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) 2008, 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.
32 * Implementation of cl_io for OSC layer.
34 * Author: Nikita Danilov <nikita.danilov@sun.com>
35 * Author: Jinshan Xiong <jinshan.xiong@whamcloud.com>
38 #define DEBUG_SUBSYSTEM S_OSC
40 #include <lustre_obdo.h>
41 #include <lustre_osc.h>
42 #include <linux/pagevec.h>
43 #include <linux/falloc.h>
45 #include "osc_internal.h"
51 /*****************************************************************************
57 static void osc_io_fini(const struct lu_env *env, const struct cl_io_slice *io)
61 void osc_read_ahead_release(const struct lu_env *env, struct cl_read_ahead *ra)
63 struct ldlm_lock *dlmlock = ra->cra_dlmlock;
64 struct osc_io *oio = ra->cra_oio;
65 struct lustre_handle lockh;
67 oio->oi_is_readahead = 0;
68 ldlm_lock2handle(dlmlock, &lockh);
69 ldlm_lock_decref(&lockh, LCK_PR);
70 LDLM_LOCK_PUT(dlmlock);
72 EXPORT_SYMBOL(osc_read_ahead_release);
74 static int osc_io_read_ahead(const struct lu_env *env,
75 const struct cl_io_slice *ios,
76 pgoff_t start, struct cl_read_ahead *ra)
78 struct osc_object *osc = cl2osc(ios->cis_obj);
79 struct osc_io *oio = cl2osc_io(env, ios);
80 struct ldlm_lock *dlmlock;
81 int result = -ENODATA;
85 oio->oi_is_readahead = true;
86 dlmlock = osc_dlmlock_at_pgoff(env, osc, start, 0);
87 if (dlmlock != NULL) {
88 LASSERT(dlmlock->l_ast_data == osc);
89 if (dlmlock->l_req_mode != LCK_PR) {
90 struct lustre_handle lockh;
91 ldlm_lock2handle(dlmlock, &lockh);
92 ldlm_lock_addref(&lockh, LCK_PR);
93 ldlm_lock_decref(&lockh, dlmlock->l_req_mode);
96 ra->cra_rpc_pages = osc_cli(osc)->cl_max_pages_per_rpc;
97 ra->cra_end_idx = cl_index(osc2cl(osc),
98 dlmlock->l_policy_data.l_extent.end);
99 ra->cra_release = osc_read_ahead_release;
100 ra->cra_dlmlock = dlmlock;
102 if (ra->cra_end_idx != CL_PAGE_EOF)
103 ra->cra_contention = true;
111 * An implementation of cl_io_operations::cio_io_submit() method for osc
112 * layer. Iterates over pages in the in-queue, prepares each for io by calling
113 * cl_page_prep() and then either submits them through osc_io_submit_page()
114 * or, if page is already submitted, changes osc flags through
115 * osc_set_async_flags().
117 int osc_io_submit(const struct lu_env *env, const struct cl_io_slice *ios,
118 enum cl_req_type crt, struct cl_2queue *queue)
120 struct cl_page *page;
122 struct client_obd *cli = NULL;
123 struct osc_object *osc = NULL; /* to keep gcc happy */
124 struct osc_page *opg;
128 struct cl_page_list *qin = &queue->c2_qin;
129 struct cl_page_list *qout = &queue->c2_qout;
130 unsigned int queued = 0;
133 unsigned int max_pages;
134 unsigned int ppc_bits; /* pages per chunk bits */
136 bool sync_queue = false;
138 LASSERT(qin->pl_nr > 0);
140 CDEBUG(D_CACHE|D_READA, "%d %d\n", qin->pl_nr, crt);
142 osc = cl2osc(ios->cis_obj);
144 max_pages = cli->cl_max_pages_per_rpc;
145 ppc_bits = cli->cl_chunkbits - PAGE_SHIFT;
148 brw_flags = osc_io_srvlock(cl2osc_io(env, ios)) ? OBD_BRW_SRVLOCK : 0;
149 brw_flags |= crt == CRT_WRITE ? OBD_BRW_WRITE : OBD_BRW_READ;
150 if (crt == CRT_READ && ios->cis_io->ci_ndelay)
151 brw_flags |= OBD_BRW_NDELAY;
153 page = cl_page_list_first(qin);
154 if (page->cp_type == CPT_TRANSIENT)
155 brw_flags |= OBD_BRW_NOCACHE;
158 * NOTE: here @page is a top-level page. This is done to avoid
159 * creation of sub-page-list.
