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) 2009, 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/
31 * Lustre Unified Target
32 * These are common function to work with last_received file
34 * Author: Mikhail Pershin <mike.pershin@intel.com>
37 #include <obd_class.h>
38 #include <lustre_fid.h>
40 #include "tgt_internal.h"
42 /** version recovery epoch */
43 #define LR_EPOCH_BITS 32
45 /* Allocate a bitmap for a chunk of reply data slots */
46 static int tgt_bitmap_chunk_alloc(struct lu_target *lut, int chunk)
50 OBD_ALLOC_LARGE(bm, BITS_TO_LONGS(LUT_REPLY_SLOTS_PER_CHUNK) *
55 spin_lock(&lut->lut_client_bitmap_lock);
57 if (lut->lut_reply_bitmap[chunk] != NULL) {
58 /* someone else already allocated the bitmap for this chunk */
59 spin_unlock(&lut->lut_client_bitmap_lock);
60 OBD_FREE_LARGE(bm, BITS_TO_LONGS(LUT_REPLY_SLOTS_PER_CHUNK) *
65 lut->lut_reply_bitmap[chunk] = bm;
67 spin_unlock(&lut->lut_client_bitmap_lock);
72 /* Look for an available reply data slot in the bitmap
74 * Allocate bitmap chunk when first used
75 * XXX algo could be improved if this routine limits performance
77 static int tgt_find_free_reply_slot(struct lu_target *lut)
84 for (chunk = 0; chunk < LUT_REPLY_SLOTS_MAX_CHUNKS; chunk++) {
85 /* allocate the bitmap chunk if necessary */
86 if (unlikely(lut->lut_reply_bitmap[chunk] == NULL)) {
87 rc = tgt_bitmap_chunk_alloc(lut, chunk);
91 bmp = lut->lut_reply_bitmap[chunk];
93 /* look for an available slot in this chunk */
95 b = find_first_zero_bit(bmp, LUT_REPLY_SLOTS_PER_CHUNK);
96 if (b >= LUT_REPLY_SLOTS_PER_CHUNK)
100 if (test_and_set_bit(b, bmp) == 0)
101 return chunk * LUT_REPLY_SLOTS_PER_CHUNK + b;
108 /* Mark the reply data slot @idx 'used' in the corresponding bitmap chunk
110 * Allocate the bitmap chunk if necessary
112 static int tgt_set_reply_slot(struct lu_target *lut, int idx)
118 chunk = idx / LUT_REPLY_SLOTS_PER_CHUNK;
119 b = idx % LUT_REPLY_SLOTS_PER_CHUNK;
121 LASSERT(chunk < LUT_REPLY_SLOTS_MAX_CHUNKS);
122 LASSERT(b < LUT_REPLY_SLOTS_PER_CHUNK);
124 /* allocate the bitmap chunk if necessary */
125 if (unlikely(lut->lut_reply_bitmap[chunk] == NULL)) {
126 rc = tgt_bitmap_chunk_alloc(lut, chunk);
131 /* mark the slot 'used' in this chunk */
132 if (test_and_set_bit(b, lut->lut_reply_bitmap[chunk]) != 0) {
133 CERROR("%s: slot %d already set in bitmap\n",
142 /* Mark the reply data slot @idx 'unused' in the corresponding bitmap chunk
145 static int tgt_clear_reply_slot(struct lu_target *lut, int idx)
150 if (lut->lut_obd->obd_stopping)
152 * in case of failover keep the bit set in order to
153 * avoid overwriting slots in reply_data which might
154 * be required by resent rpcs
157 chunk = idx / LUT_REPLY_SLOTS_PER_CHUNK;
158 b = idx % LUT_REPLY_SLOTS_PER_CHUNK;
160 LASSERT(chunk < LUT_REPLY_SLOTS_MAX_CHUNKS);
161 LASSERT(b < LUT_REPLY_SLOTS_PER_CHUNK);
163 if (lut->lut_reply_bitmap[chunk] == NULL) {
164 CERROR("%s: slot %d not allocated\n",
169 if (test_and_clear_bit(b, lut->lut_reply_bitmap[chunk]) == 0) {
170 CERROR("%s: slot %d already clear in bitmap\n",
179 /* Read header of reply_data file of target @tgt into structure @lrh */
180 static int tgt_reply_header_read(const struct lu_env *env,
181 struct lu_target *tgt,
182 struct lsd_reply_header *lrh)
185 struct lsd_reply_header buf;
186 struct tgt_thread_info *tti = tgt_th_info(env);
189 tti->tti_buf.lb_buf = &buf;
190 tti->tti_buf.lb_len = sizeof(buf);
192 rc = dt_record_read(env, tgt->lut_reply_data, &tti->tti_buf,
197 lrh->lrh_magic = le32_to_cpu(buf.lrh_magic);
198 lrh->lrh_header_size = le32_to_cpu(buf.lrh_header_size);
199 lrh->lrh_reply_size = le32_to_cpu(buf.lrh_reply_size);
201 CDEBUG(D_HA, "%s: read %s header. magic=0x%08x "
202 "header_size=%d reply_size=%d\n",
203 tgt->lut_obd->obd_name, REPLY_DATA,
204 lrh->lrh_magic, lrh->lrh_header_size, lrh->lrh_reply_size);
209 /* Write header into replay_data file of target @tgt from structure @lrh */
210 static int tgt_reply_header_write(const struct lu_env *env,
211 struct lu_target *tgt,
212 struct lsd_reply_header *lrh)
215 struct lsd_reply_header buf;
216 struct tgt_thread_info *tti = tgt_th_info(env);
218 struct dt_object *dto;
220 CDEBUG(D_HA, "%s: write %s header. magic=0x%08x "
221 "header_size=%d reply_size=%d\n",
222 tgt->lut_obd->obd_name, REPLY_DATA,
223 lrh->lrh_magic, lrh->lrh_header_size, lrh->lrh_reply_size);
225 if (tgt->lut_bottom->dd_rdonly)
228 buf.lrh_magic = cpu_to_le32(lrh->lrh_magic);
229 buf.lrh_header_size = cpu_to_le32(lrh->lrh_header_size);
230 buf.lrh_reply_size = cpu_to_le32(lrh->lrh_reply_size);
232 th = dt_trans_create(env, tgt->lut_bottom);
238 tti->tti_buf.lb_buf = &buf;
239 tti->tti_buf.lb_len = sizeof(buf);
241 rc = dt_declare_record_write(env, tgt->lut_reply_data,
242 &tti->tti_buf, tti->tti_off, th);
246 rc = dt_trans_start(env, tgt->lut_bottom, th);
250 dto = dt_object_locate(tgt->lut_reply_data, th->th_dev);
251 rc = dt_record_write(env, dto, &tti->tti_buf, &tti->tti_off, th);
253 dt_trans_stop(env, tgt->lut_bottom, th);
257 /* Write the reply data @lrd into reply_data file of target @tgt
260 static int tgt_reply_data_write(const struct lu_env *env, struct lu_target *tgt,
261 struct lsd_reply_data *lrd, loff_t off,
264 struct tgt_thread_info *tti = tgt_th_info(env);
265 struct dt_object *dto;
266 struct lsd_reply_data *buf = &tti->tti_lrd;
268 lrd->lrd_result = ptlrpc_status_hton(lrd->lrd_result);
270 buf->lrd_transno = cpu_to_le64(lrd->lrd_transno);
271 buf->lrd_xid = cpu_to_le64(lrd->lrd_xid);
272 buf->lrd_data = cpu_to_le64(lrd->lrd_data);
273 buf->lrd_result = cpu_to_le32(lrd->lrd_result);
274 buf->lrd_client_gen = cpu_to_le32(lrd->lrd_client_gen);
276 lrd->lrd_result = ptlrpc_status_ntoh(lrd->lrd_result);
279 tti->tti_buf.lb_buf = buf;
280 tti->tti_buf.lb_len = sizeof(*buf);
282 dto = dt_object_locate(tgt->lut_reply_data, th->th_dev);
283 return dt_record_write(env, dto, &tti->tti_buf, &tti->tti_off, th);
286 /* Read the reply data from reply_data file of target @tgt at offset @off
287 * into structure @lrd
289 static int tgt_reply_data_read(const struct lu_env *env, struct lu_target *tgt,
290 struct lsd_reply_data *lrd, loff_t off)
293 struct tgt_thread_info *tti = tgt_th_info(env);
294 struct lsd_reply_data *buf = &tti->tti_lrd;
297 tti->tti_buf.lb_buf = buf;
298 tti->tti_buf.lb_len = sizeof(*buf);
300 rc = dt_record_read(env, tgt->lut_reply_data, &tti->tti_buf,
305 lrd->lrd_transno = le64_to_cpu(buf->lrd_transno);
306 lrd->lrd_xid = le64_to_cpu(buf->lrd_xid);
307 lrd->lrd_data = le64_to_cpu(buf->lrd_data);
308 lrd->lrd_result = le32_to_cpu(buf->lrd_result);
309 lrd->lrd_client_gen = le32_to_cpu(buf->lrd_client_gen);
315 /* Free the in-memory reply data structure @trd and release
316 * the corresponding slot in the reply_data file of target @lut
317 * Called with ted_lcd_lock held
319 static void tgt_free_reply_data(struct lu_target *lut,
320 struct tg_export_data *ted,
321 struct tg_reply_data *trd)
323 CDEBUG(D_TRACE, "%s: free reply data %p: xid %llu, transno %llu, "
324 "client gen %u, slot idx %d\n",
325 lut == NULL ? "" : tgt_name(lut), trd, trd->trd_reply.lrd_xid,
326 trd->trd_reply.lrd_transno, trd->trd_reply.lrd_client_gen,
329 LASSERT(mutex_is_locked(&ted->ted_lcd_lock));
331 list_del(&trd->trd_list);
332 ted->ted_reply_cnt--;
333 if (lut != NULL && trd->trd_index != TRD_INDEX_MEMORY)
334 tgt_clear_reply_slot(lut, trd->trd_index);
338 /* Release the reply data @trd from target @lut
339 * The reply data with the highest transno for this export
340 * is retained to ensure correctness of target recovery
341 * Called with ted_lcd_lock held
343 static void tgt_release_reply_data(struct lu_target *lut,
344 struct tg_export_data *ted,
345 struct tg_reply_data *trd)
347 CDEBUG(D_TRACE, "%s: release reply data %p: xid %llu, transno %llu, "
348 "client gen %u, slot idx %d\n",
349 lut == NULL ? "" : tgt_name(lut), trd, trd->trd_reply.lrd_xid,
350 trd->trd_reply.lrd_transno, trd->trd_reply.lrd_client_gen,
353 LASSERT(mutex_is_locked(&ted->ted_lcd_lock));
355 /* Do not free the reply data corresponding to the
356 * highest transno of this export.
