4 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
6 * This program is free software; you can redistribute it and/or modify
7 * it under the terms of the GNU General Public License version 2 only,
8 * as published by the Free Software Foundation.
10 * This program is distributed in the hope that it will be useful, but
11 * WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13 * General Public License version 2 for more details (a copy is included
14 * in the LICENSE file that accompanied this code).
16 * You should have received a copy of the GNU General Public License
17 * version 2 along with this program; If not, see
18 * http://www.gnu.org/licenses/gpl-2.0.html
23 * Copyright (c) 2002, 2010, Oracle and/or its affiliates. All rights reserved.
24 * Use is subject to license terms.
26 * Copyright (c) 2011, 2017, Intel Corporation.
29 * This file is part of Lustre, http://www.lustre.org/
30 * Lustre is a trademark of Sun Microsystems, Inc.
33 #define DEBUG_SUBSYSTEM S_OSC
35 #include <linux/workqueue.h>
36 #include <lprocfs_status.h>
37 #include <lustre_debug.h>
38 #include <lustre_dlm.h>
39 #include <lustre_fid.h>
40 #include <lustre_ha.h>
41 #include <uapi/linux/lustre/lustre_ioctl.h>
42 #include <lustre_net.h>
43 #include <lustre_obdo.h>
45 #include <obd_cksum.h>
46 #include <obd_class.h>
47 #include <lustre_osc.h>
49 #include "osc_internal.h"
51 atomic_t osc_pool_req_count;
52 unsigned int osc_reqpool_maxreqcount;
53 struct ptlrpc_request_pool *osc_rq_pool;
55 /* max memory used for request pool, unit is MB */
56 static unsigned int osc_reqpool_mem_max = 5;
57 module_param(osc_reqpool_mem_max, uint, 0444);
59 static int osc_idle_timeout = 20;
60 module_param(osc_idle_timeout, uint, 0644);
62 #define osc_grant_args osc_brw_async_args
64 struct osc_setattr_args {
66 obd_enqueue_update_f sa_upcall;
70 struct osc_fsync_args {
71 struct osc_object *fa_obj;
73 obd_enqueue_update_f fa_upcall;
77 struct osc_ladvise_args {
79 obd_enqueue_update_f la_upcall;
83 static void osc_release_ppga(struct brw_page **ppga, size_t count);
84 static int brw_interpret(const struct lu_env *env, struct ptlrpc_request *req,
87 void osc_pack_req_body(struct ptlrpc_request *req, struct obdo *oa)
89 struct ost_body *body;
91 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
94 lustre_set_wire_obdo(&req->rq_import->imp_connect_data, &body->oa, oa);
97 static int osc_getattr(const struct lu_env *env, struct obd_export *exp,
100 struct ptlrpc_request *req;
101 struct ost_body *body;
105 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_GETATTR);
109 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_GETATTR);
111 ptlrpc_request_free(req);
115 osc_pack_req_body(req, oa);
117 ptlrpc_request_set_replen(req);
119 rc = ptlrpc_queue_wait(req);
123 body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
125 GOTO(out, rc = -EPROTO);
127 CDEBUG(D_INODE, "mode: %o\n", body->oa.o_mode);
128 lustre_get_wire_obdo(&req->rq_import->imp_connect_data, oa, &body->oa);
130 oa->o_blksize = cli_brw_size(exp->exp_obd);
131 oa->o_valid |= OBD_MD_FLBLKSZ;
135 ptlrpc_req_finished(req);
140 static int osc_setattr(const struct lu_env *env, struct obd_export *exp,
143 struct ptlrpc_request *req;
144 struct ost_body *body;
148 LASSERT(oa->o_valid & OBD_MD_FLGROUP);
150 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_SETATTR);
154 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_SETATTR);
156 ptlrpc_request_free(req);
160 osc_pack_req_body(req, oa);
162 ptlrpc_request_set_replen(req);
164 rc = ptlrpc_queue_wait(req);
168 body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
170 GOTO(out, rc = -EPROTO);
172 lustre_get_wire_obdo(&req->rq_import->imp_connect_data, oa, &body->oa);
176 ptlrpc_req_finished(req);
181 static int osc_setattr_interpret(const struct lu_env *env,
182 struct ptlrpc_request *req, void *args, int rc)
184 struct osc_setattr_args *sa = args;
185 struct ost_body *body;
192 body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
194 GOTO(out, rc = -EPROTO);
196 lustre_get_wire_obdo(&req->rq_import->imp_connect_data, sa->sa_oa,
199 rc = sa->sa_upcall(sa->sa_cookie, rc);
203 int osc_setattr_async(struct obd_export *exp, struct obdo *oa,
204 obd_enqueue_update_f upcall, void *cookie,
205 struct ptlrpc_request_set *rqset)
207 struct ptlrpc_request *req;
208 struct osc_setattr_args *sa;
213 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_SETATTR);
217 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_SETATTR);
219 ptlrpc_request_free(req);
223 osc_pack_req_body(req, oa);
225 ptlrpc_request_set_replen(req);
227 /* do mds to ost setattr asynchronously */
229 /* Do not wait for response. */
230 ptlrpcd_add_req(req);
232 req->rq_interpret_reply = osc_setattr_interpret;
234 sa = ptlrpc_req_async_args(sa, req);
236 sa->sa_upcall = upcall;
237 sa->sa_cookie = cookie;
239 if (rqset == PTLRPCD_SET)
240 ptlrpcd_add_req(req);
242 ptlrpc_set_add_req(rqset, req);
248 static int osc_ladvise_interpret(const struct lu_env *env,
249 struct ptlrpc_request *req,
252 struct osc_ladvise_args *la = arg;
253 struct ost_body *body;
259 body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
261 GOTO(out, rc = -EPROTO);
263 *la->la_oa = body->oa;
265 rc = la->la_upcall(la->la_cookie, rc);
270 * If rqset is NULL, do not wait for response. Upcall and cookie could also
271 * be NULL in this case
273 int osc_ladvise_base(struct obd_export *exp, struct obdo *oa,
274 struct ladvise_hdr *ladvise_hdr,
275 obd_enqueue_update_f upcall, void *cookie,
276 struct ptlrpc_request_set *rqset)
278 struct ptlrpc_request *req;
279 struct ost_body *body;
280 struct osc_ladvise_args *la;
282 struct lu_ladvise *req_ladvise;
283 struct lu_ladvise *ladvise = ladvise_hdr->lah_advise;
284 int num_advise = ladvise_hdr->lah_count;
285 struct ladvise_hdr *req_ladvise_hdr;
288 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_LADVISE);
292 req_capsule_set_size(&req->rq_pill, &RMF_OST_LADVISE, RCL_CLIENT,
293 num_advise * sizeof(*ladvise));
294 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_LADVISE);
296 ptlrpc_request_free(req);
299 req->rq_request_portal = OST_IO_PORTAL;
300 ptlrpc_at_set_req_timeout(req);
302 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
304 lustre_set_wire_obdo(&req->rq_import->imp_connect_data, &body->oa,
307 req_ladvise_hdr = req_capsule_client_get(&req->rq_pill,
308 &RMF_OST_LADVISE_HDR);
309 memcpy(req_ladvise_hdr, ladvise_hdr, sizeof(*ladvise_hdr));
311 req_ladvise = req_capsule_client_get(&req->rq_pill, &RMF_OST_LADVISE);
312 memcpy(req_ladvise, ladvise, sizeof(*ladvise) * num_advise);
313 ptlrpc_request_set_replen(req);
316 /* Do not wait for response. */
317 ptlrpcd_add_req(req);
321 req->rq_interpret_reply = osc_ladvise_interpret;
322 la = ptlrpc_req_async_args(la, req);
324 la->la_upcall = upcall;
325 la->la_cookie = cookie;
327 if (rqset == PTLRPCD_SET)
328 ptlrpcd_add_req(req);
330 ptlrpc_set_add_req(rqset, req);
335 static int osc_create(const struct lu_env *env, struct obd_export *exp,
338 struct ptlrpc_request *req;
339 struct ost_body *body;
344 LASSERT(oa->o_valid & OBD_MD_FLGROUP);
345 LASSERT(fid_seq_is_echo(ostid_seq(&oa->o_oi)));
347 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_CREATE);
349 GOTO(out, rc = -ENOMEM);
351 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_CREATE);
353 ptlrpc_request_free(req);
357 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
360 lustre_set_wire_obdo(&req->rq_import->imp_connect_data, &body->oa, oa);
362 ptlrpc_request_set_replen(req);
364 rc = ptlrpc_queue_wait(req);
368 body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
370 GOTO(out_req, rc = -EPROTO);
372 CDEBUG(D_INFO, "oa flags %x\n", oa->o_flags);
373 lustre_get_wire_obdo(&req->rq_import->imp_connect_data, oa, &body->oa);
375 oa->o_blksize = cli_brw_size(exp->exp_obd);
376 oa->o_valid |= OBD_MD_FLBLKSZ;
378 CDEBUG(D_HA, "transno: %lld\n",
379 lustre_msg_get_transno(req->rq_repmsg));
381 ptlrpc_req_finished(req);
386 int osc_punch_send(struct obd_export *exp, struct obdo *oa,
387 obd_enqueue_update_f upcall, void *cookie)
389 struct ptlrpc_request *req;
390 struct osc_setattr_args *sa;
391 struct obd_import *imp = class_exp2cliimp(exp);
392 struct ost_body *body;
397 req = ptlrpc_request_alloc(imp, &RQF_OST_PUNCH);
401 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_PUNCH);
403 ptlrpc_request_free(req);
407 osc_set_io_portal(req);
409 ptlrpc_at_set_req_timeout(req);
411 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
413 lustre_set_wire_obdo(&imp->imp_connect_data, &body->oa, oa);
415 ptlrpc_request_set_replen(req);
417 req->rq_interpret_reply = osc_setattr_interpret;
418 sa = ptlrpc_req_async_args(sa, req);
420 sa->sa_upcall = upcall;
421 sa->sa_cookie = cookie;
423 ptlrpcd_add_req(req);
427 EXPORT_SYMBOL(osc_punch_send);
429 static int osc_sync_interpret(const struct lu_env *env,
430 struct ptlrpc_request *req, void *args, int rc)
432 struct osc_fsync_args *fa = args;
433 struct ost_body *body;
434 struct cl_attr *attr = &osc_env_info(env)->oti_attr;
435 unsigned long valid = 0;
436 struct cl_object *obj;
442 body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
444 CERROR("can't unpack ost_body\n");
445 GOTO(out, rc = -EPROTO);
448 *fa->fa_oa = body->oa;
449 obj = osc2cl(fa->fa_obj);
451 /* Update osc object's blocks attribute */
452 cl_object_attr_lock(obj);
453 if (body->oa.o_valid & OBD_MD_FLBLOCKS) {
454 attr->cat_blocks = body->oa.o_blocks;
459 cl_object_attr_update(env, obj, attr, valid);
460 cl_object_attr_unlock(obj);
463 rc = fa->fa_upcall(fa->fa_cookie, rc);
467 int osc_sync_base(struct osc_object *obj, struct obdo *oa,
468 obd_enqueue_update_f upcall, void *cookie,
469 struct ptlrpc_request_set *rqset)
471 struct obd_export *exp = osc_export(obj);
472 struct ptlrpc_request *req;
473 struct ost_body *body;
474 struct osc_fsync_args *fa;
478 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_SYNC);
482 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_SYNC);
484 ptlrpc_request_free(req);
488 /* overload the size and blocks fields in the oa with start/end */
489 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
491 lustre_set_wire_obdo(&req->rq_import->imp_connect_data, &body->oa, oa);
493 ptlrpc_request_set_replen(req);
494 req->rq_interpret_reply = osc_sync_interpret;
496 fa = ptlrpc_req_async_args(fa, req);
499 fa->fa_upcall = upcall;
500 fa->fa_cookie = cookie;
502 if (rqset == PTLRPCD_SET)
503 ptlrpcd_add_req(req);
505 ptlrpc_set_add_req(rqset, req);
510 /* Find and cancel locally locks matched by @mode in the resource found by
511 * @objid. Found locks are added into @cancel list. Returns the amount of
512 * locks added to @cancels list. */
513 static int osc_resource_get_unused(struct obd_export *exp, struct obdo *oa,
514 struct list_head *cancels,
515 enum ldlm_mode mode, __u64 lock_flags)
517 struct ldlm_namespace *ns = exp->exp_obd->obd_namespace;
518 struct ldlm_res_id res_id;
519 struct ldlm_resource *res;
523 /* Return, i.e. cancel nothing, only if ELC is supported (flag in
524 * export) but disabled through procfs (flag in NS).