161 cl_page_list_for_each_safe(page, tmp, qin) {
162 struct osc_async_page *oap;
168 opg = osc_cl_page_osc(page, osc);
170 LASSERT(osc == oap->oap_obj);
172 if (!list_empty(&oap->oap_pending_item) ||
173 !list_empty(&oap->oap_rpc_item)) {
174 CDEBUG(D_CACHE, "Busy oap %p page %p for submit.\n",
180 result = cl_page_prep(env, io, page, crt);
183 if (result != -EALREADY)
186 * Handle -EALREADY error: for read case, the page is
187 * already in UPTODATE state; for write, the page
194 spin_lock(&oap->oap_lock);
195 oap->oap_async_flags = ASYNC_URGENT|ASYNC_READY;
196 oap->oap_async_flags |= ASYNC_COUNT_STABLE;
197 spin_unlock(&oap->oap_lock);
199 osc_page_submit(env, opg, crt, brw_flags);
200 list_add_tail(&oap->oap_pending_item, &list);
202 if (page->cp_sync_io != NULL)
203 cl_page_list_move(qout, qin, page);
205 cl_page_list_del(env, qin, page);
208 if (queued == max_pages) {
210 } else if (crt == CRT_WRITE) {
212 unsigned int next_chunks;
214 chunks = (queued + ppc - 1) >> ppc_bits;
215 /* chunk number if add another page */
216 next_chunks = (queued + ppc) >> ppc_bits;
218 /* next page will excceed write chunk limit */
219 if (chunks == osc_max_write_chunks(cli) &&
220 next_chunks > chunks)
225 result = osc_queue_sync_pages(env, io, osc, &list,
235 result = osc_queue_sync_pages(env, io, osc, &list, brw_flags);
237 /* Update c/mtime for sync write. LU-7310 */
238 if (crt == CRT_WRITE && qout->pl_nr > 0 && result == 0) {
239 struct cl_object *obj = ios->cis_obj;
240 struct cl_attr *attr = &osc_env_info(env)->oti_attr;
242 cl_object_attr_lock(obj);
243 attr->cat_mtime = attr->cat_ctime = ktime_get_real_seconds();
244 cl_object_attr_update(env, obj, attr, CAT_MTIME | CAT_CTIME);
245 cl_object_attr_unlock(obj);
248 CDEBUG(D_INFO, "%d/%d %d\n", qin->pl_nr, qout->pl_nr, result);
249 return qout->pl_nr > 0 ? 0 : result;
251 EXPORT_SYMBOL(osc_io_submit);
254 * This is called to update the attributes when modifying a specific page,
255 * both when making new pages and when doing updates to existing cached pages.
257 * Expand stripe KMS if necessary.
259 void osc_page_touch_at(const struct lu_env *env, struct cl_object *obj,
260 pgoff_t idx, size_t to)
262 struct lov_oinfo *loi = cl2osc(obj)->oo_oinfo;
263 struct cl_attr *attr = &osc_env_info(env)->oti_attr;
269 /* offset within stripe */
270 kms = cl_offset(obj, idx) + to;
272 cl_object_attr_lock(obj);
273 CDEBUG(D_INODE, "stripe KMS %sincreasing %llu->%llu %llu\n",
274 kms > loi->loi_kms ? "" : "not ", loi->loi_kms, kms,
275 loi->loi_lvb.lvb_size);
277 attr->cat_mtime = attr->cat_ctime = ktime_get_real_seconds();
278 valid = CAT_MTIME | CAT_CTIME;
279 if (kms > loi->loi_kms) {
283 if (kms > loi->loi_lvb.lvb_size) {
284 attr->cat_size = kms;
287 cl_object_attr_update(env, obj, attr, valid);
288 cl_object_attr_unlock(obj);
293 int osc_io_commit_async(const struct lu_env *env,
294 const struct cl_io_slice *ios,
295 struct cl_page_list *qin, int from, int to,
298 struct cl_io *io = ios->cis_io;
299 struct osc_io *oio = cl2osc_io(env, ios);
300 struct osc_object *osc = cl2osc(ios->cis_obj);
301 struct cl_page *page;
302 struct cl_page *last_page;
303 struct osc_page *opg;
304 struct pagevec *pvec = &osc_env_info(env)->oti_pagevec;
308 LASSERT(qin->pl_nr > 0);
310 /* Handle partial page cases */
311 last_page = cl_page_list_last(qin);
312 if (oio->oi_lockless) {
313 page = cl_page_list_first(qin);
314 if (page == last_page) {
315 cl_page_clip(env, page, from, to);
318 cl_page_clip(env, page, from, PAGE_SIZE);
320 cl_page_clip(env, last_page, 0, to);
324 ll_pagevec_init(pvec, 0);
326 while (qin->pl_nr > 0) {
327 struct osc_async_page *oap;
329 page = cl_page_list_first(qin);
330 opg = osc_cl_page_osc(page, osc);
333 LASSERTF(osc == oap->oap_obj,
334 "obj mismatch: %p / %p\n", osc, oap->oap_obj);
336 if (!list_empty(&oap->oap_rpc_item)) {
337 CDEBUG(D_CACHE, "Busy oap %p page %p for submit.\n",
343 /* The page may be already in dirty cache. */
344 if (list_empty(&oap->oap_pending_item)) {
345 result = osc_page_cache_add(env, opg, io, cb);
350 osc_page_touch_at(env, osc2cl(osc), osc_index(opg),
351 page == last_page ? to : PAGE_SIZE);
353 cl_page_list_del(env, qin, page);
355 /* if there are no more slots, do the callback & reinit */
356 if (pagevec_add(pvec, page->cp_vmpage) == 0) {
357 (*cb)(env, io, pvec);
358 pagevec_reinit(pvec);
362 /* Clean up any partially full pagevecs */
363 if (pagevec_count(pvec) != 0)
364 (*cb)(env, io, pvec);
366 /* Can't access these pages any more. Page can be in transfer and
367 * complete at any time. */
369 /* for sync write, kernel will wait for this page to be flushed before
370 * osc_io_end() is called, so release it earlier.