357 * This ensures on-disk reply data is kept and
358 * last committed transno can be restored from disk in case
361 if (trd->trd_reply.lrd_transno == ted->ted_lcd->lcd_last_transno) {
362 /* free previous retained reply */
363 if (ted->ted_reply_last != NULL)
364 tgt_free_reply_data(lut, ted, ted->ted_reply_last);
365 /* retain the reply */
366 list_del_init(&trd->trd_list);
367 ted->ted_reply_last = trd;
369 tgt_free_reply_data(lut, ted, trd);
373 static inline struct lu_buf *tti_buf_lsd(struct tgt_thread_info *tti)
375 tti->tti_buf.lb_buf = &tti->tti_lsd;
376 tti->tti_buf.lb_len = sizeof(tti->tti_lsd);
377 return &tti->tti_buf;
380 static inline struct lu_buf *tti_buf_lcd(struct tgt_thread_info *tti)
382 tti->tti_buf.lb_buf = &tti->tti_lcd;
383 tti->tti_buf.lb_len = sizeof(tti->tti_lcd);
384 return &tti->tti_buf;
388 * Allocate in-memory data for client slot related to export.
390 int tgt_client_alloc(struct obd_export *exp)
393 LASSERT(exp != exp->exp_obd->obd_self_export);
395 spin_lock_init(&exp->exp_target_data.ted_nodemap_lock);
396 INIT_LIST_HEAD(&exp->exp_target_data.ted_nodemap_member);
397 spin_lock_init(&exp->exp_target_data.ted_fmd_lock);
398 INIT_LIST_HEAD(&exp->exp_target_data.ted_fmd_list);
400 OBD_ALLOC_PTR(exp->exp_target_data.ted_lcd);
401 if (exp->exp_target_data.ted_lcd == NULL)
403 /* Mark that slot is not yet valid, 0 doesn't work here */
404 exp->exp_target_data.ted_lr_idx = -1;
405 INIT_LIST_HEAD(&exp->exp_target_data.ted_reply_list);
406 mutex_init(&exp->exp_target_data.ted_lcd_lock);
409 EXPORT_SYMBOL(tgt_client_alloc);
412 * Free in-memory data for client slot related to export.
414 void tgt_client_free(struct obd_export *exp)
416 struct tg_export_data *ted = &exp->exp_target_data;
417 struct lu_target *lut = class_exp2tgt(exp);
418 struct tg_reply_data *trd, *tmp;
420 LASSERT(exp != exp->exp_obd->obd_self_export);
422 tgt_fmd_cleanup(exp);
424 /* free reply data */
425 mutex_lock(&ted->ted_lcd_lock);
426 list_for_each_entry_safe(trd, tmp, &ted->ted_reply_list, trd_list) {
427 tgt_release_reply_data(lut, ted, trd);
429 if (ted->ted_reply_last != NULL) {
430 tgt_free_reply_data(lut, ted, ted->ted_reply_last);
431 ted->ted_reply_last = NULL;
433 mutex_unlock(&ted->ted_lcd_lock);
435 if (!hlist_unhashed(&exp->exp_gen_hash))
436 cfs_hash_del(exp->exp_obd->obd_gen_hash,
437 &ted->ted_lcd->lcd_generation,
440 OBD_FREE_PTR(ted->ted_lcd);
443 /* Target may have been freed (see LU-7430)
444 * Slot may be not yet assigned */
445 if (exp->exp_obd->u.obt.obt_magic != OBT_MAGIC ||
449 /* Clear bit when lcd is freed */
450 LASSERT(lut && lut->lut_client_bitmap);
451 if (!test_and_clear_bit(ted->ted_lr_idx, lut->lut_client_bitmap)) {
452 CERROR("%s: client %u bit already clear in bitmap\n",
453 exp->exp_obd->obd_name, ted->ted_lr_idx);
457 if (tgt_is_multimodrpcs_client(exp) && !exp->exp_obd->obd_stopping)
458 atomic_dec(&lut->lut_num_clients);
460 EXPORT_SYMBOL(tgt_client_free);
462 static inline void tgt_check_lcd(const char *obd_name, int index,
463 struct lsd_client_data *lcd)
465 size_t uuid_size = sizeof(lcd->lcd_uuid);
467 if (strnlen((char*)lcd->lcd_uuid, uuid_size) == uuid_size) {
468 lcd->lcd_uuid[uuid_size - 1] = '\0';
470 LCONSOLE_ERROR("the client UUID (%s) on %s for exports stored in last_rcvd(index = %d) is bad!\n",
471 lcd->lcd_uuid, obd_name, index);
475 static int tgt_client_data_read(const struct lu_env *env, struct lu_target *tgt,
476 struct lsd_client_data *lcd,
477 loff_t *off, int index)
479 struct tgt_thread_info *tti = tgt_th_info(env);
483 rc = dt_record_read(env, tgt->lut_last_rcvd, &tti->tti_buf, off);
485 tgt_check_lcd(tgt->lut_obd->obd_name, index, &tti->tti_lcd);
486 lcd_le_to_cpu(&tti->tti_lcd, lcd);
487 lcd->lcd_last_result = ptlrpc_status_ntoh(lcd->lcd_last_result);
488 lcd->lcd_last_close_result =
489 ptlrpc_status_ntoh(lcd->lcd_last_close_result);
492 CDEBUG(D_INFO, "%s: read lcd @%lld uuid = %s, last_transno = %llu"
493 ", last_xid = %llu, last_result = %u, last_data = %u, "
494 "last_close_transno = %llu, last_close_xid = %llu, "
495 "last_close_result = %u, rc = %d\n", tgt->lut_obd->obd_name,
496 *off, lcd->lcd_uuid, lcd->lcd_last_transno, lcd->lcd_last_xid,
497 lcd->lcd_last_result, lcd->lcd_last_data,
498 lcd->lcd_last_close_transno, lcd->lcd_last_close_xid,
499 lcd->lcd_last_close_result, rc);
503 static int tgt_client_data_write(const struct lu_env *env,
504 struct lu_target *tgt,
505 struct lsd_client_data *lcd,
506 loff_t *off, struct thandle *th)
508 struct tgt_thread_info *tti = tgt_th_info(env);
509 struct dt_object *dto;
511 lcd->lcd_last_result = ptlrpc_status_hton(lcd->lcd_last_result);
512 lcd->lcd_last_close_result =
513 ptlrpc_status_hton(lcd->lcd_last_close_result);
514 lcd_cpu_to_le(lcd, &tti->tti_lcd);
517 dto = dt_object_locate(tgt->lut_last_rcvd, th->th_dev);
518 return dt_record_write(env, dto, &tti->tti_buf, off, th);
521 struct tgt_new_client_callback {
522 struct dt_txn_commit_cb lncc_cb;
523 struct obd_export *lncc_exp;
526 static void tgt_cb_new_client(struct lu_env *env, struct thandle *th,
527 struct dt_txn_commit_cb *cb, int err)
529 struct tgt_new_client_callback *ccb;
531 ccb = container_of(cb, struct tgt_new_client_callback, lncc_cb);
533 LASSERT(ccb->lncc_exp->exp_obd);
535 CDEBUG(D_RPCTRACE, "%s: committing for initial connect of %s\n",
536 ccb->lncc_exp->exp_obd->obd_name,
537 ccb->lncc_exp->exp_client_uuid.uuid);
539 spin_lock(&ccb->lncc_exp->exp_lock);
541 ccb->lncc_exp->exp_need_sync = 0;
543 spin_unlock(&ccb->lncc_exp->exp_lock);
544 class_export_cb_put(ccb->lncc_exp);
549 int tgt_new_client_cb_add(struct thandle *th, struct obd_export *exp)
551 struct tgt_new_client_callback *ccb;
552 struct dt_txn_commit_cb *dcb;
559 ccb->lncc_exp = class_export_cb_get(exp);
562 dcb->dcb_func = tgt_cb_new_client;
563 INIT_LIST_HEAD(&dcb->dcb_linkage);
564 strlcpy(dcb->dcb_name, "tgt_cb_new_client", sizeof(dcb->dcb_name));
566 rc = dt_trans_cb_add(th, dcb);
568 class_export_cb_put(exp);
575 * Update client data in last_rcvd
577 static int tgt_client_data_update(const struct lu_env *env,
578 struct obd_export *exp)
580 struct tg_export_data *ted = &exp->exp_target_data;
581 struct lu_target *tgt = class_exp2tgt(exp);
582 struct tgt_thread_info *tti = tgt_th_info(env);
588 if (unlikely(tgt == NULL)) {
589 CDEBUG(D_ERROR, "%s: No target for connected export\n",
590 class_exp2obd(exp)->obd_name);
594 if (tgt->lut_bottom->dd_rdonly)
597 th = dt_trans_create(env, tgt->lut_bottom);
602 rc = dt_declare_record_write(env, tgt->lut_last_rcvd,
604 ted->ted_lr_off, th);
608 rc = dt_trans_start_local(env, tgt->lut_bottom, th);
612 mutex_lock(&ted->ted_lcd_lock);
615 * Until this operations will be committed the sync is needed
616 * for this export. This should be done _after_ starting the
617 * transaction so that many connecting clients will not bring
618 * server down with lots of sync writes.