526 * This distinguishes from a case when ELC is not supported originally,
527 * when we still want to cancel locks in advance and just cancel them
528 * locally, without sending any RPC. */
529 if (exp_connect_cancelset(exp) && !ns_connect_cancelset(ns))
532 ostid_build_res_name(&oa->o_oi, &res_id);
533 res = ldlm_resource_get(ns, NULL, &res_id, 0, 0);
537 LDLM_RESOURCE_ADDREF(res);
538 count = ldlm_cancel_resource_local(res, cancels, NULL, mode,
539 lock_flags, 0, NULL);
540 LDLM_RESOURCE_DELREF(res);
541 ldlm_resource_putref(res);
545 static int osc_destroy_interpret(const struct lu_env *env,
546 struct ptlrpc_request *req, void *args, int rc)
548 struct client_obd *cli = &req->rq_import->imp_obd->u.cli;
550 atomic_dec(&cli->cl_destroy_in_flight);
551 wake_up(&cli->cl_destroy_waitq);
556 static int osc_can_send_destroy(struct client_obd *cli)
558 if (atomic_inc_return(&cli->cl_destroy_in_flight) <=
559 cli->cl_max_rpcs_in_flight) {
560 /* The destroy request can be sent */
563 if (atomic_dec_return(&cli->cl_destroy_in_flight) <
564 cli->cl_max_rpcs_in_flight) {
566 * The counter has been modified between the two atomic
569 wake_up(&cli->cl_destroy_waitq);
574 static int osc_destroy(const struct lu_env *env, struct obd_export *exp,
577 struct client_obd *cli = &exp->exp_obd->u.cli;
578 struct ptlrpc_request *req;
579 struct ost_body *body;
585 CDEBUG(D_INFO, "oa NULL\n");
589 count = osc_resource_get_unused(exp, oa, &cancels, LCK_PW,
590 LDLM_FL_DISCARD_DATA);
592 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_DESTROY);
594 ldlm_lock_list_put(&cancels, l_bl_ast, count);
598 rc = ldlm_prep_elc_req(exp, req, LUSTRE_OST_VERSION, OST_DESTROY,
601 ptlrpc_request_free(req);
605 req->rq_request_portal = OST_IO_PORTAL; /* bug 7198 */
606 ptlrpc_at_set_req_timeout(req);
608 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
610 lustre_set_wire_obdo(&req->rq_import->imp_connect_data, &body->oa, oa);
612 ptlrpc_request_set_replen(req);
614 req->rq_interpret_reply = osc_destroy_interpret;
615 if (!osc_can_send_destroy(cli)) {
616 struct l_wait_info lwi = LWI_INTR(LWI_ON_SIGNAL_NOOP, NULL);
619 * Wait until the number of on-going destroy RPCs drops
620 * under max_rpc_in_flight
622 rc = l_wait_event_exclusive(cli->cl_destroy_waitq,
623 osc_can_send_destroy(cli), &lwi);
625 ptlrpc_req_finished(req);
630 /* Do not wait for response */
631 ptlrpcd_add_req(req);
635 static void osc_announce_cached(struct client_obd *cli, struct obdo *oa,
638 u64 bits = OBD_MD_FLBLOCKS | OBD_MD_FLGRANT;
640 LASSERT(!(oa->o_valid & bits));
643 spin_lock(&cli->cl_loi_list_lock);
644 if (OCD_HAS_FLAG(&cli->cl_import->imp_connect_data, GRANT_PARAM))
645 oa->o_dirty = cli->cl_dirty_grant;
647 oa->o_dirty = cli->cl_dirty_pages << PAGE_SHIFT;
648 if (unlikely(cli->cl_dirty_pages > cli->cl_dirty_max_pages)) {
649 CERROR("dirty %lu > dirty_max %lu\n",
651 cli->cl_dirty_max_pages);
653 } else if (unlikely(atomic_long_read(&obd_dirty_pages) >
654 (long)(obd_max_dirty_pages + 1))) {
655 /* The atomic_read() allowing the atomic_inc() are
656 * not covered by a lock thus they may safely race and trip
657 * this CERROR() unless we add in a small fudge factor (+1). */
658 CERROR("%s: dirty %ld > system dirty_max %ld\n",
659 cli_name(cli), atomic_long_read(&obd_dirty_pages),
660 obd_max_dirty_pages);
662 } else if (unlikely(cli->cl_dirty_max_pages - cli->cl_dirty_pages >
664 CERROR("dirty %lu - dirty_max %lu too big???\n",
665 cli->cl_dirty_pages, cli->cl_dirty_max_pages);
668 unsigned long nrpages;
669 unsigned long undirty;
671 nrpages = cli->cl_max_pages_per_rpc;
672 nrpages *= cli->cl_max_rpcs_in_flight + 1;
673 nrpages = max(nrpages, cli->cl_dirty_max_pages);
674 undirty = nrpages << PAGE_SHIFT;
675 if (OCD_HAS_FLAG(&cli->cl_import->imp_connect_data,
679 /* take extent tax into account when asking for more
681 nrextents = (nrpages + cli->cl_max_extent_pages - 1) /
682 cli->cl_max_extent_pages;
683 undirty += nrextents * cli->cl_grant_extent_tax;
685 /* Do not ask for more than OBD_MAX_GRANT - a margin for server
686 * to add extent tax, etc.
688 oa->o_undirty = min(undirty, OBD_MAX_GRANT &
689 ~(PTLRPC_MAX_BRW_SIZE * 4UL));
691 oa->o_grant = cli->cl_avail_grant + cli->cl_reserved_grant;
692 oa->o_dropped = cli->cl_lost_grant;
693 cli->cl_lost_grant = 0;
694 spin_unlock(&cli->cl_loi_list_lock);
695 CDEBUG(D_CACHE, "dirty: %llu undirty: %u dropped %u grant: %llu\n",
696 oa->o_dirty, oa->o_undirty, oa->o_dropped, oa->o_grant);
699 void osc_update_next_shrink(struct client_obd *cli)
701 cli->cl_next_shrink_grant = ktime_get_seconds() +
702 cli->cl_grant_shrink_interval;
704 CDEBUG(D_CACHE, "next time %lld to shrink grant\n",
705 cli->cl_next_shrink_grant);
708 static void __osc_update_grant(struct client_obd *cli, u64 grant)
710 spin_lock(&cli->cl_loi_list_lock);
711 cli->cl_avail_grant += grant;
712 spin_unlock(&cli->cl_loi_list_lock);
715 static void osc_update_grant(struct client_obd *cli, struct ost_body *body)
717 if (body->oa.o_valid & OBD_MD_FLGRANT) {
718 CDEBUG(D_CACHE, "got %llu extra grant\n", body->oa.o_grant);
719 __osc_update_grant(cli, body->oa.o_grant);
724 * grant thread data for shrinking space.
726 struct grant_thread_data {
727 struct list_head gtd_clients;
728 struct mutex gtd_mutex;
729 unsigned long gtd_stopped:1;
731 static struct grant_thread_data client_gtd;
733 static int osc_shrink_grant_interpret(const struct lu_env *env,
734 struct ptlrpc_request *req,
737 struct osc_grant_args *aa = args;
738 struct client_obd *cli = &req->rq_import->imp_obd->u.cli;
739 struct ost_body *body;
742 __osc_update_grant(cli, aa->aa_oa->o_grant);
746 body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
748 osc_update_grant(cli, body);
750 OBD_SLAB_FREE_PTR(aa->aa_oa, osc_obdo_kmem);
756 static void osc_shrink_grant_local(struct client_obd *cli, struct obdo *oa)
758 spin_lock(&cli->cl_loi_list_lock);
759 oa->o_grant = cli->cl_avail_grant / 4;
760 cli->cl_avail_grant -= oa->o_grant;
761 spin_unlock(&cli->cl_loi_list_lock);
762 if (!(oa->o_valid & OBD_MD_FLFLAGS)) {
763 oa->o_valid |= OBD_MD_FLFLAGS;
766 oa->o_flags |= OBD_FL_SHRINK_GRANT;
767 osc_update_next_shrink(cli);
770 /* Shrink the current grant, either from some large amount to enough for a
771 * full set of in-flight RPCs, or if we have already shrunk to that limit
772 * then to enough for a single RPC. This avoids keeping more grant than
773 * needed, and avoids shrinking the grant piecemeal. */
774 static int osc_shrink_grant(struct client_obd *cli)
776 __u64 target_bytes = (cli->cl_max_rpcs_in_flight + 1) *
777 (cli->cl_max_pages_per_rpc << PAGE_SHIFT);
779 spin_lock(&cli->cl_loi_list_lock);
780 if (cli->cl_avail_grant <= target_bytes)
781 target_bytes = cli->cl_max_pages_per_rpc << PAGE_SHIFT;
782 spin_unlock(&cli->cl_loi_list_lock);
784 return osc_shrink_grant_to_target(cli, target_bytes);
787 int osc_shrink_grant_to_target(struct client_obd *cli, __u64 target_bytes)
790 struct ost_body *body;
793 spin_lock(&cli->cl_loi_list_lock);
794 /* Don't shrink if we are already above or below the desired limit
795 * We don't want to shrink below a single RPC, as that will negatively
796 * impact block allocation and long-term performance. */
797 if (target_bytes < cli->cl_max_pages_per_rpc << PAGE_SHIFT)
798 target_bytes = cli->cl_max_pages_per_rpc << PAGE_SHIFT;
800 if (target_bytes >= cli->cl_avail_grant) {
801 spin_unlock(&cli->cl_loi_list_lock);
804 spin_unlock(&cli->cl_loi_list_lock);
810 osc_announce_cached(cli, &body->oa, 0);
812 spin_lock(&cli->cl_loi_list_lock);
813 if (target_bytes >= cli->cl_avail_grant) {
814 /* available grant has changed since target calculation */
815 spin_unlock(&cli->cl_loi_list_lock);
816 GOTO(out_free, rc = 0);
818 body->oa.o_grant = cli->cl_avail_grant - target_bytes;
819 cli->cl_avail_grant = target_bytes;
820 spin_unlock(&cli->cl_loi_list_lock);
821 if (!(body->oa.o_valid & OBD_MD_FLFLAGS)) {
822 body->oa.o_valid |= OBD_MD_FLFLAGS;
823 body->oa.o_flags = 0;
825 body->oa.o_flags |= OBD_FL_SHRINK_GRANT;
826 osc_update_next_shrink(cli);
828 rc = osc_set_info_async(NULL, cli->cl_import->imp_obd->obd_self_export,
829 sizeof(KEY_GRANT_SHRINK), KEY_GRANT_SHRINK,
830 sizeof(*body), body, NULL);
832 __osc_update_grant(cli, body->oa.o_grant);
838 static int osc_should_shrink_grant(struct client_obd *client)
840 time64_t next_shrink = client->cl_next_shrink_grant;
842 if (client->cl_import == NULL)
845 if (!OCD_HAS_FLAG(&client->cl_import->imp_connect_data, GRANT_SHRINK) ||
846 client->cl_import->imp_grant_shrink_disabled)
849 if (ktime_get_seconds() >= next_shrink - 5) {
850 /* Get the current RPC size directly, instead of going via:
851 * cli_brw_size(obd->u.cli.cl_import->imp_obd->obd_self_export)
852 * Keep comment here so that it can be found by searching. */
853 int brw_size = client->cl_max_pages_per_rpc << PAGE_SHIFT;
855 if (client->cl_import->imp_state == LUSTRE_IMP_FULL &&
856 client->cl_avail_grant > brw_size)
859 osc_update_next_shrink(client);
864 #define GRANT_SHRINK_RPC_BATCH 100
866 static struct delayed_work work;
868 static void osc_grant_work_handler(struct work_struct *data)
870 struct client_obd *cli;
872 bool init_next_shrink = true;
873 time64_t next_shrink = ktime_get_seconds() + GRANT_SHRINK_INTERVAL;
876 mutex_lock(&client_gtd.gtd_mutex);
877 list_for_each_entry(cli, &client_gtd.gtd_clients,
879 if (rpc_sent < GRANT_SHRINK_RPC_BATCH &&
880 osc_should_shrink_grant(cli)) {
881 osc_shrink_grant(cli);
885 if (!init_next_shrink) {
886 if (cli->cl_next_shrink_grant < next_shrink &&
887 cli->cl_next_shrink_grant > ktime_get_seconds())
888 next_shrink = cli->cl_next_shrink_grant;
890 init_next_shrink = false;
891 next_shrink = cli->cl_next_shrink_grant;
894 mutex_unlock(&client_gtd.gtd_mutex);
896 if (client_gtd.gtd_stopped == 1)
899 if (next_shrink > ktime_get_seconds()) {
900 time64_t delay = next_shrink - ktime_get_seconds();
902 schedule_delayed_work(&work, cfs_time_seconds(delay));
904 schedule_work(&work.work);
908 void osc_schedule_grant_work(void)
910 cancel_delayed_work_sync(&work);
911 schedule_work(&work.work);
915 * Start grant thread for returing grant to server for idle clients.
917 static int osc_start_grant_work(void)
919 client_gtd.gtd_stopped = 0;
920 mutex_init(&client_gtd.gtd_mutex);
921 INIT_LIST_HEAD(&client_gtd.gtd_clients);
923 INIT_DELAYED_WORK(&work, osc_grant_work_handler);
924 schedule_work(&work.work);
929 static void osc_stop_grant_work(void)
931 client_gtd.gtd_stopped = 1;
932 cancel_delayed_work_sync(&work);
935 static void osc_add_grant_list(struct client_obd *client)
937 mutex_lock(&client_gtd.gtd_mutex);
938 list_add(&client->cl_grant_chain, &client_gtd.gtd_clients);
939 mutex_unlock(&client_gtd.gtd_mutex);
942 static void osc_del_grant_list(struct client_obd *client)
944 if (list_empty(&client->cl_grant_chain))
947 mutex_lock(&client_gtd.gtd_mutex);
948 list_del_init(&client->cl_grant_chain);
949 mutex_unlock(&client_gtd.gtd_mutex);
952 void osc_init_grant(struct client_obd *cli, struct obd_connect_data *ocd)
955 * ocd_grant is the total grant amount we're expect to hold: if we've
956 * been evicted, it's the new avail_grant amount, cl_dirty_pages will
957 * drop to 0 as inflight RPCs fail out; otherwise, it's avail_grant +
960 * race is tolerable here: if we're evicted, but imp_state already
961 * left EVICTED state, then cl_dirty_pages must be 0 already.
