
1842 opd.opd_oa = &body->oa;
1843 opd.opd_extent.start = nb->offset;
1844 nb += ioo->ioo_bufcnt - 1;
1845 opd.opd_extent.end = nb->offset + nb->len - 1;
1846 opd.opd_timeout = prolong_timeout(req);
1848 DEBUG_REQ(D_RPCTRACE, req,
1849 "%s %s: refresh rw locks: " LPU64"/"LPU64" ("LPU64"->"LPU64")\n",
1850 obd->obd_name, cfs_current()->comm,
1851 opd.opd_resid.name[0], opd.opd_resid.name[1],
1852 opd.opd_extent.start, opd.opd_extent.end);
1854 ost_prolong_locks(&opd);
1856 CDEBUG(D_DLMTRACE, "%s: refreshed %u locks timeout for req %p.\n",
1857 obd->obd_name, opd.opd_locks, req);
1859 RETURN(opd.opd_locks);
1862 static void ost_rw_hpreq_fini(struct ptlrpc_request *req)
1864 (void)ost_rw_hpreq_check(req);
1868 * Like ost_rw_hpreq_lock_match(), but for OST_PUNCH RPCs.
1870 static int ost_punch_hpreq_lock_match(struct ptlrpc_request *req,
1871 struct ldlm_lock *lock)
1873 struct ost_body *body;
1876 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
1877 LASSERT(body != NULL);
1879 if (body->oa.o_valid & OBD_MD_FLHANDLE &&
1880 body->oa.o_handle.cookie == lock->l_handle.h_cookie)
1887 * Like ost_rw_hpreq_check(), but for OST_PUNCH RPCs.
1889 static int ost_punch_hpreq_check(struct ptlrpc_request *req)
1891 struct obd_device *obd = req->rq_export->exp_obd;
1892 struct ost_body *body;
1894 struct ost_prolong_data opd = { 0 };
1898 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
1899 LASSERT(body != NULL);
1902 LASSERT(!(oa->o_valid & OBD_MD_FLFLAGS) ||
1903 !(oa->o_flags & OBD_FL_SRVLOCK));
1906 end = start + oa->o_blocks;
1909 opd.opd_mode = LCK_PW;
1910 opd.opd_exp = req->rq_export;
1912 opd.opd_extent.start = start;
1913 opd.opd_extent.end = end;
1914 if (oa->o_blocks == OBD_OBJECT_EOF)
1915 opd.opd_extent.end = OBD_OBJECT_EOF;
1916 opd.opd_timeout = prolong_timeout(req);
1918 osc_build_res_name(oa->o_id, oa->o_seq, &opd.opd_resid);
1921 "%s: refresh locks: "LPU64"/"LPU64" ("LPU64"->"LPU64")\n",
1923 opd.opd_resid.name[0], opd.opd_resid.name[1],
1924 opd.opd_extent.start, opd.opd_extent.end);
1926 ost_prolong_locks(&opd);
1928 CDEBUG(D_DLMTRACE, "%s: refreshed %u locks timeout for req %p.\n",
1929 obd->obd_name, opd.opd_locks, req);
1931 RETURN(opd.opd_locks > 0);
1934 static void ost_punch_hpreq_fini(struct ptlrpc_request *req)
1936 (void)ost_punch_hpreq_check(req);
1939 struct ptlrpc_hpreq_ops ost_hpreq_rw = {
1940 .hpreq_lock_match = ost_rw_hpreq_lock_match,
1941 .hpreq_check = ost_rw_hpreq_check,
1942 .hpreq_fini = ost_rw_hpreq_fini
1945 struct ptlrpc_hpreq_ops ost_hpreq_punch = {
1946 .hpreq_lock_match = ost_punch_hpreq_lock_match,
1947 .hpreq_check = ost_punch_hpreq_check,
1948 .hpreq_fini = ost_punch_hpreq_fini
1951 /** Assign high priority operations to the request if needed. */
1952 static int ost_hpreq_handler(struct ptlrpc_request *req)
1955 if (req->rq_export) {
1956 int opc = lustre_msg_get_opc(req->rq_reqmsg);
1957 struct ost_body *body;
1959 if (opc == OST_READ || opc == OST_WRITE) {
1960 struct niobuf_remote *nb;
1961 struct obd_ioobj *ioo;
1962 int objcount, niocount;
1966 /* RPCs on the H-P queue can be inspected before
1967 * ost_handler() initializes their pills, so we
1968 * initialize that here. Capsule initialization is
1969 * idempotent, as is setting the pill's format (provided
1970 * it doesn't change).
