]> git.proxmox.com Git - mirror_ubuntu-eoan-kernel.git/blob - fs/ceph/mds_client.c
projects / mirror_ubuntu-eoan-kernel.git / blob
? search:


f1590256d2e68258fad9e0cb368f42108d6f16aa
[mirror_ubuntu-eoan-kernel.git] / fs / ceph / mds_client.c
1 // SPDX-License-Identifier: GPL-2.0
2 #include <linux/ceph/ceph_debug.h>
3
4 #include <linux/fs.h>
5 #include <linux/wait.h>
6 #include <linux/slab.h>
7 #include <linux/gfp.h>
8 #include <linux/sched.h>
9 #include <linux/debugfs.h>
10 #include <linux/seq_file.h>
11 #include <linux/ratelimit.h>
12
13 #include "super.h"
14 #include "mds_client.h"
15
16 #include <linux/ceph/ceph_features.h>
17 #include <linux/ceph/messenger.h>
18 #include <linux/ceph/decode.h>
19 #include <linux/ceph/pagelist.h>
20 #include <linux/ceph/auth.h>
21 #include <linux/ceph/debugfs.h>
22
23 /*
24 * A cluster of MDS (metadata server) daemons is responsible for
25 * managing the file system namespace (the directory hierarchy and
26 * inodes) and for coordinating shared access to storage. Metadata is
27 * partitioning hierarchically across a number of servers, and that
28 * partition varies over time as the cluster adjusts the distribution
29 * in order to balance load.
30 *
31 * The MDS client is primarily responsible to managing synchronous
32 * metadata requests for operations like open, unlink, and so forth.
33 * If there is a MDS failure, we find out about it when we (possibly
34 * request and) receive a new MDS map, and can resubmit affected
35 * requests.
36 *
37 * For the most part, though, we take advantage of a lossless
38 * communications channel to the MDS, and do not need to worry about
39 * timing out or resubmitting requests.
40 *
41 * We maintain a stateful "session" with each MDS we interact with.
42 * Within each session, we sent periodic heartbeat messages to ensure
43 * any capabilities or leases we have been issues remain valid. If
44 * the session times out and goes stale, our leases and capabilities
45 * are no longer valid.
46 */
47
48 struct ceph_reconnect_state {
49 int nr_caps;
50 struct ceph_pagelist *pagelist;
51 unsigned msg_version;
52 };
53
54 static void __wake_requests(struct ceph_mds_client *mdsc,
55 struct list_head *head);
56
57 static const struct ceph_connection_operations mds_con_ops;
58
59
60 /*
61 * mds reply parsing
62 */
63
64 /*
65 * parse individual inode info
66 */
67 static int parse_reply_info_in(void **p, void *end,
68 struct ceph_mds_reply_info_in *info,
69 u64 features)
70 {
71 int err = -EIO;
72
73 info->in = *p;
74 *p += sizeof(struct ceph_mds_reply_inode) +
75 sizeof(*info->in->fragtree.splits) *
76 le32_to_cpu(info->in->fragtree.nsplits);
77
78 ceph_decode_32_safe(p, end, info->symlink_len, bad);
79 ceph_decode_need(p, end, info->symlink_len, bad);
80 info->symlink = *p;
81 *p += info->symlink_len;
82
83 if (features & CEPH_FEATURE_DIRLAYOUTHASH)
84 ceph_decode_copy_safe(p, end, &info->dir_layout,
85 sizeof(info->dir_layout), bad);
86 else
87 memset(&info->dir_layout, 0, sizeof(info->dir_layout));
88
89 ceph_decode_32_safe(p, end, info->xattr_len, bad);
90 ceph_decode_need(p, end, info->xattr_len, bad);
