staging/lustre/llite: fix build erorr if CONFIG_FS_POSIX_ACL is off
[deliverable/linux.git] / drivers / staging / lustre / lustre / mdc / mdc_request.c
CommitLineData
d7e09d03
PT
1/*
2 * GPL HEADER START
3 *
4 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
5 *
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.
9 *
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).
15 *
16 * You should have received a copy of the GNU General Public License
17 * version 2 along with this program; If not, see
18 * http://www.sun.com/software/products/lustre/docs/GPLv2.pdf
19 *
20 * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara,
21 * CA 95054 USA or visit www.sun.com if you need additional information or
22 * have any questions.
23 *
24 * GPL HEADER END
25 */
26/*
27 * Copyright (c) 2001, 2010, Oracle and/or its affiliates. All rights reserved.
28 * Use is subject to license terms.
29 *
30 * Copyright (c) 2011, 2012, Intel Corporation.
31 */
32/*
33 * This file is part of Lustre, http://www.lustre.org/
34 * Lustre is a trademark of Sun Microsystems, Inc.
35 */
36
37#define DEBUG_SUBSYSTEM S_MDC
38
39# include <linux/module.h>
40# include <linux/pagemap.h>
41# include <linux/miscdevice.h>
42# include <linux/init.h>
43# include <linux/utsname.h>
44
45#include <lustre_acl.h>
46#include <obd_class.h>
47#include <lustre_fid.h>
48#include <lprocfs_status.h>
49#include <lustre_param.h>
50#include <lustre_log.h>
51
52#include "mdc_internal.h"
53
54#define REQUEST_MINOR 244
55
56struct mdc_renew_capa_args {
57 struct obd_capa *ra_oc;
58 renew_capa_cb_t ra_cb;
59};
60
61static int mdc_cleanup(struct obd_device *obd);
62
63int mdc_unpack_capa(struct obd_export *exp, struct ptlrpc_request *req,
64 const struct req_msg_field *field, struct obd_capa **oc)
65{
66 struct lustre_capa *capa;
67 struct obd_capa *c;
68 ENTRY;
69
70 /* swabbed already in mdc_enqueue */
71 capa = req_capsule_server_get(&req->rq_pill, field);
72 if (capa == NULL)
73 RETURN(-EPROTO);
74
75 c = alloc_capa(CAPA_SITE_CLIENT);
76 if (IS_ERR(c)) {
77 CDEBUG(D_INFO, "alloc capa failed!\n");
78 RETURN(PTR_ERR(c));
79 } else {
80 c->c_capa = *capa;
81 *oc = c;
82 RETURN(0);
83 }
84}
85
86static inline int mdc_queue_wait(struct ptlrpc_request *req)
87{
88 struct client_obd *cli = &req->rq_import->imp_obd->u.cli;
89 int rc;
90
91 /* mdc_enter_request() ensures that this client has no more
92 * than cl_max_rpcs_in_flight RPCs simultaneously inf light
93 * against an MDT. */
94 rc = mdc_enter_request(cli);
95 if (rc != 0)
96 return rc;
97
98 rc = ptlrpc_queue_wait(req);
99 mdc_exit_request(cli);
100
101 return rc;
102}
103
104/* Helper that implements most of mdc_getstatus and signal_completed_replay. */
105/* XXX this should become mdc_get_info("key"), sending MDS_GET_INFO RPC */
106static int send_getstatus(struct obd_import *imp, struct lu_fid *rootfid,
107 struct obd_capa **pc, int level, int msg_flags)
108{
109 struct ptlrpc_request *req;
110 struct mdt_body *body;
111 int rc;
112 ENTRY;
113
114 req = ptlrpc_request_alloc_pack(imp, &RQF_MDS_GETSTATUS,
115 LUSTRE_MDS_VERSION, MDS_GETSTATUS);
116 if (req == NULL)
117 RETURN(-ENOMEM);
118
119 mdc_pack_body(req, NULL, NULL, 0, 0, -1, 0);
120 lustre_msg_add_flags(req->rq_reqmsg, msg_flags);
121 req->rq_send_state = level;
122
123 ptlrpc_request_set_replen(req);
124
125 rc = ptlrpc_queue_wait(req);
126 if (rc)
127 GOTO(out, rc);
128
129 body = req_capsule_server_get(&req->rq_pill, &RMF_MDT_BODY);
130 if (body == NULL)
131 GOTO(out, rc = -EPROTO);
132
133 if (body->valid & OBD_MD_FLMDSCAPA) {
134 rc = mdc_unpack_capa(NULL, req, &RMF_CAPA1, pc);
135 if (rc)
136 GOTO(out, rc);
137 }
138
139 *rootfid = body->fid1;
140 CDEBUG(D_NET,
141 "root fid="DFID", last_committed="LPU64"\n",
142 PFID(rootfid),
143 lustre_msg_get_last_committed(req->rq_repmsg));
144 EXIT;
145out:
146 ptlrpc_req_finished(req);
147 return rc;
148}
149
150/* This should be mdc_get_info("rootfid") */
151int mdc_getstatus(struct obd_export *exp, struct lu_fid *rootfid,
152 struct obd_capa **pc)
153{
154 return send_getstatus(class_exp2cliimp(exp), rootfid, pc,
155 LUSTRE_IMP_FULL, 0);
156}
157
158/*
159 * This function now is known to always saying that it will receive 4 buffers
160 * from server. Even for cases when acl_size and md_size is zero, RPC header
161 * will contain 4 fields and RPC itself will contain zero size fields. This is
162 * because mdt_getattr*() _always_ returns 4 fields, but if acl is not needed
163 * and thus zero, it shrinks it, making zero size. The same story about
164 * md_size. And this is course of problem when client waits for smaller number
165 * of fields. This issue will be fixed later when client gets aware of RPC
166 * layouts. --umka
167 */
168static int mdc_getattr_common(struct obd_export *exp,
169 struct ptlrpc_request *req)
170{
171 struct req_capsule *pill = &req->rq_pill;
172 struct mdt_body *body;
173 void *eadata;
174 int rc;
175 ENTRY;
176
177 /* Request message already built. */
178 rc = ptlrpc_queue_wait(req);
179 if (rc != 0)
180 RETURN(rc);
181
182 /* sanity check for the reply */
183 body = req_capsule_server_get(pill, &RMF_MDT_BODY);
184 if (body == NULL)
185 RETURN(-EPROTO);
186
187 CDEBUG(D_NET, "mode: %o\n", body->mode);
188
189 if (body->eadatasize != 0) {
190 mdc_update_max_ea_from_body(exp, body);
191
192 eadata = req_capsule_server_sized_get(pill, &RMF_MDT_MD,
193 body->eadatasize);
194 if (eadata == NULL)
195 RETURN(-EPROTO);
196 }
197
198 if (body->valid & OBD_MD_FLRMTPERM) {
199 struct mdt_remote_perm *perm;
200
201 LASSERT(client_is_remote(exp));
202 perm = req_capsule_server_swab_get(pill, &RMF_ACL,
203 lustre_swab_mdt_remote_perm);
204 if (perm == NULL)
205 RETURN(-EPROTO);
206 }
207
208 if (body->valid & OBD_MD_FLMDSCAPA) {
209 struct lustre_capa *capa;
210 capa = req_capsule_server_get(pill, &RMF_CAPA1);
211 if (capa == NULL)
212 RETURN(-EPROTO);
213 }
214
215 RETURN(0);
216}
217
218int mdc_getattr(struct obd_export *exp, struct md_op_data *op_data,
219 struct ptlrpc_request **request)
220{
221 struct ptlrpc_request *req;
222 int rc;
223 ENTRY;
224
225 /* Single MDS without an LMV case */
226 if (op_data->op_flags & MF_GET_MDT_IDX) {
227 op_data->op_mds = 0;
228 RETURN(0);
229 }
230 *request = NULL;
231 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_MDS_GETATTR);
232 if (req == NULL)
233 RETURN(-ENOMEM);
234
235 mdc_set_capa_size(req, &RMF_CAPA1, op_data->op_capa1);
236
237 rc = ptlrpc_request_pack(req, LUSTRE_MDS_VERSION, MDS_GETATTR);
238 if (rc) {
239 ptlrpc_request_free(req);
240 RETURN(rc);
241 }
242
243 mdc_pack_body(req, &op_data->op_fid1, op_data->op_capa1,
244 op_data->op_valid, op_data->op_mode, -1, 0);
245
246 req_capsule_set_size(&req->rq_pill, &RMF_MDT_MD, RCL_SERVER,
247 op_data->op_mode);
248 if (op_data->op_valid & OBD_MD_FLRMTPERM) {
249 LASSERT(client_is_remote(exp));
250 req_capsule_set_size(&req->rq_pill, &RMF_ACL, RCL_SERVER,
251 sizeof(struct mdt_remote_perm));
252 }
253 ptlrpc_request_set_replen(req);
254
255 rc = mdc_getattr_common(exp, req);
256 if (rc)
257 ptlrpc_req_finished(req);
258 else
259 *request = req;
260 RETURN(rc);
261}
262
263int mdc_getattr_name(struct obd_export *exp, struct md_op_data *op_data,
264 struct ptlrpc_request **request)
265{
266 struct ptlrpc_request *req;
267 int rc;
268 ENTRY;
269
270 *request = NULL;
271 req = ptlrpc_request_alloc(class_exp2cliimp(exp),
272 &RQF_MDS_GETATTR_NAME);
273 if (req == NULL)
274 RETURN(-ENOMEM);
275
276 mdc_set_capa_size(req, &RMF_CAPA1, op_data->op_capa1);
277 req_capsule_set_size(&req->rq_pill, &RMF_NAME, RCL_CLIENT,
278 op_data->op_namelen + 1);
279
280 rc = ptlrpc_request_pack(req, LUSTRE_MDS_VERSION, MDS_GETATTR_NAME);
281 if (rc) {
282 ptlrpc_request_free(req);
283 RETURN(rc);
284 }
285
286 mdc_pack_body(req, &op_data->op_fid1, op_data->op_capa1,
287 op_data->op_valid, op_data->op_mode,
288 op_data->op_suppgids[0], 0);
289
290 if (op_data->op_name) {
291 char *name = req_capsule_client_get(&req->rq_pill, &RMF_NAME);
292 LASSERT(strnlen(op_data->op_name, op_data->op_namelen) ==
293 op_data->op_namelen);
294 memcpy(name, op_data->op_name, op_data->op_namelen);
295 }
296
297 req_capsule_set_size(&req->rq_pill, &RMF_MDT_MD, RCL_SERVER,
298 op_data->op_mode);
299 ptlrpc_request_set_replen(req);
300
301 rc = mdc_getattr_common(exp, req);
302 if (rc)
303 ptlrpc_req_finished(req);
304 else
305 *request = req;
306 RETURN(rc);
307}
308
309static int mdc_is_subdir(struct obd_export *exp,
310 const struct lu_fid *pfid,
311 const struct lu_fid *cfid,
312 struct ptlrpc_request **request)
313{
314 struct ptlrpc_request *req;
315 int rc;
316
317 ENTRY;
318
319 *request = NULL;
320 req = ptlrpc_request_alloc_pack(class_exp2cliimp(exp),
321 &RQF_MDS_IS_SUBDIR, LUSTRE_MDS_VERSION,
322 MDS_IS_SUBDIR);
323 if (req == NULL)
324 RETURN(-ENOMEM);
325
326 mdc_is_subdir_pack(req, pfid, cfid, 0);
327 ptlrpc_request_set_replen(req);
328
329 rc = ptlrpc_queue_wait(req);
330 if (rc && rc != -EREMOTE)
331 ptlrpc_req_finished(req);
332 else
333 *request = req;
334 RETURN(rc);
335}
336
337static int mdc_xattr_common(struct obd_export *exp,const struct req_format *fmt,
338 const struct lu_fid *fid,
339 struct obd_capa *oc, int opcode, obd_valid valid,
340 const char *xattr_name, const char *input,
341 int input_size, int output_size, int flags,
342 __u32 suppgid, struct ptlrpc_request **request)
343{
344 struct ptlrpc_request *req;
345 int xattr_namelen = 0;
346 char *tmp;
347 int rc;
348 ENTRY;
349
350 *request = NULL;
351 req = ptlrpc_request_alloc(class_exp2cliimp(exp), fmt);
352 if (req == NULL)
353 RETURN(-ENOMEM);
354
355 mdc_set_capa_size(req, &RMF_CAPA1, oc);
356 if (xattr_name) {
357 xattr_namelen = strlen(xattr_name) + 1;
358 req_capsule_set_size(&req->rq_pill, &RMF_NAME, RCL_CLIENT,
359 xattr_namelen);
360 }
361 if (input_size) {
362 LASSERT(input);
363 req_capsule_set_size(&req->rq_pill, &RMF_EADATA, RCL_CLIENT,
364 input_size);
365 }
366
367 rc = ptlrpc_request_pack(req, LUSTRE_MDS_VERSION, opcode);
368 if (rc) {
369 ptlrpc_request_free(req);
370 RETURN(rc);
371 }
372
373 if (opcode == MDS_REINT) {
374 struct mdt_rec_setxattr *rec;
375
376 CLASSERT(sizeof(struct mdt_rec_setxattr) ==
377 sizeof(struct mdt_rec_reint));
378 rec = req_capsule_client_get(&req->rq_pill, &RMF_REC_REINT);
379 rec->sx_opcode = REINT_SETXATTR;
380 /* TODO:
381 * cfs_curproc_fs{u,g}id() should replace
382 * current->fs{u,g}id for portability.