1972 req_capsule_init(&req->rq_pill, req, RCL_SERVER);
1973 if (opc == OST_READ)
1974 req_capsule_set(&req->rq_pill,
1977 req_capsule_set(&req->rq_pill,
1978 &RQF_OST_BRW_WRITE);
1980 body = req_capsule_client_get(&req->rq_pill,
1983 CERROR("Missing/short ost_body\n");
1987 objcount = req_capsule_get_size(&req->rq_pill,
1991 if (objcount == 0) {
1992 CERROR("Missing/short ioobj\n");
1996 CERROR("too many ioobjs (%d)\n", objcount);
2000 ioo = req_capsule_client_get(&req->rq_pill,
2003 CERROR("Missing/short ioobj\n");
2007 rc = ost_validate_obdo(req->rq_export, &body->oa, ioo);
2009 CERROR("invalid object ids\n");
2013 for (niocount = i = 0; i < objcount; i++) {
2014 if (ioo[i].ioo_bufcnt == 0) {
2015 CERROR("ioo[%d] has zero bufcnt\n", i);
2018 niocount += ioo[i].ioo_bufcnt;
2020 if (niocount > PTLRPC_MAX_BRW_PAGES) {
2021 DEBUG_REQ(D_RPCTRACE, req,
2022 "bulk has too many pages (%d)",
2027 nb = req_capsule_client_get(&req->rq_pill,
2028 &RMF_NIOBUF_REMOTE);
2030 CERROR("Missing/short niobuf\n");
2034 if (niocount == 0 || !(nb[0].flags & OBD_BRW_SRVLOCK))
2035 req->rq_ops = &ost_hpreq_rw;
2036 } else if (opc == OST_PUNCH) {
2037 req_capsule_init(&req->rq_pill, req, RCL_SERVER);
2038 req_capsule_set(&req->rq_pill, &RQF_OST_PUNCH);
2040 body = req_capsule_client_get(&req->rq_pill,
2043 CERROR("Missing/short ost_body\n");
2047 if (!(body->oa.o_valid & OBD_MD_FLFLAGS) ||
2048 !(body->oa.o_flags & OBD_FL_SRVLOCK))
2049 req->rq_ops = &ost_hpreq_punch;
2055 /* TODO: handle requests in a similar way as MDT: see mdt_handle_common() */
2056 int ost_handle(struct ptlrpc_request *req)
2058 struct obd_trans_info trans_info = { 0, };
2059 struct obd_trans_info *oti = &trans_info;
2060 int should_process, fail = OBD_FAIL_OST_ALL_REPLY_NET, rc = 0;
2061 struct obd_device *obd = NULL;
2064 LASSERT(current->journal_info == NULL);
2066 /* primordial rpcs don't affect server recovery */
2067 switch (lustre_msg_get_opc(req->rq_reqmsg)) {
2069 case SEC_CTX_INIT_CONT:
2074 req_capsule_init(&req->rq_pill, req, RCL_SERVER);
2076 if (lustre_msg_get_opc(req->rq_reqmsg) != OST_CONNECT) {
2077 if (!class_connected_export(req->rq_export)) {
2078 CDEBUG(D_HA,"operation %d on unconnected OST from %s\n",
2079 lustre_msg_get_opc(req->rq_reqmsg),
2080 libcfs_id2str(req->rq_peer));
2081 req->rq_status = -ENOTCONN;
2082 GOTO(out, rc = -ENOTCONN);
2085 obd = req->rq_export->exp_obd;
