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) 2002, 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_OSC
39 #include "../../include/linux/libcfs/libcfs.h"
41 #include "../include/lustre_dlm.h"
42 #include "../include/lustre_net.h"
43 #include "../include/lustre/lustre_user.h"
44 #include "../include/obd_cksum.h"
46 #include "../include/lustre_ha.h"
47 #include "../include/lprocfs_status.h"
48 #include "../include/lustre_debug.h"
49 #include "../include/lustre_param.h"
50 #include "../include/lustre_fid.h"
51 #include "../include/obd_class.h"
52 #include "../include/obd.h"
53 #include "osc_internal.h"
54 #include "osc_cl_internal.h"
56 atomic_t osc_pool_req_count;
57 unsigned int osc_reqpool_maxreqcount;
58 struct ptlrpc_request_pool *osc_rq_pool;
60 /* max memory used for request pool, unit is MB */
61 static unsigned int osc_reqpool_mem_max = 5;
62 module_param(osc_reqpool_mem_max, uint, 0444);
64 struct osc_brw_async_args {
70 struct brw_page **aa_ppga;
71 struct client_obd *aa_cli;
72 struct list_head aa_oaps;
73 struct list_head aa_exts;
74 struct cl_req *aa_clerq;
77 struct osc_async_args {
78 struct obd_info *aa_oi;
81 struct osc_setattr_args {
83 obd_enqueue_update_f sa_upcall;
87 struct osc_fsync_args {
88 struct obd_info *fa_oi;
89 obd_enqueue_update_f fa_upcall;
93 struct osc_enqueue_args {
94 struct obd_export *oa_exp;
96 obd_enqueue_update_f oa_upcall;
98 struct ost_lvb *oa_lvb;
99 struct lustre_handle *oa_lockh;
100 struct ldlm_enqueue_info *oa_ei;
101 unsigned int oa_agl:1;
104 static void osc_release_ppga(struct brw_page **ppga, u32 count);
105 static int brw_interpret(const struct lu_env *env,
106 struct ptlrpc_request *req, void *data, int rc);
108 /* Pack OSC object metadata for disk storage (LE byte order). */
109 static int osc_packmd(struct obd_export *exp, struct lov_mds_md **lmmp,
110 struct lov_stripe_md *lsm)
114 lmm_size = sizeof(**lmmp);
122 } else if (unlikely(lsm && ostid_id(&lsm->lsm_oi) == 0)) {
127 *lmmp = kzalloc(lmm_size, GFP_NOFS);
133 ostid_cpu_to_le(&lsm->lsm_oi, &(*lmmp)->lmm_oi);
138 /* Unpack OSC object metadata from disk storage (LE byte order). */
139 static int osc_unpackmd(struct obd_export *exp, struct lov_stripe_md **lsmp,
140 struct lov_mds_md *lmm, int lmm_bytes)
143 struct obd_import *imp = class_exp2cliimp(exp);
146 if (lmm_bytes < sizeof(*lmm)) {
147 CERROR("%s: lov_mds_md too small: %d, need %d\n",
148 exp->exp_obd->obd_name, lmm_bytes,
152 /* XXX LOV_MAGIC etc check? */
154 if (unlikely(ostid_id(&lmm->lmm_oi) == 0)) {
155 CERROR("%s: zero lmm_object_id: rc = %d\n",
156 exp->exp_obd->obd_name, -EINVAL);
161 lsm_size = lov_stripe_md_size(1);
166 kfree((*lsmp)->lsm_oinfo[0]);
173 *lsmp = kzalloc(lsm_size, GFP_NOFS);
174 if (unlikely(!*lsmp))
176 (*lsmp)->lsm_oinfo[0] = kzalloc(sizeof(struct lov_oinfo),
178 if (unlikely(!(*lsmp)->lsm_oinfo[0])) {
182 loi_init((*lsmp)->lsm_oinfo[0]);
183 } else if (unlikely(ostid_id(&(*lsmp)->lsm_oi) == 0)) {
188 /* XXX zero *lsmp? */
189 ostid_le_to_cpu(&lmm->lmm_oi, &(*lsmp)->lsm_oi);
192 (imp->imp_connect_data.ocd_connect_flags & OBD_CONNECT_MAXBYTES))
193 (*lsmp)->lsm_maxbytes = imp->imp_connect_data.ocd_maxbytes;
195 (*lsmp)->lsm_maxbytes = LUSTRE_STRIPE_MAXBYTES;
200 static inline void osc_pack_req_body(struct ptlrpc_request *req,
201 struct obd_info *oinfo)
203 struct ost_body *body;
205 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
208 lustre_set_wire_obdo(&req->rq_import->imp_connect_data, &body->oa,
212 static int osc_getattr_interpret(const struct lu_env *env,
213 struct ptlrpc_request *req,
214 struct osc_async_args *aa, int rc)
216 struct ost_body *body;
221 body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
223 CDEBUG(D_INODE, "mode: %o\n", body->oa.o_mode);
224 lustre_get_wire_obdo(&req->rq_import->imp_connect_data,
225 aa->aa_oi->oi_oa, &body->oa);
227 /* This should really be sent by the OST */
228 aa->aa_oi->oi_oa->o_blksize = DT_MAX_BRW_SIZE;
229 aa->aa_oi->oi_oa->o_valid |= OBD_MD_FLBLKSZ;
231 CDEBUG(D_INFO, "can't unpack ost_body\n");
233 aa->aa_oi->oi_oa->o_valid = 0;
236 rc = aa->aa_oi->oi_cb_up(aa->aa_oi, rc);
240 static int osc_getattr_async(struct obd_export *exp, struct obd_info *oinfo,
241 struct ptlrpc_request_set *set)
243 struct ptlrpc_request *req;
244 struct osc_async_args *aa;
247 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_GETATTR);
251 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_GETATTR);
253 ptlrpc_request_free(req);
257 osc_pack_req_body(req, oinfo);
259 ptlrpc_request_set_replen(req);
260 req->rq_interpret_reply = (ptlrpc_interpterer_t)osc_getattr_interpret;
262 CLASSERT(sizeof(*aa) <= sizeof(req->rq_async_args));
263 aa = ptlrpc_req_async_args(req);
266 ptlrpc_set_add_req(set, req);
270 static int osc_getattr(const struct lu_env *env, struct obd_export *exp,
271 struct obd_info *oinfo)
273 struct ptlrpc_request *req;
274 struct ost_body *body;
277 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_GETATTR);
281 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_GETATTR);
283 ptlrpc_request_free(req);
287 osc_pack_req_body(req, oinfo);
289 ptlrpc_request_set_replen(req);
291 rc = ptlrpc_queue_wait(req);
295 body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
301 CDEBUG(D_INODE, "mode: %o\n", body->oa.o_mode);
302 lustre_get_wire_obdo(&req->rq_import->imp_connect_data, oinfo->oi_oa,
305 oinfo->oi_oa->o_blksize = cli_brw_size(exp->exp_obd);
306 oinfo->oi_oa->o_valid |= OBD_MD_FLBLKSZ;
309 ptlrpc_req_finished(req);
313 static int osc_setattr(const struct lu_env *env, struct obd_export *exp,
314 struct obd_info *oinfo, struct obd_trans_info *oti)
316 struct ptlrpc_request *req;
317 struct ost_body *body;
320 LASSERT(oinfo->oi_oa->o_valid & OBD_MD_FLGROUP);
322 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_SETATTR);
326 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_SETATTR);
328 ptlrpc_request_free(req);
332 osc_pack_req_body(req, oinfo);
334 ptlrpc_request_set_replen(req);
336 rc = ptlrpc_queue_wait(req);
340 body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
346 lustre_get_wire_obdo(&req->rq_import->imp_connect_data, oinfo->oi_oa,
350 ptlrpc_req_finished(req);
354 static int osc_setattr_interpret(const struct lu_env *env,
355 struct ptlrpc_request *req,
356 struct osc_setattr_args *sa, int rc)
358 struct ost_body *body;
363 body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
369 lustre_get_wire_obdo(&req->rq_import->imp_connect_data, sa->sa_oa,
372 rc = sa->sa_upcall(sa->sa_cookie, rc);
376 int osc_setattr_async_base(struct obd_export *exp, struct obd_info *oinfo,
377 struct obd_trans_info *oti,
378 obd_enqueue_update_f upcall, void *cookie,
379 struct ptlrpc_request_set *rqset)
381 struct ptlrpc_request *req;
382 struct osc_setattr_args *sa;
385 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_SETATTR);
389 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_SETATTR);
391 ptlrpc_request_free(req);
395 if (oti && oinfo->oi_oa->o_valid & OBD_MD_FLCOOKIE)
396 oinfo->oi_oa->o_lcookie = *oti->oti_logcookies;
398 osc_pack_req_body(req, oinfo);
400 ptlrpc_request_set_replen(req);
402 /* do mds to ost setattr asynchronously */
404 /* Do not wait for response. */
405 ptlrpcd_add_req(req);
407 req->rq_interpret_reply =
408 (ptlrpc_interpterer_t)osc_setattr_interpret;
410 CLASSERT(sizeof(*sa) <= sizeof(req->rq_async_args));
411 sa = ptlrpc_req_async_args(req);
412 sa->sa_oa = oinfo->oi_oa;
413 sa->sa_upcall = upcall;
414 sa->sa_cookie = cookie;
416 if (rqset == PTLRPCD_SET)
417 ptlrpcd_add_req(req);
419 ptlrpc_set_add_req(rqset, req);
425 static int osc_setattr_async(struct obd_export *exp, struct obd_info *oinfo,
426 struct obd_trans_info *oti,
427 struct ptlrpc_request_set *rqset)
429 return osc_setattr_async_base(exp, oinfo, oti,
430 oinfo->oi_cb_up, oinfo, rqset);
433 static int osc_real_create(struct obd_export *exp, struct obdo *oa,
434 struct lov_stripe_md **ea,
435 struct obd_trans_info *oti)
437 struct ptlrpc_request *req;
438 struct ost_body *body;
439 struct lov_stripe_md *lsm;
447 rc = obd_alloc_memmd(exp, &lsm);
452 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_CREATE);
458 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_CREATE);
460 ptlrpc_request_free(req);
464 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
467 lustre_set_wire_obdo(&req->rq_import->imp_connect_data, &body->oa, oa);
469 ptlrpc_request_set_replen(req);
471 if ((oa->o_valid & OBD_MD_FLFLAGS) &&
472 oa->o_flags == OBD_FL_DELORPHAN) {
474 "delorphan from OST integration");
475 /* Don't resend the delorphan req */
476 req->rq_no_resend = req->rq_no_delay = 1;
479 rc = ptlrpc_queue_wait(req);
483 body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
489 CDEBUG(D_INFO, "oa flags %x\n", oa->o_flags);
490 lustre_get_wire_obdo(&req->rq_import->imp_connect_data, oa, &body->oa);
492 oa->o_blksize = cli_brw_size(exp->exp_obd);
493 oa->o_valid |= OBD_MD_FLBLKSZ;
495 /* XXX LOV STACKING: the lsm that is passed to us from LOV does not
496 * have valid lsm_oinfo data structs, so don't go touching that.
497 * This needs to be fixed in a big way.
499 lsm->lsm_oi = oa->o_oi;
503 oti->oti_transno = lustre_msg_get_transno(req->rq_repmsg);
505 if (oa->o_valid & OBD_MD_FLCOOKIE) {
506 if (!oti->oti_logcookies)
507 oti_alloc_cookies(oti, 1);
508 *oti->oti_logcookies = oa->o_lcookie;
512 CDEBUG(D_HA, "transno: %lld\n",
513 lustre_msg_get_transno(req->rq_repmsg));
515 ptlrpc_req_finished(req);
518 obd_free_memmd(exp, &lsm);
522 int osc_punch_base(struct obd_export *exp, struct obd_info *oinfo,
523 obd_enqueue_update_f upcall, void *cookie,
524 struct ptlrpc_request_set *rqset)
526 struct ptlrpc_request *req;
527 struct osc_setattr_args *sa;
528 struct ost_body *body;
531 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_PUNCH);
535 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_PUNCH);
537 ptlrpc_request_free(req);
540 req->rq_request_portal = OST_IO_PORTAL; /* bug 7198 */
541 ptlrpc_at_set_req_timeout(req);
543 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
545 lustre_set_wire_obdo(&req->rq_import->imp_connect_data, &body->oa,
548 ptlrpc_request_set_replen(req);
550 req->rq_interpret_reply = (ptlrpc_interpterer_t)osc_setattr_interpret;
551 CLASSERT(sizeof(*sa) <= sizeof(req->rq_async_args));
552 sa = ptlrpc_req_async_args(req);
553 sa->sa_oa = oinfo->oi_oa;
554 sa->sa_upcall = upcall;
555 sa->sa_cookie = cookie;
556 if (rqset == PTLRPCD_SET)
557 ptlrpcd_add_req(req);
559 ptlrpc_set_add_req(rqset, req);
564 static int osc_sync_interpret(const struct lu_env *env,
565 struct ptlrpc_request *req,
568 struct osc_fsync_args *fa = arg;
569 struct ost_body *body;
574 body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
576 CERROR("can't unpack ost_body\n");
581 *fa->fa_oi->oi_oa = body->oa;
583 rc = fa->fa_upcall(fa->fa_cookie, rc);
587 int osc_sync_base(struct obd_export *exp, struct obd_info *oinfo,
588 obd_enqueue_update_f upcall, void *cookie,
589 struct ptlrpc_request_set *rqset)
591 struct ptlrpc_request *req;
592 struct ost_body *body;
593 struct osc_fsync_args *fa;
596 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_SYNC);
600 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_SYNC);
602 ptlrpc_request_free(req);
606 /* overload the size and blocks fields in the oa with start/end */
607 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
609 lustre_set_wire_obdo(&req->rq_import->imp_connect_data, &body->oa,
612 ptlrpc_request_set_replen(req);
613 req->rq_interpret_reply = osc_sync_interpret;
615 CLASSERT(sizeof(*fa) <= sizeof(req->rq_async_args));
616 fa = ptlrpc_req_async_args(req);
618 fa->fa_upcall = upcall;
619 fa->fa_cookie = cookie;
621 if (rqset == PTLRPCD_SET)
622 ptlrpcd_add_req(req);
624 ptlrpc_set_add_req(rqset, req);
629 /* Find and cancel locally locks matched by @mode in the resource found by
630 * @objid. Found locks are added into @cancel list. Returns the amount of
631 * locks added to @cancels list.