383 */
384 rec->sx_fsuid = current_fsuid();
385 rec->sx_fsgid = current_fsgid();
386 rec->sx_cap = cfs_curproc_cap_pack();
387 rec->sx_suppgid1 = suppgid;
388 rec->sx_suppgid2 = -1;
389 rec->sx_fid = *fid;
390 rec->sx_valid = valid | OBD_MD_FLCTIME;
391 rec->sx_time = cfs_time_current_sec();
392 rec->sx_size = output_size;
393 rec->sx_flags = flags;
394
395 mdc_pack_capa(req, &RMF_CAPA1, oc);
396 } else {
397 mdc_pack_body(req, fid, oc, valid, output_size, suppgid, flags);
398 }
399
400 if (xattr_name) {
401 tmp = req_capsule_client_get(&req->rq_pill, &RMF_NAME);
402 memcpy(tmp, xattr_name, xattr_namelen);
403 }
404 if (input_size) {
405 tmp = req_capsule_client_get(&req->rq_pill, &RMF_EADATA);
406 memcpy(tmp, input, input_size);
407 }
408
409 if (req_capsule_has_field(&req->rq_pill, &RMF_EADATA, RCL_SERVER))
410 req_capsule_set_size(&req->rq_pill, &RMF_EADATA,
411 RCL_SERVER, output_size);
412 ptlrpc_request_set_replen(req);
413
414 /* make rpc */
415 if (opcode == MDS_REINT)
416 mdc_get_rpc_lock(exp->exp_obd->u.cli.cl_rpc_lock, NULL);
417
418 rc = ptlrpc_queue_wait(req);
419
420 if (opcode == MDS_REINT)
421 mdc_put_rpc_lock(exp->exp_obd->u.cli.cl_rpc_lock, NULL);
422
423 if (rc)
424 ptlrpc_req_finished(req);
425 else
426 *request = req;
427 RETURN(rc);
428}
429
430int mdc_setxattr(struct obd_export *exp, const struct lu_fid *fid,
431 struct obd_capa *oc, obd_valid valid, const char *xattr_name,
432 const char *input, int input_size, int output_size,
433 int flags, __u32 suppgid, struct ptlrpc_request **request)
434{
435 return mdc_xattr_common(exp, &RQF_MDS_REINT_SETXATTR,
436 fid, oc, MDS_REINT, valid, xattr_name,
437 input, input_size, output_size, flags,
438 suppgid, request);
439}
440
441int mdc_getxattr(struct obd_export *exp, const struct lu_fid *fid,
442 struct obd_capa *oc, obd_valid valid, const char *xattr_name,
443 const char *input, int input_size, int output_size,
444 int flags, struct ptlrpc_request **request)
445{
446 return mdc_xattr_common(exp, &RQF_MDS_GETXATTR,
447 fid, oc, MDS_GETXATTR, valid, xattr_name,
448 input, input_size, output_size, flags,
449 -1, request);
450}
451
452#ifdef CONFIG_FS_POSIX_ACL
453static int mdc_unpack_acl(struct ptlrpc_request *req, struct lustre_md *md)
454{
455 struct req_capsule *pill = &req->rq_pill;
456 struct mdt_body *body = md->body;
457 struct posix_acl *acl;
458 void *buf;
459 int rc;
460 ENTRY;
461
462 if (!body->aclsize)
463 RETURN(0);
464
465 buf = req_capsule_server_sized_get(pill, &RMF_ACL, body->aclsize);
466
467 if (!buf)
468 RETURN(-EPROTO);
469
470 acl = posix_acl_from_xattr(&init_user_ns, buf, body->aclsize);
471 if (IS_ERR(acl)) {
472 rc = PTR_ERR(acl);
473 CERROR("convert xattr to acl: %d\n", rc);
474 RETURN(rc);
475 }
476
477 rc = posix_acl_valid(acl);
478 if (rc) {
479 CERROR("validate acl: %d\n", rc);
480 posix_acl_release(acl);
481 RETURN(rc);
482 }
483
484 md->posix_acl = acl;
485 RETURN(0);
486}
487#else
488#define mdc_unpack_acl(req, md) 0
489#endif
490
491int mdc_get_lustre_md(struct obd_export *exp, struct ptlrpc_request *req,
492 struct obd_export *dt_exp, struct obd_export *md_exp,
493 struct lustre_md *md)
494{
495 struct req_capsule *pill = &req->rq_pill;
496 int rc;
497 ENTRY;
498
499 LASSERT(md);
500 memset(md, 0, sizeof(*md));
501
502 md->body = req_capsule_server_get(pill, &RMF_MDT_BODY);
503 LASSERT(md->body != NULL);
504
505 if (md->body->valid & OBD_MD_FLEASIZE) {
506 int lmmsize;
507 struct lov_mds_md *lmm;
508
509 if (!S_ISREG(md->body->mode)) {
510 CDEBUG(D_INFO, "OBD_MD_FLEASIZE set, should be a "
511 "regular file, but is not\n");
512 GOTO(out, rc = -EPROTO);
513 }
514
515 if (md->body->eadatasize == 0) {
516 CDEBUG(D_INFO, "OBD_MD_FLEASIZE set, "
517 "but eadatasize 0\n");
518 GOTO(out, rc = -EPROTO);
519 }
520 lmmsize = md->body->eadatasize;
521 lmm = req_capsule_server_sized_get(pill, &RMF_MDT_MD, lmmsize);
522 if (!lmm)
523 GOTO(out, rc = -EPROTO);
524
525 rc = obd_unpackmd(dt_exp, &md->lsm, lmm, lmmsize);
526 if (rc < 0)
527 GOTO(out, rc);
528
529 if (rc < sizeof(*md->lsm)) {
530 CDEBUG(D_INFO, "lsm size too small: "
531 "rc < sizeof (*md->lsm) (%d < %d)\n",
532 rc, (int)sizeof(*md->lsm));
533 GOTO(out, rc = -EPROTO);
534 }
535
536 } else if (md->body->valid & OBD_MD_FLDIREA) {
537 int lmvsize;
538 struct lov_mds_md *lmv;
539
540 if(!S_ISDIR(md->body->mode)) {
541 CDEBUG(D_INFO, "OBD_MD_FLDIREA set, should be a "
542 "directory, but is not\n");
543 GOTO(out, rc = -EPROTO);
544 }
545
546 if (md->body->eadatasize == 0) {
547 CDEBUG(D_INFO, "OBD_MD_FLDIREA is set, "
548 "but eadatasize 0\n");
549 RETURN(-EPROTO);
550 }
551 if (md->body->valid & OBD_MD_MEA) {
552 lmvsize = md->body->eadatasize;
553 lmv = req_capsule_server_sized_get(pill, &RMF_MDT_MD,
554 lmvsize);
555 if (!lmv)
556 GOTO(out, rc = -EPROTO);
557
558 rc = obd_unpackmd(md_exp, (void *)&md->mea, lmv,
559 lmvsize);
560 if (rc < 0)
561 GOTO(out, rc);
562
563 if (rc < sizeof(*md->mea)) {
564 CDEBUG(D_INFO, "size too small: "
565 "rc < sizeof(*md->mea) (%d < %d)\n",
566 rc, (int)sizeof(*md->mea));
567 GOTO(out, rc = -EPROTO);
568 }
569 }
570 }
571 rc = 0;
572
573 if (md->body->valid & OBD_MD_FLRMTPERM) {
574 /* remote permission */
575 LASSERT(client_is_remote(exp));
576 md->remote_perm = req_capsule_server_swab_get(pill, &RMF_ACL,
577 lustre_swab_mdt_remote_perm);
578 if (!md->remote_perm)
579 GOTO(out, rc = -EPROTO);
580 }
581 else if (md->body->valid & OBD_MD_FLACL) {
582 /* for ACL, it's possible that FLACL is set but aclsize is zero.
583 * only when aclsize != 0 there's an actual segment for ACL
584 * in reply buffer.
585 */
586 if (md->body->aclsize) {
587 rc = mdc_unpack_acl(req, md);
588 if (rc)
589 GOTO(out, rc);
590#ifdef CONFIG_FS_POSIX_ACL
591 } else {
592 md->posix_acl = NULL;
593#endif
594 }
595 }
596 if (md->body->valid & OBD_MD_FLMDSCAPA) {
597 struct obd_capa *oc = NULL;
598
599 rc = mdc_unpack_capa(NULL, req, &RMF_CAPA1, &oc);
600 if (rc)
601 GOTO(out, rc);
602 md->mds_capa = oc;
603 }
604
605 if (md->body->valid & OBD_MD_FLOSSCAPA) {
606 struct obd_capa *oc = NULL;
607
608 rc = mdc_unpack_capa(NULL, req, &RMF_CAPA2, &oc);
609 if (rc)
610 GOTO(out, rc);
611 md->oss_capa = oc;
612 }
613
614 EXIT;
615out:
616 if (rc) {
617 if (md->oss_capa) {
618 capa_put(md->oss_capa);
619 md->oss_capa = NULL;
620 }
621 if (md->mds_capa) {
622 capa_put(md->mds_capa);
623 md->mds_capa = NULL;
624 }
625#ifdef CONFIG_FS_POSIX_ACL
626 posix_acl_release(md->posix_acl);
627#endif
628 if (md->lsm)
629 obd_free_memmd(dt_exp, &md->lsm);
630 }
631 return rc;
632}
633
634int mdc_free_lustre_md(struct obd_export *exp, struct lustre_md *md)
635{
636 ENTRY;
637 RETURN(0);
638}
639
640/**
641 * Handles both OPEN and SETATTR RPCs for OPEN-CLOSE and SETATTR-DONE_WRITING
642 * RPC chains.
643 */
644void mdc_replay_open(struct ptlrpc_request *req)
645{
646 struct md_open_data *mod = req->rq_cb_data;
647 struct ptlrpc_request *close_req;
648 struct obd_client_handle *och;
649 struct lustre_handle old;
650 struct mdt_body *body;
651 ENTRY;
652
653 if (mod == NULL) {
654 DEBUG_REQ(D_ERROR, req,
655 "Can't properly replay without open data.");
656 EXIT;
657 return;
658 }
659
660 body = req_capsule_server_get(&req->rq_pill, &RMF_MDT_BODY);
661 LASSERT(body != NULL);
662
663 och = mod->mod_och;
664 if (och != NULL) {
665 struct lustre_handle *file_fh;
666
667 LASSERT(och->och_magic == OBD_CLIENT_HANDLE_MAGIC);
668
669 file_fh = &och->och_fh;
670 CDEBUG(D_HA, "updating handle from "LPX64" to "LPX64"\n",
671 file_fh->cookie, body->handle.cookie);
672 old = *file_fh;
673 *file_fh = body->handle;
674 }
675 close_req = mod->mod_close_req;
676 if (close_req != NULL) {
677 __u32 opc = lustre_msg_get_opc(close_req->rq_reqmsg);
678 struct mdt_ioepoch *epoch;
679
680 LASSERT(opc == MDS_CLOSE || opc == MDS_DONE_WRITING);
681 epoch = req_capsule_client_get(&close_req->rq_pill,
682 &RMF_MDT_EPOCH);
683 LASSERT(epoch);
684
685 if (och != NULL)
686 LASSERT(!memcmp(&old, &epoch->handle, sizeof(old)));
687 DEBUG_REQ(D_HA, close_req, "updating close body with new fh");
688 epoch->handle = body->handle;
689 }
690 EXIT;
691}
692
693void mdc_commit_open(struct ptlrpc_request *req)
694{
695 struct md_open_data *mod = req->rq_cb_data;
696 if (mod == NULL)
697 return;
698
699 /**
700 * No need to touch md_open_data::mod_och, it holds a reference on
701 * \var mod and will zero references to each other, \var mod will be
702 * freed after that when md_open_data::mod_och will put the reference.
703 */
704
705 /**
706 * Do not let open request to disappear as it still may be needed
707 * for close rpc to happen (it may happen on evict only, otherwise
708 * ptlrpc_request::rq_replay does not let mdc_commit_open() to be
709 * called), just mark this rpc as committed to distinguish these 2
710 * cases, see mdc_close() for details. The open request reference will
711 * be put along with freeing \var mod.
712 */
713 ptlrpc_request_addref(req);
714 spin_lock(&req->rq_lock);
715 req->rq_committed = 1;
716 spin_unlock(&req->rq_lock);
717 req->rq_cb_data = NULL;
718 obd_mod_put(mod);
719}
720
721int mdc_set_open_replay_data(struct obd_export *exp,
722 struct obd_client_handle *och,
723 struct ptlrpc_request *open_req)
724{
725 struct md_open_data *mod;
726 struct mdt_rec_create *rec;
727 struct mdt_body *body;
728 struct obd_import *imp = open_req->rq_import;
729 ENTRY;
730
731 if (!open_req->rq_replay)
732 RETURN(0);
733
734 rec = req_capsule_client_get(&open_req->rq_pill, &RMF_REC_REINT);
735 body = req_capsule_server_get(&open_req->rq_pill, &RMF_MDT_BODY);
736 LASSERT(rec != NULL);
737 /* Incoming message in my byte order (it's been swabbed). */
738 /* Outgoing messages always in my byte order. */
739 LASSERT(body != NULL);
740
741 /* Only if the import is replayable, we set replay_open data */
742 if (och && imp->imp_replayable) {
743 mod = obd_mod_alloc();
744 if (mod == NULL) {
745 DEBUG_REQ(D_ERROR, open_req,
746 "Can't allocate md_open_data");
747 RETURN(0);
748 }
749
750 /**
751 * Take a reference on \var mod, to be freed on mdc_close().
752 * It protects \var mod from being freed on eviction (commit
753 * callback is called despite rq_replay flag).
754 * Another reference for \var och.