2087 /* Check for aborted recovery. */
2088 if (obd->obd_recovering) {
2089 rc = ost_filter_recovery_request(req, obd,
2091 if (rc || !should_process)
2093 else if (should_process < 0) {
2094 req->rq_status = should_process;
2095 rc = ptlrpc_error(req);
2103 rc = ost_msg_check_version(req->rq_reqmsg);
2107 switch (lustre_msg_get_opc(req->rq_reqmsg)) {
2109 CDEBUG(D_INODE, "connect\n");
2110 req_capsule_set(&req->rq_pill, &RQF_OST_CONNECT);
2111 if (OBD_FAIL_CHECK(OBD_FAIL_OST_CONNECT_NET))
2113 rc = target_handle_connect(req);
2114 if (OBD_FAIL_CHECK(OBD_FAIL_OST_CONNECT_NET2))
2117 rc = ost_init_sec_level(req);
2119 rc = ost_connect_check_sptlrpc(req);
2123 case OST_DISCONNECT:
2124 CDEBUG(D_INODE, "disconnect\n");
2125 req_capsule_set(&req->rq_pill, &RQF_OST_DISCONNECT);
2126 if (OBD_FAIL_CHECK(OBD_FAIL_OST_DISCONNECT_NET))
2128 rc = target_handle_disconnect(req);
2131 CDEBUG(D_INODE, "create\n");
2132 req_capsule_set(&req->rq_pill, &RQF_OST_CREATE);
2133 if (OBD_FAIL_CHECK(OBD_FAIL_OST_CREATE_NET))
2135 if (OBD_FAIL_CHECK(OBD_FAIL_OST_EROFS))
2136 GOTO(out, rc = -EROFS);
2137 rc = ost_create(req->rq_export, req, oti);
2140 CDEBUG(D_INODE, "destroy\n");
2141 req_capsule_set(&req->rq_pill, &RQF_OST_DESTROY);
2142 if (OBD_FAIL_CHECK(OBD_FAIL_OST_DESTROY_NET))
2144 if (OBD_FAIL_CHECK(OBD_FAIL_OST_EROFS))
2145 GOTO(out, rc = -EROFS);
2146 rc = ost_destroy(req->rq_export, req, oti);
2149 CDEBUG(D_INODE, "getattr\n");
2150 req_capsule_set(&req->rq_pill, &RQF_OST_GETATTR);
2151 if (OBD_FAIL_CHECK(OBD_FAIL_OST_GETATTR_NET))
2153 rc = ost_getattr(req->rq_export, req);
2156 CDEBUG(D_INODE, "setattr\n");
2157 req_capsule_set(&req->rq_pill, &RQF_OST_SETATTR);
2158 if (OBD_FAIL_CHECK(OBD_FAIL_OST_SETATTR_NET))
2160 rc = ost_setattr(req->rq_export, req, oti);
2163 req_capsule_set(&req->rq_pill, &RQF_OST_BRW_WRITE);
2164 CDEBUG(D_INODE, "write\n");
2165 /* req->rq_request_portal would be nice, if it was set */
2166 if (req->rq_rqbd->rqbd_service->srv_req_portal !=OST_IO_PORTAL){
2167 CERROR("%s: deny write request from %s to portal %u\n",
2168 req->rq_export->exp_obd->obd_name,
2169 obd_export_nid2str(req->rq_export),
2170 req->rq_rqbd->rqbd_service->srv_req_portal);
2171 GOTO(out, rc = -EPROTO);
2173 if (OBD_FAIL_CHECK(OBD_FAIL_OST_BRW_NET))
2175 if (OBD_FAIL_CHECK(OBD_FAIL_OST_ENOSPC))
2176 GOTO(out, rc = -ENOSPC);
2177 if (OBD_FAIL_CHECK(OBD_FAIL_OST_EROFS))