371 * for mkwrite(), it's known there is no further pages. */
372 if (cl_io_is_sync_write(io) && oio->oi_active != NULL) {
373 osc_extent_release(env, oio->oi_active);
374 oio->oi_active = NULL;
377 CDEBUG(D_INFO, "%d %d\n", qin->pl_nr, result);
380 EXPORT_SYMBOL(osc_io_commit_async);
382 void osc_io_extent_release(const struct lu_env *env,
383 const struct cl_io_slice *ios)
385 struct osc_io *oio = cl2osc_io(env, ios);
387 if (oio->oi_active != NULL) {
388 osc_extent_release(env, oio->oi_active);
389 oio->oi_active = NULL;
392 EXPORT_SYMBOL(osc_io_extent_release);
394 static bool osc_import_not_healthy(struct obd_import *imp)
396 return imp->imp_invalid || imp->imp_deactive ||
397 !(imp->imp_state == LUSTRE_IMP_FULL ||
398 imp->imp_state == LUSTRE_IMP_IDLE);
401 int osc_io_iter_init(const struct lu_env *env, const struct cl_io_slice *ios)
403 struct osc_object *osc = cl2osc(ios->cis_obj);
404 struct obd_import *imp = osc_cli(osc)->cl_import;
405 struct osc_io *oio = osc_env_io(env);
410 spin_lock(&imp->imp_lock);
412 * check whether this OSC device is available for non-delay read,
413 * fast switching mirror if we haven't tried all mirrors.
415 if (ios->cis_io->ci_type == CIT_READ && ios->cis_io->ci_ndelay &&
416 !ios->cis_io->ci_tried_all_mirrors && osc_import_not_healthy(imp)) {
418 } else if (likely(!imp->imp_invalid)) {
419 atomic_inc(&osc->oo_nr_ios);
420 oio->oi_is_active = 1;
423 spin_unlock(&imp->imp_lock);
425 if (capable(CAP_SYS_RESOURCE))
426 oio->oi_cap_sys_resource = 1;
430 EXPORT_SYMBOL(osc_io_iter_init);
432 void osc_io_iter_fini(const struct lu_env *env,
433 const struct cl_io_slice *ios)
435 struct osc_io *oio = osc_env_io(env);
437 if (oio->oi_is_active) {
438 struct osc_object *osc = cl2osc(ios->cis_obj);
440 oio->oi_is_active = 0;
441 LASSERT(atomic_read(&osc->oo_nr_ios) > 0);
442 if (atomic_dec_and_test(&osc->oo_nr_ios))
443 wake_up_all(&osc->oo_io_waitq);
446 EXPORT_SYMBOL(osc_io_iter_fini);
448 void osc_io_rw_iter_fini(const struct lu_env *env,
449 const struct cl_io_slice *ios)
451 struct osc_io *oio = osc_env_io(env);
452 struct osc_object *osc = cl2osc(ios->cis_obj);
454 if (oio->oi_lru_reserved > 0) {
455 osc_lru_unreserve(osc_cli(osc), oio->oi_lru_reserved);
456 oio->oi_lru_reserved = 0;
458 oio->oi_write_osclock = NULL;
460 osc_io_iter_fini(env, ios);
462 EXPORT_SYMBOL(osc_io_rw_iter_fini);
464 int osc_io_fault_start(const struct lu_env *env, const struct cl_io_slice *ios)
467 struct cl_fault_io *fio;
471 fio = &io->u.ci_fault;
472 CDEBUG(D_INFO, "%lu %d %zu\n",
473 fio->ft_index, fio->ft_writable, fio->ft_nob);
475 * If mapping is writeable, adjust kms to cover this page,
476 * but do not extend kms beyond actual file size.
479 if (fio->ft_writable)
480 osc_page_touch_at(env, ios->cis_obj,
481 fio->ft_index, fio->ft_nob);
484 EXPORT_SYMBOL(osc_io_fault_start);
487 static int osc_async_upcall(void *a, int rc)
489 struct osc_async_cbargs *args = a;
492 complete(&args->opc_sync);
497 * Checks that there are no pages being written in the extent being truncated.
499 static bool trunc_check_cb(const struct lu_env *env, struct cl_io *io,
500 struct osc_page *ops , void *cbdata)
502 struct cl_page *page = ops->ops_cl.cpl_page;
503 struct osc_async_page *oap;
504 __u64 start = *(__u64 *)cbdata;
507 if (oap->oap_cmd & OBD_BRW_WRITE &&
508 !list_empty(&oap->oap_pending_item))
509 CL_PAGE_DEBUG(D_ERROR, env, page, "exists %llu/%s.\n",
510 start, current->comm);
512 if (PageLocked(page->cp_vmpage))
513 CDEBUG(D_CACHE, "page %p index %lu locked for %d.\n",
514 ops, osc_index(ops), oap->oap_cmd & OBD_BRW_RWMASK);
519 static void osc_trunc_check(const struct lu_env *env, struct cl_io *io,
520 struct osc_io *oio, __u64 size)
522 struct cl_object *clob;
526 clob = oio->oi_cl.cis_obj;
527 start = cl_index(clob, size);
528 partial = cl_offset(clob, start) < size;
531 * Complain if there are pages in the truncated region.