963 spin_lock(&cli->cl_loi_list_lock);
964 cli->cl_avail_grant = ocd->ocd_grant;
965 if (cli->cl_import->imp_state != LUSTRE_IMP_EVICTED) {
966 cli->cl_avail_grant -= cli->cl_reserved_grant;
967 if (OCD_HAS_FLAG(ocd, GRANT_PARAM))
968 cli->cl_avail_grant -= cli->cl_dirty_grant;
970 cli->cl_avail_grant -=
971 cli->cl_dirty_pages << PAGE_SHIFT;
974 if (OCD_HAS_FLAG(ocd, GRANT_PARAM)) {
978 /* overhead for each extent insertion */
979 cli->cl_grant_extent_tax = ocd->ocd_grant_tax_kb << 10;
980 /* determine the appropriate chunk size used by osc_extent. */
981 cli->cl_chunkbits = max_t(int, PAGE_SHIFT,
982 ocd->ocd_grant_blkbits);
983 /* max_pages_per_rpc must be chunk aligned */
984 chunk_mask = ~((1 << (cli->cl_chunkbits - PAGE_SHIFT)) - 1);
985 cli->cl_max_pages_per_rpc = (cli->cl_max_pages_per_rpc +
986 ~chunk_mask) & chunk_mask;
987 /* determine maximum extent size, in #pages */
988 size = (u64)ocd->ocd_grant_max_blks << ocd->ocd_grant_blkbits;
989 cli->cl_max_extent_pages = size >> PAGE_SHIFT;
990 if (cli->cl_max_extent_pages == 0)
991 cli->cl_max_extent_pages = 1;
993 cli->cl_grant_extent_tax = 0;
994 cli->cl_chunkbits = PAGE_SHIFT;
995 cli->cl_max_extent_pages = DT_MAX_BRW_PAGES;
997 spin_unlock(&cli->cl_loi_list_lock);
999 CDEBUG(D_CACHE, "%s, setting cl_avail_grant: %ld cl_lost_grant: %ld."
1000 "chunk bits: %d cl_max_extent_pages: %d\n",
1002 cli->cl_avail_grant, cli->cl_lost_grant, cli->cl_chunkbits,
1003 cli->cl_max_extent_pages);
1005 if (OCD_HAS_FLAG(ocd, GRANT_SHRINK) && list_empty(&cli->cl_grant_chain))
1006 osc_add_grant_list(cli);
1008 EXPORT_SYMBOL(osc_init_grant);
1010 /* We assume that the reason this OSC got a short read is because it read
1011 * beyond the end of a stripe file; i.e. lustre is reading a sparse file
1012 * via the LOV, and it _knows_ it's reading inside the file, it's just that
1013 * this stripe never got written at or beyond this stripe offset yet. */
1014 static void handle_short_read(int nob_read, size_t page_count,
1015 struct brw_page **pga)
1020 /* skip bytes read OK */
1021 while (nob_read > 0) {
1022 LASSERT (page_count > 0);
1024 if (pga[i]->count > nob_read) {
1025 /* EOF inside this page */
1026 ptr = kmap(pga[i]->pg) +
1027 (pga[i]->off & ~PAGE_MASK);
1028 memset(ptr + nob_read, 0, pga[i]->count - nob_read);
1035 nob_read -= pga[i]->count;
1040 /* zero remaining pages */
1041 while (page_count-- > 0) {
1042 ptr = kmap(pga[i]->pg) + (pga[i]->off & ~PAGE_MASK);
1043 memset(ptr, 0, pga[i]->count);
1049 static int check_write_rcs(struct ptlrpc_request *req,
1050 int requested_nob, int niocount,
1051 size_t page_count, struct brw_page **pga)
1056 remote_rcs = req_capsule_server_sized_get(&req->rq_pill, &RMF_RCS,
1057 sizeof(*remote_rcs) *
1059 if (remote_rcs == NULL) {
1060 CDEBUG(D_INFO, "Missing/short RC vector on BRW_WRITE reply\n");
1064 /* return error if any niobuf was in error */
1065 for (i = 0; i < niocount; i++) {
1066 if ((int)remote_rcs[i] < 0) {
1067 CDEBUG(D_INFO, "rc[%d]: %d req %p\n",
1068 i, remote_rcs[i], req);
1069 return remote_rcs[i];
1072 if (remote_rcs[i] != 0) {
1073 CDEBUG(D_INFO, "rc[%d] invalid (%d) req %p\n",
1074 i, remote_rcs[i], req);
1078 if (req->rq_bulk != NULL &&
1079 req->rq_bulk->bd_nob_transferred != requested_nob) {
1080 CERROR("Unexpected # bytes transferred: %d (requested %d)\n",
1081 req->rq_bulk->bd_nob_transferred, requested_nob);
1088 static inline int can_merge_pages(struct brw_page *p1, struct brw_page *p2)
1090 if (p1->flag != p2->flag) {
1091 unsigned mask = ~(OBD_BRW_FROM_GRANT | OBD_BRW_NOCACHE |
1092 OBD_BRW_SYNC | OBD_BRW_ASYNC |
1093 OBD_BRW_NOQUOTA | OBD_BRW_SOFT_SYNC);
1095 /* warn if we try to combine flags that we don't know to be
1096 * safe to combine */
1097 if (unlikely((p1->flag & mask) != (p2->flag & mask))) {
1098 CWARN("Saw flags 0x%x and 0x%x in the same brw, please "
1099 "report this at https://jira.whamcloud.com/\n",
1100 p1->flag, p2->flag);
1105 return (p1->off + p1->count == p2->off);
1108 #if IS_ENABLED(CONFIG_CRC_T10DIF)
1109 static int osc_checksum_bulk_t10pi(const char *obd_name, int nob,
1110 size_t pg_count, struct brw_page **pga,
1111 int opc, obd_dif_csum_fn *fn,
1115 struct ahash_request *req;
1116 /* Used Adler as the default checksum type on top of DIF tags */
1117 unsigned char cfs_alg = cksum_obd2cfs(OBD_CKSUM_T10_TOP);
1118 struct page *__page;
1119 unsigned char *buffer;
1121 unsigned int bufsize;
1123 int used_number = 0;
1129 LASSERT(pg_count > 0);
1131 __page = alloc_page(GFP_KERNEL);
1135 req = cfs_crypto_hash_init(cfs_alg, NULL, 0);
1138 CERROR("%s: unable to initialize checksum hash %s: rc = %d\n",
1139 obd_name, cfs_crypto_hash_name(cfs_alg), rc);
1143 buffer = kmap(__page);
1144 guard_start = (__u16 *)buffer;
1145 guard_number = PAGE_SIZE / sizeof(*guard_start);
1146 while (nob > 0 && pg_count > 0) {
1147 unsigned int count = pga[i]->count > nob ? nob : pga[i]->count;
1149 /* corrupt the data before we compute the checksum, to
1150 * simulate an OST->client data error */
1151 if (unlikely(i == 0 && opc == OST_READ &&
1152 OBD_FAIL_CHECK(OBD_FAIL_OSC_CHECKSUM_RECEIVE))) {
1153 unsigned char *ptr = kmap(pga[i]->pg);
1154 int off = pga[i]->off & ~PAGE_MASK;
1156 memcpy(ptr + off, "bad1", min_t(typeof(nob), 4, nob));
1161 * The left guard number should be able to hold checksums of a
1164 rc = obd_page_dif_generate_buffer(obd_name, pga[i]->pg,
1165 pga[i]->off & ~PAGE_MASK,
1167 guard_start + used_number,
1168 guard_number - used_number,
1174 used_number += used;
1175 if (used_number == guard_number) {
1176 cfs_crypto_hash_update_page(req, __page, 0,
1177 used_number * sizeof(*guard_start));
1181 nob -= pga[i]->count;
1189 if (used_number != 0)
1190 cfs_crypto_hash_update_page(req, __page, 0,
1191 used_number * sizeof(*guard_start));
1193 bufsize = sizeof(cksum);
1194 cfs_crypto_hash_final(req, (unsigned char *)&cksum, &bufsize);
1196 /* For sending we only compute the wrong checksum instead
1197 * of corrupting the data so it is still correct on a redo */
1198 if (opc == OST_WRITE && OBD_FAIL_CHECK(OBD_FAIL_OSC_CHECKSUM_SEND))
1203 __free_page(__page);
1206 #else /* !CONFIG_CRC_T10DIF */
1207 #define obd_dif_ip_fn NULL
1208 #define obd_dif_crc_fn NULL
1209 #define osc_checksum_bulk_t10pi(name, nob, pgc, pga, opc, fn, ssize, csum) \
1211 #endif /* CONFIG_CRC_T10DIF */
1213 static int osc_checksum_bulk(int nob, size_t pg_count,
1214 struct brw_page **pga, int opc,
1215 enum cksum_types cksum_type,
1219 struct ahash_request *req;
1220 unsigned int bufsize;
1221 unsigned char cfs_alg = cksum_obd2cfs(cksum_type);
1223 LASSERT(pg_count > 0);
1225 req = cfs_crypto_hash_init(cfs_alg, NULL, 0);
1227 CERROR("Unable to initialize checksum hash %s\n",
1228 cfs_crypto_hash_name(cfs_alg));
1229 return PTR_ERR(req);
1232 while (nob > 0 && pg_count > 0) {
1233 unsigned int count = pga[i]->count > nob ? nob : pga[i]->count;
1235 /* corrupt the data before we compute the checksum, to
1236 * simulate an OST->client data error */
1237 if (i == 0 && opc == OST_READ &&
1238 OBD_FAIL_CHECK(OBD_FAIL_OSC_CHECKSUM_RECEIVE)) {
1239 unsigned char *ptr = kmap(pga[i]->pg);
1240 int off = pga[i]->off & ~PAGE_MASK;
1242 memcpy(ptr + off, "bad1", min_t(typeof(nob), 4, nob));
1245 cfs_crypto_hash_update_page(req, pga[i]->pg,
1246 pga[i]->off & ~PAGE_MASK,
1248 LL_CDEBUG_PAGE(D_PAGE, pga[i]->pg, "off %d\n",
1249 (int)(pga[i]->off & ~PAGE_MASK));
1251 nob -= pga[i]->count;
1256 bufsize = sizeof(*cksum);
1257 cfs_crypto_hash_final(req, (unsigned char *)cksum, &bufsize);
1259 /* For sending we only compute the wrong checksum instead
1260 * of corrupting the data so it is still correct on a redo */
1261 if (opc == OST_WRITE && OBD_FAIL_CHECK(OBD_FAIL_OSC_CHECKSUM_SEND))
1267 static int osc_checksum_bulk_rw(const char *obd_name,
1268 enum cksum_types cksum_type,
1269 int nob, size_t pg_count,
1270 struct brw_page **pga, int opc,
1273 obd_dif_csum_fn *fn = NULL;
1274 int sector_size = 0;
1278 obd_t10_cksum2dif(cksum_type, &fn, §or_size);
1281 rc = osc_checksum_bulk_t10pi(obd_name, nob, pg_count, pga,
1282 opc, fn, sector_size, check_sum);
1284 rc = osc_checksum_bulk(nob, pg_count, pga, opc, cksum_type,
1291 osc_brw_prep_request(int cmd, struct client_obd *cli, struct obdo *oa,
1292 u32 page_count, struct brw_page **pga,
1293 struct ptlrpc_request **reqp, int resend)
1295 struct ptlrpc_request *req;
1296 struct ptlrpc_bulk_desc *desc;
1297 struct ost_body *body;
1298 struct obd_ioobj *ioobj;
1299 struct niobuf_remote *niobuf;
1300 int niocount, i, requested_nob, opc, rc, short_io_size = 0;
1301 struct osc_brw_async_args *aa;
1302 struct req_capsule *pill;
1303 struct brw_page *pg_prev;
1305 const char *obd_name = cli->cl_import->imp_obd->obd_name;
1308 if (OBD_FAIL_CHECK(OBD_FAIL_OSC_BRW_PREP_REQ))
1309 RETURN(-ENOMEM); /* Recoverable */
1310 if (OBD_FAIL_CHECK(OBD_FAIL_OSC_BRW_PREP_REQ2))
1311 RETURN(-EINVAL); /* Fatal */
1313 if ((cmd & OBD_BRW_WRITE) != 0) {
1315 req = ptlrpc_request_alloc_pool(cli->cl_import,
1317 &RQF_OST_BRW_WRITE);
1320 req = ptlrpc_request_alloc(cli->cl_import, &RQF_OST_BRW_READ);
1325 for (niocount = i = 1; i < page_count; i++) {
1326 if (!can_merge_pages(pga[i - 1], pga[i]))
1330 pill = &req->rq_pill;
1331 req_capsule_set_size(pill, &RMF_OBD_IOOBJ, RCL_CLIENT,
1333 req_capsule_set_size(pill, &RMF_NIOBUF_REMOTE, RCL_CLIENT,
1334 niocount * sizeof(*niobuf));
1336 for (i = 0; i < page_count; i++)
1337 short_io_size += pga[i]->count;
1339 /* Check if read/write is small enough to be a short io. */
1340 if (short_io_size > cli->cl_max_short_io_bytes || niocount > 1 ||
1341 !imp_connect_shortio(cli->cl_import))
1344 req_capsule_set_size(pill, &RMF_SHORT_IO, RCL_CLIENT,
1345 opc == OST_READ ? 0 : short_io_size);
1346 if (opc == OST_READ)
1347 req_capsule_set_size(pill, &RMF_SHORT_IO, RCL_SERVER,
1350 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, opc);
1352 ptlrpc_request_free(req);
1355 osc_set_io_portal(req);
1357 ptlrpc_at_set_req_timeout(req);
1358 /* ask ptlrpc not to resend on EINPROGRESS since BRWs have their own
1360 req->rq_no_retry_einprogress = 1;
1362 if (short_io_size != 0) {
1364 short_io_buf = NULL;
1368 desc = ptlrpc_prep_bulk_imp(req, page_count,
1369 cli->cl_import->imp_connect_data.ocd_brw_size >> LNET_MTU_BITS,
1370 (opc == OST_WRITE ? PTLRPC_BULK_GET_SOURCE :
1371 PTLRPC_BULK_PUT_SINK) |
1372 PTLRPC_BULK_BUF_KIOV,
1374 &ptlrpc_bulk_kiov_pin_ops);
1377 GOTO(out, rc = -ENOMEM);
1378 /* NB request now owns desc and will free it when it gets freed */
1380 body = req_capsule_client_get(pill, &RMF_OST_BODY);
1381 ioobj = req_capsule_client_get(pill, &RMF_OBD_IOOBJ);
1382 niobuf = req_capsule_client_get(pill, &RMF_NIOBUF_REMOTE);
1383 LASSERT(body != NULL && ioobj != NULL && niobuf != NULL);
1385 lustre_set_wire_obdo(&req->rq_import->imp_connect_data, &body->oa, oa);
1387 /* For READ and WRITE, we can't fill o_uid and o_gid using from_kuid()
1388 * and from_kgid(), because they are asynchronous. Fortunately, variable
1389 * oa contains valid o_uid and o_gid in these two operations.
1390 * Besides, filling o_uid and o_gid is enough for nrs-tbf, see LU-9658.