2178 GOTO(out, rc = -EROFS);
2179 rc = ost_brw_write(req, oti);
2180 LASSERT(current->journal_info == NULL);
2181 /* ost_brw_write sends its own replies */
2184 req_capsule_set(&req->rq_pill, &RQF_OST_BRW_READ);
2185 CDEBUG(D_INODE, "read\n");
2186 /* req->rq_request_portal would be nice, if it was set */
2187 if (req->rq_rqbd->rqbd_service->srv_req_portal !=OST_IO_PORTAL){
2188 CERROR("%s: deny read request from %s to portal %u\n",
2189 req->rq_export->exp_obd->obd_name,
2190 obd_export_nid2str(req->rq_export),
2191 req->rq_rqbd->rqbd_service->srv_req_portal);
2192 GOTO(out, rc = -EPROTO);
2194 if (OBD_FAIL_CHECK(OBD_FAIL_OST_BRW_NET))
2196 rc = ost_brw_read(req, oti);
2197 LASSERT(current->journal_info == NULL);
2198 /* ost_brw_read sends its own replies */
2201 CDEBUG(D_INODE, "punch\n");
2202 req_capsule_set(&req->rq_pill, &RQF_OST_PUNCH);
2203 if (OBD_FAIL_CHECK(OBD_FAIL_OST_PUNCH_NET))
2205 if (OBD_FAIL_CHECK(OBD_FAIL_OST_EROFS))
2206 GOTO(out, rc = -EROFS);
2207 rc = ost_punch(req->rq_export, req, oti);
2210 CDEBUG(D_INODE, "statfs\n");
2211 req_capsule_set(&req->rq_pill, &RQF_OST_STATFS);
2212 if (OBD_FAIL_CHECK(OBD_FAIL_OST_STATFS_NET))
2214 rc = ost_statfs(req);
2217 CDEBUG(D_INODE, "sync\n");
2218 req_capsule_set(&req->rq_pill, &RQF_OST_SYNC);
2219 if (OBD_FAIL_CHECK(OBD_FAIL_OST_SYNC_NET))
2221 rc = ost_sync(req->rq_export, req);
2224 DEBUG_REQ(D_INODE, req, "set_info");
2225 req_capsule_set(&req->rq_pill, &RQF_OBD_SET_INFO);
2226 rc = ost_set_info(req->rq_export, req);
2229 DEBUG_REQ(D_INODE, req, "get_info");
2230 req_capsule_set(&req->rq_pill, &RQF_OST_GET_INFO_GENERIC);
2231 rc = ost_get_info(req->rq_export, req);
2233 #ifdef HAVE_QUOTA_SUPPORT
2234 case OST_QUOTACHECK:
2235 CDEBUG(D_INODE, "quotacheck\n");
2236 req_capsule_set(&req->rq_pill, &RQF_OST_QUOTACHECK);
2237 if (OBD_FAIL_CHECK(OBD_FAIL_OST_QUOTACHECK_NET))
2239 rc = ost_handle_quotacheck(req);
2242 CDEBUG(D_INODE, "quotactl\n");
2243 req_capsule_set(&req->rq_pill, &RQF_OST_QUOTACTL);
2244 if (OBD_FAIL_CHECK(OBD_FAIL_OST_QUOTACTL_NET))
2246 rc = ost_handle_quotactl(req);
2248 case OST_QUOTA_ADJUST_QUNIT:
2249 CDEBUG(D_INODE, "quota_adjust_qunit\n");
2250 req_capsule_set(&req->rq_pill, &RQF_OST_QUOTA_ADJUST_QUNIT);
2251 rc = ost_handle_quota_adjust_qunit(req);
2255 DEBUG_REQ(D_INODE, req, "ping");
2256 req_capsule_set(&req->rq_pill, &RQF_OBD_PING);
2257 rc = target_handle_ping(req);