91 info->xattr_data = *p;
92 *p += info->xattr_len;
93
94 if (features & CEPH_FEATURE_MDS_INLINE_DATA) {
95 ceph_decode_64_safe(p, end, info->inline_version, bad);
96 ceph_decode_32_safe(p, end, info->inline_len, bad);
97 ceph_decode_need(p, end, info->inline_len, bad);
98 info->inline_data = *p;
99 *p += info->inline_len;
100 } else
101 info->inline_version = CEPH_INLINE_NONE;
102
103 if (features & CEPH_FEATURE_MDS_QUOTA) {
104 u8 struct_v, struct_compat;
105 u32 struct_len;
106
107 /*
108 * both struct_v and struct_compat are expected to be >= 1
109 */
110 ceph_decode_8_safe(p, end, struct_v, bad);
111 ceph_decode_8_safe(p, end, struct_compat, bad);
112 if (!struct_v || !struct_compat)
113 goto bad;
114 ceph_decode_32_safe(p, end, struct_len, bad);
115 ceph_decode_need(p, end, struct_len, bad);
116 ceph_decode_64_safe(p, end, info->max_bytes, bad);
117 ceph_decode_64_safe(p, end, info->max_files, bad);
118 } else {
119 info->max_bytes = 0;
120 info->max_files = 0;
121 }
122
123 info->pool_ns_len = 0;
124 info->pool_ns_data = NULL;
125 if (features & CEPH_FEATURE_FS_FILE_LAYOUT_V2) {
126 ceph_decode_32_safe(p, end, info->pool_ns_len, bad);
127 if (info->pool_ns_len > 0) {
128 ceph_decode_need(p, end, info->pool_ns_len, bad);
129 info->pool_ns_data = *p;
130 *p += info->pool_ns_len;
131 }
132 }
133
134 return 0;
135 bad:
136 return err;
137 }
138
139 /*
140 * parse a normal reply, which may contain a (dir+)dentry and/or a
141 * target inode.
142 */
143 static int parse_reply_info_trace(void **p, void *end,
144 struct ceph_mds_reply_info_parsed *info,
145 u64 features)
146 {
147 int err;
148
149 if (info->head->is_dentry) {
150 err = parse_reply_info_in(p, end, &info->diri, features);
151 if (err < 0)
152 goto out_bad;
153
154 if (unlikely(*p + sizeof(*info->dirfrag) > end))
155 goto bad;
156 info->dirfrag = *p;
157 *p += sizeof(*info->dirfrag) +
158 sizeof(u32)*le32_to_cpu(info->dirfrag->ndist);
159 if (unlikely(*p > end))
160 goto bad;
161
162 ceph_decode_32_safe(p, end, info->dname_len, bad);
163 ceph_decode_need(p, end, info->dname_len, bad);
164 info->dname = *p;
165 *p += info->dname_len;
166 info->dlease = *p;
167 *p += sizeof(*info->dlease);
168 }
169
170 if (info->head->is_target) {
171 err = parse_reply_info_in(p, end, &info->targeti, features);
172 if (err < 0)
173 goto out_bad;
174 }
175
176 if (unlikely(*p != end))
177 goto bad;
178 return 0;
179
180 bad:
181 err = -EIO;
182 out_bad:
183 pr_err("problem parsing mds trace %d\n", err);
184 return err;
185 }
186
187 /*
188 * parse readdir results
189 */
190 static int parse_reply_info_dir(void **p, void *end,
191 struct ceph_mds_reply_info_parsed *info,
192 u64 features)
193 {
194 u32 num, i = 0;
195 int err;
196
197 info->dir_dir = *p;
198 if (*p + sizeof(*info->dir_dir) > end)
199 goto bad;
200 *p += sizeof(*info->dir_dir) +
201 sizeof(u32)*le32_to_cpu(info->dir_dir->ndist);
202 if (*p > end)
203 goto bad;
204
205 ceph_decode_need(p, end, sizeof(num) + 2, bad);
206 num = ceph_decode_32(p);
207 {
208 u16 flags = ceph_decode_16(p);
209 info->dir_end = !!(flags & CEPH_READDIR_FRAG_END);
210 info->dir_complete = !!(flags & CEPH_READDIR_FRAG_COMPLETE);