633 static int osc_resource_get_unused(struct obd_export *exp, struct obdo *oa,
634 struct list_head *cancels,
635 enum ldlm_mode mode, __u64 lock_flags)
637 struct ldlm_namespace *ns = exp->exp_obd->obd_namespace;
638 struct ldlm_res_id res_id;
639 struct ldlm_resource *res;
642 /* Return, i.e. cancel nothing, only if ELC is supported (flag in
643 * export) but disabled through procfs (flag in NS).
645 * This distinguishes from a case when ELC is not supported originally,
646 * when we still want to cancel locks in advance and just cancel them
647 * locally, without sending any RPC.
649 if (exp_connect_cancelset(exp) && !ns_connect_cancelset(ns))
652 ostid_build_res_name(&oa->o_oi, &res_id);
653 res = ldlm_resource_get(ns, NULL, &res_id, 0, 0);
657 LDLM_RESOURCE_ADDREF(res);
658 count = ldlm_cancel_resource_local(res, cancels, NULL, mode,
659 lock_flags, 0, NULL);
660 LDLM_RESOURCE_DELREF(res);
661 ldlm_resource_putref(res);
665 static int osc_destroy_interpret(const struct lu_env *env,
666 struct ptlrpc_request *req, void *data,
669 struct client_obd *cli = &req->rq_import->imp_obd->u.cli;
671 atomic_dec(&cli->cl_destroy_in_flight);
672 wake_up(&cli->cl_destroy_waitq);
676 static int osc_can_send_destroy(struct client_obd *cli)
678 if (atomic_inc_return(&cli->cl_destroy_in_flight) <=
679 cli->cl_max_rpcs_in_flight) {
680 /* The destroy request can be sent */
683 if (atomic_dec_return(&cli->cl_destroy_in_flight) <
684 cli->cl_max_rpcs_in_flight) {
686 * The counter has been modified between the two atomic
689 wake_up(&cli->cl_destroy_waitq);
694 static int osc_create(const struct lu_env *env, struct obd_export *exp,
695 struct obdo *oa, struct lov_stripe_md **ea,
696 struct obd_trans_info *oti)
702 LASSERT(oa->o_valid & OBD_MD_FLGROUP);
704 if ((oa->o_valid & OBD_MD_FLFLAGS) &&
705 oa->o_flags == OBD_FL_RECREATE_OBJS) {
706 return osc_real_create(exp, oa, ea, oti);
709 if (!fid_seq_is_mdt(ostid_seq(&oa->o_oi)))
710 return osc_real_create(exp, oa, ea, oti);
712 /* we should not get here anymore */
718 /* Destroy requests can be async always on the client, and we don't even really
719 * care about the return code since the client cannot do anything at all about
721 * When the MDS is unlinking a filename, it saves the file objects into a
722 * recovery llog, and these object records are cancelled when the OST reports
723 * they were destroyed and sync'd to disk (i.e. transaction committed).
724 * If the client dies, or the OST is down when the object should be destroyed,
725 * the records are not cancelled, and when the OST reconnects to the MDS next,
726 * it will retrieve the llog unlink logs and then sends the log cancellation
727 * cookies to the MDS after committing destroy transactions.
729 static int osc_destroy(const struct lu_env *env, struct obd_export *exp,
730 struct obdo *oa, struct lov_stripe_md *ea,
731 struct obd_trans_info *oti, struct obd_export *md_export)
733 struct client_obd *cli = &exp->exp_obd->u.cli;
734 struct ptlrpc_request *req;
735 struct ost_body *body;
740 CDEBUG(D_INFO, "oa NULL\n");
744 count = osc_resource_get_unused(exp, oa, &cancels, LCK_PW,
745 LDLM_FL_DISCARD_DATA);
747 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_DESTROY);
749 ldlm_lock_list_put(&cancels, l_bl_ast, count);
753 rc = ldlm_prep_elc_req(exp, req, LUSTRE_OST_VERSION, OST_DESTROY,
756 ptlrpc_request_free(req);
760 req->rq_request_portal = OST_IO_PORTAL; /* bug 7198 */
761 ptlrpc_at_set_req_timeout(req);
763 if (oti && oa->o_valid & OBD_MD_FLCOOKIE)
764 oa->o_lcookie = *oti->oti_logcookies;
765 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
767 lustre_set_wire_obdo(&req->rq_import->imp_connect_data, &body->oa, oa);
769 ptlrpc_request_set_replen(req);
771 /* If osc_destroy is for destroying the unlink orphan,
772 * sent from MDT to OST, which should not be blocked here,
773 * because the process might be triggered by ptlrpcd, and
774 * it is not good to block ptlrpcd thread (b=16006
776 if (!(oa->o_flags & OBD_FL_DELORPHAN)) {
777 req->rq_interpret_reply = osc_destroy_interpret;
778 if (!osc_can_send_destroy(cli)) {
779 struct l_wait_info lwi = LWI_INTR(LWI_ON_SIGNAL_NOOP,
783 * Wait until the number of on-going destroy RPCs drops
784 * under max_rpc_in_flight
786 l_wait_event_exclusive(cli->cl_destroy_waitq,
787 osc_can_send_destroy(cli), &lwi);
791 /* Do not wait for response */
792 ptlrpcd_add_req(req);
796 static void osc_announce_cached(struct client_obd *cli, struct obdo *oa,
799 u32 bits = OBD_MD_FLBLOCKS|OBD_MD_FLGRANT;
801 LASSERT(!(oa->o_valid & bits));
804 client_obd_list_lock(&cli->cl_loi_list_lock);
805 oa->o_dirty = cli->cl_dirty;
806 if (unlikely(cli->cl_dirty - cli->cl_dirty_transit >
807 cli->cl_dirty_max)) {
808 CERROR("dirty %lu - %lu > dirty_max %lu\n",
809 cli->cl_dirty, cli->cl_dirty_transit, cli->cl_dirty_max);
811 } else if (unlikely(atomic_read(&obd_dirty_pages) -
812 atomic_read(&obd_dirty_transit_pages) >
813 (long)(obd_max_dirty_pages + 1))) {
814 /* The atomic_read() allowing the atomic_inc() are
815 * not covered by a lock thus they may safely race and trip
816 * this CERROR() unless we add in a small fudge factor (+1).
818 CERROR("dirty %d - %d > system dirty_max %d\n",
819 atomic_read(&obd_dirty_pages),
820 atomic_read(&obd_dirty_transit_pages),
821 obd_max_dirty_pages);
823 } else if (unlikely(cli->cl_dirty_max - cli->cl_dirty > 0x7fffffff)) {
824 CERROR("dirty %lu - dirty_max %lu too big???\n",
825 cli->cl_dirty, cli->cl_dirty_max);
828 long max_in_flight = (cli->cl_max_pages_per_rpc <<
830 (cli->cl_max_rpcs_in_flight + 1);
831 oa->o_undirty = max(cli->cl_dirty_max, max_in_flight);
833 oa->o_grant = cli->cl_avail_grant + cli->cl_reserved_grant;
834 oa->o_dropped = cli->cl_lost_grant;
835 cli->cl_lost_grant = 0;
836 client_obd_list_unlock(&cli->cl_loi_list_lock);
837 CDEBUG(D_CACHE, "dirty: %llu undirty: %u dropped %u grant: %llu\n",
838 oa->o_dirty, oa->o_undirty, oa->o_dropped, oa->o_grant);
842 void osc_update_next_shrink(struct client_obd *cli)
844 cli->cl_next_shrink_grant =
845 cfs_time_shift(cli->cl_grant_shrink_interval);
846 CDEBUG(D_CACHE, "next time %ld to shrink grant\n",
847 cli->cl_next_shrink_grant);
850 static void __osc_update_grant(struct client_obd *cli, u64 grant)
852 client_obd_list_lock(&cli->cl_loi_list_lock);
853 cli->cl_avail_grant += grant;
854 client_obd_list_unlock(&cli->cl_loi_list_lock);
857 static void osc_update_grant(struct client_obd *cli, struct ost_body *body)
859 if (body->oa.o_valid & OBD_MD_FLGRANT) {
860 CDEBUG(D_CACHE, "got %llu extra grant\n", body->oa.o_grant);
861 __osc_update_grant(cli, body->oa.o_grant);
865 static int osc_set_info_async(const struct lu_env *env, struct obd_export *exp,
866 u32 keylen, void *key, u32 vallen,
867 void *val, struct ptlrpc_request_set *set);
869 static int osc_shrink_grant_interpret(const struct lu_env *env,
870 struct ptlrpc_request *req,
873 struct client_obd *cli = &req->rq_import->imp_obd->u.cli;
874 struct obdo *oa = ((struct osc_brw_async_args *)aa)->aa_oa;
875 struct ost_body *body;
878 __osc_update_grant(cli, oa->o_grant);
882 body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
884 osc_update_grant(cli, body);
886 kmem_cache_free(obdo_cachep, oa);
890 static void osc_shrink_grant_local(struct client_obd *cli, struct obdo *oa)
892 client_obd_list_lock(&cli->cl_loi_list_lock);
893 oa->o_grant = cli->cl_avail_grant / 4;
894 cli->cl_avail_grant -= oa->o_grant;
895 client_obd_list_unlock(&cli->cl_loi_list_lock);
896 if (!(oa->o_valid & OBD_MD_FLFLAGS)) {
897 oa->o_valid |= OBD_MD_FLFLAGS;
900 oa->o_flags |= OBD_FL_SHRINK_GRANT;
901 osc_update_next_shrink(cli);
904 /* Shrink the current grant, either from some large amount to enough for a
905 * full set of in-flight RPCs, or if we have already shrunk to that limit
906 * then to enough for a single RPC. This avoids keeping more grant than
907 * needed, and avoids shrinking the grant piecemeal.
909 static int osc_shrink_grant(struct client_obd *cli)
911 __u64 target_bytes = (cli->cl_max_rpcs_in_flight + 1) *
912 (cli->cl_max_pages_per_rpc << PAGE_CACHE_SHIFT);
914 client_obd_list_lock(&cli->cl_loi_list_lock);
915 if (cli->cl_avail_grant <= target_bytes)
916 target_bytes = cli->cl_max_pages_per_rpc << PAGE_CACHE_SHIFT;
917 client_obd_list_unlock(&cli->cl_loi_list_lock);
919 return osc_shrink_grant_to_target(cli, target_bytes);
922 int osc_shrink_grant_to_target(struct client_obd *cli, __u64 target_bytes)
925 struct ost_body *body;
927 client_obd_list_lock(&cli->cl_loi_list_lock);
928 /* Don't shrink if we are already above or below the desired limit
929 * We don't want to shrink below a single RPC, as that will negatively
930 * impact block allocation and long-term performance.
932 if (target_bytes < cli->cl_max_pages_per_rpc << PAGE_CACHE_SHIFT)
933 target_bytes = cli->cl_max_pages_per_rpc << PAGE_CACHE_SHIFT;
935 if (target_bytes >= cli->cl_avail_grant) {
936 client_obd_list_unlock(&cli->cl_loi_list_lock);
939 client_obd_list_unlock(&cli->cl_loi_list_lock);
941 body = kzalloc(sizeof(*body), GFP_NOFS);
945 osc_announce_cached(cli, &body->oa, 0);
947 client_obd_list_lock(&cli->cl_loi_list_lock);
948 body->oa.o_grant = cli->cl_avail_grant - target_bytes;
949 cli->cl_avail_grant = target_bytes;
950 client_obd_list_unlock(&cli->cl_loi_list_lock);
951 if (!(body->oa.o_valid & OBD_MD_FLFLAGS)) {
952 body->oa.o_valid |= OBD_MD_FLFLAGS;
953 body->oa.o_flags = 0;
955 body->oa.o_flags |= OBD_FL_SHRINK_GRANT;
956 osc_update_next_shrink(cli);
958 rc = osc_set_info_async(NULL, cli->cl_import->imp_obd->obd_self_export,
959 sizeof(KEY_GRANT_SHRINK), KEY_GRANT_SHRINK,
960 sizeof(*body), body, NULL);
962 __osc_update_grant(cli, body->oa.o_grant);
967 static int osc_should_shrink_grant(struct client_obd *client)
969 unsigned long time = cfs_time_current();
970 unsigned long next_shrink = client->cl_next_shrink_grant;
972 if ((client->cl_import->imp_connect_data.ocd_connect_flags &
973 OBD_CONNECT_GRANT_SHRINK) == 0)
976 if (cfs_time_aftereq(time, next_shrink - 5 * CFS_TICK)) {
977 /* Get the current RPC size directly, instead of going via:
978 * cli_brw_size(obd->u.cli.cl_import->imp_obd->obd_self_export)
979 * Keep comment here so that it can be found by searching.