2259 /* FIXME - just reply status */
2260 case LLOG_ORIGIN_CONNECT:
2261 DEBUG_REQ(D_INODE, req, "log connect");
2262 req_capsule_set(&req->rq_pill, &RQF_LLOG_ORIGIN_CONNECT);
2263 rc = ost_llog_handle_connect(req->rq_export, req);
2264 req->rq_status = rc;
2265 rc = req_capsule_server_pack(&req->rq_pill);
2268 RETURN(ptlrpc_reply(req));
2269 case OBD_LOG_CANCEL:
2270 CDEBUG(D_INODE, "log cancel\n");
2271 req_capsule_set(&req->rq_pill, &RQF_LOG_CANCEL);
2272 if (OBD_FAIL_CHECK(OBD_FAIL_OBD_LOG_CANCEL_NET))
2274 rc = llog_origin_handle_cancel(req);
2275 if (OBD_FAIL_CHECK(OBD_FAIL_OBD_LOG_CANCEL_REP))
2277 req->rq_status = rc;
2278 rc = req_capsule_server_pack(&req->rq_pill);
2281 RETURN(ptlrpc_reply(req));
2283 CDEBUG(D_INODE, "enqueue\n");
2284 req_capsule_set(&req->rq_pill, &RQF_LDLM_ENQUEUE);
2285 if (OBD_FAIL_CHECK(OBD_FAIL_LDLM_ENQUEUE))
2287 rc = ldlm_handle_enqueue(req, ldlm_server_completion_ast,
2289 ldlm_server_glimpse_ast);
2290 fail = OBD_FAIL_OST_LDLM_REPLY_NET;
2293 CDEBUG(D_INODE, "convert\n");
2294 req_capsule_set(&req->rq_pill, &RQF_LDLM_CONVERT);
2295 if (OBD_FAIL_CHECK(OBD_FAIL_LDLM_CONVERT))
2297 rc = ldlm_handle_convert(req);
2300 CDEBUG(D_INODE, "cancel\n");
2301 req_capsule_set(&req->rq_pill, &RQF_LDLM_CANCEL);
2302 if (OBD_FAIL_CHECK(OBD_FAIL_LDLM_CANCEL))
2304 rc = ldlm_handle_cancel(req);
2306 case LDLM_BL_CALLBACK:
2307 case LDLM_CP_CALLBACK:
2308 CDEBUG(D_INODE, "callback\n");
2309 CERROR("callbacks should not happen on OST\n");
2312 CERROR("Unexpected opcode %d\n",
2313 lustre_msg_get_opc(req->rq_reqmsg));
2314 req->rq_status = -ENOTSUPP;
2315 rc = ptlrpc_error(req);
2319 LASSERT(current->journal_info == NULL);
2322 /* If we're DISCONNECTing, the export_data is already freed */
2323 if (!rc && lustre_msg_get_opc(req->rq_reqmsg) != OST_DISCONNECT)
2324 target_committed_to_req(req);
2328 oti_to_request(oti, req);
2330 target_send_reply(req, rc, fail);
2333 EXPORT_SYMBOL(ost_handle);
2335 * free per-thread pool created by ost_thread_init().
2337 static void ost_thread_done(struct ptlrpc_thread *thread)
2339 struct ost_thread_local_cache *tls; /* TLS stands for Thread-Local
2344 LASSERT(thread != NULL);
2347 * be prepared to handle partially-initialized pools (because this is
2348 * called from ost_thread_init() for cleanup.
2350 tls = thread->t_data;
2353 thread->t_data = NULL;
2359 * initialize per-thread page pool (bug 5137).