211 info->hash_order = !!(flags & CEPH_READDIR_HASH_ORDER);
212 info->offset_hash = !!(flags & CEPH_READDIR_OFFSET_HASH);
213 }
214 if (num == 0)
215 goto done;
216
217 BUG_ON(!info->dir_entries);
218 if ((unsigned long)(info->dir_entries + num) >
219 (unsigned long)info->dir_entries + info->dir_buf_size) {
220 pr_err("dir contents are larger than expected\n");
221 WARN_ON(1);
222 goto bad;
223 }
224
225 info->dir_nr = num;
226 while (num) {
227 struct ceph_mds_reply_dir_entry *rde = info->dir_entries + i;
228 /* dentry */
229 ceph_decode_need(p, end, sizeof(u32)*2, bad);
230 rde->name_len = ceph_decode_32(p);
231 ceph_decode_need(p, end, rde->name_len, bad);
232 rde->name = *p;
233 *p += rde->name_len;
234 dout("parsed dir dname '%.*s'\n", rde->name_len, rde->name);
235 rde->lease = *p;
236 *p += sizeof(struct ceph_mds_reply_lease);
237
238 /* inode */
239 err = parse_reply_info_in(p, end, &rde->inode, features);
240 if (err < 0)
241 goto out_bad;
242 /* ceph_readdir_prepopulate() will update it */
243 rde->offset = 0;
244 i++;
245 num--;
246 }
247
248 done:
249 if (*p != end)
250 goto bad;
251 return 0;
252
253 bad:
254 err = -EIO;
255 out_bad:
256 pr_err("problem parsing dir contents %d\n", err);
257 return err;
258 }
259
260 /*
261 * parse fcntl F_GETLK results
262 */
263 static int parse_reply_info_filelock(void **p, void *end,
264 struct ceph_mds_reply_info_parsed *info,
265 u64 features)
266 {
267 if (*p + sizeof(*info->filelock_reply) > end)
268 goto bad;
269
270 info->filelock_reply = *p;
271 *p += sizeof(*info->filelock_reply);
272
273 if (unlikely(*p != end))
274 goto bad;
275 return 0;
276
277 bad:
278 return -EIO;
279 }
280
281 /*
282 * parse create results
283 */
284 static int parse_reply_info_create(void **p, void *end,
285 struct ceph_mds_reply_info_parsed *info,
286 u64 features)
287 {
288 if (features & CEPH_FEATURE_REPLY_CREATE_INODE) {
289 if (*p == end) {
290 info->has_create_ino = false;
291 } else {
292 info->has_create_ino = true;
293 info->ino = ceph_decode_64(p);
294 }
295 }
296
297 if (unlikely(*p != end))
298 goto bad;
299 return 0;
300
301 bad:
302 return -EIO;
303 }
304
305 /*
306 * parse extra results
307 */
308 static int parse_reply_info_extra(void **p, void *end,
309 struct ceph_mds_reply_info_parsed *info,
310 u64 features)
311 {
312 u32 op = le32_to_cpu(info->head->op);
313
314 if (op == CEPH_MDS_OP_GETFILELOCK)
315 return parse_reply_info_filelock(p, end, info, features);
316 else if (op == CEPH_MDS_OP_READDIR || op == CEPH_MDS_OP_LSSNAP)
317 return parse_reply_info_dir(p, end, info, features);
318 else if (op == CEPH_MDS_OP_CREATE)
319 return parse_reply_info_create(p, end, info, features);
320 else
321 return -EIO;
322 }
323
324 /*
325 * parse entire mds reply
326 */
327 static int parse_reply_info(struct ceph_msg *msg,
328 struct ceph_mds_reply_info_parsed *info,
329 u64 features)
330 {
331 void *p, *end;
332 u32 len;
333 int err;
334
335 info->head = msg->front.iov_base;
336 p = msg->front.iov_base + sizeof(struct ceph_mds_reply_head);
337 end = p + msg->front.iov_len - sizeof(struct ceph_mds_reply_head);
338
339 /* trace */