755 */
756 obd_mod_get(mod);
757 obd_mod_get(mod);
758
759 spin_lock(&open_req->rq_lock);
760 och->och_mod = mod;
761 mod->mod_och = och;
762 mod->mod_open_req = open_req;
763 open_req->rq_cb_data = mod;
764 open_req->rq_commit_cb = mdc_commit_open;
765 spin_unlock(&open_req->rq_lock);
766 }
767
768 rec->cr_fid2 = body->fid1;
769 rec->cr_ioepoch = body->ioepoch;
770 rec->cr_old_handle.cookie = body->handle.cookie;
771 open_req->rq_replay_cb = mdc_replay_open;
772 if (!fid_is_sane(&body->fid1)) {
773 DEBUG_REQ(D_ERROR, open_req, "Saving replay request with "
774 "insane fid");
775 LBUG();
776 }
777
778 DEBUG_REQ(D_RPCTRACE, open_req, "Set up open replay data");
779 RETURN(0);
780}
781
782int mdc_clear_open_replay_data(struct obd_export *exp,
783 struct obd_client_handle *och)
784{
785 struct md_open_data *mod = och->och_mod;
786 ENTRY;
787
788 /**
789 * It is possible to not have \var mod in a case of eviction between
790 * lookup and ll_file_open().
791 **/
792 if (mod == NULL)
793 RETURN(0);
794
795 LASSERT(mod != LP_POISON);
796
797 mod->mod_och = NULL;
798 och->och_mod = NULL;
799 obd_mod_put(mod);
800
801 RETURN(0);
802}
803
804/* Prepares the request for the replay by the given reply */
805static void mdc_close_handle_reply(struct ptlrpc_request *req,
806 struct md_op_data *op_data, int rc) {
807 struct mdt_body *repbody;
808 struct mdt_ioepoch *epoch;
809
810 if (req && rc == -EAGAIN) {
811 repbody = req_capsule_server_get(&req->rq_pill, &RMF_MDT_BODY);
812 epoch = req_capsule_client_get(&req->rq_pill, &RMF_MDT_EPOCH);
813
814 epoch->flags |= MF_SOM_AU;
815 if (repbody->valid & OBD_MD_FLGETATTRLOCK)
816 op_data->op_flags |= MF_GETATTR_LOCK;
817 }
818}
819
820int mdc_close(struct obd_export *exp, struct md_op_data *op_data,
821 struct md_open_data *mod, struct ptlrpc_request **request)
822{
823 struct obd_device *obd = class_exp2obd(exp);
824 struct ptlrpc_request *req;
825 int rc;
826 ENTRY;
827
828 *request = NULL;
829 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_MDS_CLOSE);
830 if (req == NULL)
831 RETURN(-ENOMEM);
832
833 mdc_set_capa_size(req, &RMF_CAPA1, op_data->op_capa1);
834
835 rc = ptlrpc_request_pack(req, LUSTRE_MDS_VERSION, MDS_CLOSE);
836 if (rc) {
837 ptlrpc_request_free(req);
838 RETURN(rc);
839 }
840
841 /* To avoid a livelock (bug 7034), we need to send CLOSE RPCs to a
842 * portal whose threads are not taking any DLM locks and are therefore
843 * always progressing */
844 req->rq_request_portal = MDS_READPAGE_PORTAL;
845 ptlrpc_at_set_req_timeout(req);
846
847 /* Ensure that this close's handle is fixed up during replay. */
848 if (likely(mod != NULL)) {
849 LASSERTF(mod->mod_open_req != NULL &&
850 mod->mod_open_req->rq_type != LI_POISON,
851 "POISONED open %p!\n", mod->mod_open_req);
852
853 mod->mod_close_req = req;
854
855 DEBUG_REQ(D_HA, mod->mod_open_req, "matched open");
856 /* We no longer want to preserve this open for replay even
857 * though the open was committed. b=3632, b=3633 */
858 spin_lock(&mod->mod_open_req->rq_lock);
859 mod->mod_open_req->rq_replay = 0;
860 spin_unlock(&mod->mod_open_req->rq_lock);
861 } else {
862 CDEBUG(D_HA, "couldn't find open req; expecting close error\n");
863 }
864
865 mdc_close_pack(req, op_data);
866
867 req_capsule_set_size(&req->rq_pill, &RMF_MDT_MD, RCL_SERVER,
868 obd->u.cli.cl_max_mds_easize);
869 req_capsule_set_size(&req->rq_pill, &RMF_LOGCOOKIES, RCL_SERVER,
870 obd->u.cli.cl_max_mds_cookiesize);
871
872 ptlrpc_request_set_replen(req);
873
874 mdc_get_rpc_lock(obd->u.cli.cl_close_lock, NULL);
875 rc = ptlrpc_queue_wait(req);
876 mdc_put_rpc_lock(obd->u.cli.cl_close_lock, NULL);
877
878 if (req->rq_repmsg == NULL) {
879 CDEBUG(D_RPCTRACE, "request failed to send: %p, %d\n", req,
880 req->rq_status);
881 if (rc == 0)
882 rc = req->rq_status ?: -EIO;
883 } else if (rc == 0 || rc == -EAGAIN) {
884 struct mdt_body *body;
885
886 rc = lustre_msg_get_status(req->rq_repmsg);
887 if (lustre_msg_get_type(req->rq_repmsg) == PTL_RPC_MSG_ERR) {
888 DEBUG_REQ(D_ERROR, req, "type == PTL_RPC_MSG_ERR, err "
889 "= %d", rc);
890 if (rc > 0)
891 rc = -rc;
892 }
893 body = req_capsule_server_get(&req->rq_pill, &RMF_MDT_BODY);
894 if (body == NULL)
895 rc = -EPROTO;
896 } else if (rc == -ESTALE) {
897 /**
898 * it can be allowed error after 3633 if open was committed and
899 * server failed before close was sent. Let's check if mod
900 * exists and return no error in that case
901 */
902 if (mod) {
903 DEBUG_REQ(D_HA, req, "Reset ESTALE = %d", rc);
904 LASSERT(mod->mod_open_req != NULL);
905 if (mod->mod_open_req->rq_committed)
906 rc = 0;
907 }
908 }
909
910 if (mod) {
911 if (rc != 0)
912 mod->mod_close_req = NULL;
913 /* Since now, mod is accessed through open_req only,
914 * thus close req does not keep a reference on mod anymore. */
915 obd_mod_put(mod);
916 }
917 *request = req;
918 mdc_close_handle_reply(req, op_data, rc);
919 RETURN(rc);
920}
921
922int mdc_done_writing(struct obd_export *exp, struct md_op_data *op_data,
923 struct md_open_data *mod)
924{
925 struct obd_device *obd = class_exp2obd(exp);
926 struct ptlrpc_request *req;
927 int rc;
928 ENTRY;
929
930 req = ptlrpc_request_alloc(class_exp2cliimp(exp),
931 &RQF_MDS_DONE_WRITING);
932 if (req == NULL)
933 RETURN(-ENOMEM);
934
935 mdc_set_capa_size(req, &RMF_CAPA1, op_data->op_capa1);
936 rc = ptlrpc_request_pack(req, LUSTRE_MDS_VERSION, MDS_DONE_WRITING);
937 if (rc) {
938 ptlrpc_request_free(req);
939 RETURN(rc);
940 }
941
942 if (mod != NULL) {
943 LASSERTF(mod->mod_open_req != NULL &&
944 mod->mod_open_req->rq_type != LI_POISON,
945 "POISONED setattr %p!\n", mod->mod_open_req);
946
947 mod->mod_close_req = req;
948 DEBUG_REQ(D_HA, mod->mod_open_req, "matched setattr");
949 /* We no longer want to preserve this setattr for replay even
950 * though the open was committed. b=3632, b=3633 */
951 spin_lock(&mod->mod_open_req->rq_lock);
952 mod->mod_open_req->rq_replay = 0;
953 spin_unlock(&mod->mod_open_req->rq_lock);
954 }
955
956 mdc_close_pack(req, op_data);
957 ptlrpc_request_set_replen(req);
958
959 mdc_get_rpc_lock(obd->u.cli.cl_close_lock, NULL);
960 rc = ptlrpc_queue_wait(req);
961 mdc_put_rpc_lock(obd->u.cli.cl_close_lock, NULL);
962
963 if (rc == -ESTALE) {
964 /**
965 * it can be allowed error after 3633 if open or setattr were
966 * committed and server failed before close was sent.
967 * Let's check if mod exists and return no error in that case
968 */
969 if (mod) {
970 LASSERT(mod->mod_open_req != NULL);
971 if (mod->mod_open_req->rq_committed)
972 rc = 0;
973 }
974 }
975
976 if (mod) {
977 if (rc != 0)
978 mod->mod_close_req = NULL;
979 /* Since now, mod is accessed through setattr req only,
980 * thus DW req does not keep a reference on mod anymore. */
981 obd_mod_put(mod);
982 }
983
984 mdc_close_handle_reply(req, op_data, rc);
985 ptlrpc_req_finished(req);
986 RETURN(rc);
987}
988
989
990int mdc_readpage(struct obd_export *exp, struct md_op_data *op_data,
991 struct page **pages, struct ptlrpc_request **request)
992{
993 struct ptlrpc_request *req;
994 struct ptlrpc_bulk_desc *desc;
995 int i;
996 wait_queue_head_t waitq;
997 int resends = 0;
998 struct l_wait_info lwi;
999 int rc;
1000 ENTRY;
1001
1002 *request = NULL;
1003 init_waitqueue_head(&waitq);
1004
1005restart_bulk:
1006 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_MDS_READPAGE);
1007 if (req == NULL)
1008 RETURN(-ENOMEM);
1009
1010 mdc_set_capa_size(req, &RMF_CAPA1, op_data->op_capa1);
1011
1012 rc = ptlrpc_request_pack(req, LUSTRE_MDS_VERSION, MDS_READPAGE);
1013 if (rc) {
1014 ptlrpc_request_free(req);
1015 RETURN(rc);
1016 }
1017
1018 req->rq_request_portal = MDS_READPAGE_PORTAL;
1019 ptlrpc_at_set_req_timeout(req);
1020
1021 desc = ptlrpc_prep_bulk_imp(req, op_data->op_npages, 1, BULK_PUT_SINK,
1022 MDS_BULK_PORTAL);
1023 if (desc == NULL) {
1024 ptlrpc_request_free(req);
1025 RETURN(-ENOMEM);
1026 }
1027
1028 /* NB req now owns desc and will free it when it gets freed */
1029 for (i = 0; i < op_data->op_npages; i++)
1030 ptlrpc_prep_bulk_page_pin(desc, pages[i], 0, PAGE_CACHE_SIZE);
1031
1032 mdc_readdir_pack(req, op_data->op_offset,
1033 PAGE_CACHE_SIZE * op_data->op_npages,
1034 &op_data->op_fid1, op_data->op_capa1);
1035
1036 ptlrpc_request_set_replen(req);
1037 rc = ptlrpc_queue_wait(req);
1038 if (rc) {
1039 ptlrpc_req_finished(req);
1040 if (rc != -ETIMEDOUT)
1041 RETURN(rc);
1042
1043 resends++;
1044 if (!client_should_resend(resends, &exp->exp_obd->u.cli)) {
1045 CERROR("too many resend retries, returning error\n");
1046 RETURN(-EIO);
1047 }
1048 lwi = LWI_TIMEOUT_INTR(cfs_time_seconds(resends), NULL, NULL, NULL);
1049 l_wait_event(waitq, 0, &lwi);
1050
1051 goto restart_bulk;
1052 }
1053
1054 rc = sptlrpc_cli_unwrap_bulk_read(req, req->rq_bulk,
1055 req->rq_bulk->bd_nob_transferred);
1056 if (rc < 0) {
1057 ptlrpc_req_finished(req);
1058 RETURN(rc);
1059 }