620 rc = tgt_new_client_cb_add(th, exp);
622 /* can't add callback, do sync now */
625 spin_lock(&exp->exp_lock);
626 exp->exp_need_sync = 1;
627 spin_unlock(&exp->exp_lock);
630 tti->tti_off = ted->ted_lr_off;
631 rc = tgt_client_data_write(env, tgt, ted->ted_lcd, &tti->tti_off, th);
633 mutex_unlock(&ted->ted_lcd_lock);
637 dt_trans_stop(env, tgt->lut_bottom, th);
638 CDEBUG(D_INFO, "%s: update last_rcvd client data for UUID = %s, "
639 "last_transno = %llu: rc = %d\n", tgt->lut_obd->obd_name,
640 tgt->lut_lsd.lsd_uuid, tgt->lut_lsd.lsd_last_transno, rc);
645 static int tgt_server_data_read(const struct lu_env *env, struct lu_target *tgt)
647 struct tgt_thread_info *tti = tgt_th_info(env);
652 rc = dt_record_read(env, tgt->lut_last_rcvd, &tti->tti_buf,
655 lsd_le_to_cpu(&tti->tti_lsd, &tgt->lut_lsd);
657 CDEBUG(D_INFO, "%s: read last_rcvd server data for UUID = %s, "
658 "last_transno = %llu: rc = %d\n", tgt->lut_obd->obd_name,
659 tgt->lut_lsd.lsd_uuid, tgt->lut_lsd.lsd_last_transno, rc);
663 static int tgt_server_data_write(const struct lu_env *env,
664 struct lu_target *tgt, struct thandle *th)
666 struct tgt_thread_info *tti = tgt_th_info(env);
667 struct dt_object *dto;
674 lsd_cpu_to_le(&tgt->lut_lsd, &tti->tti_lsd);
676 dto = dt_object_locate(tgt->lut_last_rcvd, th->th_dev);
677 rc = dt_record_write(env, dto, &tti->tti_buf, &tti->tti_off, th);
679 CDEBUG(D_INFO, "%s: write last_rcvd server data for UUID = %s, "
680 "last_transno = %llu: rc = %d\n", tgt->lut_obd->obd_name,
681 tgt->lut_lsd.lsd_uuid, tgt->lut_lsd.lsd_last_transno, rc);
687 * Update server data in last_rcvd
689 int tgt_server_data_update(const struct lu_env *env, struct lu_target *tgt,
692 struct tgt_thread_info *tti = tgt_th_info(env);
699 "%s: mount_count is %llu, last_transno is %llu\n",
700 tgt->lut_lsd.lsd_uuid, tgt->lut_obd->u.obt.obt_mount_count,
701 tgt->lut_last_transno);
703 /* Always save latest transno to keep it fresh */
704 spin_lock(&tgt->lut_translock);
705 tgt->lut_lsd.lsd_last_transno = tgt->lut_last_transno;
706 spin_unlock(&tgt->lut_translock);
708 if (tgt->lut_bottom->dd_rdonly)
711 th = dt_trans_create(env, tgt->lut_bottom);
718 rc = dt_declare_record_write(env, tgt->lut_last_rcvd,
719 &tti->tti_buf, tti->tti_off, th);
723 rc = dt_trans_start(env, tgt->lut_bottom, th);
727 rc = tgt_server_data_write(env, tgt, th);
729 dt_trans_stop(env, tgt->lut_bottom, th);
731 CDEBUG(D_INFO, "%s: update last_rcvd server data for UUID = %s, "
732 "last_transno = %llu: rc = %d\n", tgt->lut_obd->obd_name,
733 tgt->lut_lsd.lsd_uuid, tgt->lut_lsd.lsd_last_transno, rc);
736 EXPORT_SYMBOL(tgt_server_data_update);
738 static int tgt_truncate_last_rcvd(const struct lu_env *env,
739 struct lu_target *tgt, loff_t size)
741 struct dt_object *dt = tgt->lut_last_rcvd;
748 if (tgt->lut_bottom->dd_rdonly)
752 attr.la_valid = LA_SIZE;
754 th = dt_trans_create(env, tgt->lut_bottom);
757 rc = dt_declare_punch(env, dt, size, OBD_OBJECT_EOF, th);
760 rc = dt_declare_attr_set(env, dt, &attr, th);
763 rc = dt_trans_start_local(env, tgt->lut_bottom, th);
767 rc = dt_punch(env, dt, size, OBD_OBJECT_EOF, th);
769 rc = dt_attr_set(env, dt, &attr, th);
772 dt_trans_stop(env, tgt->lut_bottom, th);
777 static void tgt_client_epoch_update(const struct lu_env *env,
778 struct obd_export *exp)
780 struct lsd_client_data *lcd = exp->exp_target_data.ted_lcd;
781 struct lu_target *tgt = class_exp2tgt(exp);
783 LASSERT(tgt && tgt->lut_bottom);
784 /** VBR: set client last_epoch to current epoch */
785 if (lcd->lcd_last_epoch >= tgt->lut_lsd.lsd_start_epoch)
787 lcd->lcd_last_epoch = tgt->lut_lsd.lsd_start_epoch;
788 tgt_client_data_update(env, exp);
792 * Update boot epoch when recovery ends
794 void tgt_boot_epoch_update(struct lu_target *tgt)
797 struct ptlrpc_request *req;
799 LIST_HEAD(client_list);
802 if (tgt->lut_obd->obd_stopping)
805 rc = lu_env_init(&env, LCT_DT_THREAD);
807 CERROR("%s: can't initialize environment: rc = %d\n",
808 tgt->lut_obd->obd_name, rc);
812 spin_lock(&tgt->lut_translock);
813 start_epoch = (tgt->lut_last_transno >> LR_EPOCH_BITS) + 1;
814 tgt->lut_last_transno = (__u64)start_epoch << LR_EPOCH_BITS;
815 tgt->lut_lsd.lsd_start_epoch = start_epoch;
816 spin_unlock(&tgt->lut_translock);
819 * The recovery is not yet finished and final queue can still be updated
820 * with resend requests. Move final list to separate one for processing
822 spin_lock(&tgt->lut_obd->obd_recovery_task_lock);
823 list_splice_init(&tgt->lut_obd->obd_final_req_queue, &client_list);
824 spin_unlock(&tgt->lut_obd->obd_recovery_task_lock);
827 * go through list of exports participated in recovery and
828 * set new epoch for them
830 list_for_each_entry(req, &client_list, rq_list) {
831 LASSERT(!req->rq_export->exp_delayed);
832 if (!req->rq_export->exp_vbr_failed)
833 tgt_client_epoch_update(&env, req->rq_export);
835 /** return list back at once */
836 spin_lock(&tgt->lut_obd->obd_recovery_task_lock);
837 list_splice_init(&client_list, &tgt->lut_obd->obd_final_req_queue);
838 spin_unlock(&tgt->lut_obd->obd_recovery_task_lock);
840 /** Clear MULTI RPCS incompatibility flag if
841 * - target is MDT and
842 * - there is no client to recover or the recovery was aborted
844 if (!strncmp(tgt->lut_obd->obd_type->typ_name, LUSTRE_MDT_NAME, 3) &&
845 (atomic_read(&tgt->lut_obd->obd_max_recoverable_clients) == 0 ||
846 tgt->lut_obd->obd_abort_recovery))
847 tgt->lut_lsd.lsd_feature_incompat &= ~OBD_INCOMPAT_MULTI_RPCS;
849 /** update server epoch */
850 tgt_server_data_update(&env, tgt, 1);
855 * commit callback, need to update last_committed value
857 struct tgt_last_committed_callback {
858 struct dt_txn_commit_cb llcc_cb;
859 struct lu_target *llcc_tgt;
860 struct obd_export *llcc_exp;
864 static void tgt_cb_last_committed(struct lu_env *env, struct thandle *th,
865 struct dt_txn_commit_cb *cb, int err)
867 struct tgt_last_committed_callback *ccb;
869 ccb = container_of(cb, struct tgt_last_committed_callback, llcc_cb);
871 LASSERT(ccb->llcc_exp);
872 LASSERT(ccb->llcc_tgt != NULL);
873 LASSERT(ccb->llcc_exp->exp_obd == ccb->llcc_tgt->lut_obd);
875 if (th->th_reserved_quota.qrr_count > 0) {
876 struct lu_env temp_env;
879 CDEBUG(D_QUOTA, "free quota %llu %llu\n",
880 th->th_reserved_quota.qrr_id.qid_gid,
881 th->th_reserved_quota.qrr_count);
883 rc = lu_env_init(&temp_env, LCT_DT_THREAD);
885 CERROR("%s: can't initialize environment: rc = %d\n",
886 ccb->llcc_tgt->lut_obd->obd_name, rc);
890 dt_reserve_or_free_quota(&temp_env, th->th_dev,
891 th->th_reserved_quota.qrr_type,
892 th->th_reserved_quota.qrr_id.qid_uid,
893 th->th_reserved_quota.qrr_id.qid_gid,
894 -th->th_reserved_quota.qrr_count,
896 lu_env_fini(&temp_env);
899 /* error hit, don't update last committed to provide chance to
900 * replay data after fail */
904 /* Fast path w/o spinlock, if exp_last_committed was updated
905 * with higher transno, no need to take spinlock and check,
906 * also no need to update obd_last_committed. */
907 if (ccb->llcc_transno <= ccb->llcc_exp->exp_last_committed)
909 spin_lock(&ccb->llcc_tgt->lut_translock);
910 if (ccb->llcc_transno > ccb->llcc_tgt->lut_obd->obd_last_committed)
911 ccb->llcc_tgt->lut_obd->obd_last_committed = ccb->llcc_transno;
913 if (ccb->llcc_transno > ccb->llcc_exp->exp_last_committed) {
914 ccb->llcc_exp->exp_last_committed = ccb->llcc_transno;
915 spin_unlock(&ccb->llcc_tgt->lut_translock);
917 ptlrpc_commit_replies(ccb->llcc_exp);
918 tgt_cancel_slc_locks(ccb->llcc_tgt, ccb->llcc_transno);
920 spin_unlock(&ccb->llcc_tgt->lut_translock);
923 CDEBUG(D_HA, "%s: transno %lld is committed\n",
924 ccb->llcc_tgt->lut_obd->obd_name, ccb->llcc_transno);
927 class_export_cb_put(ccb->llcc_exp);
932 * Add commit callback function, it returns a non-zero value to inform
933 * caller to use sync transaction if necessary.