2361 static int ost_thread_init(struct ptlrpc_thread *thread)
2363 struct ost_thread_local_cache *tls;
2367 LASSERT(thread != NULL);
2368 LASSERT(thread->t_data == NULL);
2369 LASSERTF(thread->t_id <= OSS_THREADS_MAX, "%u\n", thread->t_id);
2374 thread->t_data = tls;
2378 #define OST_WATCHDOG_TIMEOUT (obd_timeout * 1000)
2380 /* Sigh - really, this is an OSS, the _server_, not the _target_ */
2381 static int ost_setup(struct obd_device *obd, struct lustre_cfg* lcfg)
2383 struct ost_obd *ost = &obd->u.ost;
2384 struct lprocfs_static_vars lvars;
2385 int oss_min_threads;
2386 int oss_max_threads;
2387 int oss_min_create_threads;
2388 int oss_max_create_threads;
2392 rc = cfs_cleanup_group_info();
2396 lprocfs_ost_init_vars(&lvars);
2397 lprocfs_obd_setup(obd, lvars.obd_vars);
2399 cfs_mutex_init(&ost->ost_health_mutex);
2401 if (oss_num_threads) {
2402 /* If oss_num_threads is set, it is the min and the max. */
2403 if (oss_num_threads > OSS_THREADS_MAX)
2404 oss_num_threads = OSS_THREADS_MAX;
2405 if (oss_num_threads < OSS_THREADS_MIN)
2406 oss_num_threads = OSS_THREADS_MIN;
2407 oss_max_threads = oss_min_threads = oss_num_threads;
2409 /* Base min threads on memory and cpus */
2411 cfs_num_possible_cpus() * CFS_NUM_CACHEPAGES >>
2412 (27 - CFS_PAGE_SHIFT);
2413 if (oss_min_threads < OSS_THREADS_MIN)
2414 oss_min_threads = OSS_THREADS_MIN;
2415 /* Insure a 4x range for dynamic threads */
2416 if (oss_min_threads > OSS_THREADS_MAX / 4)
2417 oss_min_threads = OSS_THREADS_MAX / 4;
2418 oss_max_threads = min(OSS_THREADS_MAX, oss_min_threads * 4 + 1);
2422 ptlrpc_init_svc(OST_NBUFS, OST_BUFSIZE, OST_MAXREQSIZE,
2423 OST_MAXREPSIZE, OST_REQUEST_PORTAL,
2424 OSC_REPLY_PORTAL, OSS_SERVICE_WATCHDOG_FACTOR,
2425 ost_handle, LUSTRE_OSS_NAME,
2426 obd->obd_proc_entry, target_print_req,
2427 oss_min_threads, oss_max_threads,
2428 "ll_ost", LCT_DT_THREAD, NULL);
2429 if (ost->ost_service == NULL) {
2430 CERROR("failed to start service\n");
2431 GOTO(out_lprocfs, rc = -ENOMEM);
2434 rc = ptlrpc_start_threads(ost->ost_service);
2436 GOTO(out_service, rc = -EINVAL);
2438 if (oss_num_create_threads) {
2439 if (oss_num_create_threads > OSS_MAX_CREATE_THREADS)
2440 oss_num_create_threads = OSS_MAX_CREATE_THREADS;
2441 if (oss_num_create_threads < OSS_MIN_CREATE_THREADS)
2442 oss_num_create_threads = OSS_MIN_CREATE_THREADS;
2443 oss_min_create_threads = oss_max_create_threads =
2444 oss_num_create_threads;
2446 oss_min_create_threads = OSS_MIN_CREATE_THREADS;
2447 oss_max_create_threads = OSS_MAX_CREATE_THREADS;
2450 ost->ost_create_service =
2451 ptlrpc_init_svc(OST_NBUFS, OST_BUFSIZE, OST_MAXREQSIZE,
2452 OST_MAXREPSIZE, OST_CREATE_PORTAL,
2453 OSC_REPLY_PORTAL, OSS_SERVICE_WATCHDOG_FACTOR,
2454 ost_handle, "ost_create",
2455 obd->obd_proc_entry, target_print_req,
2456 oss_min_create_threads, oss_max_create_threads,
2457 "ll_ost_creat", LCT_DT_THREAD, NULL);
2458 if (ost->ost_create_service == NULL) {
2459 CERROR("failed to start OST create service\n");
2460 GOTO(out_service, rc = -ENOMEM);
2463 rc = ptlrpc_start_threads(ost->ost_create_service);
2465 GOTO(out_create, rc = -EINVAL);
2467 ost->ost_io_service =