1391 * OBD_MD_FLUID and OBD_MD_FLUID is not set in order to avoid breaking
1392 * other process logic */
1393 body->oa.o_uid = oa->o_uid;
1394 body->oa.o_gid = oa->o_gid;
1396 obdo_to_ioobj(oa, ioobj);
1397 ioobj->ioo_bufcnt = niocount;
1398 /* The high bits of ioo_max_brw tells server _maximum_ number of bulks
1399 * that might be send for this request. The actual number is decided
1400 * when the RPC is finally sent in ptlrpc_register_bulk(). It sends
1401 * "max - 1" for old client compatibility sending "0", and also so the
1402 * the actual maximum is a power-of-two number, not one less. LU-1431 */
1404 ioobj_max_brw_set(ioobj, desc->bd_md_max_brw);
1406 ioobj_max_brw_set(ioobj, 0);
1408 if (short_io_size != 0) {
1409 if ((body->oa.o_valid & OBD_MD_FLFLAGS) == 0) {
1410 body->oa.o_valid |= OBD_MD_FLFLAGS;
1411 body->oa.o_flags = 0;
1413 body->oa.o_flags |= OBD_FL_SHORT_IO;
1414 CDEBUG(D_CACHE, "Using short io for data transfer, size = %d\n",
1416 if (opc == OST_WRITE) {
1417 short_io_buf = req_capsule_client_get(pill,
1419 LASSERT(short_io_buf != NULL);
1423 LASSERT(page_count > 0);
1425 for (requested_nob = i = 0; i < page_count; i++, niobuf++) {
1426 struct brw_page *pg = pga[i];
1427 int poff = pg->off & ~PAGE_MASK;
1429 LASSERT(pg->count > 0);
1430 /* make sure there is no gap in the middle of page array */
1431 LASSERTF(page_count == 1 ||
1432 (ergo(i == 0, poff + pg->count == PAGE_SIZE) &&
1433 ergo(i > 0 && i < page_count - 1,
1434 poff == 0 && pg->count == PAGE_SIZE) &&
1435 ergo(i == page_count - 1, poff == 0)),
1436 "i: %d/%d pg: %p off: %llu, count: %u\n",
1437 i, page_count, pg, pg->off, pg->count);
1438 LASSERTF(i == 0 || pg->off > pg_prev->off,
1439 "i %d p_c %u pg %p [pri %lu ind %lu] off %llu"
1440 " prev_pg %p [pri %lu ind %lu] off %llu\n",
1442 pg->pg, page_private(pg->pg), pg->pg->index, pg->off,
1443 pg_prev->pg, page_private(pg_prev->pg),
1444 pg_prev->pg->index, pg_prev->off);
1445 LASSERT((pga[0]->flag & OBD_BRW_SRVLOCK) ==
1446 (pg->flag & OBD_BRW_SRVLOCK));
1447 if (short_io_size != 0 && opc == OST_WRITE) {
1448 unsigned char *ptr = ll_kmap_atomic(pg->pg, KM_USER0);
1450 LASSERT(short_io_size >= requested_nob + pg->count);
1451 memcpy(short_io_buf + requested_nob,
1454 ll_kunmap_atomic(ptr, KM_USER0);
1455 } else if (short_io_size == 0) {
1456 desc->bd_frag_ops->add_kiov_frag(desc, pg->pg, poff,
1459 requested_nob += pg->count;
1461 if (i > 0 && can_merge_pages(pg_prev, pg)) {
1463 niobuf->rnb_len += pg->count;
1465 niobuf->rnb_offset = pg->off;
1466 niobuf->rnb_len = pg->count;
1467 niobuf->rnb_flags = pg->flag;
1472 LASSERTF((void *)(niobuf - niocount) ==
1473 req_capsule_client_get(&req->rq_pill, &RMF_NIOBUF_REMOTE),
1474 "want %p - real %p\n", req_capsule_client_get(&req->rq_pill,
1475 &RMF_NIOBUF_REMOTE), (void *)(niobuf - niocount));
1477 osc_announce_cached(cli, &body->oa, opc == OST_WRITE ? requested_nob:0);
1479 if ((body->oa.o_valid & OBD_MD_FLFLAGS) == 0) {
1480 body->oa.o_valid |= OBD_MD_FLFLAGS;
1481 body->oa.o_flags = 0;
1483 body->oa.o_flags |= OBD_FL_RECOV_RESEND;
1486 if (osc_should_shrink_grant(cli))
1487 osc_shrink_grant_local(cli, &body->oa);
1489 /* size[REQ_REC_OFF] still sizeof (*body) */
1490 if (opc == OST_WRITE) {
1491 if (cli->cl_checksum &&
1492 !sptlrpc_flavor_has_bulk(&req->rq_flvr)) {
1493 /* store cl_cksum_type in a local variable since
1494 * it can be changed via lprocfs */
1495 enum cksum_types cksum_type = cli->cl_cksum_type;
1497 if ((body->oa.o_valid & OBD_MD_FLFLAGS) == 0)
1498 body->oa.o_flags = 0;
1500 body->oa.o_flags |= obd_cksum_type_pack(obd_name,
1502 body->oa.o_valid |= OBD_MD_FLCKSUM | OBD_MD_FLFLAGS;
1504 rc = osc_checksum_bulk_rw(obd_name, cksum_type,
1505 requested_nob, page_count,
1509 CDEBUG(D_PAGE, "failed to checksum, rc = %d\n",
1513 CDEBUG(D_PAGE, "checksum at write origin: %x\n",
1516 /* save this in 'oa', too, for later checking */
1517 oa->o_valid |= OBD_MD_FLCKSUM | OBD_MD_FLFLAGS;
1518 oa->o_flags |= obd_cksum_type_pack(obd_name,
1521 /* clear out the checksum flag, in case this is a
1522 * resend but cl_checksum is no longer set. b=11238 */
1523 oa->o_valid &= ~OBD_MD_FLCKSUM;
1525 oa->o_cksum = body->oa.o_cksum;
1526 /* 1 RC per niobuf */
1527 req_capsule_set_size(pill, &RMF_RCS, RCL_SERVER,
1528 sizeof(__u32) * niocount);
1530 if (cli->cl_checksum &&
1531 !sptlrpc_flavor_has_bulk(&req->rq_flvr)) {
1532 if ((body->oa.o_valid & OBD_MD_FLFLAGS) == 0)
1533 body->oa.o_flags = 0;
1534 body->oa.o_flags |= obd_cksum_type_pack(obd_name,
1535 cli->cl_cksum_type);
1536 body->oa.o_valid |= OBD_MD_FLCKSUM | OBD_MD_FLFLAGS;
1539 /* Client cksum has been already copied to wire obdo in previous
1540 * lustre_set_wire_obdo(), and in the case a bulk-read is being
1541 * resent due to cksum error, this will allow Server to
1542 * check+dump pages on its side */
1544 ptlrpc_request_set_replen(req);
1546 aa = ptlrpc_req_async_args(aa, req);
1548 aa->aa_requested_nob = requested_nob;
1549 aa->aa_nio_count = niocount;
1550 aa->aa_page_count = page_count;
1554 INIT_LIST_HEAD(&aa->aa_oaps);
1557 niobuf = req_capsule_client_get(pill, &RMF_NIOBUF_REMOTE);
1558 CDEBUG(D_RPCTRACE, "brw rpc %p - object "DOSTID" offset %lld<>%lld\n",
1559 req, POSTID(&oa->o_oi), niobuf[0].rnb_offset,
1560 niobuf[niocount - 1].rnb_offset + niobuf[niocount - 1].rnb_len);
1564 ptlrpc_req_finished(req);
1568 char dbgcksum_file_name[PATH_MAX];
1570 static void dump_all_bulk_pages(struct obdo *oa, __u32 page_count,
1571 struct brw_page **pga, __u32 server_cksum,
1579 /* will only keep dump of pages on first error for the same range in
1580 * file/fid, not during the resends/retries. */
1581 snprintf(dbgcksum_file_name, sizeof(dbgcksum_file_name),
1582 "%s-checksum_dump-osc-"DFID":[%llu-%llu]-%x-%x",
1583 (strncmp(libcfs_debug_file_path_arr, "NONE", 4) != 0 ?
1584 libcfs_debug_file_path_arr :
1585 LIBCFS_DEBUG_FILE_PATH_DEFAULT),
1586 oa->o_valid & OBD_MD_FLFID ? oa->o_parent_seq : 0ULL,
1587 oa->o_valid & OBD_MD_FLFID ? oa->o_parent_oid : 0,
1588 oa->o_valid & OBD_MD_FLFID ? oa->o_parent_ver : 0,
1590 pga[page_count-1]->off + pga[page_count-1]->count - 1,
1591 client_cksum, server_cksum);
1592 filp = filp_open(dbgcksum_file_name,
1593 O_CREAT | O_EXCL | O_WRONLY | O_LARGEFILE, 0600);
1597 CDEBUG(D_INFO, "%s: can't open to dump pages with "
1598 "checksum error: rc = %d\n", dbgcksum_file_name,
1601 CERROR("%s: can't open to dump pages with checksum "
1602 "error: rc = %d\n", dbgcksum_file_name, rc);
1606 for (i = 0; i < page_count; i++) {
1607 len = pga[i]->count;
1608 buf = kmap(pga[i]->pg);
1610 rc = cfs_kernel_write(filp, buf, len, &filp->f_pos);
1612 CERROR("%s: wanted to write %u but got %d "
1613 "error\n", dbgcksum_file_name, len, rc);
1618 CDEBUG(D_INFO, "%s: wrote %d bytes\n",
1619 dbgcksum_file_name, rc);
1624 rc = vfs_fsync_range(filp, 0, LLONG_MAX, 1);
1626 CERROR("%s: sync returns %d\n", dbgcksum_file_name, rc);
1627 filp_close(filp, NULL);
1631 check_write_checksum(struct obdo *oa, const struct lnet_process_id *peer,
1632 __u32 client_cksum, __u32 server_cksum,
1633 struct osc_brw_async_args *aa)
1635 const char *obd_name = aa->aa_cli->cl_import->imp_obd->obd_name;
1636 enum cksum_types cksum_type;
1637 obd_dif_csum_fn *fn = NULL;
1638 int sector_size = 0;
1643 if (server_cksum == client_cksum) {
1644 CDEBUG(D_PAGE, "checksum %x confirmed\n", client_cksum);
1648 if (aa->aa_cli->cl_checksum_dump)
1649 dump_all_bulk_pages(oa, aa->aa_page_count, aa->aa_ppga,
1650 server_cksum, client_cksum);
1652 cksum_type = obd_cksum_type_unpack(oa->o_valid & OBD_MD_FLFLAGS ?
1655 switch (cksum_type) {
1656 case OBD_CKSUM_T10IP512:
1660 case OBD_CKSUM_T10IP4K:
1664 case OBD_CKSUM_T10CRC512:
1665 fn = obd_dif_crc_fn;
1668 case OBD_CKSUM_T10CRC4K:
1669 fn = obd_dif_crc_fn;
1677 rc = osc_checksum_bulk_t10pi(obd_name, aa->aa_requested_nob,
1678 aa->aa_page_count, aa->aa_ppga,
1679 OST_WRITE, fn, sector_size,
1682 rc = osc_checksum_bulk(aa->aa_requested_nob, aa->aa_page_count,
1683 aa->aa_ppga, OST_WRITE, cksum_type,
1687 msg = "failed to calculate the client write checksum";
1688 else if (cksum_type != obd_cksum_type_unpack(aa->aa_oa->o_flags))
1689 msg = "the server did not use the checksum type specified in "
1690 "the original request - likely a protocol problem";
1691 else if (new_cksum == server_cksum)
1692 msg = "changed on the client after we checksummed it - "
1693 "likely false positive due to mmap IO (bug 11742)";
1694 else if (new_cksum == client_cksum)
1695 msg = "changed in transit before arrival at OST";
1697 msg = "changed in transit AND doesn't match the original - "
1698 "likely false positive due to mmap IO (bug 11742)";
1700 LCONSOLE_ERROR_MSG(0x132, "%s: BAD WRITE CHECKSUM: %s: from %s inode "
1701 DFID " object "DOSTID" extent [%llu-%llu], original "
1702 "client csum %x (type %x), server csum %x (type %x),"
1703 " client csum now %x\n",
1704 obd_name, msg, libcfs_nid2str(peer->nid),
1705 oa->o_valid & OBD_MD_FLFID ? oa->o_parent_seq : (__u64)0,
1706 oa->o_valid & OBD_MD_FLFID ? oa->o_parent_oid : 0,
1707 oa->o_valid & OBD_MD_FLFID ? oa->o_parent_ver : 0,
1708 POSTID(&oa->o_oi), aa->aa_ppga[0]->off,
1709 aa->aa_ppga[aa->aa_page_count - 1]->off +
1710 aa->aa_ppga[aa->aa_page_count-1]->count - 1,
1712 obd_cksum_type_unpack(aa->aa_oa->o_flags),
1713 server_cksum, cksum_type, new_cksum);
1717 /* Note rc enters this function as number of bytes transferred */
1718 static int osc_brw_fini_request(struct ptlrpc_request *req, int rc)
1720 struct osc_brw_async_args *aa = (void *)&req->rq_async_args;
1721 struct client_obd *cli = aa->aa_cli;
1722 const char *obd_name = cli->cl_import->imp_obd->obd_name;
1723 const struct lnet_process_id *peer =
1724 &req->rq_import->imp_connection->c_peer;
1725 struct ost_body *body;
1726 u32 client_cksum = 0;
1730 if (rc < 0 && rc != -EDQUOT) {
1731 DEBUG_REQ(D_INFO, req, "Failed request: rc = %d", rc);
1735 LASSERTF(req->rq_repmsg != NULL, "rc = %d\n", rc);
1736 body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
1738 DEBUG_REQ(D_INFO, req, "cannot unpack body");
1742 /* set/clear over quota flag for a uid/gid/projid */
1743 if (lustre_msg_get_opc(req->rq_reqmsg) == OST_WRITE &&
1744 body->oa.o_valid & (OBD_MD_FLALLQUOTA)) {
1745 unsigned qid[LL_MAXQUOTAS] = {
1746 body->oa.o_uid, body->oa.o_gid,
1747 body->oa.o_projid };
1749 "setdq for [%u %u %u] with valid %#llx, flags %x\n",
1750 body->oa.o_uid, body->oa.o_gid, body->oa.o_projid,
1751 body->oa.o_valid, body->oa.o_flags);
1752 osc_quota_setdq(cli, req->rq_xid, qid, body->oa.o_valid,
1756 osc_update_grant(cli, body);
1761 if (aa->aa_oa->o_valid & OBD_MD_FLCKSUM)
1762 client_cksum = aa->aa_oa->o_cksum; /* save for later */
1764 if (lustre_msg_get_opc(req->rq_reqmsg) == OST_WRITE) {
1766 CERROR("%s: unexpected positive size %d\n",
1771 if (req->rq_bulk != NULL &&
1772 sptlrpc_cli_unwrap_bulk_write(req, req->rq_bulk))
1775 if ((aa->aa_oa->o_valid & OBD_MD_FLCKSUM) && client_cksum &&
1776 check_write_checksum(&body->oa, peer, client_cksum,
1777 body->oa.o_cksum, aa))
1780 rc = check_write_rcs(req, aa->aa_requested_nob,
1781 aa->aa_nio_count, aa->aa_page_count,
1786 /* The rest of this function executes only for OST_READs */
1788 if (req->rq_bulk == NULL) {
1789 rc = req_capsule_get_size(&req->rq_pill, &RMF_SHORT_IO,
1791 LASSERT(rc == req->rq_status);
1793 /* if unwrap_bulk failed, return -EAGAIN to retry */
1794 rc = sptlrpc_cli_unwrap_bulk_read(req, req->rq_bulk, rc);
1797 GOTO(out, rc = -EAGAIN);
1799 if (rc > aa->aa_requested_nob) {
1800 CERROR("%s: unexpected size %d, requested %d\n", obd_name,
1801 rc, aa->aa_requested_nob);
1805 if (req->rq_bulk != NULL && rc != req->rq_bulk->bd_nob_transferred) {
1806 CERROR("%s: unexpected size %d, transferred %d\n", obd_name,
1807 rc, req->rq_bulk->bd_nob_transferred);
1811 if (req->rq_bulk == NULL) {
1813 int nob, pg_count, i = 0;
1816 CDEBUG(D_CACHE, "Using short io read, size %d\n", rc);
1817 pg_count = aa->aa_page_count;
1818 buf = req_capsule_server_sized_get(&req->rq_pill, &RMF_SHORT_IO,
1821 while (nob > 0 && pg_count > 0) {
1823 int count = aa->aa_ppga[i]->count > nob ?