935 static int tgt_last_commit_cb_add(struct thandle *th, struct lu_target *tgt,
936 struct obd_export *exp, __u64 transno)
938 struct tgt_last_committed_callback *ccb;
939 struct dt_txn_commit_cb *dcb;
947 ccb->llcc_exp = class_export_cb_get(exp);
948 ccb->llcc_transno = transno;
951 dcb->dcb_func = tgt_cb_last_committed;
952 INIT_LIST_HEAD(&dcb->dcb_linkage);
953 strlcpy(dcb->dcb_name, "tgt_cb_last_committed", sizeof(dcb->dcb_name));
955 rc = dt_trans_cb_add(th, dcb);
957 class_export_cb_put(exp);
961 if (exp_connect_flags(exp) & OBD_CONNECT_LIGHTWEIGHT)
962 /* report failure to force synchronous operation */
965 /* if exp_need_sync is set, return non-zero value to force
966 * a sync transaction. */
967 return rc ? rc : exp->exp_need_sync;
970 static int tgt_is_local_client(const struct lu_env *env,
971 struct obd_export *exp)
973 struct lu_target *tgt = class_exp2tgt(exp);
974 struct tgt_session_info *tsi = tgt_ses_info(env);
975 struct ptlrpc_request *req = tgt_ses_req(tsi);
977 if (exp_connect_flags(exp) & OBD_CONNECT_MDS ||
978 exp_connect_flags(exp) & OBD_CONNECT_MDS_MDS)
980 if (tgt->lut_local_recovery)
984 if (!LNetIsPeerLocal(req->rq_peer.nid))
991 * Add new client to the last_rcvd upon new connection.
993 * We use a bitmap to locate a free space in the last_rcvd file and initialize
996 int tgt_client_new(const struct lu_env *env, struct obd_export *exp)
998 struct tg_export_data *ted = &exp->exp_target_data;
999 struct lu_target *tgt = class_exp2tgt(exp);
1004 LASSERT(tgt && tgt->lut_client_bitmap != NULL);
1005 if (!strcmp(ted->ted_lcd->lcd_uuid, tgt->lut_obd->obd_uuid.uuid))
1008 if (exp_connect_flags(exp) & OBD_CONNECT_LIGHTWEIGHT)
1011 if (tgt_is_local_client(env, exp)) {
1012 LCONSOLE_WARN("%s: local client %s w/o recovery\n",
1013 exp->exp_obd->obd_name, ted->ted_lcd->lcd_uuid);
1014 exp->exp_no_recovery = 1;
1018 /* the bitmap operations can handle cl_idx > sizeof(long) * 8, so
1019 * there's no need for extra complication here
1021 idx = find_first_zero_bit(tgt->lut_client_bitmap, LR_MAX_CLIENTS);
1023 if (idx >= LR_MAX_CLIENTS ||
1024 OBD_FAIL_CHECK(OBD_FAIL_MDS_CLIENT_ADD)) {
1025 CERROR("%s: no room for %u clients - fix LR_MAX_CLIENTS\n",
1026 tgt->lut_obd->obd_name, idx);
1029 if (test_and_set_bit(idx, tgt->lut_client_bitmap)) {
1030 idx = find_next_zero_bit(tgt->lut_client_bitmap,
1031 LR_MAX_CLIENTS, idx);
1035 ted->ted_lr_idx = idx;
1036 ted->ted_lr_off = tgt->lut_lsd.lsd_client_start +
1037 idx * tgt->lut_lsd.lsd_client_size;
1039 LASSERTF(ted->ted_lr_off > 0, "ted_lr_off = %llu\n", ted->ted_lr_off);
1041 if (tgt_is_multimodrpcs_client(exp)) {
1042 /* Set MULTI RPCS incompatibility flag to prevent previous
1043 * Lustre versions to mount a target with reply_data file */
1044 atomic_inc(&tgt->lut_num_clients);
1045 if (!(tgt->lut_lsd.lsd_feature_incompat &
1046 OBD_INCOMPAT_MULTI_RPCS)) {
1047 tgt->lut_lsd.lsd_feature_incompat |=
1048 OBD_INCOMPAT_MULTI_RPCS;
1049 rc = tgt_server_data_update(env, tgt, 1);
1051 CERROR("%s: unable to set MULTI RPCS "
1052 "incompatibility flag\n",
1053 exp->exp_obd->obd_name);
1058 /* assign client slot generation */
1059 ted->ted_lcd->lcd_generation =
1060 atomic_inc_return(&tgt->lut_client_generation);
1062 ted->ted_lcd->lcd_generation = 0;
1065 CDEBUG(D_INFO, "%s: new client at index %d (%llu) with UUID '%s' "
1067 tgt->lut_obd->obd_name, ted->ted_lr_idx, ted->ted_lr_off,
1068 ted->ted_lcd->lcd_uuid, ted->ted_lcd->lcd_generation);
1070 if (OBD_FAIL_CHECK(OBD_FAIL_TGT_CLIENT_ADD))
1073 rc = tgt_client_data_update(env, exp);
1075 CERROR("%s: Failed to write client lcd at idx %d, rc %d\n",
1076 tgt->lut_obd->obd_name, idx, rc);
1080 EXPORT_SYMBOL(tgt_client_new);
1082 /* Add an existing client to the MDS in-memory state based on
1083 * a client that was previously found in the last_rcvd file and
1084 * already has an assigned slot (idx >= 0).
1086 * It should not be possible to fail adding an existing client - otherwise
1087 * mdt_init_server_data() callsite needs to be fixed.
1089 int tgt_client_add(const struct lu_env *env, struct obd_export *exp, int idx)
1091 struct tg_export_data *ted = &exp->exp_target_data;
1092 struct lu_target *tgt = class_exp2tgt(exp);
1096 LASSERT(tgt && tgt->lut_client_bitmap != NULL);
1097 LASSERTF(idx >= 0, "%d\n", idx);
1099 if (!strcmp(ted->ted_lcd->lcd_uuid, tgt->lut_obd->obd_uuid.uuid) ||
1100 exp_connect_flags(exp) & OBD_CONNECT_LIGHTWEIGHT)
1103 if (test_and_set_bit(idx, tgt->lut_client_bitmap)) {
1104 CERROR("%s: client %d: bit already set in bitmap!!\n",
1105 tgt->lut_obd->obd_name, idx);
1108 atomic_inc(&tgt->lut_num_clients);
1110 CDEBUG(D_INFO, "%s: client at idx %d with UUID '%s' added, "
1112 tgt->lut_obd->obd_name, idx, ted->ted_lcd->lcd_uuid,
1113 ted->ted_lcd->lcd_generation);
1115 ted->ted_lr_idx = idx;
1116 ted->ted_lr_off = tgt->lut_lsd.lsd_client_start +
1117 idx * tgt->lut_lsd.lsd_client_size;
1119 mutex_init(&ted->ted_lcd_lock);
1121 LASSERTF(ted->ted_lr_off > 0, "ted_lr_off = %llu\n", ted->ted_lr_off);
1126 int tgt_client_del(const struct lu_env *env, struct obd_export *exp)
1128 struct tg_export_data *ted = &exp->exp_target_data;
1129 struct lu_target *tgt = class_exp2tgt(exp);
1134 LASSERT(ted->ted_lcd);
1136 if (unlikely(tgt == NULL)) {
1137 CDEBUG(D_ERROR, "%s: No target for connected export\n",
1138 class_exp2obd(exp)->obd_name);
1142 /* XXX if lcd_uuid were a real obd_uuid, I could use obd_uuid_equals */
1143 if (!strcmp((char *)ted->ted_lcd->lcd_uuid,
1144 (char *)tgt->lut_obd->obd_uuid.uuid) ||
1145 exp_connect_flags(exp) & OBD_CONNECT_LIGHTWEIGHT ||
1146 exp->exp_no_recovery)
1149 /* Slot may be not yet assigned, use case is race between Client
1150 * reconnect and forced eviction */
1151 if (ted->ted_lr_idx < 0) {
1152 CWARN("%s: client with UUID '%s' not in bitmap\n",
1153 tgt->lut_obd->obd_name, ted->ted_lcd->lcd_uuid);
1157 CDEBUG(D_INFO, "%s: del client at idx %u, off %lld, UUID '%s'\n",
1158 tgt->lut_obd->obd_name, ted->ted_lr_idx, ted->ted_lr_off,
1159 ted->ted_lcd->lcd_uuid);
1161 /* Clear the bit _after_ zeroing out the client so we don't
1162 race with filter_client_add and zero out new clients.*/
1163 if (!test_bit(ted->ted_lr_idx, tgt->lut_client_bitmap)) {
1164 CERROR("%s: client %u: bit already clear in bitmap!!\n",
1165 tgt->lut_obd->obd_name, ted->ted_lr_idx);
1169 /* Do not erase record for recoverable client. */
1170 if (exp->exp_flags & OBD_OPT_FAILOVER)
1173 if (OBD_FAIL_CHECK(OBD_FAIL_TGT_CLIENT_DEL))
1176 /* Make sure the server's last_transno is up to date.