1060
1061 if (req->rq_bulk->bd_nob_transferred & ~LU_PAGE_MASK) {
1062 CERROR("Unexpected # bytes transferred: %d (%ld expected)\n",
1063 req->rq_bulk->bd_nob_transferred,
1064 PAGE_CACHE_SIZE * op_data->op_npages);
1065 ptlrpc_req_finished(req);
1066 RETURN(-EPROTO);
1067 }
1068
1069 *request = req;
1070 RETURN(0);
1071}
1072
1073static int mdc_statfs(const struct lu_env *env,
1074 struct obd_export *exp, struct obd_statfs *osfs,
1075 __u64 max_age, __u32 flags)
1076{
1077 struct obd_device *obd = class_exp2obd(exp);
1078 struct ptlrpc_request *req;
1079 struct obd_statfs *msfs;
1080 struct obd_import *imp = NULL;
1081 int rc;
1082 ENTRY;
1083
1084 /*
1085 * Since the request might also come from lprocfs, so we need
1086 * sync this with client_disconnect_export Bug15684
1087 */
1088 down_read(&obd->u.cli.cl_sem);
1089 if (obd->u.cli.cl_import)
1090 imp = class_import_get(obd->u.cli.cl_import);
1091 up_read(&obd->u.cli.cl_sem);
1092 if (!imp)
1093 RETURN(-ENODEV);
1094
1095 req = ptlrpc_request_alloc_pack(imp, &RQF_MDS_STATFS,
1096 LUSTRE_MDS_VERSION, MDS_STATFS);
1097 if (req == NULL)
1098 GOTO(output, rc = -ENOMEM);
1099
1100 ptlrpc_request_set_replen(req);
1101
1102 if (flags & OBD_STATFS_NODELAY) {
1103 /* procfs requests not want stay in wait for avoid deadlock */
1104 req->rq_no_resend = 1;
1105 req->rq_no_delay = 1;
1106 }
1107
1108 rc = ptlrpc_queue_wait(req);
1109 if (rc) {
1110 /* check connection error first */
1111 if (imp->imp_connect_error)
1112 rc = imp->imp_connect_error;
1113 GOTO(out, rc);
1114 }
1115
1116 msfs = req_capsule_server_get(&req->rq_pill, &RMF_OBD_STATFS);
1117 if (msfs == NULL)
1118 GOTO(out, rc = -EPROTO);
1119
1120 *osfs = *msfs;
1121 EXIT;
1122out:
1123 ptlrpc_req_finished(req);
1124output:
1125 class_import_put(imp);
1126 return rc;
1127}
1128
1129static int mdc_ioc_fid2path(struct obd_export *exp, struct getinfo_fid2path *gf)
1130{
1131 __u32 keylen, vallen;
1132 void *key;
1133 int rc;
1134
1135 if (gf->gf_pathlen > PATH_MAX)
1136 RETURN(-ENAMETOOLONG);
1137 if (gf->gf_pathlen < 2)
1138 RETURN(-EOVERFLOW);
1139
1140 /* Key is KEY_FID2PATH + getinfo_fid2path description */
1141 keylen = cfs_size_round(sizeof(KEY_FID2PATH)) + sizeof(*gf);
1142 OBD_ALLOC(key, keylen);
1143 if (key == NULL)
1144 RETURN(-ENOMEM);
1145 memcpy(key, KEY_FID2PATH, sizeof(KEY_FID2PATH));
1146 memcpy(key + cfs_size_round(sizeof(KEY_FID2PATH)), gf, sizeof(*gf));
1147
1148 CDEBUG(D_IOCTL, "path get "DFID" from "LPU64" #%d\n",
1149 PFID(&gf->gf_fid), gf->gf_recno, gf->gf_linkno);
1150
1151 if (!fid_is_sane(&gf->gf_fid))
1152 GOTO(out, rc = -EINVAL);
1153
1154 /* Val is struct getinfo_fid2path result plus path */
1155 vallen = sizeof(*gf) + gf->gf_pathlen;
1156
1157 rc = obd_get_info(NULL, exp, keylen, key, &vallen, gf, NULL);
1158 if (rc != 0 && rc != -EREMOTE)
1159 GOTO(out, rc);
1160
1161 if (vallen <= sizeof(*gf))
1162 GOTO(out, rc = -EPROTO);
1163 else if (vallen > sizeof(*gf) + gf->gf_pathlen)
1164 GOTO(out, rc = -EOVERFLOW);
1165
1166 CDEBUG(D_IOCTL, "path get "DFID" from "LPU64" #%d\n%s\n",
1167 PFID(&gf->gf_fid), gf->gf_recno, gf->gf_linkno, gf->gf_path);
1168
1169out:
1170 OBD_FREE(key, keylen);
1171 return rc;
1172}
1173
1174static int mdc_ioc_hsm_progress(struct obd_export *exp,
1175 struct hsm_progress_kernel *hpk)
1176{
1177 struct obd_import *imp = class_exp2cliimp(exp);
1178 struct hsm_progress_kernel *req_hpk;
1179 struct ptlrpc_request *req;
1180 int rc;
1181 ENTRY;
1182
1183 req = ptlrpc_request_alloc_pack(imp, &RQF_MDS_HSM_PROGRESS,
1184 LUSTRE_MDS_VERSION, MDS_HSM_PROGRESS);
1185 if (req == NULL)
1186 GOTO(out, rc = -ENOMEM);
1187
1188 mdc_pack_body(req, NULL, NULL, OBD_MD_FLRMTPERM, 0, 0, 0);
1189
1190 /* Copy hsm_progress struct */
1191 req_hpk = req_capsule_client_get(&req->rq_pill, &RMF_MDS_HSM_PROGRESS);
1192 if (req_hpk == NULL)
1193 GOTO(out, rc = -EPROTO);
1194
1195 *req_hpk = *hpk;
1196
1197 ptlrpc_request_set_replen(req);
1198
1199 rc = mdc_queue_wait(req);
1200 GOTO(out, rc);
1201out:
1202 ptlrpc_req_finished(req);
1203 return rc;
1204}
1205
1206static int mdc_ioc_hsm_ct_register(struct obd_import *imp, __u32 archives)
1207{
1208 __u32 *archive_mask;
1209 struct ptlrpc_request *req;
1210 int rc;
1211 ENTRY;
1212
1213 req = ptlrpc_request_alloc_pack(imp, &RQF_MDS_HSM_CT_REGISTER,
1214 LUSTRE_MDS_VERSION,
1215 MDS_HSM_CT_REGISTER);
1216 if (req == NULL)
1217 GOTO(out, rc = -ENOMEM);
1218
1219 mdc_pack_body(req, NULL, NULL, OBD_MD_FLRMTPERM, 0, 0, 0);
1220
1221 /* Copy hsm_progress struct */
1222 archive_mask = req_capsule_client_get(&req->rq_pill,
1223 &RMF_MDS_HSM_ARCHIVE);
1224 if (archive_mask == NULL)
1225 GOTO(out, rc = -EPROTO);
1226
1227 *archive_mask = archives;
1228
1229 ptlrpc_request_set_replen(req);
1230
1231 rc = mdc_queue_wait(req);
1232 GOTO(out, rc);
1233out:
1234 ptlrpc_req_finished(req);
1235 return rc;
1236}
1237
1238static int mdc_ioc_hsm_current_action(struct obd_export *exp,
1239 struct md_op_data *op_data)
1240{
1241 struct hsm_current_action *hca = op_data->op_data;
1242 struct hsm_current_action *req_hca;
1243 struct ptlrpc_request *req;
1244 int rc;
1245 ENTRY;
1246
1247 req = ptlrpc_request_alloc(class_exp2cliimp(exp),
1248 &RQF_MDS_HSM_ACTION);
1249 if (req == NULL)
1250 RETURN(-ENOMEM);
1251
1252 mdc_set_capa_size(req, &RMF_CAPA1, op_data->op_capa1);
1253
1254 rc = ptlrpc_request_pack(req, LUSTRE_MDS_VERSION, MDS_HSM_ACTION);
1255 if (rc) {
1256 ptlrpc_request_free(req);
1257 RETURN(rc);
1258 }
1259
1260 mdc_pack_body(req, &op_data->op_fid1, op_data->op_capa1,
1261 OBD_MD_FLRMTPERM, 0, op_data->op_suppgids[0], 0);
1262
1263 ptlrpc_request_set_replen(req);
1264
1265 rc = mdc_queue_wait(req);
1266 if (rc)
1267 GOTO(out, rc);
1268
1269 req_hca = req_capsule_server_get(&req->rq_pill,
1270 &RMF_MDS_HSM_CURRENT_ACTION);
1271 if (req_hca == NULL)
1272 GOTO(out, rc = -EPROTO);
1273
1274 *hca = *req_hca;
1275
1276 EXIT;
1277out:
1278 ptlrpc_req_finished(req);
1279 return rc;
1280}
1281
1282static int mdc_ioc_hsm_ct_unregister(struct obd_import *imp)
1283{
1284 struct ptlrpc_request *req;
1285 int rc;
1286 ENTRY;
1287
1288 req = ptlrpc_request_alloc_pack(imp, &RQF_MDS_HSM_CT_UNREGISTER,
1289 LUSTRE_MDS_VERSION,
1290 MDS_HSM_CT_UNREGISTER);
1291 if (req == NULL)
1292 GOTO(out, rc = -ENOMEM);
1293
1294 mdc_pack_body(req, NULL, NULL, OBD_MD_FLRMTPERM, 0, 0, 0);
1295
1296 ptlrpc_request_set_replen(req);
1297
1298 rc = mdc_queue_wait(req);
1299 GOTO(out, rc);
1300out:
1301 ptlrpc_req_finished(req);
1302 return rc;
1303}
1304
1305static int mdc_ioc_hsm_state_get(struct obd_export *exp,
1306 struct md_op_data *op_data)
1307{
1308 struct hsm_user_state *hus = op_data->op_data;
1309 struct hsm_user_state *req_hus;
1310 struct ptlrpc_request *req;
1311 int rc;
1312 ENTRY;
1313
1314 req = ptlrpc_request_alloc(class_exp2cliimp(exp),
1315 &RQF_MDS_HSM_STATE_GET);
1316 if (req == NULL)
1317 RETURN(-ENOMEM);
1318
1319 mdc_set_capa_size(req, &RMF_CAPA1, op_data->op_capa1);
1320
1321 rc = ptlrpc_request_pack(req, LUSTRE_MDS_VERSION, MDS_HSM_STATE_GET);
1322 if (rc != 0) {
1323 ptlrpc_request_free(req);
1324 RETURN(rc);
1325 }
1326
1327 mdc_pack_body(req, &op_data->op_fid1, op_data->op_capa1,
1328 OBD_MD_FLRMTPERM, 0, op_data->op_suppgids[0], 0);
1329
1330 ptlrpc_request_set_replen(req);
1331
1332 rc = mdc_queue_wait(req);
1333 if (rc)
1334 GOTO(out, rc);
1335
1336 req_hus = req_capsule_server_get(&req->rq_pill, &RMF_HSM_USER_STATE);
1337 if (req_hus == NULL)
1338 GOTO(out, rc = -EPROTO);
1339
1340 *hus = *req_hus;
1341
1342 EXIT;
1343out:
1344 ptlrpc_req_finished(req);
1345 return rc;
1346}
1347
1348static int mdc_ioc_hsm_state_set(struct obd_export *exp,
1349 struct md_op_data *op_data)
1350{
1351 struct hsm_state_set *hss = op_data->op_data;
1352 struct hsm_state_set *req_hss;
1353 struct ptlrpc_request *req;
1354 int rc;
1355 ENTRY;
1356
1357 req = ptlrpc_request_alloc(class_exp2cliimp(exp),
1358 &RQF_MDS_HSM_STATE_SET);
1359 if (req == NULL)
1360 RETURN(-ENOMEM);
1361
1362 mdc_set_capa_size(req, &RMF_CAPA1, op_data->op_capa1);
1363
1364 rc = ptlrpc_request_pack(req, LUSTRE_MDS_VERSION, MDS_HSM_STATE_SET);
1365 if (rc) {
1366 ptlrpc_request_free(req);
1367 RETURN(rc);
1368 }
1369
1370 mdc_pack_body(req, &op_data->op_fid1, op_data->op_capa1,
1371 OBD_MD_FLRMTPERM, 0, op_data->op_suppgids[0], 0);
1372
1373 /* Copy states */
1374 req_hss = req_capsule_client_get(&req->rq_pill, &RMF_HSM_STATE_SET);
1375 if (req_hss == NULL)
1376 GOTO(out, rc = -EPROTO);
1377 *req_hss = *hss;
1378
1379 ptlrpc_request_set_replen(req);
1380
1381 rc = mdc_queue_wait(req);
1382 GOTO(out, rc);
1383
1384 EXIT;
1385out:
1386 ptlrpc_req_finished(req);
1387 return rc;
1388}
1389
1390static int mdc_ioc_hsm_request(struct obd_export *exp,
1391 struct hsm_user_request *hur)
1392{
1393 struct obd_import *imp = class_exp2cliimp(exp);
1394 struct ptlrpc_request *req;