1824 nob : aa->aa_ppga[i]->count;
1826 CDEBUG(D_CACHE, "page %p count %d\n",
1827 aa->aa_ppga[i]->pg, count);
1828 ptr = ll_kmap_atomic(aa->aa_ppga[i]->pg, KM_USER0);
1829 memcpy(ptr + (aa->aa_ppga[i]->off & ~PAGE_MASK), buf,
1831 ll_kunmap_atomic((void *) ptr, KM_USER0);
1840 if (rc < aa->aa_requested_nob)
1841 handle_short_read(rc, aa->aa_page_count, aa->aa_ppga);
1843 if (body->oa.o_valid & OBD_MD_FLCKSUM) {
1844 static int cksum_counter;
1845 u32 server_cksum = body->oa.o_cksum;
1848 enum cksum_types cksum_type;
1849 u32 o_flags = body->oa.o_valid & OBD_MD_FLFLAGS ?
1850 body->oa.o_flags : 0;
1852 cksum_type = obd_cksum_type_unpack(o_flags);
1853 rc = osc_checksum_bulk_rw(obd_name, cksum_type, rc,
1854 aa->aa_page_count, aa->aa_ppga,
1855 OST_READ, &client_cksum);
1859 if (req->rq_bulk != NULL &&
1860 peer->nid != req->rq_bulk->bd_sender) {
1862 router = libcfs_nid2str(req->rq_bulk->bd_sender);
1865 if (server_cksum != client_cksum) {
1866 struct ost_body *clbody;
1867 u32 page_count = aa->aa_page_count;
1869 clbody = req_capsule_client_get(&req->rq_pill,
1871 if (cli->cl_checksum_dump)
1872 dump_all_bulk_pages(&clbody->oa, page_count,
1873 aa->aa_ppga, server_cksum,
1876 LCONSOLE_ERROR_MSG(0x133, "%s: BAD READ CHECKSUM: from "
1877 "%s%s%s inode "DFID" object "DOSTID
1878 " extent [%llu-%llu], client %x, "
1879 "server %x, cksum_type %x\n",
1881 libcfs_nid2str(peer->nid),
1883 clbody->oa.o_valid & OBD_MD_FLFID ?
1884 clbody->oa.o_parent_seq : 0ULL,
1885 clbody->oa.o_valid & OBD_MD_FLFID ?
1886 clbody->oa.o_parent_oid : 0,
1887 clbody->oa.o_valid & OBD_MD_FLFID ?
1888 clbody->oa.o_parent_ver : 0,
1889 POSTID(&body->oa.o_oi),
1890 aa->aa_ppga[0]->off,
1891 aa->aa_ppga[page_count-1]->off +
1892 aa->aa_ppga[page_count-1]->count - 1,
1893 client_cksum, server_cksum,
1896 aa->aa_oa->o_cksum = client_cksum;
1900 CDEBUG(D_PAGE, "checksum %x confirmed\n", client_cksum);
1903 } else if (unlikely(client_cksum)) {
1904 static int cksum_missed;
1907 if ((cksum_missed & (-cksum_missed)) == cksum_missed)
1908 CERROR("%s: checksum %u requested from %s but not sent\n",
1909 obd_name, cksum_missed,
1910 libcfs_nid2str(peer->nid));
1916 lustre_get_wire_obdo(&req->rq_import->imp_connect_data,
1917 aa->aa_oa, &body->oa);
1922 static int osc_brw_redo_request(struct ptlrpc_request *request,
1923 struct osc_brw_async_args *aa, int rc)
1925 struct ptlrpc_request *new_req;
1926 struct osc_brw_async_args *new_aa;
1927 struct osc_async_page *oap;
1930 /* The below message is checked in replay-ost-single.sh test_8ae*/
1931 DEBUG_REQ(rc == -EINPROGRESS ? D_RPCTRACE : D_ERROR, request,
1932 "redo for recoverable error %d", rc);
1934 rc = osc_brw_prep_request(lustre_msg_get_opc(request->rq_reqmsg) ==
1935 OST_WRITE ? OBD_BRW_WRITE : OBD_BRW_READ,
1936 aa->aa_cli, aa->aa_oa, aa->aa_page_count,
1937 aa->aa_ppga, &new_req, 1);
1941 list_for_each_entry(oap, &aa->aa_oaps, oap_rpc_item) {
1942 if (oap->oap_request != NULL) {
1943 LASSERTF(request == oap->oap_request,
1944 "request %p != oap_request %p\n",
1945 request, oap->oap_request);
1946 if (oap->oap_interrupted) {
1947 ptlrpc_req_finished(new_req);
1953 * New request takes over pga and oaps from old request.
1954 * Note that copying a list_head doesn't work, need to move it...
1957 new_req->rq_interpret_reply = request->rq_interpret_reply;
1958 new_req->rq_async_args = request->rq_async_args;
1959 new_req->rq_commit_cb = request->rq_commit_cb;
1960 /* cap resend delay to the current request timeout, this is similar to
1961 * what ptlrpc does (see after_reply()) */
1962 if (aa->aa_resends > new_req->rq_timeout)
1963 new_req->rq_sent = ktime_get_real_seconds() + new_req->rq_timeout;
1965 new_req->rq_sent = ktime_get_real_seconds() + aa->aa_resends;
1966 new_req->rq_generation_set = 1;
1967 new_req->rq_import_generation = request->rq_import_generation;
1969 new_aa = ptlrpc_req_async_args(new_aa, new_req);
1971 INIT_LIST_HEAD(&new_aa->aa_oaps);
1972 list_splice_init(&aa->aa_oaps, &new_aa->aa_oaps);
1973 INIT_LIST_HEAD(&new_aa->aa_exts);
1974 list_splice_init(&aa->aa_exts, &new_aa->aa_exts);
1975 new_aa->aa_resends = aa->aa_resends;
1977 list_for_each_entry(oap, &new_aa->aa_oaps, oap_rpc_item) {
1978 if (oap->oap_request) {
1979 ptlrpc_req_finished(oap->oap_request);
1980 oap->oap_request = ptlrpc_request_addref(new_req);
1984 /* XXX: This code will run into problem if we're going to support
1985 * to add a series of BRW RPCs into a self-defined ptlrpc_request_set
1986 * and wait for all of them to be finished. We should inherit request
1987 * set from old request. */
1988 ptlrpcd_add_req(new_req);
1990 DEBUG_REQ(D_INFO, new_req, "new request");
1995 * ugh, we want disk allocation on the target to happen in offset order. we'll
1996 * follow sedgewicks advice and stick to the dead simple shellsort -- it'll do
1997 * fine for our small page arrays and doesn't require allocation. its an
1998 * insertion sort that swaps elements that are strides apart, shrinking the
1999 * stride down until its '1' and the array is sorted.
2001 static void sort_brw_pages(struct brw_page **array, int num)
2004 struct brw_page *tmp;
2008 for (stride = 1; stride < num ; stride = (stride * 3) + 1)
2013 for (i = stride ; i < num ; i++) {
2016 while (j >= stride && array[j - stride]->off > tmp->off) {
2017 array[j] = array[j - stride];
2022 } while (stride > 1);
2025 static void osc_release_ppga(struct brw_page **ppga, size_t count)
2027 LASSERT(ppga != NULL);
2028 OBD_FREE(ppga, sizeof(*ppga) * count);
2031 static int brw_interpret(const struct lu_env *env,
2032 struct ptlrpc_request *req, void *args, int rc)
2034 struct osc_brw_async_args *aa = args;
2035 struct osc_extent *ext;
2036 struct osc_extent *tmp;
2037 struct client_obd *cli = aa->aa_cli;
2038 unsigned long transferred = 0;
2042 rc = osc_brw_fini_request(req, rc);
2043 CDEBUG(D_INODE, "request %p aa %p rc %d\n", req, aa, rc);
2045 * When server returns -EINPROGRESS, client should always retry
2046 * regardless of the number of times the bulk was resent already.
2048 if (osc_recoverable_error(rc) && !req->rq_no_delay) {
2049 if (req->rq_import_generation !=
2050 req->rq_import->imp_generation) {
2051 CDEBUG(D_HA, "%s: resend cross eviction for object: "
2052 ""DOSTID", rc = %d.\n",
2053 req->rq_import->imp_obd->obd_name,
2054 POSTID(&aa->aa_oa->o_oi), rc);
2055 } else if (rc == -EINPROGRESS ||
2056 client_should_resend(aa->aa_resends, aa->aa_cli)) {
2057 rc = osc_brw_redo_request(req, aa, rc);
2059 CERROR("%s: too many resent retries for object: "
2060 "%llu:%llu, rc = %d.\n",
2061 req->rq_import->imp_obd->obd_name,
2062 POSTID(&aa->aa_oa->o_oi), rc);
2067 else if (rc == -EAGAIN || rc == -EINPROGRESS)
2072 struct obdo *oa = aa->aa_oa;
2073 struct cl_attr *attr = &osc_env_info(env)->oti_attr;
2074 unsigned long valid = 0;
2075 struct cl_object *obj;
2076 struct osc_async_page *last;
2078 last = brw_page2oap(aa->aa_ppga[aa->aa_page_count - 1]);
2079 obj = osc2cl(last->oap_obj);
2081 cl_object_attr_lock(obj);
2082 if (oa->o_valid & OBD_MD_FLBLOCKS) {
2083 attr->cat_blocks = oa->o_blocks;
2084 valid |= CAT_BLOCKS;
2086 if (oa->o_valid & OBD_MD_FLMTIME) {
2087 attr->cat_mtime = oa->o_mtime;
2090 if (oa->o_valid & OBD_MD_FLATIME) {
2091 attr->cat_atime = oa->o_atime;
2094 if (oa->o_valid & OBD_MD_FLCTIME) {
2095 attr->cat_ctime = oa->o_ctime;
2099 if (lustre_msg_get_opc(req->rq_reqmsg) == OST_WRITE) {
2100 struct lov_oinfo *loi = cl2osc(obj)->oo_oinfo;
2101 loff_t last_off = last->oap_count + last->oap_obj_off +
2104 /* Change file size if this is an out of quota or
2105 * direct IO write and it extends the file size */
2106 if (loi->loi_lvb.lvb_size < last_off) {
2107 attr->cat_size = last_off;
2110 /* Extend KMS if it's not a lockless write */
2111 if (loi->loi_kms < last_off &&
2112 oap2osc_page(last)->ops_srvlock == 0) {
2113 attr->cat_kms = last_off;
2119 cl_object_attr_update(env, obj, attr, valid);
2120 cl_object_attr_unlock(obj);
2122 OBD_SLAB_FREE_PTR(aa->aa_oa, osc_obdo_kmem);
2125 if (lustre_msg_get_opc(req->rq_reqmsg) == OST_WRITE && rc == 0)
2126 osc_inc_unstable_pages(req);
2128 list_for_each_entry_safe(ext, tmp, &aa->aa_exts, oe_link) {
2129 list_del_init(&ext->oe_link);
2130 osc_extent_finish(env, ext, 1,
2131 rc && req->rq_no_delay ? -EWOULDBLOCK : rc);
2133 LASSERT(list_empty(&aa->aa_exts));
2134 LASSERT(list_empty(&aa->aa_oaps));
2136 transferred = (req->rq_bulk == NULL ? /* short io */
2137 aa->aa_requested_nob :
2138 req->rq_bulk->bd_nob_transferred);
2140 osc_release_ppga(aa->aa_ppga, aa->aa_page_count);
2141 ptlrpc_lprocfs_brw(req, transferred);
2143 spin_lock(&cli->cl_loi_list_lock);
2144 /* We need to decrement before osc_ap_completion->osc_wake_cache_waiters
2145 * is called so we know whether to go to sync BRWs or wait for more
2146 * RPCs to complete */
2147 if (lustre_msg_get_opc(req->rq_reqmsg) == OST_WRITE)
2148 cli->cl_w_in_flight--;
2150 cli->cl_r_in_flight--;
2151 osc_wake_cache_waiters(cli);
2152 spin_unlock(&cli->cl_loi_list_lock);
2154 osc_io_unplug(env, cli, NULL);
2158 static void brw_commit(struct ptlrpc_request *req)
2160 /* If osc_inc_unstable_pages (via osc_extent_finish) races with
2161 * this called via the rq_commit_cb, I need to ensure
2162 * osc_dec_unstable_pages is still called. Otherwise unstable
2163 * pages may be leaked. */
2164 spin_lock(&req->rq_lock);
2165 if (likely(req->rq_unstable)) {
2166 req->rq_unstable = 0;
2167 spin_unlock(&req->rq_lock);
2169 osc_dec_unstable_pages(req);
2171 req->rq_committed = 1;
2172 spin_unlock(&req->rq_lock);
2177 * Build an RPC by the list of extent @ext_list. The caller must ensure
2178 * that the total pages in this list are NOT over max pages per RPC.
2179 * Extents in the list must be in OES_RPC state.