2468 ptlrpc_init_svc(OST_NBUFS, OST_BUFSIZE, OST_MAXREQSIZE,
2469 OST_MAXREPSIZE, OST_IO_PORTAL,
2470 OSC_REPLY_PORTAL, OSS_SERVICE_WATCHDOG_FACTOR,
2471 ost_handle, "ost_io",
2472 obd->obd_proc_entry, target_print_req,
2473 oss_min_threads, oss_max_threads,
2474 "ll_ost_io", LCT_DT_THREAD, ost_hpreq_handler);
2475 if (ost->ost_io_service == NULL) {
2476 CERROR("failed to start OST I/O service\n");
2477 GOTO(out_create, rc = -ENOMEM);
2480 ost->ost_io_service->srv_init = ost_thread_init;
2481 ost->ost_io_service->srv_done = ost_thread_done;
2482 ost->ost_io_service->srv_cpu_affinity = 1;
2483 rc = ptlrpc_start_threads(ost->ost_io_service);
2485 GOTO(out_io, rc = -EINVAL);
2487 ping_evictor_start();
2492 ptlrpc_unregister_service(ost->ost_io_service);
2493 ost->ost_io_service = NULL;
2495 ptlrpc_unregister_service(ost->ost_create_service);
2496 ost->ost_create_service = NULL;
2498 ptlrpc_unregister_service(ost->ost_service);
2499 ost->ost_service = NULL;
2501 lprocfs_obd_cleanup(obd);
2505 static int ost_cleanup(struct obd_device *obd)
2507 struct ost_obd *ost = &obd->u.ost;
2511 ping_evictor_stop();
2513 /* there is no recovery for OST OBD, all recovery is controlled by
2515 LASSERT(obd->obd_recovering == 0);
2516 cfs_mutex_lock(&ost->ost_health_mutex);
2517 ptlrpc_unregister_service(ost->ost_service);
2518 ptlrpc_unregister_service(ost->ost_create_service);
2519 ptlrpc_unregister_service(ost->ost_io_service);
2520 ost->ost_service = NULL;
2521 ost->ost_create_service = NULL;
2522 cfs_mutex_unlock(&ost->ost_health_mutex);
2524 lprocfs_obd_cleanup(obd);
2529 static int ost_health_check(struct obd_device *obd)
2531 struct ost_obd *ost = &obd->u.ost;
2534 cfs_mutex_lock(&ost->ost_health_mutex);
2535 rc |= ptlrpc_service_health_check(ost->ost_service);
2536 rc |= ptlrpc_service_health_check(ost->ost_create_service);
2537 rc |= ptlrpc_service_health_check(ost->ost_io_service);
2538 cfs_mutex_unlock(&ost->ost_health_mutex);
2541 * health_check to return 0 on healthy
2542 * and 1 on unhealthy.
2550 struct ost_thread_local_cache *ost_tls(struct ptlrpc_request *r)
2552 return (struct ost_thread_local_cache *)(r->rq_svc_thread->t_data);
2555 /* use obd ops to offer management infrastructure */
2556 static struct obd_ops ost_obd_ops = {
2557 .o_owner = THIS_MODULE,
2558 .o_setup = ost_setup,
2559 .o_cleanup = ost_cleanup,
2560 .o_health_check = ost_health_check,
2564 static int __init ost_init(void)
2566 struct lprocfs_static_vars lvars;
2570 lprocfs_ost_init_vars(&lvars);
2571 rc = class_register_type(&ost_obd_ops, NULL, lvars.module_vars,
2572 LUSTRE_OSS_NAME, NULL);
2574 if (ost_num_threads != 0 && oss_num_threads == 0) {
2575 LCONSOLE_INFO("ost_num_threads module parameter is deprecated, "
2576 "use oss_num_threads instead or unset both for "
2577 "dynamic thread startup\n");
2578 oss_num_threads = ost_num_threads;
2584 static void /*__exit*/ ost_exit(void)
2586 class_unregister_type(LUSTRE_OSS_NAME);
2589 MODULE_AUTHOR("Sun Microsystems, Inc. <http://www.lustre.org/>");
2590 MODULE_DESCRIPTION("Lustre Object Storage Target (OST) v0.01");
2591 MODULE_LICENSE("GPL");
2593 module_init(ost_init);
2594 module_exit(ost_exit);