1177 * This should be done before zeroing client slot so last_transno will
1178 * be in server data or in client data in case of failure */
1179 rc = tgt_server_data_update(env, tgt, 0);
1181 CERROR("%s: failed to update server data, skip client %s "
1182 "zeroing, rc %d\n", tgt->lut_obd->obd_name,
1183 ted->ted_lcd->lcd_uuid, rc);
1187 memset(ted->ted_lcd->lcd_uuid, 0, sizeof ted->ted_lcd->lcd_uuid);
1188 rc = tgt_client_data_update(env, exp);
1190 CDEBUG(rc == 0 ? D_INFO : D_ERROR,
1191 "%s: zeroing out client %s at idx %u (%llu), rc %d\n",
1192 tgt->lut_obd->obd_name, ted->ted_lcd->lcd_uuid,
1193 ted->ted_lr_idx, ted->ted_lr_off, rc);
1196 EXPORT_SYMBOL(tgt_client_del);
1198 static void tgt_clean_by_tag(struct obd_export *exp, __u64 xid, __u16 tag)
1200 struct tg_export_data *ted = &exp->exp_target_data;
1201 struct lu_target *lut = class_exp2tgt(exp);
1202 struct tg_reply_data *trd, *tmp;
1207 list_for_each_entry_safe(trd, tmp, &ted->ted_reply_list, trd_list) {
1208 if (trd->trd_tag != tag)
1211 LASSERT(ergo(tgt_is_increasing_xid_client(exp),
1212 trd->trd_reply.lrd_xid <= xid));
1214 ted->ted_release_tag++;
1215 tgt_release_reply_data(lut, ted, trd);
1219 static int tgt_add_reply_data(const struct lu_env *env, struct lu_target *tgt,
1220 struct tg_export_data *ted, struct tg_reply_data *trd,
1221 struct ptlrpc_request *req,
1222 struct thandle *th, bool update_lrd_file)
1224 struct lsd_reply_data *lrd;
1228 lrd = &trd->trd_reply;
1229 /* update export last transno */
1230 mutex_lock(&ted->ted_lcd_lock);
1231 if (lrd->lrd_transno > ted->ted_lcd->lcd_last_transno)
1232 ted->ted_lcd->lcd_last_transno = lrd->lrd_transno;
1233 mutex_unlock(&ted->ted_lcd_lock);
1236 /* find a empty slot */
1237 i = tgt_find_free_reply_slot(tgt);
1238 if (unlikely(i < 0)) {
1239 CERROR("%s: couldn't find a slot for reply data: "
1240 "rc = %d\n", tgt_name(tgt), i);
1245 if (update_lrd_file) {
1248 /* write reply data to disk */
1249 off = sizeof(struct lsd_reply_header) + sizeof(*lrd) * i;
1250 rc = tgt_reply_data_write(env, tgt, lrd, off, th);
1251 if (unlikely(rc != 0)) {
1252 CERROR("%s: can't update %s file: rc = %d\n",
1253 tgt_name(tgt), REPLY_DATA, rc);
1254 GOTO(free_slot, rc);
1258 trd->trd_index = TRD_INDEX_MEMORY;
1261 /* add reply data to target export's reply list */
1262 mutex_lock(&ted->ted_lcd_lock);
1264 int exclude = tgt_is_increasing_xid_client(req->rq_export) ?
1265 MSG_REPLAY : MSG_REPLAY|MSG_RESENT;
1267 if (req->rq_obsolete) {
1269 "drop reply data update for obsolete req xid=%llu,"
1270 "transno=%llu, tag=%hu\n", req->rq_xid,
1271 lrd->lrd_transno, trd->trd_tag);
1272 mutex_unlock(&ted->ted_lcd_lock);
1273 GOTO(free_slot, rc = -EBADR);
1276 if (!(lustre_msg_get_flags(req->rq_reqmsg) & exclude))
1277 tgt_clean_by_tag(req->rq_export, req->rq_xid,
1280 list_add(&trd->trd_list, &ted->ted_reply_list);
1281 ted->ted_reply_cnt++;
1282 if (ted->ted_reply_cnt > ted->ted_reply_max)
1283 ted->ted_reply_max = ted->ted_reply_cnt;
1284 mutex_unlock(&ted->ted_lcd_lock);
1286 CDEBUG(D_TRACE, "add reply %p: xid %llu, transno %llu, "
1287 "tag %hu, client gen %u, slot idx %d\n",
1288 trd, lrd->lrd_xid, lrd->lrd_transno,
1289 trd->trd_tag, lrd->lrd_client_gen, trd->trd_index);
1295 tgt_clear_reply_slot(tgt, trd->trd_index);
1299 int tgt_mk_reply_data(const struct lu_env *env,
1300 struct lu_target *tgt,
1301 struct tg_export_data *ted,
1302 struct ptlrpc_request *req,
1308 struct tg_reply_data *trd;
1309 struct lsd_reply_data *lrd;
1310 __u64 *pre_versions = NULL;
1314 if (unlikely(trd == NULL))
1317 /* fill reply data information */
1318 lrd = &trd->trd_reply;
1319 lrd->lrd_transno = transno;
1321 lrd->lrd_xid = req->rq_xid;
1322 trd->trd_tag = lustre_msg_get_tag(req->rq_reqmsg);
1323 lrd->lrd_client_gen = ted->ted_lcd->lcd_generation;
1325 pre_versions = lustre_msg_get_versions(req->rq_repmsg);
1326 lrd->lrd_result = th->th_result;
1329 struct tgt_session_info *tsi;
1331 LASSERT(env != NULL);
1332 tsi = tgt_ses_info(env);
1333 LASSERT(tsi->tsi_xid != 0);
1335 lrd->lrd_xid = tsi->tsi_xid;
1336 lrd->lrd_result = tsi->tsi_result;
1337 lrd->lrd_client_gen = tsi->tsi_client_gen;
1340 lrd->lrd_data = opdata;
1342 trd->trd_pre_versions[0] = pre_versions[0];
1343 trd->trd_pre_versions[1] = pre_versions[1];
1344 trd->trd_pre_versions[2] = pre_versions[2];
1345 trd->trd_pre_versions[3] = pre_versions[3];
1348 rc = tgt_add_reply_data(env, tgt, ted, trd, req,
1358 EXPORT_SYMBOL(tgt_mk_reply_data);
1361 * last_rcvd & last_committed update callbacks
1363 static int tgt_last_rcvd_update(const struct lu_env *env, struct lu_target *tgt,
1364 struct dt_object *obj, __u64 opdata,
1365 struct thandle *th, struct ptlrpc_request *req)
1367 struct tgt_thread_info *tti = tgt_th_info(env);
1368 struct tgt_session_info *tsi = tgt_ses_info(env);
1369 struct obd_export *exp = tsi->tsi_exp;
1370 struct tg_export_data *ted;
1378 LASSERT(exp != NULL);
1379 ted = &exp->exp_target_data;
1381 /* Some clients don't support recovery, and they don't have last_rcvd
1383 * 1. lightweight clients.
1384 * 2. local clients on MDS which doesn't enable "localrecov".
1385 * 3. OFD connect may cause transaction before export has last_rcvd
1388 if (ted->ted_lr_idx < 0)
1392 tti->tti_transno = lustre_msg_get_transno(req->rq_reqmsg);
1394 /* From update replay, tti_transno should be set already */
1395 LASSERT(tti->tti_transno != 0);
1397 spin_lock(&tgt->lut_translock);
1398 if (th->th_result != 0) {
1399 if (tti->tti_transno != 0) {
1400 CERROR("%s: replay transno %llu failed: rc = %d\n",
1401 tgt_name(tgt), tti->tti_transno, th->th_result);
1403 } else if (tti->tti_transno == 0) {
1404 tti->tti_transno = ++tgt->lut_last_transno;
1406 /* should be replay */
1407 if (tti->tti_transno > tgt->lut_last_transno)
1408 tgt->lut_last_transno = tti->tti_transno;
1410 spin_unlock(&tgt->lut_translock);
1412 /** VBR: set new versions */
1413 if (th->th_result == 0 && obj != NULL) {
1414 struct dt_object *dto = dt_object_locate(obj, th->th_dev);
1415 dt_version_set(env, dto, tti->tti_transno, th);
1418 /* filling reply data */
1419 CDEBUG(D_INODE, "transno = %llu, last_committed = %llu\n",
1420 tti->tti_transno, tgt->lut_obd->obd_last_committed);
1423 req->rq_transno = tti->tti_transno;
1424 lustre_msg_set_transno(req->rq_repmsg, tti->tti_transno);
1427 /* if can't add callback, do sync write */
1428 th->th_sync |= !!tgt_last_commit_cb_add(th, tgt, exp, tti->tti_transno);
1431 /* store transno in the last_rcvd header */
1432 spin_lock(&tgt->lut_translock);
1433 if (tti->tti_transno > tgt->lut_lsd.lsd_last_transno) {
1434 tgt->lut_lsd.lsd_last_transno = tti->tti_transno;
1435 spin_unlock(&tgt->lut_translock);
1436 /* Although current connection doesn't have slot
1437 * in the last_rcvd, we still want to maintain
1438 * the in-memory lsd_client_data structure in order to
1439 * properly handle reply reconstruction. */
1440 rc = tgt_server_data_write(env, tgt, th);
1442 spin_unlock(&tgt->lut_translock);
1444 } else if (ted->ted_lr_off == 0) {
1445 CERROR("%s: client idx %d has offset %lld\n",
1446 tgt_name(tgt), ted->ted_lr_idx, ted->ted_lr_off);
1450 /* Target that supports multiple reply data */
1451 if (tgt_is_multimodrpcs_client(exp)) {
1452 return tgt_mk_reply_data(env, tgt, ted, req, opdata, th,
1453 !!(req != NULL), tti->tti_transno);
1456 /* Enough for update replay, let's return */
1460 mutex_lock(&ted->ted_lcd_lock);
1461 LASSERT(ergo(tti->tti_transno == 0, th->th_result != 0));
1462 if (lustre_msg_get_opc(req->rq_reqmsg) == MDS_CLOSE) {
1463 transno_p = &ted->ted_lcd->lcd_last_close_transno;
1464 ted->ted_lcd->lcd_last_close_xid = req->rq_xid;
1465 ted->ted_lcd->lcd_last_close_result = th->th_result;
1467 /* VBR: save versions in last_rcvd for reconstruct. */
1468 __u64 *pre_versions = lustre_msg_get_versions(req->rq_repmsg);
1471 ted->ted_lcd->lcd_pre_versions[0] = pre_versions[0];
1472 ted->ted_lcd->lcd_pre_versions[1] = pre_versions[1];
1473 ted->ted_lcd->lcd_pre_versions[2] = pre_versions[2];
1474 ted->ted_lcd->lcd_pre_versions[3] = pre_versions[3];
1476 transno_p = &ted->ted_lcd->lcd_last_transno;
1477 ted->ted_lcd->lcd_last_xid = req->rq_xid;
1478 ted->ted_lcd->lcd_last_result = th->th_result;
1479 /* XXX: lcd_last_data is __u32 but intent_dispostion is __u64,
1480 * see struct ldlm_reply->lock_policy_res1; */
1481 ted->ted_lcd->lcd_last_data = opdata;
1484 /* Update transno in slot only if non-zero number, i.e. no errors */
1485 if (likely(tti->tti_transno != 0)) {
1486 /* Don't overwrite bigger transaction number with lower one.