1395 struct hsm_request *req_hr;
1396 struct hsm_user_item *req_hui;
1397 char *req_opaque;
1398 int rc;
1399 ENTRY;
1400
1401 req = ptlrpc_request_alloc(imp, &RQF_MDS_HSM_REQUEST);
1402 if (req == NULL)
1403 GOTO(out, rc = -ENOMEM);
1404
1405 req_capsule_set_size(&req->rq_pill, &RMF_MDS_HSM_USER_ITEM, RCL_CLIENT,
1406 hur->hur_request.hr_itemcount
1407 * sizeof(struct hsm_user_item));
1408 req_capsule_set_size(&req->rq_pill, &RMF_GENERIC_DATA, RCL_CLIENT,
1409 hur->hur_request.hr_data_len);
1410
1411 rc = ptlrpc_request_pack(req, LUSTRE_MDS_VERSION, MDS_HSM_REQUEST);
1412 if (rc) {
1413 ptlrpc_request_free(req);
1414 RETURN(rc);
1415 }
1416
1417 mdc_pack_body(req, NULL, NULL, OBD_MD_FLRMTPERM, 0, 0, 0);
1418
1419 /* Copy hsm_request struct */
1420 req_hr = req_capsule_client_get(&req->rq_pill, &RMF_MDS_HSM_REQUEST);
1421 if (req_hr == NULL)
1422 GOTO(out, rc = -EPROTO);
1423 *req_hr = hur->hur_request;
1424
1425 /* Copy hsm_user_item structs */
1426 req_hui = req_capsule_client_get(&req->rq_pill, &RMF_MDS_HSM_USER_ITEM);
1427 if (req_hui == NULL)
1428 GOTO(out, rc = -EPROTO);
1429 memcpy(req_hui, hur->hur_user_item,
1430 hur->hur_request.hr_itemcount * sizeof(struct hsm_user_item));
1431
1432 /* Copy opaque field */
1433 req_opaque = req_capsule_client_get(&req->rq_pill, &RMF_GENERIC_DATA);
1434 if (req_opaque == NULL)
1435 GOTO(out, rc = -EPROTO);
1436 memcpy(req_opaque, hur_data(hur), hur->hur_request.hr_data_len);
1437
1438 ptlrpc_request_set_replen(req);
1439
1440 rc = mdc_queue_wait(req);
1441 GOTO(out, rc);
1442
1443out:
1444 ptlrpc_req_finished(req);
1445 return rc;
1446}
1447
1448static struct kuc_hdr *changelog_kuc_hdr(char *buf, int len, int flags)
1449{
1450 struct kuc_hdr *lh = (struct kuc_hdr *)buf;
1451
1452 LASSERT(len <= CR_MAXSIZE);
1453
1454 lh->kuc_magic = KUC_MAGIC;
1455 lh->kuc_transport = KUC_TRANSPORT_CHANGELOG;
1456 lh->kuc_flags = flags;
1457 lh->kuc_msgtype = CL_RECORD;
1458 lh->kuc_msglen = len;
1459 return lh;
1460}
1461
1462#define D_CHANGELOG 0
1463
1464struct changelog_show {
1465 __u64 cs_startrec;
1466 __u32 cs_flags;
1467 struct file *cs_fp;
1468 char *cs_buf;
1469 struct obd_device *cs_obd;
1470};
1471
e377988e 1472static int changelog_kkuc_cb(const struct lu_env *env, struct llog_handle *llh,
d7e09d03
PT
1473 struct llog_rec_hdr *hdr, void *data)
1474{
1475 struct changelog_show *cs = data;
1476 struct llog_changelog_rec *rec = (struct llog_changelog_rec *)hdr;
1477 struct kuc_hdr *lh;
1478 int len, rc;
1479 ENTRY;
1480
e377988e
AD
1481 if (rec->cr_hdr.lrh_type != CHANGELOG_REC) {
1482 rc = -EINVAL;
1483 CERROR("%s: not a changelog rec %x/%d: rc = %d\n",
1484 cs->cs_obd->obd_name, rec->cr_hdr.lrh_type,
1485 rec->cr.cr_type, rc);
1486 RETURN(rc);
d7e09d03
PT
1487 }
1488
1489 if (rec->cr.cr_index < cs->cs_startrec) {
1490 /* Skip entries earlier than what we are interested in */
1491 CDEBUG(D_CHANGELOG, "rec="LPU64" start="LPU64"\n",
1492 rec->cr.cr_index, cs->cs_startrec);
1493 RETURN(0);
1494 }
1495
1496 CDEBUG(D_CHANGELOG, LPU64" %02d%-5s "LPU64" 0x%x t="DFID" p="DFID
1497 " %.*s\n", rec->cr.cr_index, rec->cr.cr_type,
1498 changelog_type2str(rec->cr.cr_type), rec->cr.cr_time,
1499 rec->cr.cr_flags & CLF_FLAGMASK,
1500 PFID(&rec->cr.cr_tfid), PFID(&rec->cr.cr_pfid),
1501 rec->cr.cr_namelen, changelog_rec_name(&rec->cr));
1502
1503 len = sizeof(*lh) + changelog_rec_size(&rec->cr) + rec->cr.cr_namelen;
1504
1505 /* Set up the message */
1506 lh = changelog_kuc_hdr(cs->cs_buf, len, cs->cs_flags);
1507 memcpy(lh + 1, &rec->cr, len - sizeof(*lh));
1508
1509 rc = libcfs_kkuc_msg_put(cs->cs_fp, lh);
1510 CDEBUG(D_CHANGELOG, "kucmsg fp %p len %d rc %d\n", cs->cs_fp, len,rc);
1511
1512 RETURN(rc);
1513}
1514
1515static int mdc_changelog_send_thread(void *csdata)
1516{
1517 struct changelog_show *cs = csdata;
1518 struct llog_ctxt *ctxt = NULL;
1519 struct llog_handle *llh = NULL;
1520 struct kuc_hdr *kuch;
1521 int rc;
1522
1523 CDEBUG(D_CHANGELOG, "changelog to fp=%p start "LPU64"\n",
1524 cs->cs_fp, cs->cs_startrec);
1525
1526 OBD_ALLOC(cs->cs_buf, CR_MAXSIZE);
1527 if (cs->cs_buf == NULL)
1528 GOTO(out, rc = -ENOMEM);
1529
1530 /* Set up the remote catalog handle */
1531 ctxt = llog_get_context(cs->cs_obd, LLOG_CHANGELOG_REPL_CTXT);
1532 if (ctxt == NULL)
1533 GOTO(out, rc = -ENOENT);
1534 rc = llog_open(NULL, ctxt, &llh, NULL, CHANGELOG_CATALOG,
1535 LLOG_OPEN_EXISTS);
1536 if (rc) {
1537 CERROR("%s: fail to open changelog catalog: rc = %d\n",
1538 cs->cs_obd->obd_name, rc);
1539 GOTO(out, rc);
1540 }
1541 rc = llog_init_handle(NULL, llh, LLOG_F_IS_CAT, NULL);
1542 if (rc) {
1543 CERROR("llog_init_handle failed %d\n", rc);
1544 GOTO(out, rc);
1545 }
1546
e377988e 1547 rc = llog_cat_process(NULL, llh, changelog_kkuc_cb, cs, 0, 0);
d7e09d03
PT
1548
1549 /* Send EOF no matter what our result */
1550 if ((kuch = changelog_kuc_hdr(cs->cs_buf, sizeof(*kuch),
1551 cs->cs_flags))) {
1552 kuch->kuc_msgtype = CL_EOF;
1553 libcfs_kkuc_msg_put(cs->cs_fp, kuch);
1554 }
1555
1556out:
1557 fput(cs->cs_fp);
1558 if (llh)
1559 llog_cat_close(NULL, llh);
1560 if (ctxt)
1561 llog_ctxt_put(ctxt);
1562 if (cs->cs_buf)
1563 OBD_FREE(cs->cs_buf, CR_MAXSIZE);
1564 OBD_FREE_PTR(cs);
1565 return rc;
1566}
1567
1568static int mdc_ioc_changelog_send(struct obd_device *obd,
1569 struct ioc_changelog *icc)
1570{
1571 struct changelog_show *cs;
1572 int rc;
1573
1574 /* Freed in mdc_changelog_send_thread */
1575 OBD_ALLOC_PTR(cs);
1576 if (!cs)
1577 return -ENOMEM;
1578
1579 cs->cs_obd = obd;
1580 cs->cs_startrec = icc->icc_recno;
1581 /* matching fput in mdc_changelog_send_thread */
1582 cs->cs_fp = fget(icc->icc_id);
1583 cs->cs_flags = icc->icc_flags;
1584
1585 /*
1586 * New thread because we should return to user app before
1587 * writing into our pipe
1588 */
1589 rc = PTR_ERR(kthread_run(mdc_changelog_send_thread, cs,
1590 "mdc_clg_send_thread"));
1591 if (!IS_ERR_VALUE(rc)) {
1592 CDEBUG(D_CHANGELOG, "start changelog thread\n");
1593 return 0;
1594 }
1595
1596 CERROR("Failed to start changelog thread: %d\n", rc);
1597 OBD_FREE_PTR(cs);
1598 return rc;
1599}
1600
1601static int mdc_ioc_hsm_ct_start(struct obd_export *exp,
1602 struct lustre_kernelcomm *lk);
1603
1604static int mdc_quotacheck(struct obd_device *unused, struct obd_export *exp,
1605 struct obd_quotactl *oqctl)
1606{
1607 struct client_obd *cli = &exp->exp_obd->u.cli;
1608 struct ptlrpc_request *req;
1609 struct obd_quotactl *body;
1610 int rc;
1611 ENTRY;
1612
1613 req = ptlrpc_request_alloc_pack(class_exp2cliimp(exp),
1614 &RQF_MDS_QUOTACHECK, LUSTRE_MDS_VERSION,
1615 MDS_QUOTACHECK);
1616 if (req == NULL)
1617 RETURN(-ENOMEM);
1618
1619 body = req_capsule_client_get(&req->rq_pill, &RMF_OBD_QUOTACTL);
1620 *body = *oqctl;
1621
1622 ptlrpc_request_set_replen(req);
1623
1624 /* the next poll will find -ENODATA, that means quotacheck is
1625 * going on */
1626 cli->cl_qchk_stat = -ENODATA;
1627 rc = ptlrpc_queue_wait(req);
1628 if (rc)
1629 cli->cl_qchk_stat = rc;
1630 ptlrpc_req_finished(req);
1631 RETURN(rc);
1632}
1633
1634static int mdc_quota_poll_check(struct obd_export *exp,
1635 struct if_quotacheck *qchk)
1636{
1637 struct client_obd *cli = &exp->exp_obd->u.cli;
1638 int rc;
1639 ENTRY;
1640
1641 qchk->obd_uuid = cli->cl_target_uuid;
1642 memcpy(qchk->obd_type, LUSTRE_MDS_NAME, strlen(LUSTRE_MDS_NAME));
1643
1644 rc = cli->cl_qchk_stat;
1645 /* the client is not the previous one */
1646 if (rc == CL_NOT_QUOTACHECKED)
1647 rc = -EINTR;
1648 RETURN(rc);
1649}
1650
1651static int mdc_quotactl(struct obd_device *unused, struct obd_export *exp,
1652 struct obd_quotactl *oqctl)
1653{
1654 struct ptlrpc_request *req;
1655 struct obd_quotactl *oqc;
1656 int rc;
1657 ENTRY;
1658
1659 req = ptlrpc_request_alloc_pack(class_exp2cliimp(exp),
1660 &RQF_MDS_QUOTACTL, LUSTRE_MDS_VERSION,
1661 MDS_QUOTACTL);
1662 if (req == NULL)
1663 RETURN(-ENOMEM);
1664
1665 oqc = req_capsule_client_get(&req->rq_pill, &RMF_OBD_QUOTACTL);
1666 *oqc = *oqctl;
1667
1668 ptlrpc_request_set_replen(req);
1669 ptlrpc_at_set_req_timeout(req);
1670 req->rq_no_resend = 1;
1671
1672 rc = ptlrpc_queue_wait(req);
1673 if (rc)
1674 CERROR("ptlrpc_queue_wait failed, rc: %d\n", rc);
1675
1676 if (req->rq_repmsg &&
1677 (oqc = req_capsule_server_get(&req->rq_pill, &RMF_OBD_QUOTACTL))) {
1678 *oqctl = *oqc;
1679 } else if (!rc) {
1680 CERROR ("Can't unpack obd_quotactl\n");
1681 rc = -EPROTO;
1682 }
1683 ptlrpc_req_finished(req);
1684
1685 RETURN(rc);
1686}
1687
1688static int mdc_ioc_swap_layouts(struct obd_export *exp,
1689 struct md_op_data *op_data)
1690{
1691 LIST_HEAD(cancels);
1692 struct ptlrpc_request *req;
1693 int rc, count;
1694 struct mdc_swap_layouts *msl, *payload;
1695 ENTRY;
1696
1697 msl = op_data->op_data;
1698
1699 /* When the MDT will get the MDS_SWAP_LAYOUTS RPC the
1700 * first thing it will do is to cancel the 2 layout
1701 * locks hold by this client.