340 ceph_decode_32_safe(&p, end, len, bad);
341 if (len > 0) {
342 ceph_decode_need(&p, end, len, bad);
343 err = parse_reply_info_trace(&p, p+len, info, features);
344 if (err < 0)
345 goto out_bad;
346 }
347
348 /* extra */
349 ceph_decode_32_safe(&p, end, len, bad);
350 if (len > 0) {
351 ceph_decode_need(&p, end, len, bad);
352 err = parse_reply_info_extra(&p, p+len, info, features);
353 if (err < 0)
354 goto out_bad;
355 }
356
357 /* snap blob */
358 ceph_decode_32_safe(&p, end, len, bad);
359 info->snapblob_len = len;
360 info->snapblob = p;
361 p += len;
362
363 if (p != end)
364 goto bad;
365 return 0;
366
367 bad:
368 err = -EIO;
369 out_bad:
370 pr_err("mds parse_reply err %d\n", err);
371 return err;
372 }
373
374 static void destroy_reply_info(struct ceph_mds_reply_info_parsed *info)
375 {
376 if (!info->dir_entries)
377 return;
378 free_pages((unsigned long)info->dir_entries, get_order(info->dir_buf_size));
379 }
380
381
382 /*
383 * sessions
384 */
385 const char *ceph_session_state_name(int s)
386 {
387 switch (s) {
388 case CEPH_MDS_SESSION_NEW: return "new";
389 case CEPH_MDS_SESSION_OPENING: return "opening";
390 case CEPH_MDS_SESSION_OPEN: return "open";
391 case CEPH_MDS_SESSION_HUNG: return "hung";
392 case CEPH_MDS_SESSION_CLOSING: return "closing";
393 case CEPH_MDS_SESSION_RESTARTING: return "restarting";
394 case CEPH_MDS_SESSION_RECONNECTING: return "reconnecting";
395 case CEPH_MDS_SESSION_REJECTED: return "rejected";
396 default: return "???";
397 }
398 }
399
400 static struct ceph_mds_session *get_session(struct ceph_mds_session *s)
401 {
402 if (refcount_inc_not_zero(&s->s_ref)) {
403 dout("mdsc get_session %p %d -> %d\n", s,
404 refcount_read(&s->s_ref)-1, refcount_read(&s->s_ref));
405 return s;
406 } else {
407 dout("mdsc get_session %p 0 -- FAIL\n", s);
408 return NULL;
409 }
410 }
411
412 void ceph_put_mds_session(struct ceph_mds_session *s)
413 {
414 dout("mdsc put_session %p %d -> %d\n", s,
415 refcount_read(&s->s_ref), refcount_read(&s->s_ref)-1);
416 if (refcount_dec_and_test(&s->s_ref)) {
417 if (s->s_auth.authorizer)
418 ceph_auth_destroy_authorizer(s->s_auth.authorizer);
419 kfree(s);
420 }
421 }
422
423 /*
424 * called under mdsc->mutex
425 */
426 struct ceph_mds_session *__ceph_lookup_mds_session(struct ceph_mds_client *mdsc,
427 int mds)
428 {
429 struct ceph_mds_session *session;
430
431 if (mds >= mdsc->max_sessions || !mdsc->sessions[mds])
432 return NULL;
433 session = mdsc->sessions[mds];
434 dout("lookup_mds_session %p %d\n", session,
435 refcount_read(&session->s_ref));
436 get_session(session);
437 return session;
438 }
439
440 static bool __have_session(struct ceph_mds_client *mdsc, int mds)
441 {
442 if (mds >= mdsc->max_sessions || !mdsc->sessions[mds])
443 return false;
444 else
445 return true;
446 }
447
448 static int __verify_registered_session(struct ceph_mds_client *mdsc,
449 struct ceph_mds_session *s)
450 {
451 if (s->s_mds >= mdsc->max_sessions ||
452 mdsc->sessions[s->s_mds] != s)
453 return -ENOENT;
454 return 0;
455 }
456
457 /*
458 * create+register a new session for given mds.
459 * called under mdsc->mutex.