1487 * That is not sign of problem in all cases, but in any case
1488 * this value should be monotonically increased only. */
1489 if (*transno_p > tti->tti_transno) {
1490 if (!tgt->lut_no_reconstruct) {
1491 CERROR("%s: trying to overwrite bigger transno:"
1492 "on-disk: %llu, new: %llu replay: "
1493 "%d. See LU-617.\n", tgt_name(tgt),
1494 *transno_p, tti->tti_transno,
1495 req_is_replay(req));
1496 if (req_is_replay(req)) {
1497 spin_lock(&req->rq_export->exp_lock);
1498 req->rq_export->exp_vbr_failed = 1;
1499 spin_unlock(&req->rq_export->exp_lock);
1501 mutex_unlock(&ted->ted_lcd_lock);
1502 RETURN(req_is_replay(req) ? -EOVERFLOW : 0);
1505 *transno_p = tti->tti_transno;
1510 tti->tti_off = ted->ted_lr_off;
1511 if (CFS_FAIL_CHECK(OBD_FAIL_TGT_RCVD_EIO))
1514 rc = tgt_client_data_write(env, tgt, ted->ted_lcd,
1517 mutex_unlock(&ted->ted_lcd_lock);
1521 mutex_unlock(&ted->ted_lcd_lock);
1526 * last_rcvd update for echo client simulation.
1527 * It updates last_rcvd client slot and version of object in
1528 * simple way but with all locks to simulate all drawbacks
1530 static int tgt_last_rcvd_update_echo(const struct lu_env *env,
1531 struct lu_target *tgt,
1532 struct dt_object *obj,
1534 struct obd_export *exp)
1536 struct tgt_thread_info *tti = tgt_th_info(env);
1537 struct tg_export_data *ted = &exp->exp_target_data;
1542 tti->tti_transno = 0;
1544 spin_lock(&tgt->lut_translock);
1545 if (th->th_result == 0)
1546 tti->tti_transno = ++tgt->lut_last_transno;
1547 spin_unlock(&tgt->lut_translock);
1549 /** VBR: set new versions */
1550 if (th->th_result == 0 && obj != NULL)
1551 dt_version_set(env, obj, tti->tti_transno, th);
1553 /* if can't add callback, do sync write */
1554 th->th_sync |= !!tgt_last_commit_cb_add(th, tgt, exp,
1557 LASSERT(ted->ted_lr_off > 0);
1559 mutex_lock(&ted->ted_lcd_lock);
1560 LASSERT(ergo(tti->tti_transno == 0, th->th_result != 0));
1561 ted->ted_lcd->lcd_last_transno = tti->tti_transno;
1562 ted->ted_lcd->lcd_last_result = th->th_result;
1564 tti->tti_off = ted->ted_lr_off;
1565 rc = tgt_client_data_write(env, tgt, ted->ted_lcd, &tti->tti_off, th);
1566 mutex_unlock(&ted->ted_lcd_lock);
1570 static int tgt_clients_data_init(const struct lu_env *env,
1571 struct lu_target *tgt,
1572 unsigned long last_size)
1574 struct obd_device *obd = tgt->lut_obd;
1575 struct lr_server_data *lsd = &tgt->lut_lsd;
1576 struct lsd_client_data *lcd = NULL;
1577 struct tg_export_data *ted;
1580 loff_t off = lsd->lsd_client_start;
1581 __u32 generation = 0;
1582 struct cfs_hash *hash = NULL;
1586 if (tgt->lut_bottom->dd_rdonly)
1589 BUILD_BUG_ON(offsetof(struct lsd_client_data, lcd_padding) +
1590 sizeof(lcd->lcd_padding) != LR_CLIENT_SIZE);
1596 hash = cfs_hash_getref(tgt->lut_obd->obd_gen_hash);
1598 GOTO(err_out, rc = -ENODEV);
1600 for (cl_idx = 0; off < last_size; cl_idx++) {
1601 struct obd_export *exp;
1604 /* Don't assume off is incremented properly by
1605 * read_record(), in case sizeof(*lcd)
1606 * isn't the same as fsd->lsd_client_size. */
1607 off = lsd->lsd_client_start + cl_idx * lsd->lsd_client_size;
1608 rc = tgt_client_data_read(env, tgt, lcd, &off, cl_idx);
1610 CERROR("%s: error reading last_rcvd %s idx %d off "
1611 "%llu: rc = %d\n", tgt_name(tgt), LAST_RCVD,
1614 break; /* read error shouldn't cause startup to fail */
1617 if (lcd->lcd_uuid[0] == '\0') {
1618 CDEBUG(D_INFO, "skipping zeroed client at offset %d\n",
1623 last_transno = lcd_last_transno(lcd);
1625 /* These exports are cleaned up by disconnect, so they
1626 * need to be set up like real exports as connect does.
1628 CDEBUG(D_HA, "RCVRNG CLIENT uuid: %s idx: %d lr: %llu"
1629 " srv lr: %llu lx: %llu gen %u\n", lcd->lcd_uuid,
1630 cl_idx, last_transno, lsd->lsd_last_transno,
1631 lcd_last_xid(lcd), lcd->lcd_generation);
1633 exp = class_new_export(obd, (struct obd_uuid *)lcd->lcd_uuid);
1635 if (PTR_ERR(exp) == -EALREADY) {
1636 /* export already exists, zero out this one */
1637 CERROR("%s: Duplicate export %s!\n",
1638 tgt_name(tgt), lcd->lcd_uuid);
1641 GOTO(err_out, rc = PTR_ERR(exp));
1644 ted = &exp->exp_target_data;
1645 *ted->ted_lcd = *lcd;
1647 rc = tgt_client_add(env, exp, cl_idx);
1648 LASSERTF(rc == 0, "rc = %d\n", rc); /* can't fail existing */
1649 /* VBR: set export last committed version */
1650 exp->exp_last_committed = last_transno;
1651 spin_lock(&exp->exp_lock);
1652 exp->exp_connecting = 0;
1653 exp->exp_in_recovery = 0;
1654 spin_unlock(&exp->exp_lock);
1655 atomic_inc(&obd->obd_max_recoverable_clients);
1657 if (tgt->lut_lsd.lsd_feature_incompat &
1658 OBD_INCOMPAT_MULTI_RPCS &&
1659 lcd->lcd_generation != 0) {
1660 /* compute the highest valid client generation */
1661 generation = max(generation, lcd->lcd_generation);
1662 /* fill client_generation <-> export hash table */
1663 rc = cfs_hash_add_unique(hash, &lcd->lcd_generation,
1664 &exp->exp_gen_hash);
1666 CERROR("%s: duplicate export for client "
1668 tgt_name(tgt), lcd->lcd_generation);
1669 class_export_put(exp);
1674 class_export_put(exp);
1676 rc = rev_import_init(exp);
1678 class_unlink_export(exp);
1682 /* Need to check last_rcvd even for duplicated exports. */
1683 CDEBUG(D_OTHER, "client at idx %d has last_transno = %llu\n",
1684 cl_idx, last_transno);
1686 spin_lock(&tgt->lut_translock);
1687 tgt->lut_last_transno = max(last_transno,
1688 tgt->lut_last_transno);
1689 spin_unlock(&tgt->lut_translock);
1692 /* record highest valid client generation */
1693 atomic_set(&tgt->lut_client_generation, generation);
1697 cfs_hash_putref(hash);
1702 struct server_compat_data {
1709 static struct server_compat_data tgt_scd[] = {
1710 [LDD_F_SV_TYPE_MDT] = {
1711 .rocompat = OBD_ROCOMPAT_LOVOBJID,
1712 .incompat = OBD_INCOMPAT_MDT | OBD_INCOMPAT_COMMON_LR |
1713 OBD_INCOMPAT_FID | OBD_INCOMPAT_IAM_DIR |
1714 OBD_INCOMPAT_LMM_VER | OBD_INCOMPAT_MULTI_OI |
1715 OBD_INCOMPAT_MULTI_RPCS,
1716 .rocinit = OBD_ROCOMPAT_LOVOBJID,
1717 .incinit = OBD_INCOMPAT_MDT | OBD_INCOMPAT_COMMON_LR |
1718 OBD_INCOMPAT_MULTI_OI,
1720 [LDD_F_SV_TYPE_OST] = {
1721 .rocompat = OBD_ROCOMPAT_IDX_IN_IDIF,
1722 .incompat = OBD_INCOMPAT_OST | OBD_INCOMPAT_COMMON_LR |
1724 .rocinit = OBD_ROCOMPAT_IDX_IN_IDIF,
1725 .incinit = OBD_INCOMPAT_OST | OBD_INCOMPAT_COMMON_LR,
1729 int tgt_server_data_init(const struct lu_env *env, struct lu_target *tgt)
1731 struct tgt_thread_info *tti = tgt_th_info(env);
1732 struct lr_server_data *lsd = &tgt->lut_lsd;
1733 unsigned long last_rcvd_size;
1737 rc = dt_attr_get(env, tgt->lut_last_rcvd, &tti->tti_attr);