1702 * So the client must cancel its layout locks on the 2 fids
1703 * with the request RPC to avoid extra RPC round trips
1704 */
1705 count = mdc_resource_get_unused(exp, &op_data->op_fid1, &cancels,
1706 LCK_CR, MDS_INODELOCK_LAYOUT);
1707 count += mdc_resource_get_unused(exp, &op_data->op_fid2, &cancels,
1708 LCK_CR, MDS_INODELOCK_LAYOUT);
1709
1710 req = ptlrpc_request_alloc(class_exp2cliimp(exp),
1711 &RQF_MDS_SWAP_LAYOUTS);
1712 if (req == NULL) {
1713 ldlm_lock_list_put(&cancels, l_bl_ast, count);
1714 RETURN(-ENOMEM);
1715 }
1716
1717 mdc_set_capa_size(req, &RMF_CAPA1, op_data->op_capa1);
1718 mdc_set_capa_size(req, &RMF_CAPA2, op_data->op_capa2);
1719
1720 rc = mdc_prep_elc_req(exp, req, MDS_SWAP_LAYOUTS, &cancels, count);
1721 if (rc) {
1722 ptlrpc_request_free(req);
1723 RETURN(rc);
1724 }
1725
1726 mdc_swap_layouts_pack(req, op_data);
1727
1728 payload = req_capsule_client_get(&req->rq_pill, &RMF_SWAP_LAYOUTS);
1729 LASSERT(payload);
1730
1731 *payload = *msl;
1732
1733 ptlrpc_request_set_replen(req);
1734
1735 rc = ptlrpc_queue_wait(req);
1736 if (rc)
1737 GOTO(out, rc);
1738 EXIT;
1739
1740out:
1741 ptlrpc_req_finished(req);
1742 return rc;
1743}
1744
1745static int mdc_iocontrol(unsigned int cmd, struct obd_export *exp, int len,
1746 void *karg, void *uarg)
1747{
1748 struct obd_device *obd = exp->exp_obd;
1749 struct obd_ioctl_data *data = karg;
1750 struct obd_import *imp = obd->u.cli.cl_import;
1751 struct llog_ctxt *ctxt;
1752 int rc;
1753 ENTRY;
1754
1755 if (!try_module_get(THIS_MODULE)) {
1756 CERROR("Can't get module. Is it alive?");
1757 return -EINVAL;
1758 }
1759 switch (cmd) {
1760 case OBD_IOC_CHANGELOG_SEND:
1761 rc = mdc_ioc_changelog_send(obd, karg);
1762 GOTO(out, rc);
1763 case OBD_IOC_CHANGELOG_CLEAR: {
1764 struct ioc_changelog *icc = karg;
1765 struct changelog_setinfo cs =
1766 {.cs_recno = icc->icc_recno, .cs_id = icc->icc_id};
1767 rc = obd_set_info_async(NULL, exp, strlen(KEY_CHANGELOG_CLEAR),
1768 KEY_CHANGELOG_CLEAR, sizeof(cs), &cs,
1769 NULL);
1770 GOTO(out, rc);
1771 }
1772 case OBD_IOC_FID2PATH:
1773 rc = mdc_ioc_fid2path(exp, karg);
1774 GOTO(out, rc);
1775 case LL_IOC_HSM_CT_START:
1776 rc = mdc_ioc_hsm_ct_start(exp, karg);
1777 GOTO(out, rc);
1778 case LL_IOC_HSM_PROGRESS:
1779 rc = mdc_ioc_hsm_progress(exp, karg);
1780 GOTO(out, rc);
1781 case LL_IOC_HSM_STATE_GET:
1782 rc = mdc_ioc_hsm_state_get(exp, karg);
1783 GOTO(out, rc);
1784 case LL_IOC_HSM_STATE_SET:
1785 rc = mdc_ioc_hsm_state_set(exp, karg);
1786 case LL_IOC_HSM_ACTION:
1787 rc = mdc_ioc_hsm_current_action(exp, karg);
1788 GOTO(out, rc);
1789 case LL_IOC_HSM_REQUEST:
1790 rc = mdc_ioc_hsm_request(exp, karg);
1791 GOTO(out, rc);
1792 case OBD_IOC_CLIENT_RECOVER:
1793 rc = ptlrpc_recover_import(imp, data->ioc_inlbuf1, 0);
1794 if (rc < 0)
1795 GOTO(out, rc);
1796 GOTO(out, rc = 0);
1797 case IOC_OSC_SET_ACTIVE:
1798 rc = ptlrpc_set_import_active(imp, data->ioc_offset);
1799 GOTO(out, rc);
1800 case OBD_IOC_PARSE: {
1801 ctxt = llog_get_context(exp->exp_obd, LLOG_CONFIG_REPL_CTXT);
1802 rc = class_config_parse_llog(NULL, ctxt, data->ioc_inlbuf1,
1803 NULL);
1804 llog_ctxt_put(ctxt);
1805 GOTO(out, rc);
1806 }
1807 case OBD_IOC_LLOG_INFO:
1808 case OBD_IOC_LLOG_PRINT: {
1809 ctxt = llog_get_context(obd, LLOG_CONFIG_REPL_CTXT);
1810 rc = llog_ioctl(NULL, ctxt, cmd, data);
1811 llog_ctxt_put(ctxt);
1812 GOTO(out, rc);
1813 }
1814 case OBD_IOC_POLL_QUOTACHECK:
1815 rc = mdc_quota_poll_check(exp, (struct if_quotacheck *)karg);
1816 GOTO(out, rc);
1817 case OBD_IOC_PING_TARGET:
1818 rc = ptlrpc_obd_ping(obd);
1819 GOTO(out, rc);
1820 /*
1821 * Normally IOC_OBD_STATFS, OBD_IOC_QUOTACTL iocontrol are handled by
1822 * LMV instead of MDC. But when the cluster is upgraded from 1.8,
1823 * there'd be no LMV layer thus we might be called here. Eventually
1824 * this code should be removed.
1825 * bz20731, LU-592.
1826 */
1827 case IOC_OBD_STATFS: {
1828 struct obd_statfs stat_buf = {0};
1829
1830 if (*((__u32 *) data->ioc_inlbuf2) != 0)
1831 GOTO(out, rc = -ENODEV);
1832
1833 /* copy UUID */
1834 if (copy_to_user(data->ioc_pbuf2, obd2cli_tgt(obd),
1835 min((int) data->ioc_plen2,
1836 (int) sizeof(struct obd_uuid))))
1837 GOTO(out, rc = -EFAULT);
1838
1839 rc = mdc_statfs(NULL, obd->obd_self_export, &stat_buf,
1840 cfs_time_shift_64(-OBD_STATFS_CACHE_SECONDS),
1841 0);
1842 if (rc != 0)
1843 GOTO(out, rc);
1844
1845 if (copy_to_user(data->ioc_pbuf1, &stat_buf,
1846 min((int) data->ioc_plen1,
1847 (int) sizeof(stat_buf))))
1848 GOTO(out, rc = -EFAULT);
1849
1850 GOTO(out, rc = 0);
1851 }
1852 case OBD_IOC_QUOTACTL: {
1853 struct if_quotactl *qctl = karg;
1854 struct obd_quotactl *oqctl;
1855
1856 OBD_ALLOC_PTR(oqctl);
1857 if (!oqctl)
1858 RETURN(-ENOMEM);
1859
1860 QCTL_COPY(oqctl, qctl);
1861 rc = obd_quotactl(exp, oqctl);
1862 if (rc == 0) {
1863 QCTL_COPY(qctl, oqctl);
1864 qctl->qc_valid = QC_MDTIDX;
1865 qctl->obd_uuid = obd->u.cli.cl_target_uuid;
1866 }
1867 OBD_FREE_PTR(oqctl);
1868 break;
1869 }
1870 case LL_IOC_GET_CONNECT_FLAGS: {
1871 if (copy_to_user(uarg,
1872 exp_connect_flags_ptr(exp),
1873 sizeof(__u64)))
1874 GOTO(out, rc = -EFAULT);
1875 else
1876 GOTO(out, rc = 0);
1877 }
1878 case LL_IOC_LOV_SWAP_LAYOUTS: {
1879 rc = mdc_ioc_swap_layouts(exp, karg);
1880 break;
1881 }
1882 default:
1883 CERROR("mdc_ioctl(): unrecognised ioctl %#x\n", cmd);
1884 GOTO(out, rc = -ENOTTY);
1885 }
1886out:
1887 module_put(THIS_MODULE);
1888
1889 return rc;
1890}
1891
1892int mdc_get_info_rpc(struct obd_export *exp,
1893 obd_count keylen, void *key,
1894 int vallen, void *val)
1895{
1896 struct obd_import *imp = class_exp2cliimp(exp);
1897 struct ptlrpc_request *req;
1898 char *tmp;
1899 int rc = -EINVAL;
1900 ENTRY;
1901
1902 req = ptlrpc_request_alloc(imp, &RQF_MDS_GET_INFO);
1903 if (req == NULL)
1904 RETURN(-ENOMEM);
1905
1906 req_capsule_set_size(&req->rq_pill, &RMF_GETINFO_KEY,
1907 RCL_CLIENT, keylen);
1908 req_capsule_set_size(&req->rq_pill, &RMF_GETINFO_VALLEN,
1909 RCL_CLIENT, sizeof(__u32));
1910
1911 rc = ptlrpc_request_pack(req, LUSTRE_MDS_VERSION, MDS_GET_INFO);
1912 if (rc) {
1913 ptlrpc_request_free(req);
1914 RETURN(rc);
1915 }
1916
1917 tmp = req_capsule_client_get(&req->rq_pill, &RMF_GETINFO_KEY);
1918 memcpy(tmp, key, keylen);
1919 tmp = req_capsule_client_get(&req->rq_pill, &RMF_GETINFO_VALLEN);
1920 memcpy(tmp, &vallen, sizeof(__u32));
1921
1922 req_capsule_set_size(&req->rq_pill, &RMF_GETINFO_VAL,
1923 RCL_SERVER, vallen);
1924 ptlrpc_request_set_replen(req);
1925
1926 rc = ptlrpc_queue_wait(req);
1927 /* -EREMOTE means the get_info result is partial, and it needs to
1928 * continue on another MDT, see fid2path part in lmv_iocontrol */
1929 if (rc == 0 || rc == -EREMOTE) {
1930 tmp = req_capsule_server_get(&req->rq_pill, &RMF_GETINFO_VAL);
1931 memcpy(val, tmp, vallen);
1932 if (ptlrpc_rep_need_swab(req)) {
1933 if (KEY_IS(KEY_FID2PATH))
1934 lustre_swab_fid2path(val);
1935 }
1936 }
1937 ptlrpc_req_finished(req);
1938
1939 RETURN(rc);
1940}
1941
1942static void lustre_swab_hai(struct hsm_action_item *h)
1943{
1944 __swab32s(&h->hai_len);
1945 __swab32s(&h->hai_action);
1946 lustre_swab_lu_fid(&h->hai_fid);
1947 lustre_swab_lu_fid(&h->hai_dfid);
1948 __swab64s(&h->hai_cookie);
1949 __swab64s(&h->hai_extent.offset);
1950 __swab64s(&h->hai_extent.length);
1951 __swab64s(&h->hai_gid);
1952}
1953
1954static void lustre_swab_hal(struct hsm_action_list *h)
1955{
1956 struct hsm_action_item *hai;
1957 int i;
1958
1959 __swab32s(&h->hal_version);
1960 __swab32s(&h->hal_count);
1961 __swab32s(&h->hal_archive_id);
1962 __swab64s(&h->hal_flags);
1963 hai = hai_zero(h);
1964 for (i = 0; i < h->hal_count; i++) {
1965 lustre_swab_hai(hai);
1966 hai = hai_next(hai);
1967 }
1968}
1969
1970static void lustre_swab_kuch(struct kuc_hdr *l)
1971{
1972 __swab16s(&l->kuc_magic);
1973 /* __u8 l->kuc_transport */
1974 __swab16s(&l->kuc_msgtype);
1975 __swab16s(&l->kuc_msglen);
1976}
1977
1978static int mdc_ioc_hsm_ct_start(struct obd_export *exp,
1979 struct lustre_kernelcomm *lk)
1980{
1981 struct obd_import *imp = class_exp2cliimp(exp);
1982 __u32 archive = lk->lk_data;
1983 int rc = 0;
1984
1985 if (lk->lk_group != KUC_GRP_HSM) {
1986 CERROR("Bad copytool group %d\n", lk->lk_group);
1987 return -EINVAL;
1988 }
1989
1990 CDEBUG(D_HSM, "CT start r%d w%d u%d g%d f%#x\n", lk->lk_rfd, lk->lk_wfd,
1991 lk->lk_uid, lk->lk_group, lk->lk_flags);
1992
1993 if (lk->lk_flags & LK_FLG_STOP) {
1994 rc = libcfs_kkuc_group_rem(lk->lk_uid, lk->lk_group);
1995 /* Unregister with the coordinator */
1996 if (rc == 0)
1997 rc = mdc_ioc_hsm_ct_unregister(imp);
1998 } else {
1999 struct file *fp = fget(lk->lk_wfd);
2000
2001 rc = libcfs_kkuc_group_add(fp, lk->lk_uid, lk->lk_group,
2002 lk->lk_data);
2003 if (rc && fp)
2004 fput(fp);
2005 if (rc == 0)
2006 rc = mdc_ioc_hsm_ct_register(imp, archive);
2007 }
2008
2009 return rc;
2010}
2011
2012/**
2013 * Send a message to any listening copytools
2014 * @param val KUC message (kuc_hdr + hsm_action_list)
2015 * @param len total length of message
2016 */
2017static int mdc_hsm_copytool_send(int len, void *val)
2018{
2019 struct kuc_hdr *lh = (struct kuc_hdr *)val;
2020 struct hsm_action_list *hal = (struct hsm_action_list *)(lh + 1);
2021 int rc;
2022 ENTRY;
2023
2024 if (len < sizeof(*lh) + sizeof(*hal)) {
2025 CERROR("Short HSM message %d < %d\n", len,
2026 (int) (sizeof(*lh) + sizeof(*hal)));
2027 RETURN(-EPROTO);
2028 }
2029 if (lh->kuc_magic == __swab16(KUC_MAGIC)) {
2030 lustre_swab_kuch(lh);
2031 lustre_swab_hal(hal);
2032 } else if (lh->kuc_magic != KUC_MAGIC) {
2033 CERROR("Bad magic %x!=%x\n", lh->kuc_magic, KUC_MAGIC);
2034 RETURN(-EPROTO);
2035 }
2036
2037 CDEBUG(D_HSM, " Received message mg=%x t=%d m=%d l=%d actions=%d "
2038 "on %s\n",
2039 lh->kuc_magic, lh->kuc_transport, lh->kuc_msgtype,
2040 lh->kuc_msglen, hal->hal_count, hal->hal_fsname);
2041
2042 /* Broadcast to HSM listeners */
2043 rc = libcfs_kkuc_group_put(KUC_GRP_HSM, lh);
2044
2045 RETURN(rc);
2046}
2047
2048/**
2049 * callback function passed to kuc for re-registering each HSM copytool
2050 * running on MDC, after MDT shutdown/recovery.
2051 * @param data archive id served by the copytool
2052 * @param cb_arg callback argument (obd_import)
2053 */
2054static int mdc_hsm_ct_reregister(__u32 data, void *cb_arg)
2055{
2056 struct obd_import *imp = (struct obd_import *)cb_arg;
2057 __u32 archive = data;
2058 int rc;
2059
2060 CDEBUG(D_HA, "recover copytool registration to MDT (archive=%#x)\n",
2061 archive);
2062 rc = mdc_ioc_hsm_ct_register(imp, archive);
2063
2064 /* ignore error if the copytool is already registered */
2065 return ((rc != 0) && (rc != -EEXIST)) ? rc : 0;
2066}
2067
2068/**
2069 * Re-establish all kuc contexts with MDT
2070 * after MDT shutdown/recovery.