533 osc_page_gang_lookup(env, io, cl2osc(clob),
534 start + partial, CL_PAGE_EOF,
535 trunc_check_cb, (void *)&size);
538 static int osc_io_setattr_start(const struct lu_env *env,
539 const struct cl_io_slice *slice)
541 struct cl_io *io = slice->cis_io;
542 struct osc_io *oio = cl2osc_io(env, slice);
543 struct cl_object *obj = slice->cis_obj;
544 struct lov_oinfo *loi = cl2osc(obj)->oo_oinfo;
545 struct cl_attr *attr = &osc_env_info(env)->oti_attr;
546 struct obdo *oa = &oio->oi_oa;
547 struct osc_async_cbargs *cbargs = &oio->oi_cbarg;
548 unsigned int ia_avalid = io->u.ci_setattr.sa_avalid;
549 enum op_xvalid ia_xvalid = io->u.ci_setattr.sa_xvalid;
551 __u64 size = io->u.ci_setattr.sa_attr.lvb_size;
552 __u64 end = OBD_OBJECT_EOF;
553 bool io_is_falloc = false;
556 /* truncate cache dirty pages first */
557 if (cl_io_is_trunc(io)) {
558 result = osc_cache_truncate_start(env, cl2osc(obj), size,
560 } else if (cl_io_is_fallocate(io)) {
562 size = io->u.ci_setattr.sa_falloc_offset;
563 end = io->u.ci_setattr.sa_falloc_end;
566 if (result == 0 && oio->oi_lockless == 0) {
567 cl_object_attr_lock(obj);
568 result = cl_object_attr_get(env, obj, attr);
570 struct ost_lvb *lvb = &io->u.ci_setattr.sa_attr;
571 unsigned int cl_valid = 0;
573 if (ia_avalid & ATTR_SIZE) {
576 io->u.ci_setattr.sa_attr.lvb_size;
577 attr->cat_kms = attr->cat_size;
579 attr->cat_size = size;
580 attr->cat_kms = size;
582 cl_valid = (CAT_SIZE | CAT_KMS);
584 if (ia_avalid & ATTR_MTIME_SET) {
585 attr->cat_mtime = lvb->lvb_mtime;
586 cl_valid |= CAT_MTIME;
588 if (ia_avalid & ATTR_ATIME_SET) {
589 attr->cat_atime = lvb->lvb_atime;
590 cl_valid |= CAT_ATIME;
592 if (ia_xvalid & OP_XVALID_CTIME_SET) {
593 attr->cat_ctime = lvb->lvb_ctime;
594 cl_valid |= CAT_CTIME;
596 result = cl_object_attr_update(env, obj, attr,
599 cl_object_attr_unlock(obj);
601 memset(oa, 0, sizeof(*oa));
603 oa->o_oi = loi->loi_oi;
604 obdo_set_parent_fid(oa, io->u.ci_setattr.sa_parent_fid);
605 oa->o_stripe_idx = io->u.ci_setattr.sa_stripe_index;
606 oa->o_layout = io->u.ci_setattr.sa_layout;
607 oa->o_valid |= OBD_MD_FLID | OBD_MD_FLGROUP |
609 if (ia_avalid & ATTR_CTIME) {
610 oa->o_valid |= OBD_MD_FLCTIME;
611 oa->o_ctime = attr->cat_ctime;
613 if (ia_avalid & ATTR_ATIME) {
614 oa->o_valid |= OBD_MD_FLATIME;
615 oa->o_atime = attr->cat_atime;
617 if (ia_avalid & ATTR_MTIME) {
618 oa->o_valid |= OBD_MD_FLMTIME;
619 oa->o_mtime = attr->cat_mtime;
621 if (ia_avalid & ATTR_SIZE) {
625 oa->o_valid |= OBD_MD_FLSIZE | OBD_MD_FLBLOCKS;
628 oa->o_blocks = OBD_OBJECT_EOF;
629 oa->o_valid |= OBD_MD_FLSIZE | OBD_MD_FLBLOCKS;
632 if (oio->oi_lockless) {
633 oa->o_flags = OBD_FL_SRVLOCK;
634 oa->o_valid |= OBD_MD_FLFLAGS;
637 if (io->ci_layout_version > 0) {
638 /* verify layout version */
639 oa->o_valid |= OBD_MD_LAYOUT_VERSION;
640 oa->o_layout_version = io->ci_layout_version;
643 LASSERT(oio->oi_lockless == 0);
646 if (ia_xvalid & OP_XVALID_FLAGS) {
647 oa->o_flags = io->u.ci_setattr.sa_attr_flags;
648 oa->o_valid |= OBD_MD_FLFLAGS;
651 init_completion(&cbargs->opc_sync);
654 int falloc_mode = io->u.ci_setattr.sa_falloc_mode;
656 result = osc_fallocate_base(osc_export(cl2osc(obj)),
657 oa, osc_async_upcall,
658 cbargs, falloc_mode);
659 } else if (ia_avalid & ATTR_SIZE) {
660 result = osc_punch_send(osc_export(cl2osc(obj)),
661 oa, osc_async_upcall, cbargs);
663 result = osc_setattr_async(osc_export(cl2osc(obj)),
664 oa, osc_async_upcall,
665 cbargs, PTLRPCD_SET);
667 cbargs->opc_rpc_sent = result == 0;
673 void osc_io_setattr_end(const struct lu_env *env,
674 const struct cl_io_slice *slice)
676 struct cl_io *io = slice->cis_io;
677 struct osc_io *oio = cl2osc_io(env, slice);
678 struct cl_object *obj = slice->cis_obj;