981 int brw_size = client->cl_max_pages_per_rpc << PAGE_CACHE_SHIFT;
983 if (client->cl_import->imp_state == LUSTRE_IMP_FULL &&
984 client->cl_avail_grant > brw_size)
987 osc_update_next_shrink(client);
992 static int osc_grant_shrink_grant_cb(struct timeout_item *item, void *data)
994 struct client_obd *client;
996 list_for_each_entry(client, &item->ti_obd_list, cl_grant_shrink_list) {
997 if (osc_should_shrink_grant(client))
998 osc_shrink_grant(client);
1003 static int osc_add_shrink_grant(struct client_obd *client)
1007 rc = ptlrpc_add_timeout_client(client->cl_grant_shrink_interval,
1009 osc_grant_shrink_grant_cb, NULL,
1010 &client->cl_grant_shrink_list);
1012 CERROR("add grant client %s error %d\n",
1013 client->cl_import->imp_obd->obd_name, rc);
1016 CDEBUG(D_CACHE, "add grant client %s\n",
1017 client->cl_import->imp_obd->obd_name);
1018 osc_update_next_shrink(client);
1022 static int osc_del_shrink_grant(struct client_obd *client)
1024 return ptlrpc_del_timeout_client(&client->cl_grant_shrink_list,
1028 static void osc_init_grant(struct client_obd *cli, struct obd_connect_data *ocd)
1031 * ocd_grant is the total grant amount we're expect to hold: if we've
1032 * been evicted, it's the new avail_grant amount, cl_dirty will drop
1033 * to 0 as inflight RPCs fail out; otherwise, it's avail_grant + dirty.
1035 * race is tolerable here: if we're evicted, but imp_state already
1036 * left EVICTED state, then cl_dirty must be 0 already.
1038 client_obd_list_lock(&cli->cl_loi_list_lock);
1039 if (cli->cl_import->imp_state == LUSTRE_IMP_EVICTED)
1040 cli->cl_avail_grant = ocd->ocd_grant;
1042 cli->cl_avail_grant = ocd->ocd_grant - cli->cl_dirty;
1044 if (cli->cl_avail_grant < 0) {
1045 CWARN("%s: available grant < 0: avail/ocd/dirty %ld/%u/%ld\n",
1046 cli->cl_import->imp_obd->obd_name, cli->cl_avail_grant,
1047 ocd->ocd_grant, cli->cl_dirty);
1048 /* workaround for servers which do not have the patch from
1051 cli->cl_avail_grant = ocd->ocd_grant;
1054 /* determine the appropriate chunk size used by osc_extent. */
1055 cli->cl_chunkbits = max_t(int, PAGE_CACHE_SHIFT, ocd->ocd_blocksize);
1056 client_obd_list_unlock(&cli->cl_loi_list_lock);
1058 CDEBUG(D_CACHE, "%s, setting cl_avail_grant: %ld cl_lost_grant: %ld chunk bits: %d\n",
1059 cli->cl_import->imp_obd->obd_name,
1060 cli->cl_avail_grant, cli->cl_lost_grant, cli->cl_chunkbits);
1062 if (ocd->ocd_connect_flags & OBD_CONNECT_GRANT_SHRINK &&
1063 list_empty(&cli->cl_grant_shrink_list))
1064 osc_add_shrink_grant(cli);
1067 /* We assume that the reason this OSC got a short read is because it read
1068 * beyond the end of a stripe file; i.e. lustre is reading a sparse file
1069 * via the LOV, and it _knows_ it's reading inside the file, it's just that
1070 * this stripe never got written at or beyond this stripe offset yet.
1072 static void handle_short_read(int nob_read, u32 page_count,
1073 struct brw_page **pga)
1078 /* skip bytes read OK */
1079 while (nob_read > 0) {
1080 LASSERT(page_count > 0);
1082 if (pga[i]->count > nob_read) {
1083 /* EOF inside this page */
1084 ptr = kmap(pga[i]->pg) +
1085 (pga[i]->off & ~CFS_PAGE_MASK);
1086 memset(ptr + nob_read, 0, pga[i]->count - nob_read);
1093 nob_read -= pga[i]->count;
1098 /* zero remaining pages */
1099 while (page_count-- > 0) {
1100 ptr = kmap(pga[i]->pg) + (pga[i]->off & ~CFS_PAGE_MASK);
1101 memset(ptr, 0, pga[i]->count);
1107 static int check_write_rcs(struct ptlrpc_request *req,
1108 int requested_nob, int niocount,
1109 u32 page_count, struct brw_page **pga)
1114 remote_rcs = req_capsule_server_sized_get(&req->rq_pill, &RMF_RCS,
1115 sizeof(*remote_rcs) *
1118 CDEBUG(D_INFO, "Missing/short RC vector on BRW_WRITE reply\n");
1122 /* return error if any niobuf was in error */
1123 for (i = 0; i < niocount; i++) {
1124 if ((int)remote_rcs[i] < 0)
1125 return remote_rcs[i];
1127 if (remote_rcs[i] != 0) {
1128 CDEBUG(D_INFO, "rc[%d] invalid (%d) req %p\n",
1129 i, remote_rcs[i], req);
1134 if (req->rq_bulk->bd_nob_transferred != requested_nob) {
1135 CERROR("Unexpected # bytes transferred: %d (requested %d)\n",
1136 req->rq_bulk->bd_nob_transferred, requested_nob);
1143 static inline int can_merge_pages(struct brw_page *p1, struct brw_page *p2)
1145 if (p1->flag != p2->flag) {
1146 unsigned mask = ~(OBD_BRW_FROM_GRANT | OBD_BRW_NOCACHE |
1147 OBD_BRW_SYNC | OBD_BRW_ASYNC|OBD_BRW_NOQUOTA);
1149 /* warn if we try to combine flags that we don't know to be
1152 if (unlikely((p1->flag & mask) != (p2->flag & mask))) {
1153 CWARN("Saw flags 0x%x and 0x%x in the same brw, please report this at http://bugs.whamcloud.com/\n",
1154 p1->flag, p2->flag);
1159 return (p1->off + p1->count == p2->off);
1162 static u32 osc_checksum_bulk(int nob, u32 pg_count,
1163 struct brw_page **pga, int opc,
1164 enum cksum_type cksum_type)
1168 struct cfs_crypto_hash_desc *hdesc;
1169 unsigned int bufsize;
1171 unsigned char cfs_alg = cksum_obd2cfs(cksum_type);
1173 LASSERT(pg_count > 0);
1175 hdesc = cfs_crypto_hash_init(cfs_alg, NULL, 0);
1176 if (IS_ERR(hdesc)) {
1177 CERROR("Unable to initialize checksum hash %s\n",
1178 cfs_crypto_hash_name(cfs_alg));
1179 return PTR_ERR(hdesc);
1182 while (nob > 0 && pg_count > 0) {
1183 int count = pga[i]->count > nob ? nob : pga[i]->count;
1185 /* corrupt the data before we compute the checksum, to
1186 * simulate an OST->client data error
1188 if (i == 0 && opc == OST_READ &&
1189 OBD_FAIL_CHECK(OBD_FAIL_OSC_CHECKSUM_RECEIVE)) {
1190 unsigned char *ptr = kmap(pga[i]->pg);
1191 int off = pga[i]->off & ~CFS_PAGE_MASK;
1193 memcpy(ptr + off, "bad1", min(4, nob));
1196 cfs_crypto_hash_update_page(hdesc, pga[i]->pg,
1197 pga[i]->off & ~CFS_PAGE_MASK,
1200 "page %p map %p index %lu flags %lx count %u priv %0lx: off %d\n",
1201 pga[i]->pg, pga[i]->pg->mapping, pga[i]->pg->index,
1202 (long)pga[i]->pg->flags, page_count(pga[i]->pg),
1203 page_private(pga[i]->pg),
1204 (int)(pga[i]->off & ~CFS_PAGE_MASK));
1206 nob -= pga[i]->count;
1212 err = cfs_crypto_hash_final(hdesc, (unsigned char *)&cksum, &bufsize);
1215 cfs_crypto_hash_final(hdesc, NULL, NULL);
1217 /* For sending we only compute the wrong checksum instead
1218 * of corrupting the data so it is still correct on a redo
1220 if (opc == OST_WRITE && OBD_FAIL_CHECK(OBD_FAIL_OSC_CHECKSUM_SEND))
1226 static int osc_brw_prep_request(int cmd, struct client_obd *cli,
1228 struct lov_stripe_md *lsm, u32 page_count,
1229 struct brw_page **pga,
1230 struct ptlrpc_request **reqp,
1234 struct ptlrpc_request *req;
1235 struct ptlrpc_bulk_desc *desc;
1236 struct ost_body *body;
1237 struct obd_ioobj *ioobj;
1238 struct niobuf_remote *niobuf;
1239 int niocount, i, requested_nob, opc, rc;
1240 struct osc_brw_async_args *aa;
1241 struct req_capsule *pill;
1242 struct brw_page *pg_prev;
1244 if (OBD_FAIL_CHECK(OBD_FAIL_OSC_BRW_PREP_REQ))
1245 return -ENOMEM; /* Recoverable */
1246 if (OBD_FAIL_CHECK(OBD_FAIL_OSC_BRW_PREP_REQ2))
1247 return -EINVAL; /* Fatal */
1249 if ((cmd & OBD_BRW_WRITE) != 0) {
1251 req = ptlrpc_request_alloc_pool(cli->cl_import,
1253 &RQF_OST_BRW_WRITE);
1256 req = ptlrpc_request_alloc(cli->cl_import, &RQF_OST_BRW_READ);
1261 for (niocount = i = 1; i < page_count; i++) {
1262 if (!can_merge_pages(pga[i - 1], pga[i]))
1266 pill = &req->rq_pill;
1267 req_capsule_set_size(pill, &RMF_OBD_IOOBJ, RCL_CLIENT,
1269 req_capsule_set_size(pill, &RMF_NIOBUF_REMOTE, RCL_CLIENT,
1270 niocount * sizeof(*niobuf));
1272 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, opc);
1274 ptlrpc_request_free(req);
1277 req->rq_request_portal = OST_IO_PORTAL; /* bug 7198 */
1278 ptlrpc_at_set_req_timeout(req);
1279 /* ask ptlrpc not to resend on EINPROGRESS since BRWs have their own
1282 req->rq_no_retry_einprogress = 1;
1284 desc = ptlrpc_prep_bulk_imp(req, page_count,
1285 cli->cl_import->imp_connect_data.ocd_brw_size >> LNET_MTU_BITS,
1286 opc == OST_WRITE ? BULK_GET_SOURCE : BULK_PUT_SINK,
1293 /* NB request now owns desc and will free it when it gets freed */
1295 body = req_capsule_client_get(pill, &RMF_OST_BODY);
1296 ioobj = req_capsule_client_get(pill, &RMF_OBD_IOOBJ);
1297 niobuf = req_capsule_client_get(pill, &RMF_NIOBUF_REMOTE);
1298 LASSERT(body && ioobj && niobuf);
1300 lustre_set_wire_obdo(&req->rq_import->imp_connect_data, &body->oa, oa);
1302 obdo_to_ioobj(oa, ioobj);
1303 ioobj->ioo_bufcnt = niocount;
1304 /* The high bits of ioo_max_brw tells server _maximum_ number of bulks
1305 * that might be send for this request. The actual number is decided
1306 * when the RPC is finally sent in ptlrpc_register_bulk(). It sends
1307 * "max - 1" for old client compatibility sending "0", and also so the
1308 * the actual maximum is a power-of-two number, not one less. LU-1431
1310 ioobj_max_brw_set(ioobj, desc->bd_md_max_brw);
1311 LASSERT(page_count > 0);
1313 for (requested_nob = i = 0; i < page_count; i++, niobuf++) {
1314 struct brw_page *pg = pga[i];
1315 int poff = pg->off & ~CFS_PAGE_MASK;
1317 LASSERT(pg->count > 0);
1318 /* make sure there is no gap in the middle of page array */
1319 LASSERTF(page_count == 1 ||
1320 (ergo(i == 0, poff + pg->count == PAGE_CACHE_SIZE) &&
1321 ergo(i > 0 && i < page_count - 1,
1322 poff == 0 && pg->count == PAGE_CACHE_SIZE) &&
1323 ergo(i == page_count - 1, poff == 0)),
1324 "i: %d/%d pg: %p off: %llu, count: %u\n",
1325 i, page_count, pg, pg->off, pg->count);
1326 LASSERTF(i == 0 || pg->off > pg_prev->off,
1327 "i %d p_c %u pg %p [pri %lu ind %lu] off %llu prev_pg %p [pri %lu ind %lu] off %llu\n",
1329 pg->pg, page_private(pg->pg), pg->pg->index, pg->off,
1330 pg_prev->pg, page_private(pg_prev->pg),
1331 pg_prev->pg->index, pg_prev->off);
1332 LASSERT((pga[0]->flag & OBD_BRW_SRVLOCK) ==
1333 (pg->flag & OBD_BRW_SRVLOCK));
1335 ptlrpc_prep_bulk_page_pin(desc, pg->pg, poff, pg->count);
1336 requested_nob += pg->count;
1338 if (i > 0 && can_merge_pages(pg_prev, pg)) {
1340 niobuf->len += pg->count;
1342 niobuf->offset = pg->off;
1343 niobuf->len = pg->count;
1344 niobuf->flags = pg->flag;
1349 LASSERTF((void *)(niobuf - niocount) ==
1350 req_capsule_client_get(&req->rq_pill, &RMF_NIOBUF_REMOTE),
1351 "want %p - real %p\n", req_capsule_client_get(&req->rq_pill,
1352 &RMF_NIOBUF_REMOTE), (void *)(niobuf - niocount));
1354 osc_announce_cached(cli, &body->oa, opc == OST_WRITE ? requested_nob:0);
1356 if ((body->oa.o_valid & OBD_MD_FLFLAGS) == 0) {
1357 body->oa.o_valid |= OBD_MD_FLFLAGS;
1358 body->oa.o_flags = 0;
1360 body->oa.o_flags |= OBD_FL_RECOV_RESEND;
1363 if (osc_should_shrink_grant(cli))
1364 osc_shrink_grant_local(cli, &body->oa);
1366 /* size[REQ_REC_OFF] still sizeof (*body) */
1367 if (opc == OST_WRITE) {
1368 if (cli->cl_checksum &&
1369 !sptlrpc_flavor_has_bulk(&req->rq_flvr)) {
1370 /* store cl_cksum_type in a local variable since
1371 * it can be changed via lprocfs
1373 enum cksum_type cksum_type = cli->cl_cksum_type;
1375 if ((body->oa.o_valid & OBD_MD_FLFLAGS) == 0) {
1376 oa->o_flags &= OBD_FL_LOCAL_MASK;
1377 body->oa.o_flags = 0;
1379 body->oa.o_flags |= cksum_type_pack(cksum_type);
1380 body->oa.o_valid |= OBD_MD_FLCKSUM | OBD_MD_FLFLAGS;
1381 body->oa.o_cksum = osc_checksum_bulk(requested_nob,
1385 CDEBUG(D_PAGE, "checksum at write origin: %x\n",
1387 /* save this in 'oa', too, for later checking */
1388 oa->o_valid |= OBD_MD_FLCKSUM | OBD_MD_FLFLAGS;
1389 oa->o_flags |= cksum_type_pack(cksum_type);
1391 /* clear out the checksum flag, in case this is a
1392 * resend but cl_checksum is no longer set. b=11238
1394 oa->o_valid &= ~OBD_MD_FLCKSUM;
1396 oa->o_cksum = body->oa.o_cksum;
1397 /* 1 RC per niobuf */
1398 req_capsule_set_size(pill, &RMF_RCS, RCL_SERVER,
1399 sizeof(__u32) * niocount);
1401 if (cli->cl_checksum &&
1402 !sptlrpc_flavor_has_bulk(&req->rq_flvr)) {
1403 if ((body->oa.o_valid & OBD_MD_FLFLAGS) == 0)
1404 body->oa.o_flags = 0;
1405 body->oa.o_flags |= cksum_type_pack(cli->cl_cksum_type);
1406 body->oa.o_valid |= OBD_MD_FLCKSUM | OBD_MD_FLFLAGS;
1409 ptlrpc_request_set_replen(req);
1411 CLASSERT(sizeof(*aa) <= sizeof(req->rq_async_args));
1412 aa = ptlrpc_req_async_args(req);
1414 aa->aa_requested_nob = requested_nob;
1415 aa->aa_nio_count = niocount;
1416 aa->aa_page_count = page_count;
1420 INIT_LIST_HEAD(&aa->aa_oaps);
1426 ptlrpc_req_finished(req);
1430 static int check_write_checksum(struct obdo *oa, const lnet_process_id_t *peer,
1431 __u32 client_cksum, __u32 server_cksum, int nob,
1432 u32 page_count, struct brw_page **pga,
1433 enum cksum_type client_cksum_type)
1437 enum cksum_type cksum_type;
1439 if (server_cksum == client_cksum) {
1440 CDEBUG(D_PAGE, "checksum %x confirmed\n", client_cksum);
1444 cksum_type = cksum_type_unpack(oa->o_valid & OBD_MD_FLFLAGS ?