1741 last_rcvd_size = (unsigned long)tti->tti_attr.la_size;
1743 /* ensure padding in the struct is the correct size */
1744 BUILD_BUG_ON(offsetof(struct lr_server_data, lsd_padding) +
1745 sizeof(lsd->lsd_padding) != LR_SERVER_SIZE);
1747 rc = server_name2index(tgt_name(tgt), &index, NULL);
1749 CERROR("%s: Can not get index from name: rc = %d\n",
1753 /* server_name2index() returns type */
1755 if (type != LDD_F_SV_TYPE_MDT && type != LDD_F_SV_TYPE_OST) {
1756 CERROR("%s: unknown target type %x\n", tgt_name(tgt), type);
1760 /* last_rcvd on OST doesn't provide reconstruct support because there
1761 * may be up to 8 in-flight write requests per single slot in
1762 * last_rcvd client data
1764 tgt->lut_no_reconstruct = (type == LDD_F_SV_TYPE_OST);
1766 if (last_rcvd_size == 0) {
1767 LCONSOLE_WARN("%s: new disk, initializing\n", tgt_name(tgt));
1769 memcpy(lsd->lsd_uuid, tgt->lut_obd->obd_uuid.uuid,
1770 sizeof(lsd->lsd_uuid));
1771 lsd->lsd_last_transno = 0;
1772 lsd->lsd_mount_count = 0;
1773 lsd->lsd_server_size = LR_SERVER_SIZE;
1774 lsd->lsd_client_start = LR_CLIENT_START;
1775 lsd->lsd_client_size = LR_CLIENT_SIZE;
1776 lsd->lsd_subdir_count = OBJ_SUBDIR_COUNT;
1777 lsd->lsd_osd_index = index;
1778 lsd->lsd_feature_rocompat = tgt_scd[type].rocinit;
1779 lsd->lsd_feature_incompat = tgt_scd[type].incinit;
1781 rc = tgt_server_data_read(env, tgt);
1783 CERROR("%s: error reading LAST_RCVD: rc= %d\n",
1787 if (strcmp(lsd->lsd_uuid, tgt->lut_obd->obd_uuid.uuid)) {
1788 if (tgt->lut_bottom->dd_rdonly) {
1789 /* Such difference may be caused by mounting
1790 * up snapshot with new fsname under rd_only
1791 * mode. But even if it was NOT, it will not
1792 * damage the system because of "rd_only". */
1793 memcpy(lsd->lsd_uuid,
1794 tgt->lut_obd->obd_uuid.uuid,
1795 sizeof(lsd->lsd_uuid));
1797 LCONSOLE_ERROR_MSG(0x157, "Trying to start "
1798 "OBD %s using the wrong "
1799 "disk %s. Were the /dev/ "
1800 "assignments rearranged?\n",
1801 tgt->lut_obd->obd_uuid.uuid,
1807 if (lsd->lsd_osd_index != index) {
1808 LCONSOLE_ERROR_MSG(0x157,
1809 "%s: index %d in last rcvd is different with the index %d in config log, It might be disk corruption!\n",
1811 lsd->lsd_osd_index, index);
1816 if (lsd->lsd_feature_incompat & ~tgt_scd[type].incompat) {
1817 CERROR("%s: unsupported incompat filesystem feature(s) %x\n",
1819 lsd->lsd_feature_incompat & ~tgt_scd[type].incompat);
1823 if (type == LDD_F_SV_TYPE_MDT)
1824 lsd->lsd_feature_incompat |= OBD_INCOMPAT_FID;
1826 if (lsd->lsd_feature_rocompat & ~tgt_scd[type].rocompat) {
1827 CERROR("%s: unsupported read-only filesystem feature(s) %x\n",
1829 lsd->lsd_feature_rocompat & ~tgt_scd[type].rocompat);
1832 /** Interop: evict all clients at first boot with 1.8 last_rcvd */
1833 if (type == LDD_F_SV_TYPE_MDT &&
1834 !(lsd->lsd_feature_compat & OBD_COMPAT_20)) {
1835 if (last_rcvd_size > lsd->lsd_client_start) {
1836 LCONSOLE_WARN("%s: mounting at first time on 1.8 FS, "
1837 "remove all clients for interop needs\n",
1839 rc = tgt_truncate_last_rcvd(env, tgt,
1840 lsd->lsd_client_start);
1843 last_rcvd_size = lsd->lsd_client_start;
1845 /** set 2.0 flag to upgrade/downgrade between 1.8 and 2.0 */
1846 lsd->lsd_feature_compat |= OBD_COMPAT_20;
1849 spin_lock(&tgt->lut_translock);
1850 tgt->lut_last_transno = lsd->lsd_last_transno;
1851 spin_unlock(&tgt->lut_translock);
1853 lsd->lsd_mount_count++;
1855 CDEBUG(D_INODE, "=======,=BEGIN DUMPING LAST_RCVD========\n");
1856 CDEBUG(D_INODE, "%s: server last_transno: %llu\n",
1857 tgt_name(tgt), tgt->lut_last_transno);
1858 CDEBUG(D_INODE, "%s: server mount_count: %llu\n",
1859 tgt_name(tgt), lsd->lsd_mount_count);
1860 CDEBUG(D_INODE, "%s: server data size: %u\n",
1861 tgt_name(tgt), lsd->lsd_server_size);
1862 CDEBUG(D_INODE, "%s: per-client data start: %u\n",
1863 tgt_name(tgt), lsd->lsd_client_start);
1864 CDEBUG(D_INODE, "%s: per-client data size: %u\n",
1865 tgt_name(tgt), lsd->lsd_client_size);
1866 CDEBUG(D_INODE, "%s: last_rcvd size: %lu\n",
1867 tgt_name(tgt), last_rcvd_size);
1868 CDEBUG(D_INODE, "%s: server subdir_count: %u\n",
1869 tgt_name(tgt), lsd->lsd_subdir_count);
1870 CDEBUG(D_INODE, "%s: last_rcvd clients: %lu\n", tgt_name(tgt),
1871 last_rcvd_size <= lsd->lsd_client_start ? 0 :
1872 (last_rcvd_size - lsd->lsd_client_start) /
1873 lsd->lsd_client_size);
1874 CDEBUG(D_INODE, "========END DUMPING LAST_RCVD========\n");
1876 if (lsd->lsd_server_size == 0 || lsd->lsd_client_start == 0 ||
1877 lsd->lsd_client_size == 0) {
1878 CERROR("%s: bad last_rcvd contents!\n", tgt_name(tgt));
1882 if (!tgt->lut_obd->obd_replayable)
1883 CWARN("%s: recovery support OFF\n", tgt_name(tgt));
1885 rc = tgt_clients_data_init(env, tgt, last_rcvd_size);
1887 GOTO(err_client, rc);
1889 spin_lock(&tgt->lut_translock);
1890 /* obd_last_committed is used for compatibility
1891 * with other lustre recovery code */
1892 tgt->lut_obd->obd_last_committed = tgt->lut_last_transno;
1893 spin_unlock(&tgt->lut_translock);
1895 tgt->lut_obd->u.obt.obt_mount_count = lsd->lsd_mount_count;
1896 tgt->lut_obd->u.obt.obt_instance = (__u32)lsd->lsd_mount_count;
1898 /* save it, so mount count and last_transno is current */
1899 rc = tgt_server_data_update(env, tgt, 0);
1901 GOTO(err_client, rc);
1906 class_disconnect_exports(tgt->lut_obd);
1910 /* add credits for last_rcvd update */
1911 int tgt_txn_start_cb(const struct lu_env *env, struct thandle *th,
1914 struct lu_target *tgt = cookie;
1915 struct tgt_session_info *tsi;
1916 struct tgt_thread_info *tti = tgt_th_info(env);
1917 struct dt_object *dto;
1920 /* For readonly case, the caller should have got failure
1921 * when start the transaction. If the logic comes here,
1922 * there must be something wrong. */
1923 if (unlikely(tgt->lut_bottom->dd_rdonly)) {
1928 /* if there is no session, then this transaction is not result of
1929 * request processing but some local operation */
1930 if (env->le_ses == NULL)
1933 LASSERT(tgt->lut_last_rcvd);
1934 tsi = tgt_ses_info(env);
1935 /* OFD may start transaction without export assigned */
1936 if (tsi->tsi_exp == NULL)
1939 if (tgt_is_multimodrpcs_client(tsi->tsi_exp)) {
1941 * Use maximum possible file offset for declaration to ensure
1942 * ZFS will reserve enough credits for a write anywhere in this
1943 * file, since we don't know where in the file the write will be
1944 * because a replay slot has not been assigned. This should be
1945 * replaced by dmu_tx_hold_append() when available.