2181 int osc_build_rpc(const struct lu_env *env, struct client_obd *cli,
2182 struct list_head *ext_list, int cmd)
2184 struct ptlrpc_request *req = NULL;
2185 struct osc_extent *ext;
2186 struct brw_page **pga = NULL;
2187 struct osc_brw_async_args *aa = NULL;
2188 struct obdo *oa = NULL;
2189 struct osc_async_page *oap;
2190 struct osc_object *obj = NULL;
2191 struct cl_req_attr *crattr = NULL;
2192 loff_t starting_offset = OBD_OBJECT_EOF;
2193 loff_t ending_offset = 0;
2197 bool soft_sync = false;
2198 bool interrupted = false;
2199 bool ndelay = false;
2203 __u32 layout_version = 0;
2204 LIST_HEAD(rpc_list);
2205 struct ost_body *body;
2207 LASSERT(!list_empty(ext_list));
2209 /* add pages into rpc_list to build BRW rpc */
2210 list_for_each_entry(ext, ext_list, oe_link) {
2211 LASSERT(ext->oe_state == OES_RPC);
2212 mem_tight |= ext->oe_memalloc;
2213 grant += ext->oe_grants;
2214 page_count += ext->oe_nr_pages;
2215 layout_version = MAX(layout_version, ext->oe_layout_version);
2220 soft_sync = osc_over_unstable_soft_limit(cli);
2222 mpflag = cfs_memory_pressure_get_and_set();
2224 OBD_ALLOC(pga, sizeof(*pga) * page_count);
2226 GOTO(out, rc = -ENOMEM);
2228 OBD_SLAB_ALLOC_PTR_GFP(oa, osc_obdo_kmem, GFP_NOFS);
2230 GOTO(out, rc = -ENOMEM);
2233 list_for_each_entry(ext, ext_list, oe_link) {
2234 list_for_each_entry(oap, &ext->oe_pages, oap_pending_item) {
2236 oap->oap_brw_flags |= OBD_BRW_MEMALLOC;
2238 oap->oap_brw_flags |= OBD_BRW_SOFT_SYNC;
2239 pga[i] = &oap->oap_brw_page;
2240 pga[i]->off = oap->oap_obj_off + oap->oap_page_off;
2243 list_add_tail(&oap->oap_rpc_item, &rpc_list);
2244 if (starting_offset == OBD_OBJECT_EOF ||
2245 starting_offset > oap->oap_obj_off)
2246 starting_offset = oap->oap_obj_off;
2248 LASSERT(oap->oap_page_off == 0);
2249 if (ending_offset < oap->oap_obj_off + oap->oap_count)
2250 ending_offset = oap->oap_obj_off +
2253 LASSERT(oap->oap_page_off + oap->oap_count ==
2255 if (oap->oap_interrupted)
2262 /* first page in the list */
2263 oap = list_entry(rpc_list.next, typeof(*oap), oap_rpc_item);
2265 crattr = &osc_env_info(env)->oti_req_attr;
2266 memset(crattr, 0, sizeof(*crattr));
2267 crattr->cra_type = (cmd & OBD_BRW_WRITE) ? CRT_WRITE : CRT_READ;
2268 crattr->cra_flags = ~0ULL;
2269 crattr->cra_page = oap2cl_page(oap);
2270 crattr->cra_oa = oa;
2271 cl_req_attr_set(env, osc2cl(obj), crattr);
2273 if (cmd == OBD_BRW_WRITE) {
2274 oa->o_grant_used = grant;
2275 if (layout_version > 0) {
2276 CDEBUG(D_LAYOUT, DFID": write with layout version %u\n",
2277 PFID(&oa->o_oi.oi_fid), layout_version);
2279 oa->o_layout_version = layout_version;
2280 oa->o_valid |= OBD_MD_LAYOUT_VERSION;
2284 sort_brw_pages(pga, page_count);
2285 rc = osc_brw_prep_request(cmd, cli, oa, page_count, pga, &req, 0);
2287 CERROR("prep_req failed: %d\n", rc);
2291 req->rq_commit_cb = brw_commit;
2292 req->rq_interpret_reply = brw_interpret;
2293 req->rq_memalloc = mem_tight != 0;
2294 oap->oap_request = ptlrpc_request_addref(req);
2295 if (interrupted && !req->rq_intr)
2296 ptlrpc_mark_interrupted(req);
2298 req->rq_no_resend = req->rq_no_delay = 1;
2299 /* probably set a shorter timeout value.
2300 * to handle ETIMEDOUT in brw_interpret() correctly. */
2301 /* lustre_msg_set_timeout(req, req->rq_timeout / 2); */
2304 /* Need to update the timestamps after the request is built in case
2305 * we race with setattr (locally or in queue at OST). If OST gets
2306 * later setattr before earlier BRW (as determined by the request xid),
2307 * the OST will not use BRW timestamps. Sadly, there is no obvious
2308 * way to do this in a single call. bug 10150 */
2309 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
2310 crattr->cra_oa = &body->oa;
2311 crattr->cra_flags = OBD_MD_FLMTIME | OBD_MD_FLCTIME | OBD_MD_FLATIME;
2312 cl_req_attr_set(env, osc2cl(obj), crattr);
2313 lustre_msg_set_jobid(req->rq_reqmsg, crattr->cra_jobid);
2315 aa = ptlrpc_req_async_args(aa, req);
2316 INIT_LIST_HEAD(&aa->aa_oaps);
2317 list_splice_init(&rpc_list, &aa->aa_oaps);
2318 INIT_LIST_HEAD(&aa->aa_exts);
2319 list_splice_init(ext_list, &aa->aa_exts);
2321 spin_lock(&cli->cl_loi_list_lock);
2322 starting_offset >>= PAGE_SHIFT;
2323 if (cmd == OBD_BRW_READ) {
2324 cli->cl_r_in_flight++;
2325 lprocfs_oh_tally_log2(&cli->cl_read_page_hist, page_count);
2326 lprocfs_oh_tally(&cli->cl_read_rpc_hist, cli->cl_r_in_flight);
2327 lprocfs_oh_tally_log2(&cli->cl_read_offset_hist,
2328 starting_offset + 1);
2330 cli->cl_w_in_flight++;
2331 lprocfs_oh_tally_log2(&cli->cl_write_page_hist, page_count);
2332 lprocfs_oh_tally(&cli->cl_write_rpc_hist, cli->cl_w_in_flight);
2333 lprocfs_oh_tally_log2(&cli->cl_write_offset_hist,
2334 starting_offset + 1);
2336 spin_unlock(&cli->cl_loi_list_lock);
2338 DEBUG_REQ(D_INODE, req, "%d pages, aa %p, now %ur/%uw in flight",
2339 page_count, aa, cli->cl_r_in_flight,
2340 cli->cl_w_in_flight);
2341 OBD_FAIL_TIMEOUT(OBD_FAIL_OSC_DELAY_IO, cfs_fail_val);
2343 ptlrpcd_add_req(req);
2349 cfs_memory_pressure_restore(mpflag);
2352 LASSERT(req == NULL);
2355 OBD_SLAB_FREE_PTR(oa, osc_obdo_kmem);
2357 OBD_FREE(pga, sizeof(*pga) * page_count);
2358 /* this should happen rarely and is pretty bad, it makes the
2359 * pending list not follow the dirty order */
2360 while (!list_empty(ext_list)) {
2361 ext = list_entry(ext_list->next, struct osc_extent,
2363 list_del_init(&ext->oe_link);
2364 osc_extent_finish(env, ext, 0, rc);
2370 static int osc_set_lock_data(struct ldlm_lock *lock, void *data)
2374 LASSERT(lock != NULL);
2376 lock_res_and_lock(lock);
2378 if (lock->l_ast_data == NULL)
2379 lock->l_ast_data = data;
2380 if (lock->l_ast_data == data)
2383 unlock_res_and_lock(lock);
2388 int osc_enqueue_fini(struct ptlrpc_request *req, osc_enqueue_upcall_f upcall,
2389 void *cookie, struct lustre_handle *lockh,
2390 enum ldlm_mode mode, __u64 *flags, bool speculative,
2393 bool intent = *flags & LDLM_FL_HAS_INTENT;
2397 /* The request was created before ldlm_cli_enqueue call. */
2398 if (intent && errcode == ELDLM_LOCK_ABORTED) {
2399 struct ldlm_reply *rep;
2401 rep = req_capsule_server_get(&req->rq_pill, &RMF_DLM_REP);
2402 LASSERT(rep != NULL);
2404 rep->lock_policy_res1 =
2405 ptlrpc_status_ntoh(rep->lock_policy_res1);
2406 if (rep->lock_policy_res1)
2407 errcode = rep->lock_policy_res1;
2409 *flags |= LDLM_FL_LVB_READY;
2410 } else if (errcode == ELDLM_OK) {
2411 *flags |= LDLM_FL_LVB_READY;
2414 /* Call the update callback. */
2415 rc = (*upcall)(cookie, lockh, errcode);
2417 /* release the reference taken in ldlm_cli_enqueue() */
2418 if (errcode == ELDLM_LOCK_MATCHED)
2420 if (errcode == ELDLM_OK && lustre_handle_is_used(lockh))
2421 ldlm_lock_decref(lockh, mode);
2426 int osc_enqueue_interpret(const struct lu_env *env, struct ptlrpc_request *req,
2429 struct osc_enqueue_args *aa = args;
2430 struct ldlm_lock *lock;
2431 struct lustre_handle *lockh = &aa->oa_lockh;
2432 enum ldlm_mode mode = aa->oa_mode;
2433 struct ost_lvb *lvb = aa->oa_lvb;
2434 __u32 lvb_len = sizeof(*lvb);
2439 /* ldlm_cli_enqueue is holding a reference on the lock, so it must
2441 lock = ldlm_handle2lock(lockh);
2442 LASSERTF(lock != NULL,
2443 "lockh %#llx, req %p, aa %p - client evicted?\n",
2444 lockh->cookie, req, aa);
2446 /* Take an additional reference so that a blocking AST that
2447 * ldlm_cli_enqueue_fini() might post for a failed lock, is guaranteed
2448 * to arrive after an upcall has been executed by
2449 * osc_enqueue_fini(). */
2450 ldlm_lock_addref(lockh, mode);
2452 /* Let cl_lock_state_wait fail with -ERESTARTSYS to unuse sublocks. */
2453 OBD_FAIL_TIMEOUT(OBD_FAIL_LDLM_ENQUEUE_HANG, 2);
2455 /* Let CP AST to grant the lock first. */
2456 OBD_FAIL_TIMEOUT(OBD_FAIL_OSC_CP_ENQ_RACE, 1);
2458 if (aa->oa_speculative) {
2459 LASSERT(aa->oa_lvb == NULL);
2460 LASSERT(aa->oa_flags == NULL);
2461 aa->oa_flags = &flags;
2464 /* Complete obtaining the lock procedure. */
2465 rc = ldlm_cli_enqueue_fini(aa->oa_exp, req, aa->oa_type, 1,
2466 aa->oa_mode, aa->oa_flags, lvb, lvb_len,
2468 /* Complete osc stuff. */
2469 rc = osc_enqueue_fini(req, aa->oa_upcall, aa->oa_cookie, lockh, mode,
2470 aa->oa_flags, aa->oa_speculative, rc);
2472 OBD_FAIL_TIMEOUT(OBD_FAIL_OSC_CP_CANCEL_RACE, 10);
2474 ldlm_lock_decref(lockh, mode);
2475 LDLM_LOCK_PUT(lock);
2479 struct ptlrpc_request_set *PTLRPCD_SET = (void *)1;
2481 /* When enqueuing asynchronously, locks are not ordered, we can obtain a lock
2482 * from the 2nd OSC before a lock from the 1st one. This does not deadlock with
2483 * other synchronous requests, however keeping some locks and trying to obtain
2484 * others may take a considerable amount of time in a case of ost failure; and
2485 * when other sync requests do not get released lock from a client, the client
2486 * is evicted from the cluster -- such scenarious make the life difficult, so
2487 * release locks just after they are obtained. */
2488 int osc_enqueue_base(struct obd_export *exp, struct ldlm_res_id *res_id,
2489 __u64 *flags, union ldlm_policy_data *policy,
2490 struct ost_lvb *lvb, osc_enqueue_upcall_f upcall,
2491 void *cookie, struct ldlm_enqueue_info *einfo,
2492 struct ptlrpc_request_set *rqset, int async,
2495 struct obd_device *obd = exp->exp_obd;
2496 struct lustre_handle lockh = { 0 };
2497 struct ptlrpc_request *req = NULL;
2498 int intent = *flags & LDLM_FL_HAS_INTENT;
2499 __u64 match_flags = *flags;
2500 enum ldlm_mode mode;
2504 /* Filesystem lock extents are extended to page boundaries so that
2505 * dealing with the page cache is a little smoother. */
2506 policy->l_extent.start -= policy->l_extent.start & ~PAGE_MASK;
2507 policy->l_extent.end |= ~PAGE_MASK;
2509 /* Next, search for already existing extent locks that will cover us */
2510 /* If we're trying to read, we also search for an existing PW lock. The
2511 * VFS and page cache already protect us locally, so lots of readers/
2512 * writers can share a single PW lock.