1446 new_cksum = osc_checksum_bulk(nob, page_count, pga, OST_WRITE,
1449 if (cksum_type != client_cksum_type)
1450 msg = "the server did not use the checksum type specified in the original request - likely a protocol problem"
1452 else if (new_cksum == server_cksum)
1453 msg = "changed on the client after we checksummed it - likely false positive due to mmap IO (bug 11742)"
1455 else if (new_cksum == client_cksum)
1456 msg = "changed in transit before arrival at OST";
1458 msg = "changed in transit AND doesn't match the original - likely false positive due to mmap IO (bug 11742)"
1461 LCONSOLE_ERROR_MSG(0x132, "BAD WRITE CHECKSUM: %s: from %s inode "DFID
1462 " object "DOSTID" extent [%llu-%llu]\n",
1463 msg, libcfs_nid2str(peer->nid),
1464 oa->o_valid & OBD_MD_FLFID ? oa->o_parent_seq : (__u64)0,
1465 oa->o_valid & OBD_MD_FLFID ? oa->o_parent_oid : 0,
1466 oa->o_valid & OBD_MD_FLFID ? oa->o_parent_ver : 0,
1467 POSTID(&oa->o_oi), pga[0]->off,
1468 pga[page_count-1]->off + pga[page_count-1]->count - 1);
1469 CERROR("original client csum %x (type %x), server csum %x (type %x), client csum now %x\n",
1470 client_cksum, client_cksum_type,
1471 server_cksum, cksum_type, new_cksum);
1475 /* Note rc enters this function as number of bytes transferred */
1476 static int osc_brw_fini_request(struct ptlrpc_request *req, int rc)
1478 struct osc_brw_async_args *aa = (void *)&req->rq_async_args;
1479 const lnet_process_id_t *peer =
1480 &req->rq_import->imp_connection->c_peer;
1481 struct client_obd *cli = aa->aa_cli;
1482 struct ost_body *body;
1483 __u32 client_cksum = 0;
1485 if (rc < 0 && rc != -EDQUOT) {
1486 DEBUG_REQ(D_INFO, req, "Failed request with rc = %d\n", rc);
1490 LASSERTF(req->rq_repmsg, "rc = %d\n", rc);
1491 body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
1493 DEBUG_REQ(D_INFO, req, "Can't unpack body\n");
1497 /* set/clear over quota flag for a uid/gid */
1498 if (lustre_msg_get_opc(req->rq_reqmsg) == OST_WRITE &&
1499 body->oa.o_valid & (OBD_MD_FLUSRQUOTA | OBD_MD_FLGRPQUOTA)) {
1500 unsigned int qid[MAXQUOTAS] = { body->oa.o_uid, body->oa.o_gid };
1502 CDEBUG(D_QUOTA, "setdq for [%u %u] with valid %#llx, flags %x\n",
1503 body->oa.o_uid, body->oa.o_gid, body->oa.o_valid,
1505 osc_quota_setdq(cli, qid, body->oa.o_valid, body->oa.o_flags);
1508 osc_update_grant(cli, body);
1513 if (aa->aa_oa->o_valid & OBD_MD_FLCKSUM)
1514 client_cksum = aa->aa_oa->o_cksum; /* save for later */
1516 if (lustre_msg_get_opc(req->rq_reqmsg) == OST_WRITE) {
1518 CERROR("Unexpected +ve rc %d\n", rc);
1521 LASSERT(req->rq_bulk->bd_nob == aa->aa_requested_nob);
1523 if (sptlrpc_cli_unwrap_bulk_write(req, req->rq_bulk))
1526 if ((aa->aa_oa->o_valid & OBD_MD_FLCKSUM) && client_cksum &&
1527 check_write_checksum(&body->oa, peer, client_cksum,
1528 body->oa.o_cksum, aa->aa_requested_nob,
1529 aa->aa_page_count, aa->aa_ppga,
1530 cksum_type_unpack(aa->aa_oa->o_flags)))
1533 rc = check_write_rcs(req, aa->aa_requested_nob,
1535 aa->aa_page_count, aa->aa_ppga);
1539 /* The rest of this function executes only for OST_READs */
1541 /* if unwrap_bulk failed, return -EAGAIN to retry */
1542 rc = sptlrpc_cli_unwrap_bulk_read(req, req->rq_bulk, rc);
1548 if (rc > aa->aa_requested_nob) {
1549 CERROR("Unexpected rc %d (%d requested)\n", rc,
1550 aa->aa_requested_nob);
1554 if (rc != req->rq_bulk->bd_nob_transferred) {
1555 CERROR("Unexpected rc %d (%d transferred)\n",
1556 rc, req->rq_bulk->bd_nob_transferred);
1560 if (rc < aa->aa_requested_nob)
1561 handle_short_read(rc, aa->aa_page_count, aa->aa_ppga);
1563 if (body->oa.o_valid & OBD_MD_FLCKSUM) {
1564 static int cksum_counter;
1565 __u32 server_cksum = body->oa.o_cksum;
1568 enum cksum_type cksum_type;
1570 cksum_type = cksum_type_unpack(body->oa.o_valid&OBD_MD_FLFLAGS ?
1571 body->oa.o_flags : 0);
1572 client_cksum = osc_checksum_bulk(rc, aa->aa_page_count,
1573 aa->aa_ppga, OST_READ,
1576 if (peer->nid != req->rq_bulk->bd_sender) {
1578 router = libcfs_nid2str(req->rq_bulk->bd_sender);
1581 if (server_cksum != client_cksum) {
1582 LCONSOLE_ERROR_MSG(0x133, "%s: BAD READ CHECKSUM: from %s%s%s inode " DFID " object " DOSTID " extent [%llu-%llu]\n",
1583 req->rq_import->imp_obd->obd_name,
1584 libcfs_nid2str(peer->nid),
1586 body->oa.o_valid & OBD_MD_FLFID ?
1587 body->oa.o_parent_seq : (__u64)0,
1588 body->oa.o_valid & OBD_MD_FLFID ?
1589 body->oa.o_parent_oid : 0,
1590 body->oa.o_valid & OBD_MD_FLFID ?
1591 body->oa.o_parent_ver : 0,
1592 POSTID(&body->oa.o_oi),
1593 aa->aa_ppga[0]->off,
1594 aa->aa_ppga[aa->aa_page_count-1]->off +
1595 aa->aa_ppga[aa->aa_page_count-1]->count -
1597 CERROR("client %x, server %x, cksum_type %x\n",
1598 client_cksum, server_cksum, cksum_type);
1600 aa->aa_oa->o_cksum = client_cksum;
1604 CDEBUG(D_PAGE, "checksum %x confirmed\n", client_cksum);
1607 } else if (unlikely(client_cksum)) {
1608 static int cksum_missed;
1611 if ((cksum_missed & (-cksum_missed)) == cksum_missed)
1612 CERROR("Checksum %u requested from %s but not sent\n",
1613 cksum_missed, libcfs_nid2str(peer->nid));
1619 lustre_get_wire_obdo(&req->rq_import->imp_connect_data,
1620 aa->aa_oa, &body->oa);
1625 static int osc_brw_redo_request(struct ptlrpc_request *request,
1626 struct osc_brw_async_args *aa, int rc)
1628 struct ptlrpc_request *new_req;
1629 struct osc_brw_async_args *new_aa;
1630 struct osc_async_page *oap;
1632 DEBUG_REQ(rc == -EINPROGRESS ? D_RPCTRACE : D_ERROR, request,
1633 "redo for recoverable error %d", rc);
1635 rc = osc_brw_prep_request(lustre_msg_get_opc(request->rq_reqmsg) ==
1636 OST_WRITE ? OBD_BRW_WRITE : OBD_BRW_READ,
1637 aa->aa_cli, aa->aa_oa,
1638 NULL /* lsm unused by osc currently */,
1639 aa->aa_page_count, aa->aa_ppga,
1644 list_for_each_entry(oap, &aa->aa_oaps, oap_rpc_item) {
1645 if (oap->oap_request) {
1646 LASSERTF(request == oap->oap_request,
1647 "request %p != oap_request %p\n",
1648 request, oap->oap_request);
1649 if (oap->oap_interrupted) {
1650 ptlrpc_req_finished(new_req);
1655 /* New request takes over pga and oaps from old request.
1656 * Note that copying a list_head doesn't work, need to move it...
1659 new_req->rq_interpret_reply = request->rq_interpret_reply;
1660 new_req->rq_async_args = request->rq_async_args;
1661 /* cap resend delay to the current request timeout, this is similar to
1662 * what ptlrpc does (see after_reply())
1664 if (aa->aa_resends > new_req->rq_timeout)
1665 new_req->rq_sent = ktime_get_real_seconds() + new_req->rq_timeout;
1667 new_req->rq_sent = ktime_get_real_seconds() + aa->aa_resends;
1668 new_req->rq_generation_set = 1;
1669 new_req->rq_import_generation = request->rq_import_generation;
1671 new_aa = ptlrpc_req_async_args(new_req);
1673 INIT_LIST_HEAD(&new_aa->aa_oaps);
1674 list_splice_init(&aa->aa_oaps, &new_aa->aa_oaps);
1675 INIT_LIST_HEAD(&new_aa->aa_exts);
1676 list_splice_init(&aa->aa_exts, &new_aa->aa_exts);
1677 new_aa->aa_resends = aa->aa_resends;
1679 list_for_each_entry(oap, &new_aa->aa_oaps, oap_rpc_item) {
1680 if (oap->oap_request) {
1681 ptlrpc_req_finished(oap->oap_request);
1682 oap->oap_request = ptlrpc_request_addref(new_req);
1686 /* XXX: This code will run into problem if we're going to support
1687 * to add a series of BRW RPCs into a self-defined ptlrpc_request_set
1688 * and wait for all of them to be finished. We should inherit request
1689 * set from old request.
1691 ptlrpcd_add_req(new_req);
1693 DEBUG_REQ(D_INFO, new_req, "new request");
1698 * ugh, we want disk allocation on the target to happen in offset order. we'll
1699 * follow sedgewicks advice and stick to the dead simple shellsort -- it'll do
1700 * fine for our small page arrays and doesn't require allocation. its an
1701 * insertion sort that swaps elements that are strides apart, shrinking the
1702 * stride down until its '1' and the array is sorted.