2071 */
2072static int mdc_kuc_reregister(struct obd_import *imp)
2073{
2074 /* re-register HSM agents */
2075 return libcfs_kkuc_group_foreach(KUC_GRP_HSM, mdc_hsm_ct_reregister,
2076 (void *)imp);
2077}
2078
2079int mdc_set_info_async(const struct lu_env *env,
2080 struct obd_export *exp,
2081 obd_count keylen, void *key,
2082 obd_count vallen, void *val,
2083 struct ptlrpc_request_set *set)
2084{
2085 struct obd_import *imp = class_exp2cliimp(exp);
2086 int rc;
2087 ENTRY;
2088
2089 if (KEY_IS(KEY_READ_ONLY)) {
2090 if (vallen != sizeof(int))
2091 RETURN(-EINVAL);
2092
2093 spin_lock(&imp->imp_lock);
2094 if (*((int *)val)) {
2095 imp->imp_connect_flags_orig |= OBD_CONNECT_RDONLY;
2096 imp->imp_connect_data.ocd_connect_flags |=
2097 OBD_CONNECT_RDONLY;
2098 } else {
2099 imp->imp_connect_flags_orig &= ~OBD_CONNECT_RDONLY;
2100 imp->imp_connect_data.ocd_connect_flags &=
2101 ~OBD_CONNECT_RDONLY;
2102 }
2103 spin_unlock(&imp->imp_lock);
2104
2105 rc = do_set_info_async(imp, MDS_SET_INFO, LUSTRE_MDS_VERSION,
2106 keylen, key, vallen, val, set);
2107 RETURN(rc);
2108 }
2109 if (KEY_IS(KEY_SPTLRPC_CONF)) {
2110 sptlrpc_conf_client_adapt(exp->exp_obd);
2111 RETURN(0);
2112 }
2113 if (KEY_IS(KEY_FLUSH_CTX)) {
2114 sptlrpc_import_flush_my_ctx(imp);
2115 RETURN(0);
2116 }
2117 if (KEY_IS(KEY_MDS_CONN)) {
2118 /* mds-mds import */
2119 spin_lock(&imp->imp_lock);
2120 imp->imp_server_timeout = 1;
2121 spin_unlock(&imp->imp_lock);
2122 imp->imp_client->cli_request_portal = MDS_MDS_PORTAL;
2123 CDEBUG(D_OTHER, "%s: timeout / 2\n", exp->exp_obd->obd_name);
2124 RETURN(0);
2125 }
2126 if (KEY_IS(KEY_CHANGELOG_CLEAR)) {
2127 rc = do_set_info_async(imp, MDS_SET_INFO, LUSTRE_MDS_VERSION,
2128 keylen, key, vallen, val, set);
2129 RETURN(rc);
2130 }
2131 if (KEY_IS(KEY_HSM_COPYTOOL_SEND)) {
2132 rc = mdc_hsm_copytool_send(vallen, val);
2133 RETURN(rc);
2134 }
2135
2136 CERROR("Unknown key %s\n", (char *)key);
2137 RETURN(-EINVAL);
2138}
2139
2140int mdc_get_info(const struct lu_env *env, struct obd_export *exp,
2141 __u32 keylen, void *key, __u32 *vallen, void *val,
2142 struct lov_stripe_md *lsm)
2143{
2144 int rc = -EINVAL;
2145
2146 if (KEY_IS(KEY_MAX_EASIZE)) {
2147 int mdsize, *max_easize;
2148
2149 if (*vallen != sizeof(int))
2150 RETURN(-EINVAL);
2151 mdsize = *(int*)val;
2152 if (mdsize > exp->exp_obd->u.cli.cl_max_mds_easize)
2153 exp->exp_obd->u.cli.cl_max_mds_easize = mdsize;
2154 max_easize = val;
2155 *max_easize = exp->exp_obd->u.cli.cl_max_mds_easize;
2156 RETURN(0);
2157 } else if (KEY_IS(KEY_CONN_DATA)) {
2158 struct obd_import *imp = class_exp2cliimp(exp);
2159 struct obd_connect_data *data = val;
2160
2161 if (*vallen != sizeof(*data))
2162 RETURN(-EINVAL);
2163
2164 *data = imp->imp_connect_data;
2165 RETURN(0);
2166 } else if (KEY_IS(KEY_TGT_COUNT)) {
2167 *((int *)val) = 1;
2168 RETURN(0);
2169 }
2170
2171 rc = mdc_get_info_rpc(exp, keylen, key, *vallen, val);
2172
2173 RETURN(rc);
2174}
2175
2176static int mdc_pin(struct obd_export *exp, const struct lu_fid *fid,
2177 struct obd_capa *oc, struct obd_client_handle *handle,
2178 int flags)
2179{
2180 struct ptlrpc_request *req;
2181 struct mdt_body *body;
2182 int rc;
2183 ENTRY;
2184
2185 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_MDS_PIN);
2186 if (req == NULL)
2187 RETURN(-ENOMEM);
2188
2189 mdc_set_capa_size(req, &RMF_CAPA1, oc);
2190
2191 rc = ptlrpc_request_pack(req, LUSTRE_MDS_VERSION, MDS_PIN);
2192 if (rc) {
2193 ptlrpc_request_free(req);
2194 RETURN(rc);
2195 }
2196
2197 mdc_pack_body(req, fid, oc, 0, 0, -1, flags);
2198
2199 ptlrpc_request_set_replen(req);
2200
2201 mdc_get_rpc_lock(exp->exp_obd->u.cli.cl_rpc_lock, NULL);
2202 rc = ptlrpc_queue_wait(req);
2203 mdc_put_rpc_lock(exp->exp_obd->u.cli.cl_rpc_lock, NULL);
2204 if (rc) {
2205 CERROR("Pin failed: %d\n", rc);
2206 GOTO(err_out, rc);
2207 }
2208
2209 body = req_capsule_server_get(&req->rq_pill, &RMF_MDT_BODY);
2210 if (body == NULL)
2211 GOTO(err_out, rc = -EPROTO);
2212
2213 handle->och_fh = body->handle;
2214 handle->och_magic = OBD_CLIENT_HANDLE_MAGIC;
2215
2216 handle->och_mod = obd_mod_alloc();
2217 if (handle->och_mod == NULL) {
2218 DEBUG_REQ(D_ERROR, req, "can't allocate md_open_data");
2219 GOTO(err_out, rc = -ENOMEM);
2220 }
2221 handle->och_mod->mod_open_req = req; /* will be dropped by unpin */
2222
2223 RETURN(0);
2224
2225err_out:
2226 ptlrpc_req_finished(req);
2227 RETURN(rc);
2228}
2229
2230static int mdc_unpin(struct obd_export *exp, struct obd_client_handle *handle,
2231 int flag)
2232{
2233 struct ptlrpc_request *req;
2234 struct mdt_body *body;
2235 int rc;
2236 ENTRY;
2237
2238 req = ptlrpc_request_alloc_pack(class_exp2cliimp(exp), &RQF_MDS_UNPIN,
2239 LUSTRE_MDS_VERSION, MDS_UNPIN);
2240 if (req == NULL)
2241 RETURN(-ENOMEM);
2242
2243 body = req_capsule_client_get(&req->rq_pill, &RMF_MDT_BODY);
2244 body->handle = handle->och_fh;
2245 body->flags = flag;
2246
2247 ptlrpc_request_set_replen(req);
2248
2249 mdc_get_rpc_lock(exp->exp_obd->u.cli.cl_rpc_lock, NULL);
2250 rc = ptlrpc_queue_wait(req);
2251 mdc_put_rpc_lock(exp->exp_obd->u.cli.cl_rpc_lock, NULL);
2252
2253 if (rc != 0)
2254 CERROR("Unpin failed: %d\n", rc);
2255
2256 ptlrpc_req_finished(req);
2257 ptlrpc_req_finished(handle->och_mod->mod_open_req);
2258
2259 obd_mod_put(handle->och_mod);
2260 RETURN(rc);
2261}
2262
2263int mdc_sync(struct obd_export *exp, const struct lu_fid *fid,
2264 struct obd_capa *oc, struct ptlrpc_request **request)
2265{
2266 struct ptlrpc_request *req;
2267 int rc;
2268 ENTRY;
2269
2270 *request = NULL;
2271 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_MDS_SYNC);
2272 if (req == NULL)
2273 RETURN(-ENOMEM);
2274
2275 mdc_set_capa_size(req, &RMF_CAPA1, oc);
2276
2277 rc = ptlrpc_request_pack(req, LUSTRE_MDS_VERSION, MDS_SYNC);
2278 if (rc) {
2279 ptlrpc_request_free(req);
2280 RETURN(rc);
2281 }
2282
2283 mdc_pack_body(req, fid, oc, 0, 0, -1, 0);
2284
2285 ptlrpc_request_set_replen(req);
2286
2287 rc = ptlrpc_queue_wait(req);
2288 if (rc)
2289 ptlrpc_req_finished(req);
2290 else
2291 *request = req;
2292 RETURN(rc);
2293}
2294
2295static int mdc_import_event(struct obd_device *obd, struct obd_import *imp,
2296 enum obd_import_event event)
2297{
2298 int rc = 0;
2299
2300 LASSERT(imp->imp_obd == obd);
2301
2302 switch (event) {
2303 case IMP_EVENT_DISCON: {
2304#if 0
2305 /* XXX Pass event up to OBDs stack. used only for FLD now */
2306 rc = obd_notify_observer(obd, obd, OBD_NOTIFY_DISCON, NULL);
2307#endif
2308 break;
2309 }
2310 case IMP_EVENT_INACTIVE: {
2311 struct client_obd *cli = &obd->u.cli;
2312 /*
2313 * Flush current sequence to make client obtain new one
2314 * from server in case of disconnect/reconnect.
2315 */
2316 if (cli->cl_seq != NULL)
2317 seq_client_flush(cli->cl_seq);
2318
2319 rc = obd_notify_observer(obd, obd, OBD_NOTIFY_INACTIVE, NULL);
2320 break;
2321 }
2322 case IMP_EVENT_INVALIDATE: {
2323 struct ldlm_namespace *ns = obd->obd_namespace;
2324
2325 ldlm_namespace_cleanup(ns, LDLM_FL_LOCAL_ONLY);
2326
2327 break;
2328 }
2329 case IMP_EVENT_ACTIVE:
2330 rc = obd_notify_observer(obd, obd, OBD_NOTIFY_ACTIVE, NULL);
2331 /* restore re-establish kuc registration after reconnecting */
2332 if (rc == 0)
2333 rc = mdc_kuc_reregister(imp);
2334 break;
2335 case IMP_EVENT_OCD:
2336 rc = obd_notify_observer(obd, obd, OBD_NOTIFY_OCD, NULL);
2337 break;
2338 case IMP_EVENT_DEACTIVATE:
2339 case IMP_EVENT_ACTIVATE:
2340 break;
2341 default:
2342 CERROR("Unknown import event %x\n", event);
2343 LBUG();
2344 }
2345 RETURN(rc);
2346}
2347
2348int mdc_fid_alloc(struct obd_export *exp, struct lu_fid *fid,
2349 struct md_op_data *op_data)
2350{
2351 struct client_obd *cli = &exp->exp_obd->u.cli;
2352 struct lu_client_seq *seq = cli->cl_seq;
2353 ENTRY;
2354 RETURN(seq_client_alloc_fid(NULL, seq, fid));
2355}
2356
2357struct obd_uuid *mdc_get_uuid(struct obd_export *exp) {
2358 struct client_obd *cli = &exp->exp_obd->u.cli;
2359 return &cli->cl_target_uuid;
2360}
2361
2362/**
2363 * Determine whether the lock can be canceled before replaying it during
2364 * recovery, non zero value will be return if the lock can be canceled,
2365 * or zero returned for not
2366 */
2367static int mdc_cancel_for_recovery(struct ldlm_lock *lock)
2368{
2369 if (lock->l_resource->lr_type != LDLM_IBITS)
2370 RETURN(0);
2371
2372 /* FIXME: if we ever get into a situation where there are too many
2373 * opened files with open locks on a single node, then we really
2374 * should replay these open locks to reget it */
2375 if (lock->l_policy_data.l_inodebits.bits & MDS_INODELOCK_OPEN)
2376 RETURN(0);
2377
2378 RETURN(1);
2379}
2380
2381static int mdc_resource_inode_free(struct ldlm_resource *res)
2382{
2383 if (res->lr_lvb_inode)
2384 res->lr_lvb_inode = NULL;
2385
2386 return 0;
2387}
2388
2389struct ldlm_valblock_ops inode_lvbo = {
2390 lvbo_free: mdc_resource_inode_free
2391};
2392
2393static int mdc_setup(struct obd_device *obd, struct lustre_cfg *cfg)
2394{
2395 struct client_obd *cli = &obd->u.cli;
2396 struct lprocfs_static_vars lvars = { 0 };
2397 int rc;
2398 ENTRY;
2399
2400 OBD_ALLOC(cli->cl_rpc_lock, sizeof (*cli->cl_rpc_lock));
2401 if (!cli->cl_rpc_lock)
2402 RETURN(-ENOMEM);
2403 mdc_init_rpc_lock(cli->cl_rpc_lock);
2404
2405 ptlrpcd_addref();
2406
2407 OBD_ALLOC(cli->cl_close_lock, sizeof (*cli->cl_close_lock));
2408 if (!cli->cl_close_lock)
2409 GOTO(err_rpc_lock, rc = -ENOMEM);
2410 mdc_init_rpc_lock(cli->cl_close_lock);
2411
2412 rc = client_obd_setup(obd, cfg);
2413 if (rc)
2414 GOTO(err_close_lock, rc);
2415 lprocfs_mdc_init_vars(&lvars);
2416 lprocfs_obd_setup(obd, lvars.obd_vars);
2417 sptlrpc_lprocfs_cliobd_attach(obd);