460 */
461 static struct ceph_mds_session *register_session(struct ceph_mds_client *mdsc,
462 int mds)
463 {
464 struct ceph_mds_session *s;
465
466 if (mds >= mdsc->mdsmap->m_num_mds)
467 return ERR_PTR(-EINVAL);
468
469 s = kzalloc(sizeof(*s), GFP_NOFS);
470 if (!s)
471 return ERR_PTR(-ENOMEM);
472
473 if (mds >= mdsc->max_sessions) {
474 int newmax = 1 << get_count_order(mds + 1);
475 struct ceph_mds_session **sa;
476
477 dout("%s: realloc to %d\n", __func__, newmax);
478 sa = kcalloc(newmax, sizeof(void *), GFP_NOFS);
479 if (!sa)
480 goto fail_realloc;
481 if (mdsc->sessions) {
482 memcpy(sa, mdsc->sessions,
483 mdsc->max_sessions * sizeof(void *));
484 kfree(mdsc->sessions);
485 }
486 mdsc->sessions = sa;
487 mdsc->max_sessions = newmax;
488 }
489
490 dout("%s: mds%d\n", __func__, mds);
491 s->s_mdsc = mdsc;
492 s->s_mds = mds;
493 s->s_state = CEPH_MDS_SESSION_NEW;
494 s->s_ttl = 0;
495 s->s_seq = 0;
496 mutex_init(&s->s_mutex);
497
498 ceph_con_init(&s->s_con, s, &mds_con_ops, &mdsc->fsc->client->msgr);
499
500 spin_lock_init(&s->s_gen_ttl_lock);
501 s->s_cap_gen = 0;
502 s->s_cap_ttl = jiffies - 1;
503
504 spin_lock_init(&s->s_cap_lock);
505 s->s_renew_requested = 0;
506 s->s_renew_seq = 0;
507 INIT_LIST_HEAD(&s->s_caps);
508 s->s_nr_caps = 0;
509 s->s_trim_caps = 0;
510 refcount_set(&s->s_ref, 1);
511 INIT_LIST_HEAD(&s->s_waiting);
512 INIT_LIST_HEAD(&s->s_unsafe);
513 s->s_num_cap_releases = 0;
514 s->s_cap_reconnect = 0;
515 s->s_cap_iterator = NULL;
516 INIT_LIST_HEAD(&s->s_cap_releases);
517 INIT_LIST_HEAD(&s->s_cap_flushing);
518
519 mdsc->sessions[mds] = s;
520 atomic_inc(&mdsc->num_sessions);
521 refcount_inc(&s->s_ref); /* one ref to sessions[], one to caller */
522
523 ceph_con_open(&s->s_con, CEPH_ENTITY_TYPE_MDS, mds,
524 ceph_mdsmap_get_addr(mdsc->mdsmap, mds));
525
526 return s;
527
528 fail_realloc:
529 kfree(s);
530 return ERR_PTR(-ENOMEM);
531 }
532
533 /*
534 * called under mdsc->mutex
535 */
536 static void __unregister_session(struct ceph_mds_client *mdsc,
537 struct ceph_mds_session *s)
538 {
539 dout("__unregister_session mds%d %p\n", s->s_mds, s);
540 BUG_ON(mdsc->sessions[s->s_mds] != s);
541 mdsc->sessions[s->s_mds] = NULL;
542 ceph_con_close(&s->s_con);
543 ceph_put_mds_session(s);
544 atomic_dec(&mdsc->num_sessions);
545 }
546
547 /*
548 * drop session refs in request.
549 *
550 * should be last request ref, or hold mdsc->mutex
551 */
552 static void put_request_session(struct ceph_mds_request *req)
553 {
554 if (req->r_session) {
555 ceph_put_mds_session(req->r_session);
556 req->r_session = NULL;
557 }
558 }
559
560 void ceph_mdsc_release_request(struct kref *kref)
561 {
562 struct ceph_mds_request *req = container_of(kref,
563 struct ceph_mds_request,
564 r_kref);
565 destroy_reply_info(&req->r_reply_info);
566 if (req->r_request)
567 ceph_msg_put(req->r_request);
568 if (req->r_reply)
569 ceph_msg_put(req->r_reply);
570 if (req->r_inode) {
571 ceph_put_cap_refs(ceph_inode(req->r_inode), CEPH_CAP_PIN);
572 iput(req->r_inode);
573 }
574 if (req->r_parent)
575 ceph_put_cap_refs(ceph_inode(req->r_parent), CEPH_CAP_PIN);
576 iput(req->r_target_inode);
577 if (req->r_dentry)
578 dput(req->r_dentry);
579 if (req->r_old_dentry)
580 dput(req->r_old_dentry);
581 if (req->r_old_dentry_dir) {
582 /*
583 * track (and drop pins for) r_old_dentry_dir
584 * separately, since r_old_dentry's d_parent may have
585 * changed between the dir mutex being dropped and
586 * this request being freed.
587 */
588 ceph_put_cap_refs(ceph_inode(req->r_old_dentry_dir),
589 CEPH_CAP_PIN);
590 iput(req->r_old_dentry_dir);
591 }
592 kfree(req->r_path1);
593 kfree(req->r_path2);
594 if (req->r_pagelist)
595 ceph_pagelist_release(req->r_pagelist);
596 put_request_session(req);
597 ceph_unreserve_caps(req->r_mdsc, &req->r_caps_reservation);
598 kfree(req);
599 }
600
601 DEFINE_RB_FUNCS(request, struct ceph_mds_request, r_tid, r_node)
602
603 /*
604 * lookup session, bump ref if found.