2514 * There are problems with conversion deadlocks, so instead of
2515 * converting a read lock to a write lock, we'll just enqueue a new
2518 * At some point we should cancel the read lock instead of making them
2519 * send us a blocking callback, but there are problems with canceling
2520 * locks out from other users right now, too. */
2521 mode = einfo->ei_mode;
2522 if (einfo->ei_mode == LCK_PR)
2524 /* Normal lock requests must wait for the LVB to be ready before
2525 * matching a lock; speculative lock requests do not need to,
2526 * because they will not actually use the lock. */
2528 match_flags |= LDLM_FL_LVB_READY;
2530 match_flags |= LDLM_FL_BLOCK_GRANTED;
2531 mode = ldlm_lock_match(obd->obd_namespace, match_flags, res_id,
2532 einfo->ei_type, policy, mode, &lockh, 0);
2534 struct ldlm_lock *matched;
2536 if (*flags & LDLM_FL_TEST_LOCK)
2539 matched = ldlm_handle2lock(&lockh);
2541 /* This DLM lock request is speculative, and does not
2542 * have an associated IO request. Therefore if there
2543 * is already a DLM lock, it wll just inform the
2544 * caller to cancel the request for this stripe.*/
2545 lock_res_and_lock(matched);
2546 if (ldlm_extent_equal(&policy->l_extent,
2547 &matched->l_policy_data.l_extent))
2551 unlock_res_and_lock(matched);
2553 ldlm_lock_decref(&lockh, mode);
2554 LDLM_LOCK_PUT(matched);
2556 } else if (osc_set_lock_data(matched, einfo->ei_cbdata)) {
2557 *flags |= LDLM_FL_LVB_READY;
2559 /* We already have a lock, and it's referenced. */
2560 (*upcall)(cookie, &lockh, ELDLM_LOCK_MATCHED);
2562 ldlm_lock_decref(&lockh, mode);
2563 LDLM_LOCK_PUT(matched);
2566 ldlm_lock_decref(&lockh, mode);
2567 LDLM_LOCK_PUT(matched);
2571 if (*flags & (LDLM_FL_TEST_LOCK | LDLM_FL_MATCH_LOCK))
2575 req = ptlrpc_request_alloc(class_exp2cliimp(exp),
2576 &RQF_LDLM_ENQUEUE_LVB);
2580 rc = ldlm_prep_enqueue_req(exp, req, NULL, 0);
2582 ptlrpc_request_free(req);
2586 req_capsule_set_size(&req->rq_pill, &RMF_DLM_LVB, RCL_SERVER,
2588 ptlrpc_request_set_replen(req);
2591 /* users of osc_enqueue() can pass this flag for ldlm_lock_match() */
2592 *flags &= ~LDLM_FL_BLOCK_GRANTED;
2594 rc = ldlm_cli_enqueue(exp, &req, einfo, res_id, policy, flags, lvb,
2595 sizeof(*lvb), LVB_T_OST, &lockh, async);
2598 struct osc_enqueue_args *aa;
2599 aa = ptlrpc_req_async_args(aa, req);
2601 aa->oa_mode = einfo->ei_mode;
2602 aa->oa_type = einfo->ei_type;
2603 lustre_handle_copy(&aa->oa_lockh, &lockh);
2604 aa->oa_upcall = upcall;
2605 aa->oa_cookie = cookie;
2606 aa->oa_speculative = speculative;
2608 aa->oa_flags = flags;
2611 /* speculative locks are essentially to enqueue
2612 * a DLM lock in advance, so we don't care
2613 * about the result of the enqueue. */
2615 aa->oa_flags = NULL;
2618 req->rq_interpret_reply = osc_enqueue_interpret;
2619 if (rqset == PTLRPCD_SET)
2620 ptlrpcd_add_req(req);
2622 ptlrpc_set_add_req(rqset, req);
2623 } else if (intent) {
2624 ptlrpc_req_finished(req);
2629 rc = osc_enqueue_fini(req, upcall, cookie, &lockh, einfo->ei_mode,
2630 flags, speculative, rc);
2632 ptlrpc_req_finished(req);
2637 int osc_match_base(const struct lu_env *env, struct obd_export *exp,
2638 struct ldlm_res_id *res_id, enum ldlm_type type,
2639 union ldlm_policy_data *policy, enum ldlm_mode mode,
2640 __u64 *flags, struct osc_object *obj,
2641 struct lustre_handle *lockh, int unref)
2643 struct obd_device *obd = exp->exp_obd;
2644 __u64 lflags = *flags;
2648 if (OBD_FAIL_CHECK(OBD_FAIL_OSC_MATCH))
2651 /* Filesystem lock extents are extended to page boundaries so that
2652 * dealing with the page cache is a little smoother */
2653 policy->l_extent.start -= policy->l_extent.start & ~PAGE_MASK;
2654 policy->l_extent.end |= ~PAGE_MASK;
2656 /* Next, search for already existing extent locks that will cover us */
2657 /* If we're trying to read, we also search for an existing PW lock. The
2658 * VFS and page cache already protect us locally, so lots of readers/
2659 * writers can share a single PW lock. */
2663 rc = ldlm_lock_match(obd->obd_namespace, lflags,
2664 res_id, type, policy, rc, lockh, unref);
2665 if (rc == 0 || lflags & LDLM_FL_TEST_LOCK)
2669 struct ldlm_lock *lock = ldlm_handle2lock(lockh);
2671 LASSERT(lock != NULL);
2672 if (osc_set_lock_data(lock, obj)) {
2673 lock_res_and_lock(lock);
2674 if (!ldlm_is_lvb_cached(lock)) {
2675 LASSERT(lock->l_ast_data == obj);
2676 osc_lock_lvb_update(env, obj, lock, NULL);
2677 ldlm_set_lvb_cached(lock);
2679 unlock_res_and_lock(lock);
2681 ldlm_lock_decref(lockh, rc);
2684 LDLM_LOCK_PUT(lock);
2689 static int osc_statfs_interpret(const struct lu_env *env,
2690 struct ptlrpc_request *req, void *args, int rc)
2692 struct osc_async_args *aa = args;
2693 struct obd_statfs *msfs;
2698 * The request has in fact never been sent due to issues at
2699 * a higher level (LOV). Exit immediately since the caller
2700 * is aware of the problem and takes care of the clean up.
2704 if ((rc == -ENOTCONN || rc == -EAGAIN) &&
2705 (aa->aa_oi->oi_flags & OBD_STATFS_NODELAY))
2711 msfs = req_capsule_server_get(&req->rq_pill, &RMF_OBD_STATFS);
2713 GOTO(out, rc = -EPROTO);
2715 *aa->aa_oi->oi_osfs = *msfs;
2717 rc = aa->aa_oi->oi_cb_up(aa->aa_oi, rc);
2722 static int osc_statfs_async(struct obd_export *exp,
2723 struct obd_info *oinfo, time64_t max_age,
2724 struct ptlrpc_request_set *rqset)
2726 struct obd_device *obd = class_exp2obd(exp);
2727 struct ptlrpc_request *req;
2728 struct osc_async_args *aa;
2732 if (obd->obd_osfs_age >= max_age) {
2734 "%s: use %p cache blocks %llu/%llu objects %llu/%llu\n",
2735 obd->obd_name, &obd->obd_osfs,
2736 obd->obd_osfs.os_bavail, obd->obd_osfs.os_blocks,
2737 obd->obd_osfs.os_ffree, obd->obd_osfs.os_files);
2738 spin_lock(&obd->obd_osfs_lock);
2739 memcpy(oinfo->oi_osfs, &obd->obd_osfs, sizeof(*oinfo->oi_osfs));
2740 spin_unlock(&obd->obd_osfs_lock);
2741 oinfo->oi_flags |= OBD_STATFS_FROM_CACHE;
2742 if (oinfo->oi_cb_up)
2743 oinfo->oi_cb_up(oinfo, 0);
2748 /* We could possibly pass max_age in the request (as an absolute
2749 * timestamp or a "seconds.usec ago") so the target can avoid doing
2750 * extra calls into the filesystem if that isn't necessary (e.g.
2751 * during mount that would help a bit). Having relative timestamps
2752 * is not so great if request processing is slow, while absolute
2753 * timestamps are not ideal because they need time synchronization. */
2754 req = ptlrpc_request_alloc(obd->u.cli.cl_import, &RQF_OST_STATFS);
2758 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_STATFS);
2760 ptlrpc_request_free(req);
2763 ptlrpc_request_set_replen(req);
2764 req->rq_request_portal = OST_CREATE_PORTAL;
2765 ptlrpc_at_set_req_timeout(req);
2767 if (oinfo->oi_flags & OBD_STATFS_NODELAY) {
2768 /* procfs requests not want stat in wait for avoid deadlock */
2769 req->rq_no_resend = 1;
2770 req->rq_no_delay = 1;
2773 req->rq_interpret_reply = osc_statfs_interpret;
2774 aa = ptlrpc_req_async_args(aa, req);
2777 ptlrpc_set_add_req(rqset, req);
2781 static int osc_statfs(const struct lu_env *env, struct obd_export *exp,
2782 struct obd_statfs *osfs, time64_t max_age, __u32 flags)
2784 struct obd_device *obd = class_exp2obd(exp);
2785 struct obd_statfs *msfs;
2786 struct ptlrpc_request *req;
2787 struct obd_import *imp = NULL;
2792 /*Since the request might also come from lprocfs, so we need
2793 *sync this with client_disconnect_export Bug15684*/
2794 down_read(&obd->u.cli.cl_sem);
2795 if (obd->u.cli.cl_import)
2796 imp = class_import_get(obd->u.cli.cl_import);
2797 up_read(&obd->u.cli.cl_sem);
2801 /* We could possibly pass max_age in the request (as an absolute
2802 * timestamp or a "seconds.usec ago") so the target can avoid doing
2803 * extra calls into the filesystem if that isn't necessary (e.g.
2804 * during mount that would help a bit). Having relative timestamps
2805 * is not so great if request processing is slow, while absolute
2806 * timestamps are not ideal because they need time synchronization. */
2807 req = ptlrpc_request_alloc(imp, &RQF_OST_STATFS);
2809 class_import_put(imp);
2814 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_STATFS);
2816 ptlrpc_request_free(req);
2819 ptlrpc_request_set_replen(req);
2820 req->rq_request_portal = OST_CREATE_PORTAL;
2821 ptlrpc_at_set_req_timeout(req);
2823 if (flags & OBD_STATFS_NODELAY) {
2824 /* procfs requests not want stat in wait for avoid deadlock */
2825 req->rq_no_resend = 1;
2826 req->rq_no_delay = 1;
2829 rc = ptlrpc_queue_wait(req);
2833 msfs = req_capsule_server_get(&req->rq_pill, &RMF_OBD_STATFS);
2835 GOTO(out, rc = -EPROTO);
2841 ptlrpc_req_finished(req);
2845 static int osc_iocontrol(unsigned int cmd, struct obd_export *exp, int len,
2846 void *karg, void __user *uarg)
2848 struct obd_device *obd = exp->exp_obd;
2849 struct obd_ioctl_data *data = karg;
2853 if (!try_module_get(THIS_MODULE)) {
2854 CERROR("%s: cannot get module '%s'\n", obd->obd_name,
2855 module_name(THIS_MODULE));
2859 case OBD_IOC_CLIENT_RECOVER:
2860 rc = ptlrpc_recover_import(obd->u.cli.cl_import,
2861 data->ioc_inlbuf1, 0);
2865 case IOC_OSC_SET_ACTIVE:
2866 rc = ptlrpc_set_import_active(obd->u.cli.cl_import,
2871 CDEBUG(D_INODE, "%s: unrecognised ioctl %#x by %s: rc = %d\n",
2872 obd->obd_name, cmd, current_comm(), rc);
2876 module_put(THIS_MODULE);
2880 int osc_set_info_async(const struct lu_env *env, struct obd_export *exp,
2881 u32 keylen, void *key, u32 vallen, void *val,
2882 struct ptlrpc_request_set *set)
2884 struct ptlrpc_request *req;
2885 struct obd_device *obd = exp->exp_obd;
2886 struct obd_import *imp = class_exp2cliimp(exp);
2891 OBD_FAIL_TIMEOUT(OBD_FAIL_OSC_SHUTDOWN, 10);
2893 if (KEY_IS(KEY_CHECKSUM)) {
2894 if (vallen != sizeof(int))
2896 exp->exp_obd->u.cli.cl_checksum = (*(int *)val) ? 1 : 0;
2900 if (KEY_IS(KEY_SPTLRPC_CONF)) {
2901 sptlrpc_conf_client_adapt(obd);
2905 if (KEY_IS(KEY_FLUSH_CTX)) {
2906 sptlrpc_import_flush_my_ctx(imp);
2910 if (KEY_IS(KEY_CACHE_LRU_SHRINK)) {
2911 struct client_obd *cli = &obd->u.cli;
2912 long nr = atomic_long_read(&cli->cl_lru_in_list) >> 1;
2913 long target = *(long *)val;
2915 nr = osc_lru_shrink(env, cli, min(nr, target), true);
2920 if (!set && !KEY_IS(KEY_GRANT_SHRINK))
2923 /* We pass all other commands directly to OST. Since nobody calls osc
2924 methods directly and everybody is supposed to go through LOV, we
2925 assume lov checked invalid values for us.
2926 The only recognised values so far are evict_by_nid and mds_conn.
2927 Even if something bad goes through, we'd get a -EINVAL from OST
2930 req = ptlrpc_request_alloc(imp, KEY_IS(KEY_GRANT_SHRINK) ?
2931 &RQF_OST_SET_GRANT_INFO :
2936 req_capsule_set_size(&req->rq_pill, &RMF_SETINFO_KEY,
2937 RCL_CLIENT, keylen);
2938 if (!KEY_IS(KEY_GRANT_SHRINK))
2939 req_capsule_set_size(&req->rq_pill, &RMF_SETINFO_VAL,
2940 RCL_CLIENT, vallen);
2941 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_SET_INFO);
2943 ptlrpc_request_free(req);
2947 tmp = req_capsule_client_get(&req->rq_pill, &RMF_SETINFO_KEY);
2948 memcpy(tmp, key, keylen);
2949 tmp = req_capsule_client_get(&req->rq_pill, KEY_IS(KEY_GRANT_SHRINK) ?
2952 memcpy(tmp, val, vallen);
2954 if (KEY_IS(KEY_GRANT_SHRINK)) {
2955 struct osc_grant_args *aa;
2958 aa = ptlrpc_req_async_args(aa, req);
2959 OBD_SLAB_ALLOC_PTR_GFP(oa, osc_obdo_kmem, GFP_NOFS);
2961 ptlrpc_req_finished(req);
2964 *oa = ((struct ost_body *)val)->oa;
2966 req->rq_interpret_reply = osc_shrink_grant_interpret;
2969 ptlrpc_request_set_replen(req);
2970 if (!KEY_IS(KEY_GRANT_SHRINK)) {
2971 LASSERT(set != NULL);
2972 ptlrpc_set_add_req(set, req);
2973 ptlrpc_check_set(NULL, set);
2975 ptlrpcd_add_req(req);
2980 EXPORT_SYMBOL(osc_set_info_async);
2982 int osc_reconnect(const struct lu_env *env, struct obd_export *exp,
2983 struct obd_device *obd, struct obd_uuid *cluuid,
2984 struct obd_connect_data *data, void *localdata)
2986 struct client_obd *cli = &obd->u.cli;
2988 if (data != NULL && (data->ocd_connect_flags & OBD_CONNECT_GRANT)) {
2992 spin_lock(&cli->cl_loi_list_lock);
2993 grant = cli->cl_avail_grant + cli->cl_reserved_grant;
2994 if (data->ocd_connect_flags & OBD_CONNECT_GRANT_PARAM) {
2995 /* restore ocd_grant_blkbits as client page bits */
2996 data->ocd_grant_blkbits = PAGE_SHIFT;
2997 grant += cli->cl_dirty_grant;
2999 grant += cli->cl_dirty_pages << PAGE_SHIFT;
3001 data->ocd_grant = grant ? : 2 * cli_brw_size(obd);
3002 lost_grant = cli->cl_lost_grant;
3003 cli->cl_lost_grant = 0;
3004 spin_unlock(&cli->cl_loi_list_lock);
3006 CDEBUG(D_RPCTRACE, "ocd_connect_flags: %#llx ocd_version: %d"
3007 " ocd_grant: %d, lost: %ld.\n", data->ocd_connect_flags,
3008 data->ocd_version, data->ocd_grant, lost_grant);
3013 EXPORT_SYMBOL(osc_reconnect);
3015 int osc_disconnect(struct obd_export *exp)
3017 struct obd_device *obd = class_exp2obd(exp);
3020 rc = client_disconnect_export(exp);
3022 * Initially we put del_shrink_grant before disconnect_export, but it
3023 * causes the following problem if setup (connect) and cleanup
3024 * (disconnect) are tangled together.