2418 ptlrpc_lprocfs_register_obd(obd);
2419
2420 ns_register_cancel(obd->obd_namespace, mdc_cancel_for_recovery);
2421
2422 obd->obd_namespace->ns_lvbo = &inode_lvbo;
2423
2424 rc = obd_llog_init(obd, &obd->obd_olg, obd, NULL);
2425 if (rc) {
2426 mdc_cleanup(obd);
2427 CERROR("failed to setup llogging subsystems\n");
2428 }
2429
2430 RETURN(rc);
2431
2432err_close_lock:
2433 OBD_FREE(cli->cl_close_lock, sizeof (*cli->cl_close_lock));
2434err_rpc_lock:
2435 OBD_FREE(cli->cl_rpc_lock, sizeof (*cli->cl_rpc_lock));
2436 ptlrpcd_decref();
2437 RETURN(rc);
2438}
2439
2440/* Initialize the default and maximum LOV EA and cookie sizes. This allows
2441 * us to make MDS RPCs with large enough reply buffers to hold the
2442 * maximum-sized (= maximum striped) EA and cookie without having to
2443 * calculate this (via a call into the LOV + OSCs) each time we make an RPC. */
2444static int mdc_init_ea_size(struct obd_export *exp, int easize,
2445 int def_easize, int cookiesize)
2446{
2447 struct obd_device *obd = exp->exp_obd;
2448 struct client_obd *cli = &obd->u.cli;
2449 ENTRY;
2450
2451 if (cli->cl_max_mds_easize < easize)
2452 cli->cl_max_mds_easize = easize;
2453
2454 if (cli->cl_default_mds_easize < def_easize)
2455 cli->cl_default_mds_easize = def_easize;
2456
2457 if (cli->cl_max_mds_cookiesize < cookiesize)
2458 cli->cl_max_mds_cookiesize = cookiesize;
2459
2460 RETURN(0);
2461}
2462
2463static int mdc_precleanup(struct obd_device *obd, enum obd_cleanup_stage stage)
2464{
2465 int rc = 0;
2466 ENTRY;
2467
2468 switch (stage) {
2469 case OBD_CLEANUP_EARLY:
2470 break;
2471 case OBD_CLEANUP_EXPORTS:
2472 /* Failsafe, ok if racy */
2473 if (obd->obd_type->typ_refcnt <= 1)
2474 libcfs_kkuc_group_rem(0, KUC_GRP_HSM);
2475
2476 obd_cleanup_client_import(obd);
2477 ptlrpc_lprocfs_unregister_obd(obd);
2478 lprocfs_obd_cleanup(obd);
2479
2480 rc = obd_llog_finish(obd, 0);
2481 if (rc != 0)
2482 CERROR("failed to cleanup llogging subsystems\n");
2483 break;
2484 }
2485 RETURN(rc);
2486}
2487
2488static int mdc_cleanup(struct obd_device *obd)
2489{
2490 struct client_obd *cli = &obd->u.cli;
2491
2492 OBD_FREE(cli->cl_rpc_lock, sizeof (*cli->cl_rpc_lock));
2493 OBD_FREE(cli->cl_close_lock, sizeof (*cli->cl_close_lock));
2494
2495 ptlrpcd_decref();
2496
2497 return client_obd_cleanup(obd);
2498}
2499
2500
2501static int mdc_llog_init(struct obd_device *obd, struct obd_llog_group *olg,
2502 struct obd_device *tgt, int *index)
2503{
2504 struct llog_ctxt *ctxt;
2505 int rc;
2506
2507 ENTRY;
2508
2509 LASSERT(olg == &obd->obd_olg);
2510
2511 rc = llog_setup(NULL, obd, olg, LLOG_CHANGELOG_REPL_CTXT, tgt,
2512 &llog_client_ops);
2513 if (rc)
2514 RETURN(rc);
2515
2516 ctxt = llog_group_get_ctxt(olg, LLOG_CHANGELOG_REPL_CTXT);
2517 llog_initiator_connect(ctxt);
2518 llog_ctxt_put(ctxt);
2519
2520 RETURN(0);
2521}
2522
2523static int mdc_llog_finish(struct obd_device *obd, int count)
2524{
2525 struct llog_ctxt *ctxt;
2526
2527 ENTRY;
2528
2529 ctxt = llog_get_context(obd, LLOG_CHANGELOG_REPL_CTXT);
2530 if (ctxt)
2531 llog_cleanup(NULL, ctxt);
2532
2533 RETURN(0);
2534}
2535
2536static int mdc_process_config(struct obd_device *obd, obd_count len, void *buf)
2537{
2538 struct lustre_cfg *lcfg = buf;
2539 struct lprocfs_static_vars lvars = { 0 };
2540 int rc = 0;
2541
2542 lprocfs_mdc_init_vars(&lvars);
2543 switch (lcfg->lcfg_command) {
2544 default:
2545 rc = class_process_proc_param(PARAM_MDC, lvars.obd_vars,
2546 lcfg, obd);
2547 if (rc > 0)
2548 rc = 0;
2549 break;
2550 }
2551 return(rc);
2552}
2553
2554
2555/* get remote permission for current user on fid */
2556int mdc_get_remote_perm(struct obd_export *exp, const struct lu_fid *fid,
2557 struct obd_capa *oc, __u32 suppgid,
2558 struct ptlrpc_request **request)
2559{
2560 struct ptlrpc_request *req;
2561 int rc;
2562 ENTRY;
2563
2564 LASSERT(client_is_remote(exp));
2565
2566 *request = NULL;
2567 req = ptlrpc_request_alloc(class_exp2cliimp(exp), &RQF_MDS_GETATTR);
2568 if (req == NULL)
2569 RETURN(-ENOMEM);
2570
2571 mdc_set_capa_size(req, &RMF_CAPA1, oc);
2572
2573 rc = ptlrpc_request_pack(req, LUSTRE_MDS_VERSION, MDS_GETATTR);
2574 if (rc) {
2575 ptlrpc_request_free(req);
2576 RETURN(rc);
2577 }
2578
2579 mdc_pack_body(req, fid, oc, OBD_MD_FLRMTPERM, 0, suppgid, 0);
2580
2581 req_capsule_set_size(&req->rq_pill, &RMF_ACL, RCL_SERVER,
2582 sizeof(struct mdt_remote_perm));
2583
2584 ptlrpc_request_set_replen(req);
2585
2586 rc = ptlrpc_queue_wait(req);
2587 if (rc)
2588 ptlrpc_req_finished(req);
2589 else
2590 *request = req;
2591 RETURN(rc);
2592}
2593
2594static int mdc_interpret_renew_capa(const struct lu_env *env,
2595 struct ptlrpc_request *req, void *args,
2596 int status)
2597{
2598 struct mdc_renew_capa_args *ra = args;
2599 struct mdt_body *body = NULL;
2600 struct lustre_capa *capa;
2601 ENTRY;
2602
2603 if (status)
2604 GOTO(out, capa = ERR_PTR(status));
2605
2606 body = req_capsule_server_get(&req->rq_pill, &RMF_MDT_BODY);
2607 if (body == NULL)
2608 GOTO(out, capa = ERR_PTR(-EFAULT));
2609
2610 if ((body->valid & OBD_MD_FLOSSCAPA) == 0)
2611 GOTO(out, capa = ERR_PTR(-ENOENT));
2612
2613 capa = req_capsule_server_get(&req->rq_pill, &RMF_CAPA2);
2614 if (!capa)
2615 GOTO(out, capa = ERR_PTR(-EFAULT));
2616 EXIT;
2617out:
2618 ra->ra_cb(ra->ra_oc, capa);
2619 return 0;
2620}
2621
2622static int mdc_renew_capa(struct obd_export *exp, struct obd_capa *oc,
2623 renew_capa_cb_t cb)
2624{
2625 struct ptlrpc_request *req;
2626 struct mdc_renew_capa_args *ra;
2627 ENTRY;
2628
2629 req = ptlrpc_request_alloc_pack(class_exp2cliimp(exp), &RQF_MDS_GETATTR,
2630 LUSTRE_MDS_VERSION, MDS_GETATTR);
2631 if (req == NULL)
2632 RETURN(-ENOMEM);
2633
2634 /* NB, OBD_MD_FLOSSCAPA is set here, but it doesn't necessarily mean the
2635 * capa to renew is oss capa.
2636 */
2637 mdc_pack_body(req, &oc->c_capa.lc_fid, oc, OBD_MD_FLOSSCAPA, 0, -1, 0);
2638 ptlrpc_request_set_replen(req);
2639
2640 CLASSERT(sizeof(*ra) <= sizeof(req->rq_async_args));
2641 ra = ptlrpc_req_async_args(req);
2642 ra->ra_oc = oc;
2643 ra->ra_cb = cb;
2644 req->rq_interpret_reply = mdc_interpret_renew_capa;
2645 ptlrpcd_add_req(req, PDL_POLICY_LOCAL, -1);
2646 RETURN(0);
2647}
2648
2649static int mdc_connect(const struct lu_env *env,
2650 struct obd_export **exp,
2651 struct obd_device *obd, struct obd_uuid *cluuid,
2652 struct obd_connect_data *data,
2653 void *localdata)
2654{
2655 struct obd_import *imp = obd->u.cli.cl_import;
2656
2657 /* mds-mds import features */
2658 if (data && (data->ocd_connect_flags & OBD_CONNECT_MDS_MDS)) {
2659 spin_lock(&imp->imp_lock);
2660 imp->imp_server_timeout = 1;
2661 spin_unlock(&imp->imp_lock);
2662 imp->imp_client->cli_request_portal = MDS_MDS_PORTAL;
2663 CDEBUG(D_OTHER, "%s: Set 'mds' portal and timeout\n",
2664 obd->obd_name);
2665 }
2666
2667 return client_connect_import(env, exp, obd, cluuid, data, NULL);
2668}
2669
2670struct obd_ops mdc_obd_ops = {
2671 .o_owner = THIS_MODULE,
2672 .o_setup = mdc_setup,
2673 .o_precleanup = mdc_precleanup,
2674 .o_cleanup = mdc_cleanup,
2675 .o_add_conn = client_import_add_conn,
2676 .o_del_conn = client_import_del_conn,
2677 .o_connect = mdc_connect,
2678 .o_disconnect = client_disconnect_export,
2679 .o_iocontrol = mdc_iocontrol,
2680 .o_set_info_async = mdc_set_info_async,
2681 .o_statfs = mdc_statfs,
2682 .o_pin = mdc_pin,
2683 .o_unpin = mdc_unpin,
2684 .o_fid_init = client_fid_init,
2685 .o_fid_fini = client_fid_fini,
2686 .o_fid_alloc = mdc_fid_alloc,
2687 .o_import_event = mdc_import_event,
2688 .o_llog_init = mdc_llog_init,
2689 .o_llog_finish = mdc_llog_finish,
2690 .o_get_info = mdc_get_info,
2691 .o_process_config = mdc_process_config,
2692 .o_get_uuid = mdc_get_uuid,
2693 .o_quotactl = mdc_quotactl,
2694 .o_quotacheck = mdc_quotacheck
2695};
2696
2697struct md_ops mdc_md_ops = {
2698 .m_getstatus = mdc_getstatus,
2699 .m_null_inode = mdc_null_inode,
2700 .m_find_cbdata = mdc_find_cbdata,
2701 .m_close = mdc_close,
2702 .m_create = mdc_create,
2703 .m_done_writing = mdc_done_writing,
2704 .m_enqueue = mdc_enqueue,
2705 .m_getattr = mdc_getattr,
2706 .m_getattr_name = mdc_getattr_name,
2707 .m_intent_lock = mdc_intent_lock,
2708 .m_link = mdc_link,
2709 .m_is_subdir = mdc_is_subdir,
2710 .m_rename = mdc_rename,
2711 .m_setattr = mdc_setattr,
2712 .m_setxattr = mdc_setxattr,
2713 .m_getxattr = mdc_getxattr,
2714 .m_sync = mdc_sync,
2715 .m_readpage = mdc_readpage,
2716 .m_unlink = mdc_unlink,
2717 .m_cancel_unused = mdc_cancel_unused,
2718 .m_init_ea_size = mdc_init_ea_size,
2719 .m_set_lock_data = mdc_set_lock_data,
2720 .m_lock_match = mdc_lock_match,
2721 .m_get_lustre_md = mdc_get_lustre_md,
2722 .m_free_lustre_md = mdc_free_lustre_md,
2723 .m_set_open_replay_data = mdc_set_open_replay_data,
2724 .m_clear_open_replay_data = mdc_clear_open_replay_data,
2725 .m_renew_capa = mdc_renew_capa,
2726 .m_unpack_capa = mdc_unpack_capa,
2727 .m_get_remote_perm = mdc_get_remote_perm,
2728 .m_intent_getattr_async = mdc_intent_getattr_async,
2729 .m_revalidate_lock = mdc_revalidate_lock
2730};
2731
2732int __init mdc_init(void)
2733{
2734 int rc;
2735 struct lprocfs_static_vars lvars = { 0 };
2736 lprocfs_mdc_init_vars(&lvars);
2737
2738 rc = class_register_type(&mdc_obd_ops, &mdc_md_ops, lvars.module_vars,
2739 LUSTRE_MDC_NAME, NULL);
2740 RETURN(rc);
2741}
2742
2743static void /*__exit*/ mdc_exit(void)
2744{
2745 class_unregister_type(LUSTRE_MDC_NAME);
2746}
2747
2748MODULE_AUTHOR("Sun Microsystems, Inc. <http://www.lustre.org/>");
2749MODULE_DESCRIPTION("Lustre Metadata Client");
2750MODULE_LICENSE("GPL");
2751
2752module_init(mdc_init);
2753module_exit(mdc_exit);
This page took 0.182108 seconds and 5 git commands to generate.