605 *
606 * called under mdsc->mutex.
607 */
608 static struct ceph_mds_request *
609 lookup_get_request(struct ceph_mds_client *mdsc, u64 tid)
610 {
611 struct ceph_mds_request *req;
612
613 req = lookup_request(&mdsc->request_tree, tid);
614 if (req)
615 ceph_mdsc_get_request(req);
616
617 return req;
618 }
619
620 /*
621 * Register an in-flight request, and assign a tid. Link to directory
622 * are modifying (if any).
623 *
624 * Called under mdsc->mutex.
625 */
626 static void __register_request(struct ceph_mds_client *mdsc,
627 struct ceph_mds_request *req,
628 struct inode *dir)
629 {
630 int ret = 0;
631
632 req->r_tid = ++mdsc->last_tid;
633 if (req->r_num_caps) {
634 ret = ceph_reserve_caps(mdsc, &req->r_caps_reservation,
635 req->r_num_caps);
636 if (ret < 0) {
637 pr_err("__register_request %p "
638 "failed to reserve caps: %d\n", req, ret);
639 /* set req->r_err to fail early from __do_request */
640 req->r_err = ret;
641 return;
642 }
643 }
644 dout("__register_request %p tid %lld\n", req, req->r_tid);
645 ceph_mdsc_get_request(req);
646 insert_request(&mdsc->request_tree, req);
647
648 req->r_uid = current_fsuid();
649 req->r_gid = current_fsgid();
650
651 if (mdsc->oldest_tid == 0 && req->r_op != CEPH_MDS_OP_SETFILELOCK)
652 mdsc->oldest_tid = req->r_tid;
653
654 if (dir) {
655 ihold(dir);
656 req->r_unsafe_dir = dir;
657 }
658 }
659
660 static void __unregister_request(struct ceph_mds_client *mdsc,
661 struct ceph_mds_request *req)
662 {
663 dout("__unregister_request %p tid %lld\n", req, req->r_tid);
664
665 /* Never leave an unregistered request on an unsafe list! */
666 list_del_init(&req->r_unsafe_item);
667
668 if (req->r_tid == mdsc->oldest_tid) {
669 struct rb_node *p = rb_next(&req->r_node);
670 mdsc->oldest_tid = 0;
671 while (p) {
672 struct ceph_mds_request *next_req =
673 rb_entry(p, struct ceph_mds_request, r_node);
674 if (next_req->r_op != CEPH_MDS_OP_SETFILELOCK) {
675 mdsc->oldest_tid = next_req->r_tid;
676 break;
677 }
678 p = rb_next(p);
679 }
680 }
681
682 erase_request(&mdsc->request_tree, req);
683
684 if (req->r_unsafe_dir &&
685 test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags)) {
686 struct ceph_inode_info *ci = ceph_inode(req->r_unsafe_dir);
687 spin_lock(&ci->i_unsafe_lock);
688 list_del_init(&req->r_unsafe_dir_item);
689 spin_unlock(&ci->i_unsafe_lock);
690 }
691 if (req->r_target_inode &&
692 test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags)) {
693 struct ceph_inode_info *ci = ceph_inode(req->r_target_inode);
694 spin_lock(&ci->i_unsafe_lock);
695 list_del_init(&req->r_unsafe_target_item);
696 spin_unlock(&ci->i_unsafe_lock);
697 }
698
699 if (req->r_unsafe_dir) {
700 iput(req->r_unsafe_dir);
701 req->r_unsafe_dir = NULL;
702 }
703
704 complete_all(&req->r_safe_completion);
705
706 ceph_mdsc_put_request(req);
707 }
708
709 /*
710 * Walk back up the dentry tree until we hit a dentry representing a
711 * non-snapshot inode. We do this using the rcu_read_lock (which must be held
712 * when calling this) to ensure that the objects won't disappear while we're
713 * working with them. Once we hit a candidate dentry, we attempt to take a
714 * reference to it, and return that as the result.