1704 static void sort_brw_pages(struct brw_page **array, int num)
1707 struct brw_page *tmp;
1711 for (stride = 1; stride < num ; stride = (stride * 3) + 1)
1716 for (i = stride ; i < num ; i++) {
1719 while (j >= stride && array[j - stride]->off > tmp->off) {
1720 array[j] = array[j - stride];
1725 } while (stride > 1);
1728 static void osc_release_ppga(struct brw_page **ppga, u32 count)
1734 static int brw_interpret(const struct lu_env *env,
1735 struct ptlrpc_request *req, void *data, int rc)
1737 struct osc_brw_async_args *aa = data;
1738 struct osc_extent *ext;
1739 struct osc_extent *tmp;
1740 struct cl_object *obj = NULL;
1741 struct client_obd *cli = aa->aa_cli;
1743 rc = osc_brw_fini_request(req, rc);
1744 CDEBUG(D_INODE, "request %p aa %p rc %d\n", req, aa, rc);
1745 /* When server return -EINPROGRESS, client should always retry
1746 * regardless of the number of times the bulk was resent already.
1748 if (osc_recoverable_error(rc)) {
1749 if (req->rq_import_generation !=
1750 req->rq_import->imp_generation) {
1751 CDEBUG(D_HA, "%s: resend cross eviction for object: " DOSTID ", rc = %d.\n",
1752 req->rq_import->imp_obd->obd_name,
1753 POSTID(&aa->aa_oa->o_oi), rc);
1754 } else if (rc == -EINPROGRESS ||
1755 client_should_resend(aa->aa_resends, aa->aa_cli)) {
1756 rc = osc_brw_redo_request(req, aa, rc);
1758 CERROR("%s: too many resent retries for object: %llu:%llu, rc = %d.\n",
1759 req->rq_import->imp_obd->obd_name,
1760 POSTID(&aa->aa_oa->o_oi), rc);
1765 else if (rc == -EAGAIN || rc == -EINPROGRESS)
1769 list_for_each_entry_safe(ext, tmp, &aa->aa_exts, oe_link) {
1770 if (!obj && rc == 0) {
1771 obj = osc2cl(ext->oe_obj);
1775 list_del_init(&ext->oe_link);
1776 osc_extent_finish(env, ext, 1, rc);
1778 LASSERT(list_empty(&aa->aa_exts));
1779 LASSERT(list_empty(&aa->aa_oaps));
1782 struct obdo *oa = aa->aa_oa;
1783 struct cl_attr *attr = &osc_env_info(env)->oti_attr;
1784 unsigned long valid = 0;
1787 if (oa->o_valid & OBD_MD_FLBLOCKS) {
1788 attr->cat_blocks = oa->o_blocks;
1789 valid |= CAT_BLOCKS;
1791 if (oa->o_valid & OBD_MD_FLMTIME) {
1792 attr->cat_mtime = oa->o_mtime;
1795 if (oa->o_valid & OBD_MD_FLATIME) {
1796 attr->cat_atime = oa->o_atime;
1799 if (oa->o_valid & OBD_MD_FLCTIME) {
1800 attr->cat_ctime = oa->o_ctime;
1804 cl_object_attr_lock(obj);
1805 cl_object_attr_set(env, obj, attr, valid);
1806 cl_object_attr_unlock(obj);
1808 cl_object_put(env, obj);
1810 kmem_cache_free(obdo_cachep, aa->aa_oa);
1812 cl_req_completion(env, aa->aa_clerq, rc < 0 ? rc :
1813 req->rq_bulk->bd_nob_transferred);
1814 osc_release_ppga(aa->aa_ppga, aa->aa_page_count);
1815 ptlrpc_lprocfs_brw(req, req->rq_bulk->bd_nob_transferred);
1817 client_obd_list_lock(&cli->cl_loi_list_lock);
1818 /* We need to decrement before osc_ap_completion->osc_wake_cache_waiters
1819 * is called so we know whether to go to sync BRWs or wait for more
1822 if (lustre_msg_get_opc(req->rq_reqmsg) == OST_WRITE)
1823 cli->cl_w_in_flight--;
1825 cli->cl_r_in_flight--;
1826 osc_wake_cache_waiters(cli);
1827 client_obd_list_unlock(&cli->cl_loi_list_lock);
1829 osc_io_unplug(env, cli, NULL);
1834 * Build an RPC by the list of extent @ext_list. The caller must ensure
1835 * that the total pages in this list are NOT over max pages per RPC.
1836 * Extents in the list must be in OES_RPC state.
1838 int osc_build_rpc(const struct lu_env *env, struct client_obd *cli,
1839 struct list_head *ext_list, int cmd)
1841 struct ptlrpc_request *req = NULL;
1842 struct osc_extent *ext;
1843 struct brw_page **pga = NULL;
1844 struct osc_brw_async_args *aa = NULL;
1845 struct obdo *oa = NULL;
1846 struct osc_async_page *oap;
1847 struct osc_async_page *tmp;
1848 struct cl_req *clerq = NULL;
1849 enum cl_req_type crt = (cmd & OBD_BRW_WRITE) ? CRT_WRITE : CRT_READ;
1850 struct ldlm_lock *lock = NULL;
1851 struct cl_req_attr *crattr = NULL;
1852 u64 starting_offset = OBD_OBJECT_EOF;
1853 u64 ending_offset = 0;
1859 struct ost_body *body;
1860 LIST_HEAD(rpc_list);
1862 LASSERT(!list_empty(ext_list));
1864 /* add pages into rpc_list to build BRW rpc */
1865 list_for_each_entry(ext, ext_list, oe_link) {
1866 LASSERT(ext->oe_state == OES_RPC);
1867 mem_tight |= ext->oe_memalloc;
1868 list_for_each_entry(oap, &ext->oe_pages, oap_pending_item) {
1870 list_add_tail(&oap->oap_rpc_item, &rpc_list);
1871 if (starting_offset > oap->oap_obj_off)
1872 starting_offset = oap->oap_obj_off;
1874 LASSERT(oap->oap_page_off == 0);
1875 if (ending_offset < oap->oap_obj_off + oap->oap_count)
1876 ending_offset = oap->oap_obj_off +
1879 LASSERT(oap->oap_page_off + oap->oap_count ==
1885 mpflag = cfs_memory_pressure_get_and_set();
1887 crattr = kzalloc(sizeof(*crattr), GFP_NOFS);
1893 pga = kcalloc(page_count, sizeof(*pga), GFP_NOFS);
1899 oa = kmem_cache_zalloc(obdo_cachep, GFP_NOFS);
1906 list_for_each_entry(oap, &rpc_list, oap_rpc_item) {
1907 struct cl_page *page = oap2cl_page(oap);
1910 clerq = cl_req_alloc(env, page, crt,
1911 1 /* only 1-object rpcs for now */);
1912 if (IS_ERR(clerq)) {
1913 rc = PTR_ERR(clerq);
1916 lock = oap->oap_ldlm_lock;
1919 oap->oap_brw_flags |= OBD_BRW_MEMALLOC;
1920 pga[i] = &oap->oap_brw_page;
1921 pga[i]->off = oap->oap_obj_off + oap->oap_page_off;
1922 CDEBUG(0, "put page %p index %lu oap %p flg %x to pga\n",
1923 pga[i]->pg, page_index(oap->oap_page), oap,
1926 cl_req_page_add(env, clerq, page);
1929 /* always get the data for the obdo for the rpc */
1931 crattr->cra_oa = oa;
1932 cl_req_attr_set(env, clerq, crattr, ~0ULL);
1934 oa->o_handle = lock->l_remote_handle;
1935 oa->o_valid |= OBD_MD_FLHANDLE;
1938 rc = cl_req_prep(env, clerq);
1940 CERROR("cl_req_prep failed: %d\n", rc);
1944 sort_brw_pages(pga, page_count);
1945 rc = osc_brw_prep_request(cmd, cli, oa, NULL, page_count,
1948 CERROR("prep_req failed: %d\n", rc);
1952 req->rq_interpret_reply = brw_interpret;
1955 req->rq_memalloc = 1;
1957 /* Need to update the timestamps after the request is built in case
1958 * we race with setattr (locally or in queue at OST). If OST gets
1959 * later setattr before earlier BRW (as determined by the request xid),
1960 * the OST will not use BRW timestamps. Sadly, there is no obvious
1961 * way to do this in a single call. bug 10150
1963 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
1964 crattr->cra_oa = &body->oa;
1965 cl_req_attr_set(env, clerq, crattr,
1966 OBD_MD_FLMTIME|OBD_MD_FLCTIME|OBD_MD_FLATIME);
1968 lustre_msg_set_jobid(req->rq_reqmsg, crattr->cra_jobid);
1970 CLASSERT(sizeof(*aa) <= sizeof(req->rq_async_args));
1971 aa = ptlrpc_req_async_args(req);
1972 INIT_LIST_HEAD(&aa->aa_oaps);
1973 list_splice_init(&rpc_list, &aa->aa_oaps);
1974 INIT_LIST_HEAD(&aa->aa_exts);
1975 list_splice_init(ext_list, &aa->aa_exts);
1976 aa->aa_clerq = clerq;
1978 /* queued sync pages can be torn down while the pages
1979 * were between the pending list and the rpc
1982 list_for_each_entry(oap, &aa->aa_oaps, oap_rpc_item) {
1983 /* only one oap gets a request reference */
1986 if (oap->oap_interrupted && !req->rq_intr) {
1987 CDEBUG(D_INODE, "oap %p in req %p interrupted\n",
1989 ptlrpc_mark_interrupted(req);
1993 tmp->oap_request = ptlrpc_request_addref(req);
1995 client_obd_list_lock(&cli->cl_loi_list_lock);
1996 starting_offset >>= PAGE_CACHE_SHIFT;
1997 if (cmd == OBD_BRW_READ) {
1998 cli->cl_r_in_flight++;
1999 lprocfs_oh_tally_log2(&cli->cl_read_page_hist, page_count);
2000 lprocfs_oh_tally(&cli->cl_read_rpc_hist, cli->cl_r_in_flight);
2001 lprocfs_oh_tally_log2(&cli->cl_read_offset_hist,
2002 starting_offset + 1);
2004 cli->cl_w_in_flight++;
2005 lprocfs_oh_tally_log2(&cli->cl_write_page_hist, page_count);
2006 lprocfs_oh_tally(&cli->cl_write_rpc_hist, cli->cl_w_in_flight);
2007 lprocfs_oh_tally_log2(&cli->cl_write_offset_hist,
2008 starting_offset + 1);
2010 client_obd_list_unlock(&cli->cl_loi_list_lock);
2012 DEBUG_REQ(D_INODE, req, "%d pages, aa %p. now %dr/%dw in flight",
2013 page_count, aa, cli->cl_r_in_flight,
2014 cli->cl_w_in_flight);
2016 ptlrpcd_add_req(req);
2021 cfs_memory_pressure_restore(mpflag);
2029 kmem_cache_free(obdo_cachep, oa);
2031 /* this should happen rarely and is pretty bad, it makes the
2032 * pending list not follow the dirty order
2034 while (!list_empty(ext_list)) {
2035 ext = list_entry(ext_list->next, struct osc_extent,
2037 list_del_init(&ext->oe_link);
2038 osc_extent_finish(env, ext, 0, rc);
2040 if (clerq && !IS_ERR(clerq))
2041 cl_req_completion(env, clerq, rc);
2046 static int osc_set_lock_data_with_check(struct ldlm_lock *lock,
2047 struct ldlm_enqueue_info *einfo)
2049 void *data = einfo->ei_cbdata;
2052 LASSERT(lock->l_blocking_ast == einfo->ei_cb_bl);
2053 LASSERT(lock->l_resource->lr_type == einfo->ei_type);
2054 LASSERT(lock->l_completion_ast == einfo->ei_cb_cp);
2055 LASSERT(lock->l_glimpse_ast == einfo->ei_cb_gl);
2057 lock_res_and_lock(lock);
2058 spin_lock(&osc_ast_guard);
2060 if (!lock->l_ast_data)
2061 lock->l_ast_data = data;
2062 if (lock->l_ast_data == data)
2065 spin_unlock(&osc_ast_guard);
2066 unlock_res_and_lock(lock);
2071 static int osc_set_data_with_check(struct lustre_handle *lockh,
2072 struct ldlm_enqueue_info *einfo)
2074 struct ldlm_lock *lock = ldlm_handle2lock(lockh);
2078 set = osc_set_lock_data_with_check(lock, einfo);
2079 LDLM_LOCK_PUT(lock);
2081 CERROR("lockh %p, data %p - client evicted?\n",
2082 lockh, einfo->ei_cbdata);
2086 /* find any ldlm lock of the inode in osc
2091 static int osc_find_cbdata(struct obd_export *exp, struct lov_stripe_md *lsm,
2092 ldlm_iterator_t replace, void *data)
2094 struct ldlm_res_id res_id;
2095 struct obd_device *obd = class_exp2obd(exp);
2098 ostid_build_res_name(&lsm->lsm_oi, &res_id);
2099 rc = ldlm_resource_iterate(obd->obd_namespace, &res_id, replace, data);
2100 if (rc == LDLM_ITER_STOP)
2102 if (rc == LDLM_ITER_CONTINUE)
2107 static int osc_enqueue_fini(struct ptlrpc_request *req, struct ost_lvb *lvb,
2108 obd_enqueue_update_f upcall, void *cookie,
2109 __u64 *flags, int agl, int rc)
2111 int intent = *flags & LDLM_FL_HAS_INTENT;
2114 /* The request was created before ldlm_cli_enqueue call. */
2115 if (rc == ELDLM_LOCK_ABORTED) {
2116 struct ldlm_reply *rep;
2118 rep = req_capsule_server_get(&req->rq_pill,
2121 rep->lock_policy_res1 =
2122 ptlrpc_status_ntoh(rep->lock_policy_res1);
2123 if (rep->lock_policy_res1)
2124 rc = rep->lock_policy_res1;
2128 if ((intent != 0 && rc == ELDLM_LOCK_ABORTED && agl == 0) ||
2130 *flags |= LDLM_FL_LVB_READY;
2131 CDEBUG(D_INODE, "got kms %llu blocks %llu mtime %llu\n",
2132 lvb->lvb_size, lvb->lvb_blocks, lvb->lvb_mtime);
2135 /* Call the update callback. */
2136 rc = (*upcall)(cookie, rc);
2140 static int osc_enqueue_interpret(const struct lu_env *env,
2141 struct ptlrpc_request *req,
2142 struct osc_enqueue_args *aa, int rc)
2144 struct ldlm_lock *lock;
2145 struct lustre_handle handle;
2147 struct ost_lvb *lvb;
2149 __u64 *flags = aa->oa_flags;
2151 /* Make a local copy of a lock handle and a mode, because aa->oa_*
2152 * might be freed anytime after lock upcall has been called.