3025 * connect p1 disconnect p2
3026 * ptlrpc_connect_import
3027 * ............... class_manual_cleanup
3030 * ptlrpc_connect_interrupt
3032 * add this client to shrink list
3034 * Bang! grant shrink thread trigger the shrink. BUG18662
3036 osc_del_grant_list(&obd->u.cli);
3039 EXPORT_SYMBOL(osc_disconnect);
3041 int osc_ldlm_resource_invalidate(struct cfs_hash *hs, struct cfs_hash_bd *bd,
3042 struct hlist_node *hnode, void *arg)
3044 struct lu_env *env = arg;
3045 struct ldlm_resource *res = cfs_hash_object(hs, hnode);
3046 struct ldlm_lock *lock;
3047 struct osc_object *osc = NULL;
3051 list_for_each_entry(lock, &res->lr_granted, l_res_link) {
3052 if (lock->l_ast_data != NULL && osc == NULL) {
3053 osc = lock->l_ast_data;
3054 cl_object_get(osc2cl(osc));
3057 /* clear LDLM_FL_CLEANED flag to make sure it will be canceled
3058 * by the 2nd round of ldlm_namespace_clean() call in
3059 * osc_import_event(). */
3060 ldlm_clear_cleaned(lock);
3065 osc_object_invalidate(env, osc);
3066 cl_object_put(env, osc2cl(osc));
3071 EXPORT_SYMBOL(osc_ldlm_resource_invalidate);
3073 static int osc_import_event(struct obd_device *obd,
3074 struct obd_import *imp,
3075 enum obd_import_event event)
3077 struct client_obd *cli;
3081 LASSERT(imp->imp_obd == obd);
3084 case IMP_EVENT_DISCON: {
3086 spin_lock(&cli->cl_loi_list_lock);
3087 cli->cl_avail_grant = 0;
3088 cli->cl_lost_grant = 0;
3089 spin_unlock(&cli->cl_loi_list_lock);
3092 case IMP_EVENT_INACTIVE: {
3093 rc = obd_notify_observer(obd, obd, OBD_NOTIFY_INACTIVE);
3096 case IMP_EVENT_INVALIDATE: {
3097 struct ldlm_namespace *ns = obd->obd_namespace;
3101 ldlm_namespace_cleanup(ns, LDLM_FL_LOCAL_ONLY);
3103 env = cl_env_get(&refcheck);
3105 osc_io_unplug(env, &obd->u.cli, NULL);
3107 cfs_hash_for_each_nolock(ns->ns_rs_hash,
3108 osc_ldlm_resource_invalidate,
3110 cl_env_put(env, &refcheck);
3112 ldlm_namespace_cleanup(ns, LDLM_FL_LOCAL_ONLY);
3117 case IMP_EVENT_ACTIVE: {
3118 rc = obd_notify_observer(obd, obd, OBD_NOTIFY_ACTIVE);
3121 case IMP_EVENT_OCD: {
3122 struct obd_connect_data *ocd = &imp->imp_connect_data;
3124 if (ocd->ocd_connect_flags & OBD_CONNECT_GRANT)
3125 osc_init_grant(&obd->u.cli, ocd);
3128 if (ocd->ocd_connect_flags & OBD_CONNECT_REQPORTAL)
3129 imp->imp_client->cli_request_portal =OST_REQUEST_PORTAL;
3131 rc = obd_notify_observer(obd, obd, OBD_NOTIFY_OCD);
3134 case IMP_EVENT_DEACTIVATE: {
3135 rc = obd_notify_observer(obd, obd, OBD_NOTIFY_DEACTIVATE);
3138 case IMP_EVENT_ACTIVATE: {
3139 rc = obd_notify_observer(obd, obd, OBD_NOTIFY_ACTIVATE);
3143 CERROR("Unknown import event %d\n", event);
3150 * Determine whether the lock can be canceled before replaying the lock
3151 * during recovery, see bug16774 for detailed information.
3153 * \retval zero the lock can't be canceled
3154 * \retval other ok to cancel
3156 static int osc_cancel_weight(struct ldlm_lock *lock)
3159 * Cancel all unused and granted extent lock.
3161 if (lock->l_resource->lr_type == LDLM_EXTENT &&
3162 ldlm_is_granted(lock) &&
3163 osc_ldlm_weigh_ast(lock) == 0)
3169 static int brw_queue_work(const struct lu_env *env, void *data)
3171 struct client_obd *cli = data;
3173 CDEBUG(D_CACHE, "Run writeback work for client obd %p.\n", cli);
3175 osc_io_unplug(env, cli, NULL);
3179 int osc_setup_common(struct obd_device *obd, struct lustre_cfg *lcfg)
3181 struct client_obd *cli = &obd->u.cli;
3187 rc = ptlrpcd_addref();
3191 rc = client_obd_setup(obd, lcfg);
3193 GOTO(out_ptlrpcd, rc);
3196 handler = ptlrpcd_alloc_work(cli->cl_import, brw_queue_work, cli);
3197 if (IS_ERR(handler))
3198 GOTO(out_ptlrpcd_work, rc = PTR_ERR(handler));
3199 cli->cl_writeback_work = handler;
3201 handler = ptlrpcd_alloc_work(cli->cl_import, lru_queue_work, cli);
3202 if (IS_ERR(handler))
3203 GOTO(out_ptlrpcd_work, rc = PTR_ERR(handler));
3204 cli->cl_lru_work = handler;
3206 rc = osc_quota_setup(obd);
3208 GOTO(out_ptlrpcd_work, rc);
3210 cli->cl_grant_shrink_interval = GRANT_SHRINK_INTERVAL;
3211 osc_update_next_shrink(cli);
3216 if (cli->cl_writeback_work != NULL) {
3217 ptlrpcd_destroy_work(cli->cl_writeback_work);
3218 cli->cl_writeback_work = NULL;
3220 if (cli->cl_lru_work != NULL) {
3221 ptlrpcd_destroy_work(cli->cl_lru_work);
3222 cli->cl_lru_work = NULL;
3224 client_obd_cleanup(obd);
3229 EXPORT_SYMBOL(osc_setup_common);
3231 int osc_setup(struct obd_device *obd, struct lustre_cfg *lcfg)
3233 struct client_obd *cli = &obd->u.cli;
3241 rc = osc_setup_common(obd, lcfg);
3245 rc = osc_tunables_init(obd);
3250 * We try to control the total number of requests with a upper limit
3251 * osc_reqpool_maxreqcount. There might be some race which will cause
3252 * over-limit allocation, but it is fine.
3254 req_count = atomic_read(&osc_pool_req_count);
3255 if (req_count < osc_reqpool_maxreqcount) {
3256 adding = cli->cl_max_rpcs_in_flight + 2;
3257 if (req_count + adding > osc_reqpool_maxreqcount)
3258 adding = osc_reqpool_maxreqcount - req_count;
3260 added = ptlrpc_add_rqs_to_pool(osc_rq_pool, adding);
3261 atomic_add(added, &osc_pool_req_count);
3264 ns_register_cancel(obd->obd_namespace, osc_cancel_weight);
3266 spin_lock(&osc_shrink_lock);
3267 list_add_tail(&cli->cl_shrink_list, &osc_shrink_list);
3268 spin_unlock(&osc_shrink_lock);
3269 cli->cl_import->imp_idle_timeout = osc_idle_timeout;
3270 cli->cl_import->imp_idle_debug = D_HA;
3275 int osc_precleanup_common(struct obd_device *obd)
3277 struct client_obd *cli = &obd->u.cli;
3281 * for echo client, export may be on zombie list, wait for
3282 * zombie thread to cull it, because cli.cl_import will be
3283 * cleared in client_disconnect_export():
3284 * class_export_destroy() -> obd_cleanup() ->
3285 * echo_device_free() -> echo_client_cleanup() ->
3286 * obd_disconnect() -> osc_disconnect() ->
3287 * client_disconnect_export()
3289 obd_zombie_barrier();
3290 if (cli->cl_writeback_work) {
3291 ptlrpcd_destroy_work(cli->cl_writeback_work);
3292 cli->cl_writeback_work = NULL;
3295 if (cli->cl_lru_work) {
3296 ptlrpcd_destroy_work(cli->cl_lru_work);
3297 cli->cl_lru_work = NULL;
3300 obd_cleanup_client_import(obd);
3303 EXPORT_SYMBOL(osc_precleanup_common);
3305 static int osc_precleanup(struct obd_device *obd)
3309 osc_precleanup_common(obd);
3311 ptlrpc_lprocfs_unregister_obd(obd);
3315 int osc_cleanup_common(struct obd_device *obd)
3317 struct client_obd *cli = &obd->u.cli;
3322 spin_lock(&osc_shrink_lock);
3323 list_del(&cli->cl_shrink_list);
3324 spin_unlock(&osc_shrink_lock);
3327 if (cli->cl_cache != NULL) {
3328 LASSERT(atomic_read(&cli->cl_cache->ccc_users) > 0);
3329 spin_lock(&cli->cl_cache->ccc_lru_lock);
3330 list_del_init(&cli->cl_lru_osc);
3331 spin_unlock(&cli->cl_cache->ccc_lru_lock);
3332 cli->cl_lru_left = NULL;
3333 cl_cache_decref(cli->cl_cache);
3334 cli->cl_cache = NULL;
3337 /* free memory of osc quota cache */
3338 osc_quota_cleanup(obd);
3340 rc = client_obd_cleanup(obd);
3345 EXPORT_SYMBOL(osc_cleanup_common);
3347 static const struct obd_ops osc_obd_ops = {
3348 .o_owner = THIS_MODULE,
3349 .o_setup = osc_setup,
3350 .o_precleanup = osc_precleanup,
3351 .o_cleanup = osc_cleanup_common,
3352 .o_add_conn = client_import_add_conn,
3353 .o_del_conn = client_import_del_conn,
3354 .o_connect = client_connect_import,
3355 .o_reconnect = osc_reconnect,
3356 .o_disconnect = osc_disconnect,
3357 .o_statfs = osc_statfs,
3358 .o_statfs_async = osc_statfs_async,
3359 .o_create = osc_create,
3360 .o_destroy = osc_destroy,
3361 .o_getattr = osc_getattr,
3362 .o_setattr = osc_setattr,
3363 .o_iocontrol = osc_iocontrol,
3364 .o_set_info_async = osc_set_info_async,
3365 .o_import_event = osc_import_event,
3366 .o_quotactl = osc_quotactl,
3369 static struct shrinker *osc_cache_shrinker;
3370 LIST_HEAD(osc_shrink_list);
3371 DEFINE_SPINLOCK(osc_shrink_lock);
3373 #ifndef HAVE_SHRINKER_COUNT
3374 static int osc_cache_shrink(SHRINKER_ARGS(sc, nr_to_scan, gfp_mask))
3376 struct shrink_control scv = {
3377 .nr_to_scan = shrink_param(sc, nr_to_scan),
3378 .gfp_mask = shrink_param(sc, gfp_mask)
3380 (void)osc_cache_shrink_scan(shrinker, &scv);
3382 return osc_cache_shrink_count(shrinker, &scv);
3386 static int __init osc_init(void)
3388 unsigned int reqpool_size;
3389 unsigned int reqsize;
3391 DEF_SHRINKER_VAR(osc_shvar, osc_cache_shrink,
3392 osc_cache_shrink_count, osc_cache_shrink_scan);
3395 /* print an address of _any_ initialized kernel symbol from this
3396 * module, to allow debugging with gdb that doesn't support data
3397 * symbols from modules.*/
3398 CDEBUG(D_INFO, "Lustre OSC module (%p).\n", &osc_caches);
3400 rc = lu_kmem_init(osc_caches);
3404 rc = class_register_type(&osc_obd_ops, NULL, true, NULL,
3405 LUSTRE_OSC_NAME, &osc_device_type);
3409 osc_cache_shrinker = set_shrinker(DEFAULT_SEEKS, &osc_shvar);
3411 /* This is obviously too much memory, only prevent overflow here */
3412 if (osc_reqpool_mem_max >= 1 << 12 || osc_reqpool_mem_max == 0)
3413 GOTO(out_type, rc = -EINVAL);
3415 reqpool_size = osc_reqpool_mem_max << 20;
3418 while (reqsize < OST_IO_MAXREQSIZE)
3419 reqsize = reqsize << 1;
3422 * We don't enlarge the request count in OSC pool according to
3423 * cl_max_rpcs_in_flight. The allocation from the pool will only be
3424 * tried after normal allocation failed. So a small OSC pool won't
3425 * cause much performance degression in most of cases.
3427 osc_reqpool_maxreqcount = reqpool_size / reqsize;
3429 atomic_set(&osc_pool_req_count, 0);
3430 osc_rq_pool = ptlrpc_init_rq_pool(0, OST_IO_MAXREQSIZE,
3431 ptlrpc_add_rqs_to_pool);
3433 if (osc_rq_pool == NULL)
3434 GOTO(out_type, rc = -ENOMEM);
3436 rc = osc_start_grant_work();
3438 GOTO(out_req_pool, rc);
3443 ptlrpc_free_rq_pool(osc_rq_pool);
3445 class_unregister_type(LUSTRE_OSC_NAME);
3447 lu_kmem_fini(osc_caches);
3452 static void __exit osc_exit(void)
3454 osc_stop_grant_work();
3455 remove_shrinker(osc_cache_shrinker);
3456 class_unregister_type(LUSTRE_OSC_NAME);
3457 lu_kmem_fini(osc_caches);
3458 ptlrpc_free_rq_pool(osc_rq_pool);
3461 MODULE_AUTHOR("OpenSFS, Inc. <http://www.lustre.org/>");
3462 MODULE_DESCRIPTION("Lustre Object Storage Client (OSC)");
3463 MODULE_VERSION(LUSTRE_VERSION_STRING);
3464 MODULE_LICENSE("GPL");
3466 module_init(osc_init);
3467 module_exit(osc_exit);