679 struct osc_async_cbargs *cbargs = &oio->oi_cbarg;
680 struct cl_attr *attr = &osc_env_info(env)->oti_attr;
681 struct obdo *oa = &oio->oi_oa;
682 unsigned int cl_valid = 0;
685 if (cbargs->opc_rpc_sent) {
686 wait_for_completion(&cbargs->opc_sync);
687 result = io->ci_result = cbargs->opc_rc;
691 if (oio->oi_lockless) {
692 /* lockless truncate */
693 struct osc_device *osd = lu2osc_dev(obj->co_lu.lo_dev);
695 LASSERT(cl_io_is_trunc(io));
696 LASSERT(cl_io_is_trunc(io) || cl_io_is_fallocate(io));
697 /* XXX: Need a lock. */
698 osd->od_stats.os_lockless_truncates++;
702 if (cl_io_is_trunc(io)) {
703 __u64 size = io->u.ci_setattr.sa_attr.lvb_size;
706 cl_object_attr_lock(obj);
707 if (oa->o_valid & OBD_MD_FLBLOCKS) {
708 attr->cat_blocks = oa->o_blocks;
709 cl_valid |= CAT_BLOCKS;
712 cl_object_attr_update(env, obj, attr, cl_valid);
713 cl_object_attr_unlock(obj);
715 osc_trunc_check(env, io, oio, size);
716 osc_cache_truncate_end(env, oio->oi_trunc);
717 oio->oi_trunc = NULL;
720 if (cl_io_is_fallocate(io)) {
722 cl_object_attr_lock(obj);
724 if (oa->o_valid & OBD_MD_FLBLOCKS) {
725 attr->cat_blocks = oa->o_blocks;
726 cl_valid |= CAT_BLOCKS;
729 cl_object_attr_update(env, obj, attr, cl_valid);
730 cl_object_attr_unlock(obj);
734 EXPORT_SYMBOL(osc_io_setattr_end);
736 struct osc_data_version_args {
737 struct osc_io *dva_oio;
741 osc_data_version_interpret(const struct lu_env *env, struct ptlrpc_request *req,
744 struct osc_data_version_args *dva = args;
745 struct osc_io *oio = dva->dva_oio;
746 const struct ost_body *body;
752 body = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
754 GOTO(out, rc = -EPROTO);
756 lustre_get_wire_obdo(&req->rq_import->imp_connect_data, &oio->oi_oa,
760 oio->oi_cbarg.opc_rc = rc;
761 complete(&oio->oi_cbarg.opc_sync);
766 static int osc_io_data_version_start(const struct lu_env *env,
767 const struct cl_io_slice *slice)
769 struct cl_data_version_io *dv = &slice->cis_io->u.ci_data_version;
770 struct osc_io *oio = cl2osc_io(env, slice);
771 struct obdo *oa = &oio->oi_oa;
772 struct osc_async_cbargs *cbargs = &oio->oi_cbarg;
773 struct osc_object *obj = cl2osc(slice->cis_obj);
774 struct lov_oinfo *loi = obj->oo_oinfo;
775 struct obd_export *exp = osc_export(obj);
776 struct ptlrpc_request *req;
777 struct ost_body *body;
778 struct osc_data_version_args *dva;
782 memset(oa, 0, sizeof(*oa));
783 oa->o_oi = loi->loi_oi;
784 oa->o_valid = OBD_MD_FLID | OBD_MD_FLGROUP;
786 if (dv->dv_flags & (LL_DV_RD_FLUSH | LL_DV_WR_FLUSH)) {
787 oa->o_valid |= OBD_MD_FLFLAGS;
788 oa->o_flags |= OBD_FL_SRVLOCK;
789 if (dv->dv_flags & LL_DV_WR_FLUSH)
790 oa->o_flags |= OBD_FL_FLUSH;
793 init_completion(&cbargs->opc_sync);
795 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_GETATTR);
799 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_GETATTR);
801 ptlrpc_request_free(req);
805 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
806 lustre_set_wire_obdo(&req->rq_import->imp_connect_data, &body->oa, oa);
808 ptlrpc_request_set_replen(req);
809 req->rq_interpret_reply = osc_data_version_interpret;
810 dva = ptlrpc_req_async_args(dva, req);
813 ptlrpcd_add_req(req);
818 static void osc_io_data_version_end(const struct lu_env *env,
819 const struct cl_io_slice *slice)
821 struct cl_data_version_io *dv = &slice->cis_io->u.ci_data_version;
822 struct osc_io *oio = cl2osc_io(env, slice);
823 struct osc_async_cbargs *cbargs = &oio->oi_cbarg;
826 wait_for_completion(&cbargs->opc_sync);
828 if (cbargs->opc_rc != 0) {
829 slice->cis_io->ci_result = cbargs->opc_rc;
831 slice->cis_io->ci_result = 0;
832 if (!(oio->oi_oa.o_valid &
833 (OBD_MD_LAYOUT_VERSION | OBD_MD_FLDATAVERSION)))
834 slice->cis_io->ci_result = -ENOTSUPP;