2154 lustre_handle_copy(&handle, aa->oa_lockh);
2155 mode = aa->oa_ei->ei_mode;
2157 /* ldlm_cli_enqueue is holding a reference on the lock, so it must
2160 lock = ldlm_handle2lock(&handle);
2162 /* Take an additional reference so that a blocking AST that
2163 * ldlm_cli_enqueue_fini() might post for a failed lock, is guaranteed
2164 * to arrive after an upcall has been executed by
2165 * osc_enqueue_fini().
2167 ldlm_lock_addref(&handle, mode);
2169 /* Let CP AST to grant the lock first. */
2170 OBD_FAIL_TIMEOUT(OBD_FAIL_OSC_CP_ENQ_RACE, 1);
2172 if (aa->oa_agl && rc == ELDLM_LOCK_ABORTED) {
2177 lvb_len = sizeof(*aa->oa_lvb);
2180 /* Complete obtaining the lock procedure. */
2181 rc = ldlm_cli_enqueue_fini(aa->oa_exp, req, aa->oa_ei->ei_type, 1,
2182 mode, flags, lvb, lvb_len, &handle, rc);
2183 /* Complete osc stuff. */
2184 rc = osc_enqueue_fini(req, aa->oa_lvb, aa->oa_upcall, aa->oa_cookie,
2185 flags, aa->oa_agl, rc);
2187 OBD_FAIL_TIMEOUT(OBD_FAIL_OSC_CP_CANCEL_RACE, 10);
2189 /* Release the lock for async request. */
2190 if (lustre_handle_is_used(&handle) && rc == ELDLM_OK)
2192 * Releases a reference taken by ldlm_cli_enqueue(), if it is
2193 * not already released by
2194 * ldlm_cli_enqueue_fini()->failed_lock_cleanup()
2196 ldlm_lock_decref(&handle, mode);
2198 LASSERTF(lock, "lockh %p, req %p, aa %p - client evicted?\n",
2199 aa->oa_lockh, req, aa);
2200 ldlm_lock_decref(&handle, mode);
2201 LDLM_LOCK_PUT(lock);
2205 struct ptlrpc_request_set *PTLRPCD_SET = (void *)1;
2207 /* When enqueuing asynchronously, locks are not ordered, we can obtain a lock
2208 * from the 2nd OSC before a lock from the 1st one. This does not deadlock with
2209 * other synchronous requests, however keeping some locks and trying to obtain
2210 * others may take a considerable amount of time in a case of ost failure; and
2211 * when other sync requests do not get released lock from a client, the client
2212 * is excluded from the cluster -- such scenarious make the life difficult, so
2213 * release locks just after they are obtained.
2215 int osc_enqueue_base(struct obd_export *exp, struct ldlm_res_id *res_id,
2216 __u64 *flags, ldlm_policy_data_t *policy,
2217 struct ost_lvb *lvb, int kms_valid,
2218 obd_enqueue_update_f upcall, void *cookie,
2219 struct ldlm_enqueue_info *einfo,
2220 struct lustre_handle *lockh,
2221 struct ptlrpc_request_set *rqset, int async, int agl)
2223 struct obd_device *obd = exp->exp_obd;
2224 struct ptlrpc_request *req = NULL;
2225 int intent = *flags & LDLM_FL_HAS_INTENT;
2226 __u64 match_lvb = (agl != 0 ? 0 : LDLM_FL_LVB_READY);
2227 enum ldlm_mode mode;
2230 /* Filesystem lock extents are extended to page boundaries so that
2231 * dealing with the page cache is a little smoother.
2233 policy->l_extent.start -= policy->l_extent.start & ~CFS_PAGE_MASK;
2234 policy->l_extent.end |= ~CFS_PAGE_MASK;
2237 * kms is not valid when either object is completely fresh (so that no
2238 * locks are cached), or object was evicted. In the latter case cached
2239 * lock cannot be used, because it would prime inode state with
2240 * potentially stale LVB.
2245 /* Next, search for already existing extent locks that will cover us */
2246 /* If we're trying to read, we also search for an existing PW lock. The
2247 * VFS and page cache already protect us locally, so lots of readers/
2248 * writers can share a single PW lock.
2250 * There are problems with conversion deadlocks, so instead of
2251 * converting a read lock to a write lock, we'll just enqueue a new
2254 * At some point we should cancel the read lock instead of making them
2255 * send us a blocking callback, but there are problems with canceling
2256 * locks out from other users right now, too.
2258 mode = einfo->ei_mode;
2259 if (einfo->ei_mode == LCK_PR)
2261 mode = ldlm_lock_match(obd->obd_namespace, *flags | match_lvb, res_id,
2262 einfo->ei_type, policy, mode, lockh, 0);
2264 struct ldlm_lock *matched = ldlm_handle2lock(lockh);
2266 if ((agl != 0) && !(matched->l_flags & LDLM_FL_LVB_READY)) {
2267 /* For AGL, if enqueue RPC is sent but the lock is not
2268 * granted, then skip to process this strpe.
2269 * Return -ECANCELED to tell the caller.
2271 ldlm_lock_decref(lockh, mode);
2272 LDLM_LOCK_PUT(matched);
2276 if (osc_set_lock_data_with_check(matched, einfo)) {
2277 *flags |= LDLM_FL_LVB_READY;
2278 /* addref the lock only if not async requests and PW
2279 * lock is matched whereas we asked for PR.
2281 if (!rqset && einfo->ei_mode != mode)
2282 ldlm_lock_addref(lockh, LCK_PR);
2284 /* I would like to be able to ASSERT here that
2285 * rss <= kms, but I can't, for reasons which
2286 * are explained in lov_enqueue()
2290 /* We already have a lock, and it's referenced.
2292 * At this point, the cl_lock::cll_state is CLS_QUEUING,
2293 * AGL upcall may change it to CLS_HELD directly.
2295 (*upcall)(cookie, ELDLM_OK);
2297 if (einfo->ei_mode != mode)
2298 ldlm_lock_decref(lockh, LCK_PW);
2300 /* For async requests, decref the lock. */
2301 ldlm_lock_decref(lockh, einfo->ei_mode);
2302 LDLM_LOCK_PUT(matched);
2306 ldlm_lock_decref(lockh, mode);
2307 LDLM_LOCK_PUT(matched);
2314 req = ptlrpc_request_alloc(class_exp2cliimp(exp),
2315 &RQF_LDLM_ENQUEUE_LVB);
2319 rc = ldlm_prep_enqueue_req(exp, req, &cancels, 0);
2321 ptlrpc_request_free(req);
2325 req_capsule_set_size(&req->rq_pill, &RMF_DLM_LVB, RCL_SERVER,
2327 ptlrpc_request_set_replen(req);
2330 /* users of osc_enqueue() can pass this flag for ldlm_lock_match() */
2331 *flags &= ~LDLM_FL_BLOCK_GRANTED;
2333 rc = ldlm_cli_enqueue(exp, &req, einfo, res_id, policy, flags, lvb,
2334 sizeof(*lvb), LVB_T_OST, lockh, async);
2337 struct osc_enqueue_args *aa;
2339 CLASSERT (sizeof(*aa) <= sizeof(req->rq_async_args));
2340 aa = ptlrpc_req_async_args(req);
2343 aa->oa_flags = flags;
2344 aa->oa_upcall = upcall;
2345 aa->oa_cookie = cookie;
2347 aa->oa_lockh = lockh;
2350 req->rq_interpret_reply =
2351 (ptlrpc_interpterer_t)osc_enqueue_interpret;
2352 if (rqset == PTLRPCD_SET)
2353 ptlrpcd_add_req(req);
2355 ptlrpc_set_add_req(rqset, req);
2356 } else if (intent) {
2357 ptlrpc_req_finished(req);
2362 rc = osc_enqueue_fini(req, lvb, upcall, cookie, flags, agl, rc);
2364 ptlrpc_req_finished(req);
2369 int osc_match_base(struct obd_export *exp, struct ldlm_res_id *res_id,
2370 __u32 type, ldlm_policy_data_t *policy, __u32 mode,
2371 __u64 *flags, void *data, struct lustre_handle *lockh,
2374 struct obd_device *obd = exp->exp_obd;
2375 __u64 lflags = *flags;
2378 if (OBD_FAIL_CHECK(OBD_FAIL_OSC_MATCH))
2381 /* Filesystem lock extents are extended to page boundaries so that
2382 * dealing with the page cache is a little smoother
2384 policy->l_extent.start -= policy->l_extent.start & ~CFS_PAGE_MASK;
2385 policy->l_extent.end |= ~CFS_PAGE_MASK;
2387 /* Next, search for already existing extent locks that will cover us */
2388 /* If we're trying to read, we also search for an existing PW lock. The
2389 * VFS and page cache already protect us locally, so lots of readers/
2390 * writers can share a single PW lock.
2395 rc = ldlm_lock_match(obd->obd_namespace, lflags,
2396 res_id, type, policy, rc, lockh, unref);
2399 if (!osc_set_data_with_check(lockh, data)) {
2400 if (!(lflags & LDLM_FL_TEST_LOCK))
2401 ldlm_lock_decref(lockh, rc);
2405 if (!(lflags & LDLM_FL_TEST_LOCK) && mode != rc) {
2406 ldlm_lock_addref(lockh, LCK_PR);
2407 ldlm_lock_decref(lockh, LCK_PW);
2414 int osc_cancel_base(struct lustre_handle *lockh, __u32 mode)
2416 if (unlikely(mode == LCK_GROUP))
2417 ldlm_lock_decref_and_cancel(lockh, mode);
2419 ldlm_lock_decref(lockh, mode);
2424 static int osc_statfs_interpret(const struct lu_env *env,
2425 struct ptlrpc_request *req,
2426 struct osc_async_args *aa, int rc)
2428 struct obd_statfs *msfs;
2431 /* The request has in fact never been sent
2432 * due to issues at a higher level (LOV).
2433 * Exit immediately since the caller is
2434 * aware of the problem and takes care
2439 if ((rc == -ENOTCONN || rc == -EAGAIN) &&
2440 (aa->aa_oi->oi_flags & OBD_STATFS_NODELAY)) {
2448 msfs = req_capsule_server_get(&req->rq_pill, &RMF_OBD_STATFS);
2454 *aa->aa_oi->oi_osfs = *msfs;
2456 rc = aa->aa_oi->oi_cb_up(aa->aa_oi, rc);
2460 static int osc_statfs_async(struct obd_export *exp,
2461 struct obd_info *oinfo, __u64 max_age,
2462 struct ptlrpc_request_set *rqset)
2464 struct obd_device *obd = class_exp2obd(exp);
2465 struct ptlrpc_request *req;
2466 struct osc_async_args *aa;
2469 /* We could possibly pass max_age in the request (as an absolute
2470 * timestamp or a "seconds.usec ago") so the target can avoid doing
2471 * extra calls into the filesystem if that isn't necessary (e.g.
2472 * during mount that would help a bit). Having relative timestamps
2473 * is not so great if request processing is slow, while absolute
2474 * timestamps are not ideal because they need time synchronization.
2476 req = ptlrpc_request_alloc(obd->u.cli.cl_import, &RQF_OST_STATFS);
2480 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_STATFS);
2482 ptlrpc_request_free(req);
2485 ptlrpc_request_set_replen(req);
2486 req->rq_request_portal = OST_CREATE_PORTAL;
2487 ptlrpc_at_set_req_timeout(req);
2489 if (oinfo->oi_flags & OBD_STATFS_NODELAY) {
2490 /* procfs requests not want stat in wait for avoid deadlock */
2491 req->rq_no_resend = 1;
2492 req->rq_no_delay = 1;
2495 req->rq_interpret_reply = (ptlrpc_interpterer_t)osc_statfs_interpret;
2496 CLASSERT (sizeof(*aa) <= sizeof(req->rq_async_args));
2497 aa = ptlrpc_req_async_args(req);
2500 ptlrpc_set_add_req(rqset, req);
2504 static int osc_statfs(const struct lu_env *env, struct obd_export *exp,
2505 struct obd_statfs *osfs, __u64 max_age, __u32 flags)
2507 struct obd_device *obd = class_exp2obd(exp);
2508 struct obd_statfs *msfs;
2509 struct ptlrpc_request *req;
2510 struct obd_import *imp = NULL;
2513 /* Since the request might also come from lprocfs, so we need
2514 * sync this with client_disconnect_export Bug15684
2516 down_read(&obd->u.cli.cl_sem);
2517 if (obd->u.cli.cl_import)
2518 imp = class_import_get(obd->u.cli.cl_import);
2519 up_read(&obd->u.cli.cl_sem);
2523 /* We could possibly pass max_age in the request (as an absolute
2524 * timestamp or a "seconds.usec ago") so the target can avoid doing
2525 * extra calls into the filesystem if that isn't necessary (e.g.
2526 * during mount that would help a bit). Having relative timestamps
2527 * is not so great if request processing is slow, while absolute
2528 * timestamps are not ideal because they need time synchronization.