1947 tti->tti_buf.lb_buf = NULL;
1948 tti->tti_buf.lb_len = sizeof(struct lsd_reply_data);
1949 dto = dt_object_locate(tgt->lut_reply_data, th->th_dev);
1950 rc = dt_declare_record_write(env, dto, &tti->tti_buf, -1, th);
1954 dto = dt_object_locate(tgt->lut_last_rcvd, th->th_dev);
1956 tti->tti_off = tsi->tsi_exp->exp_target_data.ted_lr_off;
1957 rc = dt_declare_record_write(env, dto, &tti->tti_buf,
1963 if (tsi->tsi_vbr_obj != NULL &&
1964 !lu_object_remote(&tsi->tsi_vbr_obj->do_lu)) {
1965 dto = dt_object_locate(tsi->tsi_vbr_obj, th->th_dev);
1966 rc = dt_declare_version_set(env, dto, th);
1972 /* Update last_rcvd records with latests transaction data */
1973 int tgt_txn_stop_cb(const struct lu_env *env, struct thandle *th,
1976 struct lu_target *tgt = cookie;
1977 struct tgt_session_info *tsi;
1978 struct tgt_thread_info *tti = tgt_th_info(env);
1979 struct dt_object *obj = NULL;
1983 if (env->le_ses == NULL)
1986 tsi = tgt_ses_info(env);
1987 /* OFD may start transaction without export assigned */
1988 if (tsi->tsi_exp == NULL)
1991 echo_client = (tgt_ses_req(tsi) == NULL && tsi->tsi_xid == 0);
1993 if (tti->tti_has_trans && !echo_client) {
1994 if (tti->tti_mult_trans == 0) {
1995 CDEBUG(D_HA, "More than one transaction %llu\n",
1999 /* we need another transno to be assigned */
2000 tti->tti_transno = 0;
2001 } else if (th->th_result == 0) {
2002 tti->tti_has_trans = 1;
2005 if (tsi->tsi_vbr_obj != NULL &&
2006 !lu_object_remote(&tsi->tsi_vbr_obj->do_lu)) {
2007 obj = tsi->tsi_vbr_obj;
2010 if (unlikely(echo_client)) /* echo client special case */
2011 rc = tgt_last_rcvd_update_echo(env, tgt, obj, th,
2014 rc = tgt_last_rcvd_update(env, tgt, obj, tsi->tsi_opdata, th,
2019 int tgt_reply_data_init(const struct lu_env *env, struct lu_target *tgt)
2021 struct tgt_thread_info *tti = tgt_th_info(env);
2022 struct lsd_reply_data *lrd = &tti->tti_lrd;
2023 unsigned long reply_data_size;
2025 struct lsd_reply_header *lrh = NULL;
2026 struct tg_reply_data *trd = NULL;
2029 struct cfs_hash *hash = NULL;
2030 struct obd_export *exp;
2031 struct tg_export_data *ted;
2032 int reply_data_recovered = 0;
2034 rc = dt_attr_get(env, tgt->lut_reply_data, &tti->tti_attr);
2037 reply_data_size = (unsigned long)tti->tti_attr.la_size;
2041 GOTO(out, rc = -ENOMEM);
2043 if (reply_data_size == 0) {
2044 CDEBUG(D_INFO, "%s: new reply_data file, initializing\n",
2046 lrh->lrh_magic = LRH_MAGIC;
2047 lrh->lrh_header_size = sizeof(struct lsd_reply_header);
2048 lrh->lrh_reply_size = sizeof(struct lsd_reply_data);
2049 rc = tgt_reply_header_write(env, tgt, lrh);
2051 CERROR("%s: error writing %s: rc = %d\n",
2052 tgt_name(tgt), REPLY_DATA, rc);
2056 rc = tgt_reply_header_read(env, tgt, lrh);
2058 CERROR("%s: error reading %s: rc = %d\n",
2059 tgt_name(tgt), REPLY_DATA, rc);
2062 if (lrh->lrh_magic != LRH_MAGIC ||
2063 lrh->lrh_header_size != sizeof(struct lsd_reply_header) ||
2064 lrh->lrh_reply_size != sizeof(struct lsd_reply_data)) {
2065 CERROR("%s: invalid header in %s\n",
2066 tgt_name(tgt), REPLY_DATA);
2067 GOTO(out, rc = -EINVAL);
2070 hash = cfs_hash_getref(tgt->lut_obd->obd_gen_hash);
2072 GOTO(out, rc = -ENODEV);
2076 GOTO(out, rc = -ENOMEM);
2078 /* Load reply_data from disk */
2079 for (idx = 0, off = sizeof(struct lsd_reply_header);
2080 off < reply_data_size;
2081 idx++, off += sizeof(struct lsd_reply_data)) {
2082 rc = tgt_reply_data_read(env, tgt, lrd, off);
2084 CERROR("%s: error reading %s: rc = %d\n",
2085 tgt_name(tgt), REPLY_DATA, rc);
2089 exp = cfs_hash_lookup(hash, &lrd->lrd_client_gen);
2091 /* old reply data from a disconnected client */
2094 ted = &exp->exp_target_data;
2095 mutex_lock(&ted->ted_lcd_lock);
2097 /* create in-memory reply_data and link it to
2098 * target export's reply list */
2099 rc = tgt_set_reply_slot(tgt, idx);
2101 mutex_unlock(&ted->ted_lcd_lock);
2104 trd->trd_reply = *lrd;
2105 trd->trd_pre_versions[0] = 0;
2106 trd->trd_pre_versions[1] = 0;
2107 trd->trd_pre_versions[2] = 0;
2108 trd->trd_pre_versions[3] = 0;
2109 trd->trd_index = idx;
2111 list_add(&trd->trd_list, &ted->ted_reply_list);
2112 ted->ted_reply_cnt++;
2113 if (ted->ted_reply_cnt > ted->ted_reply_max)
2114 ted->ted_reply_max = ted->ted_reply_cnt;
2116 CDEBUG(D_HA, "%s: restore reply %p: xid %llu, "
2117 "transno %llu, client gen %u, slot idx %d\n",
2118 tgt_name(tgt), trd, lrd->lrd_xid,
2119 lrd->lrd_transno, lrd->lrd_client_gen,
2122 /* update export last committed transation */
2123 exp->exp_last_committed = max(exp->exp_last_committed,
2125 /* Update lcd_last_transno as well for check in
2126 * tgt_release_reply_data() or the latest client
2127 * transno can be lost.
2129 ted->ted_lcd->lcd_last_transno =
2130 max(ted->ted_lcd->lcd_last_transno,
2131 exp->exp_last_committed);
2133 mutex_unlock(&ted->ted_lcd_lock);
2134 class_export_put(exp);
2136 /* update target last committed transaction */
2137 spin_lock(&tgt->lut_translock);
2138 tgt->lut_last_transno = max(tgt->lut_last_transno,
2140 spin_unlock(&tgt->lut_translock);
2142 reply_data_recovered++;
2146 GOTO(out, rc = -ENOMEM);
2148 CDEBUG(D_INFO, "%s: %d reply data have been recovered\n",
2149 tgt_name(tgt), reply_data_recovered);
2152 spin_lock(&tgt->lut_translock);
2153 /* obd_last_committed is used for compatibility
2154 * with other lustre recovery code */
2155 tgt->lut_obd->obd_last_committed = tgt->lut_last_transno;
2156 spin_unlock(&tgt->lut_translock);
2162 cfs_hash_putref(hash);
2170 static int tgt_check_lookup_req(struct ptlrpc_request *req, int lookup,
2171 struct tg_reply_data *trd)
2173 struct tg_export_data *ted = &req->rq_export->exp_target_data;
2174 struct lu_target *lut = class_exp2tgt(req->rq_export);
2175 __u16 tag = lustre_msg_get_tag(req->rq_reqmsg);
2177 struct tg_reply_data *reply;
2178 bool check_increasing;
2183 check_increasing = tgt_is_increasing_xid_client(req->rq_export) &&
2184 !(lustre_msg_get_flags(req->rq_reqmsg) & MSG_REPLAY);
2185 if (!lookup && !check_increasing)
2188 list_for_each_entry(reply, &ted->ted_reply_list, trd_list) {
2189 if (lookup && reply->trd_reply.lrd_xid == req->rq_xid) {
2194 } else if (check_increasing && reply->trd_tag == tag &&
2195 reply->trd_reply.lrd_xid > req->rq_xid) {
2197 CERROR("%s: busy tag=%u req_xid=%llu, trd=%p: xid=%llu transno=%llu client_gen=%u slot_idx=%d: rc = %d\n",
2198 tgt_name(lut), tag, req->rq_xid, trd,
2199 reply->trd_reply.lrd_xid,
2200 reply->trd_reply.lrd_transno,
2201 reply->trd_reply.lrd_client_gen,
2202 reply->trd_index, rc);
2210 /* Look for a reply data matching specified request @req
2211 * A copy is returned in @trd if the pointer is not NULL
2213 int tgt_lookup_reply(struct ptlrpc_request *req, struct tg_reply_data *trd)
2215 struct tg_export_data *ted = &req->rq_export->exp_target_data;
2217 bool not_replay = !(lustre_msg_get_flags(req->rq_reqmsg) & MSG_REPLAY);
2219 mutex_lock(&ted->ted_lcd_lock);
2220 if (not_replay && req->rq_xid <= req->rq_export->exp_last_xid) {
2221 /* A check for the last_xid is needed here in case there is
2222 * no reply data is left in the list. It may happen if another
2223 * RPC on another slot increased the last_xid between our
2224 * process_req_last_xid & tgt_lookup_reply calls */
2227 found = tgt_check_lookup_req(req, 1, trd);
2229 mutex_unlock(&ted->ted_lcd_lock);
2231 CDEBUG(D_TRACE, "%s: lookup reply xid %llu, found %d last_xid %llu\n",
2232 tgt_name(class_exp2tgt(req->rq_export)), req->rq_xid, found,
2233 req->rq_export->exp_last_xid);
2237 EXPORT_SYMBOL(tgt_lookup_reply);
2239 int tgt_handle_received_xid(struct obd_export *exp, __u64 rcvd_xid)
2241 struct tg_export_data *ted = &exp->exp_target_data;
2242 struct lu_target *lut = class_exp2tgt(exp);
2243 struct tg_reply_data *trd, *tmp;
2246 list_for_each_entry_safe(trd, tmp, &ted->ted_reply_list, trd_list) {
2247 if (trd->trd_reply.lrd_xid > rcvd_xid)
2249 ted->ted_release_xid++;
2250 tgt_release_reply_data(lut, ted, trd);
2256 int tgt_handle_tag(struct ptlrpc_request *req)
2258 return tgt_check_lookup_req(req, 0, NULL);