836 if (oio->oi_oa.o_valid & OBD_MD_LAYOUT_VERSION)
837 dv->dv_layout_version = oio->oi_oa.o_layout_version;
838 if (oio->oi_oa.o_valid & OBD_MD_FLDATAVERSION)
839 dv->dv_data_version = oio->oi_oa.o_data_version;
845 int osc_io_read_start(const struct lu_env *env,
846 const struct cl_io_slice *slice)
848 struct cl_object *obj = slice->cis_obj;
849 struct cl_attr *attr = &osc_env_info(env)->oti_attr;
853 if (!slice->cis_io->ci_noatime) {
854 cl_object_attr_lock(obj);
855 attr->cat_atime = ktime_get_real_seconds();
856 rc = cl_object_attr_update(env, obj, attr, CAT_ATIME);
857 cl_object_attr_unlock(obj);
862 EXPORT_SYMBOL(osc_io_read_start);
864 int osc_io_write_start(const struct lu_env *env,
865 const struct cl_io_slice *slice)
867 struct cl_object *obj = slice->cis_obj;
868 struct cl_attr *attr = &osc_env_info(env)->oti_attr;
872 OBD_FAIL_TIMEOUT(OBD_FAIL_OSC_DELAY_SETTIME, 1);
873 cl_object_attr_lock(obj);
874 attr->cat_mtime = attr->cat_ctime = ktime_get_real_seconds();
875 rc = cl_object_attr_update(env, obj, attr, CAT_MTIME | CAT_CTIME);
876 cl_object_attr_unlock(obj);
880 EXPORT_SYMBOL(osc_io_write_start);
882 int osc_fsync_ost(const struct lu_env *env, struct osc_object *obj,
883 struct cl_fsync_io *fio)
885 struct osc_io *oio = osc_env_io(env);
886 struct obdo *oa = &oio->oi_oa;
887 struct lov_oinfo *loi = obj->oo_oinfo;
888 struct osc_async_cbargs *cbargs = &oio->oi_cbarg;
892 memset(oa, 0, sizeof(*oa));
893 oa->o_oi = loi->loi_oi;
894 oa->o_valid = OBD_MD_FLID | OBD_MD_FLGROUP;
896 /* reload size abd blocks for start and end of sync range */
897 oa->o_size = fio->fi_start;
898 oa->o_blocks = fio->fi_end;
899 oa->o_valid |= OBD_MD_FLSIZE | OBD_MD_FLBLOCKS;
901 obdo_set_parent_fid(oa, fio->fi_fid);
903 init_completion(&cbargs->opc_sync);
905 rc = osc_sync_base(obj, oa, osc_async_upcall, cbargs, PTLRPCD_SET);
908 EXPORT_SYMBOL(osc_fsync_ost);
910 int osc_io_fsync_start(const struct lu_env *env,
911 const struct cl_io_slice *slice)
913 struct cl_io *io = slice->cis_io;
914 struct cl_fsync_io *fio = &io->u.ci_fsync;
915 struct cl_object *obj = slice->cis_obj;
916 struct osc_object *osc = cl2osc(obj);
917 pgoff_t start = cl_index(obj, fio->fi_start);
918 pgoff_t end = cl_index(obj, fio->fi_end);
922 if (fio->fi_end == OBD_OBJECT_EOF)
925 result = osc_cache_writeback_range(env, osc, start, end, 0,
926 fio->fi_mode == CL_FSYNC_DISCARD);
928 fio->fi_nr_written += result;
931 if (fio->fi_mode == CL_FSYNC_ALL) {
934 /* we have to wait for writeback to finish before we can
935 * send OST_SYNC RPC. This is bad because it causes extents
936 * to be written osc by osc. However, we usually start
937 * writeback before CL_FSYNC_ALL so this won't have any real
939 rc = osc_cache_wait_range(env, osc, start, end);
942 rc = osc_fsync_ost(env, osc, fio);
950 void osc_io_fsync_end(const struct lu_env *env,
951 const struct cl_io_slice *slice)
953 struct cl_fsync_io *fio = &slice->cis_io->u.ci_fsync;
954 struct cl_object *obj = slice->cis_obj;
955 pgoff_t start = cl_index(obj, fio->fi_start);
956 pgoff_t end = cl_index(obj, fio->fi_end);
959 if (fio->fi_mode == CL_FSYNC_LOCAL) {
960 result = osc_cache_wait_range(env, cl2osc(obj), start, end);
961 } else if (fio->fi_mode == CL_FSYNC_ALL) {
962 struct osc_io *oio = cl2osc_io(env, slice);
963 struct osc_async_cbargs *cbargs = &oio->oi_cbarg;
965 wait_for_completion(&cbargs->opc_sync);
967 result = cbargs->opc_rc;
969 slice->cis_io->ci_result = result;
971 EXPORT_SYMBOL(osc_io_fsync_end);
973 static int osc_io_ladvise_start(const struct lu_env *env,
974 const struct cl_io_slice *slice)
977 struct cl_io *io = slice->cis_io;
978 struct osc_io *oio = cl2osc_io(env, slice);
979 struct cl_object *obj = slice->cis_obj;
980 struct lov_oinfo *loi = cl2osc(obj)->oo_oinfo;