2530 req = ptlrpc_request_alloc(imp, &RQF_OST_STATFS);
2532 class_import_put(imp);
2537 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_STATFS);
2539 ptlrpc_request_free(req);
2542 ptlrpc_request_set_replen(req);
2543 req->rq_request_portal = OST_CREATE_PORTAL;
2544 ptlrpc_at_set_req_timeout(req);
2546 if (flags & OBD_STATFS_NODELAY) {
2547 /* procfs requests not want stat in wait for avoid deadlock */
2548 req->rq_no_resend = 1;
2549 req->rq_no_delay = 1;
2552 rc = ptlrpc_queue_wait(req);
2556 msfs = req_capsule_server_get(&req->rq_pill, &RMF_OBD_STATFS);
2565 ptlrpc_req_finished(req);
2569 /* Retrieve object striping information.
2571 * @lmmu is a pointer to an in-core struct with lmm_ost_count indicating
2572 * the maximum number of OST indices which will fit in the user buffer.
2573 * lmm_magic must be LOV_MAGIC (we only use 1 slot here).
2575 static int osc_getstripe(struct lov_stripe_md *lsm,
2576 struct lov_user_md __user *lump)
2578 /* we use lov_user_md_v3 because it is larger than lov_user_md_v1 */
2579 struct lov_user_md_v3 lum, *lumk;
2580 struct lov_user_ost_data_v1 *lmm_objects;
2581 int rc = 0, lum_size;
2586 /* we only need the header part from user space to get lmm_magic and
2587 * lmm_stripe_count, (the header part is common to v1 and v3)
2589 lum_size = sizeof(struct lov_user_md_v1);
2590 if (copy_from_user(&lum, lump, lum_size))
2593 if ((lum.lmm_magic != LOV_USER_MAGIC_V1) &&
2594 (lum.lmm_magic != LOV_USER_MAGIC_V3))
2597 /* lov_user_md_vX and lov_mds_md_vX must have the same size */
2598 LASSERT(sizeof(struct lov_user_md_v1) == sizeof(struct lov_mds_md_v1));
2599 LASSERT(sizeof(struct lov_user_md_v3) == sizeof(struct lov_mds_md_v3));
2600 LASSERT(sizeof(lum.lmm_objects[0]) == sizeof(lumk->lmm_objects[0]));
2602 /* we can use lov_mds_md_size() to compute lum_size
2603 * because lov_user_md_vX and lov_mds_md_vX have the same size
2605 if (lum.lmm_stripe_count > 0) {
2606 lum_size = lov_mds_md_size(lum.lmm_stripe_count, lum.lmm_magic);
2607 lumk = kzalloc(lum_size, GFP_NOFS);
2611 if (lum.lmm_magic == LOV_USER_MAGIC_V1)
2613 &(((struct lov_user_md_v1 *)lumk)->lmm_objects[0]);
2615 lmm_objects = &(lumk->lmm_objects[0]);
2616 lmm_objects->l_ost_oi = lsm->lsm_oi;
2618 lum_size = lov_mds_md_size(0, lum.lmm_magic);
2622 lumk->lmm_oi = lsm->lsm_oi;
2623 lumk->lmm_stripe_count = 1;
2625 if (copy_to_user(lump, lumk, lum_size))
2634 static int osc_iocontrol(unsigned int cmd, struct obd_export *exp, int len,
2635 void *karg, void __user *uarg)
2637 struct obd_device *obd = exp->exp_obd;
2638 struct obd_ioctl_data *data = karg;
2641 if (!try_module_get(THIS_MODULE)) {
2642 CERROR("%s: cannot get module '%s'\n", obd->obd_name,
2643 module_name(THIS_MODULE));
2647 case OBD_IOC_LOV_GET_CONFIG: {
2649 struct lov_desc *desc;
2650 struct obd_uuid uuid;
2654 if (obd_ioctl_getdata(&buf, &len, uarg)) {
2659 data = (struct obd_ioctl_data *)buf;
2661 if (sizeof(*desc) > data->ioc_inllen1) {
2662 obd_ioctl_freedata(buf, len);
2667 if (data->ioc_inllen2 < sizeof(uuid)) {
2668 obd_ioctl_freedata(buf, len);
2673 desc = (struct lov_desc *)data->ioc_inlbuf1;
2674 desc->ld_tgt_count = 1;
2675 desc->ld_active_tgt_count = 1;
2676 desc->ld_default_stripe_count = 1;
2677 desc->ld_default_stripe_size = 0;
2678 desc->ld_default_stripe_offset = 0;
2679 desc->ld_pattern = 0;
2680 memcpy(&desc->ld_uuid, &obd->obd_uuid, sizeof(uuid));
2682 memcpy(data->ioc_inlbuf2, &obd->obd_uuid, sizeof(uuid));
2684 err = copy_to_user(uarg, buf, len);
2687 obd_ioctl_freedata(buf, len);
2690 case LL_IOC_LOV_SETSTRIPE:
2691 err = obd_alloc_memmd(exp, karg);
2695 case LL_IOC_LOV_GETSTRIPE:
2696 err = osc_getstripe(karg, uarg);
2698 case OBD_IOC_CLIENT_RECOVER:
2699 err = ptlrpc_recover_import(obd->u.cli.cl_import,
2700 data->ioc_inlbuf1, 0);
2704 case IOC_OSC_SET_ACTIVE:
2705 err = ptlrpc_set_import_active(obd->u.cli.cl_import,
2708 case OBD_IOC_POLL_QUOTACHECK:
2709 err = osc_quota_poll_check(exp, karg);
2711 case OBD_IOC_PING_TARGET:
2712 err = ptlrpc_obd_ping(obd);
2715 CDEBUG(D_INODE, "unrecognised ioctl %#x by %s\n",
2716 cmd, current_comm());
2721 module_put(THIS_MODULE);
2725 static int osc_get_info(const struct lu_env *env, struct obd_export *exp,
2726 u32 keylen, void *key, __u32 *vallen, void *val,
2727 struct lov_stripe_md *lsm)
2729 if (!vallen || !val)
2732 if (KEY_IS(KEY_LOCK_TO_STRIPE)) {
2733 __u32 *stripe = val;
2734 *vallen = sizeof(*stripe);
2737 } else if (KEY_IS(KEY_LAST_ID)) {
2738 struct ptlrpc_request *req;
2743 req = ptlrpc_request_alloc(class_exp2cliimp(exp),
2744 &RQF_OST_GET_INFO_LAST_ID);
2748 req_capsule_set_size(&req->rq_pill, &RMF_SETINFO_KEY,
2749 RCL_CLIENT, keylen);
2750 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_GET_INFO);
2752 ptlrpc_request_free(req);
2756 tmp = req_capsule_client_get(&req->rq_pill, &RMF_SETINFO_KEY);
2757 memcpy(tmp, key, keylen);
2759 req->rq_no_delay = req->rq_no_resend = 1;
2760 ptlrpc_request_set_replen(req);
2761 rc = ptlrpc_queue_wait(req);
2765 reply = req_capsule_server_get(&req->rq_pill, &RMF_OBD_ID);
2771 *((u64 *)val) = *reply;
2773 ptlrpc_req_finished(req);
2775 } else if (KEY_IS(KEY_FIEMAP)) {
2776 struct ll_fiemap_info_key *fm_key = key;
2777 struct ldlm_res_id res_id;
2778 ldlm_policy_data_t policy;
2779 struct lustre_handle lockh;
2780 enum ldlm_mode mode = 0;
2781 struct ptlrpc_request *req;
2782 struct ll_user_fiemap *reply;
2786 if (!(fm_key->fiemap.fm_flags & FIEMAP_FLAG_SYNC))
2789 policy.l_extent.start = fm_key->fiemap.fm_start &
2792 if (OBD_OBJECT_EOF - fm_key->fiemap.fm_length <=
2793 fm_key->fiemap.fm_start + PAGE_CACHE_SIZE - 1)
2794 policy.l_extent.end = OBD_OBJECT_EOF;
2796 policy.l_extent.end = (fm_key->fiemap.fm_start +
2797 fm_key->fiemap.fm_length +
2798 PAGE_CACHE_SIZE - 1) & CFS_PAGE_MASK;
2800 ostid_build_res_name(&fm_key->oa.o_oi, &res_id);
2801 mode = ldlm_lock_match(exp->exp_obd->obd_namespace,
2802 LDLM_FL_BLOCK_GRANTED |
2804 &res_id, LDLM_EXTENT, &policy,
2805 LCK_PR | LCK_PW, &lockh, 0);
2806 if (mode) { /* lock is cached on client */
2807 if (mode != LCK_PR) {
2808 ldlm_lock_addref(&lockh, LCK_PR);
2809 ldlm_lock_decref(&lockh, LCK_PW);
2811 } else { /* no cached lock, needs acquire lock on server side */
2812 fm_key->oa.o_valid |= OBD_MD_FLFLAGS;
2813 fm_key->oa.o_flags |= OBD_FL_SRVLOCK;
2817 req = ptlrpc_request_alloc(class_exp2cliimp(exp),
2818 &RQF_OST_GET_INFO_FIEMAP);
2824 req_capsule_set_size(&req->rq_pill, &RMF_FIEMAP_KEY,
2825 RCL_CLIENT, keylen);
2826 req_capsule_set_size(&req->rq_pill, &RMF_FIEMAP_VAL,
2827 RCL_CLIENT, *vallen);
2828 req_capsule_set_size(&req->rq_pill, &RMF_FIEMAP_VAL,
2829 RCL_SERVER, *vallen);
2831 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_GET_INFO);
2833 ptlrpc_request_free(req);
2837 tmp = req_capsule_client_get(&req->rq_pill, &RMF_FIEMAP_KEY);
2838 memcpy(tmp, key, keylen);
2839 tmp = req_capsule_client_get(&req->rq_pill, &RMF_FIEMAP_VAL);
2840 memcpy(tmp, val, *vallen);
2842 ptlrpc_request_set_replen(req);
2843 rc = ptlrpc_queue_wait(req);
2847 reply = req_capsule_server_get(&req->rq_pill, &RMF_FIEMAP_VAL);
2853 memcpy(val, reply, *vallen);
2855 ptlrpc_req_finished(req);
2858 ldlm_lock_decref(&lockh, LCK_PR);
2865 static int osc_set_info_async(const struct lu_env *env, struct obd_export *exp,
2866 u32 keylen, void *key, u32 vallen,
2867 void *val, struct ptlrpc_request_set *set)
2869 struct ptlrpc_request *req;
2870 struct obd_device *obd = exp->exp_obd;
2871 struct obd_import *imp = class_exp2cliimp(exp);
2875 OBD_FAIL_TIMEOUT(OBD_FAIL_OSC_SHUTDOWN, 10);
2877 if (KEY_IS(KEY_CHECKSUM)) {
2878 if (vallen != sizeof(int))
2880 exp->exp_obd->u.cli.cl_checksum = (*(int *)val) ? 1 : 0;
2884 if (KEY_IS(KEY_SPTLRPC_CONF)) {
2885 sptlrpc_conf_client_adapt(obd);
2889 if (KEY_IS(KEY_FLUSH_CTX)) {
2890 sptlrpc_import_flush_my_ctx(imp);
2894 if (KEY_IS(KEY_CACHE_SET)) {
2895 struct client_obd *cli = &obd->u.cli;
2897 LASSERT(!cli->cl_cache); /* only once */
2898 cli->cl_cache = val;
2899 atomic_inc(&cli->cl_cache->ccc_users);
2900 cli->cl_lru_left = &cli->cl_cache->ccc_lru_left;
2902 /* add this osc into entity list */
2903 LASSERT(list_empty(&cli->cl_lru_osc));
2904 spin_lock(&cli->cl_cache->ccc_lru_lock);
2905 list_add(&cli->cl_lru_osc, &cli->cl_cache->ccc_lru);
2906 spin_unlock(&cli->cl_cache->ccc_lru_lock);
2911 if (KEY_IS(KEY_CACHE_LRU_SHRINK)) {
2912 struct client_obd *cli = &obd->u.cli;
2913 int nr = atomic_read(&cli->cl_lru_in_list) >> 1;
2914 int target = *(int *)val;
2916 nr = osc_lru_shrink(cli, min(nr, target));
2921 if (!set && !KEY_IS(KEY_GRANT_SHRINK))
2924 /* We pass all other commands directly to OST. Since nobody calls osc
2925 * methods directly and everybody is supposed to go through LOV, we
2926 * assume lov checked invalid values for us.
2927 * The only recognised values so far are evict_by_nid and mds_conn.
2928 * Even if something bad goes through, we'd get a -EINVAL from OST
2932 req = ptlrpc_request_alloc(imp, KEY_IS(KEY_GRANT_SHRINK) ?
2933 &RQF_OST_SET_GRANT_INFO :
2938 req_capsule_set_size(&req->rq_pill, &RMF_SETINFO_KEY,
2939 RCL_CLIENT, keylen);
2940 if (!KEY_IS(KEY_GRANT_SHRINK))
2941 req_capsule_set_size(&req->rq_pill, &RMF_SETINFO_VAL,
2942 RCL_CLIENT, vallen);
2943 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_SET_INFO);
2945 ptlrpc_request_free(req);
2949 tmp = req_capsule_client_get(&req->rq_pill, &RMF_SETINFO_KEY);
2950 memcpy(tmp, key, keylen);
2951 tmp = req_capsule_client_get(&req->rq_pill, KEY_IS(KEY_GRANT_SHRINK) ?