981 struct cl_ladvise_io *lio = &io->u.ci_ladvise;
982 struct obdo *oa = &oio->oi_oa;
983 struct osc_async_cbargs *cbargs = &oio->oi_cbarg;
984 struct lu_ladvise *ladvise;
985 struct ladvise_hdr *ladvise_hdr;
990 /* TODO: add multiple ladvise support in CLIO */
991 buf_size = offsetof(typeof(*ladvise_hdr), lah_advise[num_advise]);
992 if (osc_env_info(env)->oti_ladvise_buf.lb_len < buf_size)
993 lu_buf_realloc(&osc_env_info(env)->oti_ladvise_buf, buf_size);
995 ladvise_hdr = osc_env_info(env)->oti_ladvise_buf.lb_buf;
996 if (ladvise_hdr == NULL)
999 memset(ladvise_hdr, 0, buf_size);
1000 ladvise_hdr->lah_magic = LADVISE_MAGIC;
1001 ladvise_hdr->lah_count = num_advise;
1002 ladvise_hdr->lah_flags = lio->li_flags;
1004 memset(oa, 0, sizeof(*oa));
1005 oa->o_oi = loi->loi_oi;
1006 oa->o_valid = OBD_MD_FLID;
1007 obdo_set_parent_fid(oa, lio->li_fid);
1009 ladvise = ladvise_hdr->lah_advise;
1010 ladvise->lla_start = lio->li_start;
1011 ladvise->lla_end = lio->li_end;
1012 ladvise->lla_advice = lio->li_advice;
1014 if (lio->li_flags & LF_ASYNC) {
1015 result = osc_ladvise_base(osc_export(cl2osc(obj)), oa,
1016 ladvise_hdr, NULL, NULL, NULL);
1018 init_completion(&cbargs->opc_sync);
1019 result = osc_ladvise_base(osc_export(cl2osc(obj)), oa,
1020 ladvise_hdr, osc_async_upcall,
1021 cbargs, PTLRPCD_SET);
1022 cbargs->opc_rpc_sent = result == 0;
1027 static void osc_io_ladvise_end(const struct lu_env *env,
1028 const struct cl_io_slice *slice)
1030 struct cl_io *io = slice->cis_io;
1031 struct osc_io *oio = cl2osc_io(env, slice);
1032 struct osc_async_cbargs *cbargs = &oio->oi_cbarg;
1034 struct cl_ladvise_io *lio = &io->u.ci_ladvise;
1036 if ((!(lio->li_flags & LF_ASYNC)) && cbargs->opc_rpc_sent) {
1037 wait_for_completion(&cbargs->opc_sync);
1038 result = cbargs->opc_rc;
1040 slice->cis_io->ci_result = result;
1043 void osc_io_end(const struct lu_env *env, const struct cl_io_slice *slice)
1045 struct osc_io *oio = cl2osc_io(env, slice);
1047 if (oio->oi_active) {
1048 osc_extent_release(env, oio->oi_active);
1049 oio->oi_active = NULL;
1052 EXPORT_SYMBOL(osc_io_end);
1054 struct osc_lseek_args {
1055 struct osc_io *lsa_oio;
1058 static int osc_lseek_interpret(const struct lu_env *env,
1059 struct ptlrpc_request *req,
1062 struct ost_body *reply;
1063 struct osc_lseek_args *lsa = arg;
1064 struct osc_io *oio = lsa->lsa_oio;
1065 struct cl_io *io = oio->oi_cl.cis_io;
1066 struct cl_lseek_io *lsio = &io->u.ci_lseek;
1073 reply = req_capsule_server_get(&req->rq_pill, &RMF_OST_BODY);
1075 GOTO(out, rc = -EPROTO);
1077 lsio->ls_result = reply->oa.o_size;
1079 osc_async_upcall(&oio->oi_cbarg, rc);
1083 int osc_io_lseek_start(const struct lu_env *env,
1084 const struct cl_io_slice *slice)
1086 struct cl_io *io = slice->cis_io;
1087 struct osc_io *oio = cl2osc_io(env, slice);
1088 struct cl_object *obj = slice->cis_obj;
1089 struct lov_oinfo *loi = cl2osc(obj)->oo_oinfo;
1090 struct cl_lseek_io *lsio = &io->u.ci_lseek;
1091 struct obdo *oa = &oio->oi_oa;
1092 struct osc_async_cbargs *cbargs = &oio->oi_cbarg;
1093 struct obd_export *exp = osc_export(cl2osc(obj));
1094 struct ptlrpc_request *req;
1095 struct ost_body *body;
1096 struct osc_lseek_args *lsa;
1101 /* No negative values at this point */
1102 LASSERT(lsio->ls_start >= 0);
1103 LASSERT(lsio->ls_whence == SEEK_HOLE || lsio->ls_whence == SEEK_DATA);
1105 /* with IO lock taken we have object size in LVB and can check
1106 * boundaries prior sending LSEEK RPC
1108 if (lsio->ls_start >= loi->loi_lvb.lvb_size) {
1109 /* consider area beyond end of object as hole */
1110 if (lsio->ls_whence == SEEK_HOLE)
1111 lsio->ls_result = lsio->ls_start;
1113 lsio->ls_result = -ENXIO;
1117 /* if LSEEK RPC is not supported by server, consider whole stripe
1118 * object is data with hole after end of object
1120 if (!exp_connect_lseek(exp)) {
1121 if (lsio->ls_whence == SEEK_HOLE)
1122 lsio->ls_result = loi->loi_lvb.lvb_size;
1124 lsio->ls_result = lsio->ls_start;
1128 memset(oa, 0, sizeof(*oa));
1129 oa->o_oi = loi->loi_oi;
1130 oa->o_valid = OBD_MD_FLID | OBD_MD_FLGROUP;
1131 oa->o_size = lsio->ls_start;
1132 oa->o_mode = lsio->ls_whence;
1133 if (oio->oi_lockless) {
1134 oa->o_flags = OBD_FL_SRVLOCK;
1135 oa->o_valid |= OBD_MD_FLFLAGS;
1138 init_completion(&cbargs->opc_sync);
1139 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_OST_SEEK);
1143 rc = ptlrpc_request_pack(req, LUSTRE_OST_VERSION, OST_SEEK);
1145 ptlrpc_request_free(req);
1149 body = req_capsule_client_get(&req->rq_pill, &RMF_OST_BODY);
1150 lustre_set_wire_obdo(&req->rq_import->imp_connect_data, &body->oa, oa);
1151 ptlrpc_request_set_replen(req);
1152 req->rq_interpret_reply = osc_lseek_interpret;
1153 lsa = ptlrpc_req_async_args(lsa, req);
1156 ptlrpcd_add_req(req);
1157 cbargs->opc_rpc_sent = 1;
1161 EXPORT_SYMBOL(osc_io_lseek_start);
1163 void osc_io_lseek_end(const struct lu_env *env,
1164 const struct cl_io_slice *slice)
1166 struct osc_io *oio = cl2osc_io(env, slice);
1167 struct osc_async_cbargs *cbargs = &oio->oi_cbarg;
1170 if (cbargs->opc_rpc_sent) {
1171 wait_for_completion(&cbargs->opc_sync);
1172 rc = cbargs->opc_rc;
1174 slice->cis_io->ci_result = rc;
1176 EXPORT_SYMBOL(osc_io_lseek_end);
1178 int osc_io_lru_reserve(const struct lu_env *env,
1179 const struct cl_io_slice *ios,
1180 loff_t pos, size_t bytes)
1182 struct osc_object *osc = cl2osc(ios->cis_obj);
1183 struct osc_io *oio = osc_env_io(env);
1184 unsigned long npages = 0;
1189 page_offset = pos & ~PAGE_MASK;
1192 if (bytes > PAGE_SIZE - page_offset)
1193 bytes -= (PAGE_SIZE - page_offset);
1197 npages += (bytes + PAGE_SIZE - 1) >> PAGE_SHIFT;
1198 oio->oi_lru_reserved = osc_lru_reserve(osc_cli(osc), npages);
1202 EXPORT_SYMBOL(osc_io_lru_reserve);
1204 static const struct cl_io_operations osc_io_ops = {
1207 .cio_iter_init = osc_io_iter_init,
1208 .cio_iter_fini = osc_io_rw_iter_fini,
1209 .cio_start = osc_io_read_start,
1210 .cio_fini = osc_io_fini
1213 .cio_iter_init = osc_io_iter_init,
1214 .cio_iter_fini = osc_io_rw_iter_fini,
1215 .cio_start = osc_io_write_start,
1216 .cio_end = osc_io_end,
1217 .cio_fini = osc_io_fini
1220 .cio_iter_init = osc_io_iter_init,
1221 .cio_iter_fini = osc_io_iter_fini,
1222 .cio_start = osc_io_setattr_start,
1223 .cio_end = osc_io_setattr_end
1225 [CIT_DATA_VERSION] = {
1226 .cio_start = osc_io_data_version_start,
1227 .cio_end = osc_io_data_version_end,
1230 .cio_iter_init = osc_io_iter_init,
1231 .cio_iter_fini = osc_io_iter_fini,
1232 .cio_start = osc_io_fault_start,
1233 .cio_end = osc_io_end,
1234 .cio_fini = osc_io_fini
1237 .cio_start = osc_io_fsync_start,
1238 .cio_end = osc_io_fsync_end,
1239 .cio_fini = osc_io_fini
1242 .cio_start = osc_io_ladvise_start,
1243 .cio_end = osc_io_ladvise_end,
1244 .cio_fini = osc_io_fini
1247 .cio_start = osc_io_lseek_start,
1248 .cio_end = osc_io_lseek_end,
1249 .cio_fini = osc_io_fini
1252 .cio_fini = osc_io_fini
1255 .cio_read_ahead = osc_io_read_ahead,
1256 .cio_lru_reserve = osc_io_lru_reserve,
1257 .cio_submit = osc_io_submit,
1258 .cio_commit_async = osc_io_commit_async,
1259 .cio_extent_release = osc_io_extent_release
1262 /*****************************************************************************
1264 * Transfer operations.
1268 int osc_io_init(const struct lu_env *env,
1269 struct cl_object *obj, struct cl_io *io)
1271 struct osc_io *oio = osc_env_io(env);
1273 CL_IO_SLICE_CLEAN(oio, oi_cl);
1274 cl_io_slice_add(io, &oio->oi_cl, obj, &osc_io_ops);