2954 memcpy(tmp, val, vallen);
2956 if (KEY_IS(KEY_GRANT_SHRINK)) {
2957 struct osc_brw_async_args *aa;
2960 CLASSERT(sizeof(*aa) <= sizeof(req->rq_async_args));
2961 aa = ptlrpc_req_async_args(req);
2962 oa = kmem_cache_zalloc(obdo_cachep, GFP_NOFS);
2964 ptlrpc_req_finished(req);
2967 *oa = ((struct ost_body *)val)->oa;
2969 req->rq_interpret_reply = osc_shrink_grant_interpret;
2972 ptlrpc_request_set_replen(req);
2973 if (!KEY_IS(KEY_GRANT_SHRINK)) {
2975 ptlrpc_set_add_req(set, req);
2976 ptlrpc_check_set(NULL, set);
2978 ptlrpcd_add_req(req);
2984 static int osc_reconnect(const struct lu_env *env,
2985 struct obd_export *exp, struct obd_device *obd,
2986 struct obd_uuid *cluuid,
2987 struct obd_connect_data *data,
2990 struct client_obd *cli = &obd->u.cli;
2992 if (data && (data->ocd_connect_flags & OBD_CONNECT_GRANT)) {
2995 client_obd_list_lock(&cli->cl_loi_list_lock);
2996 data->ocd_grant = (cli->cl_avail_grant + cli->cl_dirty) ?:
2997 2 * cli_brw_size(obd);
2998 lost_grant = cli->cl_lost_grant;
2999 cli->cl_lost_grant = 0;
3000 client_obd_list_unlock(&cli->cl_loi_list_lock);
3002 CDEBUG(D_RPCTRACE, "ocd_connect_flags: %#llx ocd_version: %d ocd_grant: %d, lost: %ld.\n",
3003 data->ocd_connect_flags,
3004 data->ocd_version, data->ocd_grant, lost_grant);
3010 static int osc_disconnect(struct obd_export *exp)
3012 struct obd_device *obd = class_exp2obd(exp);
3015 rc = client_disconnect_export(exp);
3017 * Initially we put del_shrink_grant before disconnect_export, but it
3018 * causes the following problem if setup (connect) and cleanup
3019 * (disconnect) are tangled together.
3020 * connect p1 disconnect p2
3021 * ptlrpc_connect_import
3022 * ............... class_manual_cleanup
3025 * ptlrpc_connect_interrupt
3027 * add this client to shrink list
3029 * Bang! pinger trigger the shrink.
3030 * So the osc should be disconnected from the shrink list, after we
3031 * are sure the import has been destroyed. BUG18662
3033 if (!obd->u.cli.cl_import)
3034 osc_del_shrink_grant(&obd->u.cli);
3038 static int osc_import_event(struct obd_device *obd,
3039 struct obd_import *imp,
3040 enum obd_import_event event)
3042 struct client_obd *cli;
3045 LASSERT(imp->imp_obd == obd);
3048 case IMP_EVENT_DISCON: {
3050 client_obd_list_lock(&cli->cl_loi_list_lock);
3051 cli->cl_avail_grant = 0;
3052 cli->cl_lost_grant = 0;
3053 client_obd_list_unlock(&cli->cl_loi_list_lock);
3056 case IMP_EVENT_INACTIVE: {
3057 rc = obd_notify_observer(obd, obd, OBD_NOTIFY_INACTIVE, NULL);
3060 case IMP_EVENT_INVALIDATE: {
3061 struct ldlm_namespace *ns = obd->obd_namespace;
3065 env = cl_env_get(&refcheck);
3069 /* all pages go to failing rpcs due to the invalid
3072 osc_io_unplug(env, cli, NULL);
3074 ldlm_namespace_cleanup(ns, LDLM_FL_LOCAL_ONLY);
3075 cl_env_put(env, &refcheck);
3080 case IMP_EVENT_ACTIVE: {
3081 rc = obd_notify_observer(obd, obd, OBD_NOTIFY_ACTIVE, NULL);
3084 case IMP_EVENT_OCD: {
3085 struct obd_connect_data *ocd = &imp->imp_connect_data;
3087 if (ocd->ocd_connect_flags & OBD_CONNECT_GRANT)
3088 osc_init_grant(&obd->u.cli, ocd);
3091 if (ocd->ocd_connect_flags & OBD_CONNECT_REQPORTAL)
3092 imp->imp_client->cli_request_portal = OST_REQUEST_PORTAL;
3094 rc = obd_notify_observer(obd, obd, OBD_NOTIFY_OCD, NULL);
3097 case IMP_EVENT_DEACTIVATE: {
3098 rc = obd_notify_observer(obd, obd, OBD_NOTIFY_DEACTIVATE, NULL);
3101 case IMP_EVENT_ACTIVATE: {
3102 rc = obd_notify_observer(obd, obd, OBD_NOTIFY_ACTIVATE, NULL);
3106 CERROR("Unknown import event %d\n", event);
3113 * Determine whether the lock can be canceled before replaying the lock
3114 * during recovery, see bug16774 for detailed information.
3116 * \retval zero the lock can't be canceled
3117 * \retval other ok to cancel
3119 static int osc_cancel_for_recovery(struct ldlm_lock *lock)
3121 check_res_locked(lock->l_resource);
3124 * Cancel all unused extent lock in granted mode LCK_PR or LCK_CR.
3126 * XXX as a future improvement, we can also cancel unused write lock
3127 * if it doesn't have dirty data and active mmaps.
3129 if (lock->l_resource->lr_type == LDLM_EXTENT &&
3130 (lock->l_granted_mode == LCK_PR ||
3131 lock->l_granted_mode == LCK_CR) &&
3132 (osc_dlm_lock_pageref(lock) == 0))
3138 static int brw_queue_work(const struct lu_env *env, void *data)
3140 struct client_obd *cli = data;
3142 CDEBUG(D_CACHE, "Run writeback work for client obd %p.\n", cli);
3144 osc_io_unplug(env, cli, NULL);
3148 int osc_setup(struct obd_device *obd, struct lustre_cfg *lcfg)
3150 struct lprocfs_static_vars lvars = { NULL };
3151 struct client_obd *cli = &obd->u.cli;
3158 rc = ptlrpcd_addref();
3162 rc = client_obd_setup(obd, lcfg);
3166 handler = ptlrpcd_alloc_work(cli->cl_import, brw_queue_work, cli);
3167 if (IS_ERR(handler)) {
3168 rc = PTR_ERR(handler);
3169 goto out_client_setup;
3171 cli->cl_writeback_work = handler;
3173 rc = osc_quota_setup(obd);
3175 goto out_ptlrpcd_work;
3177 cli->cl_grant_shrink_interval = GRANT_SHRINK_INTERVAL;
3178 lprocfs_osc_init_vars(&lvars);
3179 if (lprocfs_obd_setup(obd, lvars.obd_vars, lvars.sysfs_vars) == 0) {
3180 lproc_osc_attach_seqstat(obd);
3181 sptlrpc_lprocfs_cliobd_attach(obd);
3182 ptlrpc_lprocfs_register_obd(obd);
3186 * We try to control the total number of requests with a upper limit
3187 * osc_reqpool_maxreqcount. There might be some race which will cause
3188 * over-limit allocation, but it is fine.
3190 req_count = atomic_read(&osc_pool_req_count);
3191 if (req_count < osc_reqpool_maxreqcount) {
3192 adding = cli->cl_max_rpcs_in_flight + 2;
3193 if (req_count + adding > osc_reqpool_maxreqcount)
3194 adding = osc_reqpool_maxreqcount - req_count;
3196 added = ptlrpc_add_rqs_to_pool(osc_rq_pool, adding);
3197 atomic_add(added, &osc_pool_req_count);
3200 INIT_LIST_HEAD(&cli->cl_grant_shrink_list);
3201 ns_register_cancel(obd->obd_namespace, osc_cancel_for_recovery);
3205 ptlrpcd_destroy_work(handler);
3207 client_obd_cleanup(obd);
3213 static int osc_precleanup(struct obd_device *obd, enum obd_cleanup_stage stage)
3216 case OBD_CLEANUP_EARLY: {
3217 struct obd_import *imp;
3219 imp = obd->u.cli.cl_import;
3220 CDEBUG(D_HA, "Deactivating import %s\n", obd->obd_name);
3221 /* ptlrpc_abort_inflight to stop an mds_lov_synchronize */
3222 ptlrpc_deactivate_import(imp);
3223 spin_lock(&imp->imp_lock);
3224 imp->imp_pingable = 0;
3225 spin_unlock(&imp->imp_lock);
3228 case OBD_CLEANUP_EXPORTS: {
3229 struct client_obd *cli = &obd->u.cli;
3231 * for echo client, export may be on zombie list, wait for
3232 * zombie thread to cull it, because cli.cl_import will be
3233 * cleared in client_disconnect_export():
3234 * class_export_destroy() -> obd_cleanup() ->
3235 * echo_device_free() -> echo_client_cleanup() ->
3236 * obd_disconnect() -> osc_disconnect() ->
3237 * client_disconnect_export()
3239 obd_zombie_barrier();
3240 if (cli->cl_writeback_work) {
3241 ptlrpcd_destroy_work(cli->cl_writeback_work);
3242 cli->cl_writeback_work = NULL;
3244 obd_cleanup_client_import(obd);
3245 ptlrpc_lprocfs_unregister_obd(obd);
3246 lprocfs_obd_cleanup(obd);
3253 static int osc_cleanup(struct obd_device *obd)
3255 struct client_obd *cli = &obd->u.cli;
3259 if (cli->cl_cache) {
3260 LASSERT(atomic_read(&cli->cl_cache->ccc_users) > 0);
3261 spin_lock(&cli->cl_cache->ccc_lru_lock);
3262 list_del_init(&cli->cl_lru_osc);
3263 spin_unlock(&cli->cl_cache->ccc_lru_lock);
3264 cli->cl_lru_left = NULL;
3265 atomic_dec(&cli->cl_cache->ccc_users);
3266 cli->cl_cache = NULL;
3269 /* free memory of osc quota cache */
3270 osc_quota_cleanup(obd);
3272 rc = client_obd_cleanup(obd);
3278 int osc_process_config_base(struct obd_device *obd, struct lustre_cfg *lcfg)
3280 struct lprocfs_static_vars lvars = { NULL };
3283 lprocfs_osc_init_vars(&lvars);
3285 switch (lcfg->lcfg_command) {
3287 rc = class_process_proc_param(PARAM_OSC, lvars.obd_vars,
3297 static int osc_process_config(struct obd_device *obd, u32 len, void *buf)
3299 return osc_process_config_base(obd, buf);
3302 static struct obd_ops osc_obd_ops = {
3303 .owner = THIS_MODULE,
3305 .precleanup = osc_precleanup,
3306 .cleanup = osc_cleanup,
3307 .add_conn = client_import_add_conn,
3308 .del_conn = client_import_del_conn,
3309 .connect = client_connect_import,
3310 .reconnect = osc_reconnect,
3311 .disconnect = osc_disconnect,
3312 .statfs = osc_statfs,
3313 .statfs_async = osc_statfs_async,
3314 .packmd = osc_packmd,
3315 .unpackmd = osc_unpackmd,
3316 .create = osc_create,
3317 .destroy = osc_destroy,
3318 .getattr = osc_getattr,
3319 .getattr_async = osc_getattr_async,
3320 .setattr = osc_setattr,
3321 .setattr_async = osc_setattr_async,
3322 .find_cbdata = osc_find_cbdata,
3323 .iocontrol = osc_iocontrol,
3324 .get_info = osc_get_info,
3325 .set_info_async = osc_set_info_async,
3326 .import_event = osc_import_event,
3327 .process_config = osc_process_config,
3328 .quotactl = osc_quotactl,
3329 .quotacheck = osc_quotacheck,
3332 extern struct lu_kmem_descr osc_caches[];
3333 extern spinlock_t osc_ast_guard;
3334 extern struct lock_class_key osc_ast_guard_class;
3336 static int __init osc_init(void)
3338 struct lprocfs_static_vars lvars = { NULL };
3339 unsigned int reqpool_size;
3340 unsigned int reqsize;
3343 /* print an address of _any_ initialized kernel symbol from this
3344 * module, to allow debugging with gdb that doesn't support data
3345 * symbols from modules.
3347 CDEBUG(D_INFO, "Lustre OSC module (%p).\n", &osc_caches);
3349 rc = lu_kmem_init(osc_caches);
3353 lprocfs_osc_init_vars(&lvars);
3355 rc = class_register_type(&osc_obd_ops, NULL,
3356 LUSTRE_OSC_NAME, &osc_device_type);
3360 spin_lock_init(&osc_ast_guard);
3361 lockdep_set_class(&osc_ast_guard, &osc_ast_guard_class);
3363 /* This is obviously too much memory, only prevent overflow here */
3364 if (osc_reqpool_mem_max >= 1 << 12 || osc_reqpool_mem_max == 0) {
3369 reqpool_size = osc_reqpool_mem_max << 20;
3372 while (reqsize < OST_MAXREQSIZE)
3373 reqsize = reqsize << 1;
3376 * We don't enlarge the request count in OSC pool according to
3377 * cl_max_rpcs_in_flight. The allocation from the pool will only be
3378 * tried after normal allocation failed. So a small OSC pool won't
3379 * cause much performance degression in most of cases.
3381 osc_reqpool_maxreqcount = reqpool_size / reqsize;
3383 atomic_set(&osc_pool_req_count, 0);
3384 osc_rq_pool = ptlrpc_init_rq_pool(0, OST_MAXREQSIZE,
3385 ptlrpc_add_rqs_to_pool);
3393 class_unregister_type(LUSTRE_OSC_NAME);
3395 lu_kmem_fini(osc_caches);
3399 static void /*__exit*/ osc_exit(void)
3401 class_unregister_type(LUSTRE_OSC_NAME);
3402 lu_kmem_fini(osc_caches);
3403 ptlrpc_free_rq_pool(osc_rq_pool);
3406 MODULE_AUTHOR("OpenSFS, Inc. <http://www.lustre.org/>");
3407 MODULE_DESCRIPTION("Lustre Object Storage Client (OSC)");
3408 MODULE_LICENSE("GPL");
3409 MODULE_VERSION(LUSTRE_VERSION_STRING);
3411 module_init(osc_init);
3412 module_exit(osc_exit);