715 */
716 static struct inode *get_nonsnap_parent(struct dentry *dentry)
717 {
718 struct inode *inode = NULL;
719
720 while (dentry && !IS_ROOT(dentry)) {
721 inode = d_inode_rcu(dentry);
722 if (!inode || ceph_snap(inode) == CEPH_NOSNAP)
723 break;
724 dentry = dentry->d_parent;
725 }
726 if (inode)
727 inode = igrab(inode);
728 return inode;
729 }
730
731 /*
732 * Choose mds to send request to next. If there is a hint set in the
733 * request (e.g., due to a prior forward hint from the mds), use that.
734 * Otherwise, consult frag tree and/or caps to identify the
735 * appropriate mds. If all else fails, choose randomly.
736 *
737 * Called under mdsc->mutex.
738 */
739 static int __choose_mds(struct ceph_mds_client *mdsc,
740 struct ceph_mds_request *req)
741 {
742 struct inode *inode;
743 struct ceph_inode_info *ci;
744 struct ceph_cap *cap;
745 int mode = req->r_direct_mode;
746 int mds = -1;
747 u32 hash = req->r_direct_hash;
748 bool is_hash = test_bit(CEPH_MDS_R_DIRECT_IS_HASH, &req->r_req_flags);
749
750 /*
751 * is there a specific mds we should try? ignore hint if we have
752 * no session and the mds is not up (active or recovering).
753 */
754 if (req->r_resend_mds >= 0 &&
755 (__have_session(mdsc, req->r_resend_mds) ||
756 ceph_mdsmap_get_state(mdsc->mdsmap, req->r_resend_mds) > 0)) {
757 dout("choose_mds using resend_mds mds%d\n",
758 req->r_resend_mds);
759 return req->r_resend_mds;
760 }
761
762 if (mode == USE_RANDOM_MDS)
763 goto random;
764
765 inode = NULL;
766 if (req->r_inode) {
767 if (ceph_snap(req->r_inode) != CEPH_SNAPDIR) {
768 inode = req->r_inode;
769 ihold(inode);
770 } else {
771 /* req->r_dentry is non-null for LSSNAP request */
772 rcu_read_lock();
773 inode = get_nonsnap_parent(req->r_dentry);
774 rcu_read_unlock();
775 dout("__choose_mds using snapdir's parent %p\n", inode);
776 }
777 } else if (req->r_dentry) {
778 /* ignore race with rename; old or new d_parent is okay */
779 struct dentry *parent;
780 struct inode *dir;
781
782 rcu_read_lock();
783 parent = req->r_dentry->d_parent;
784 dir = req->r_parent ? : d_inode_rcu(parent);
785
786 if (!dir || dir->i_sb != mdsc->fsc->sb) {
787 /* not this fs or parent went negative */
788 inode = d_inode(req->r_dentry);
789 if (inode)
790 ihold(inode);
791 } else if (ceph_snap(dir) != CEPH_NOSNAP) {
792 /* direct snapped/virtual snapdir requests
793 * based on parent dir inode */
794 inode = get_nonsnap_parent(parent);
795 dout("__choose_mds using nonsnap parent %p\n", inode);
796 } else {
797 /* dentry target */
798 inode = d_inode(req->r_dentry);
799 if (!inode || mode == USE_AUTH_MDS) {
800 /* dir + name */
801 inode = igrab(dir);
802 hash = ceph_dentry_hash(dir, req->r_dentry);
803 is_hash = true;
804 } else {
805 ihold(inode);
806 }
807 }
808 rcu_read_unlock();
809 }
810
811 dout("__choose_mds %p is_hash=%d (%d) mode %d\n", inode, (int)is_hash,
812 (int)hash, mode);
813 if (!inode)
814 goto random;
815 ci = ceph_inode(inode);
816
817 if (is_hash && S_ISDIR(inode->i_mode)) {
818 struct ceph_inode_frag frag;
819 int found;
820
821 ceph_choose_frag(ci, hash, &frag, &found);
822 if (found) {
823 if (mode == USE_ANY_MDS && frag.ndist > 0) {
824 u8 r;
825
826 /* choose a random replica */
827 get_random_bytes(&r, 1);
828 r %= frag.ndist;
829 mds = frag.dist[r];
830 dout("choose_mds %p %llx.%llx "
831 "frag %u mds%d (%d/%d)\n",
832 inode, ceph_vinop(inode),
833 frag.frag, mds,
834 (int)r, frag.ndist);
835 if (ceph_mdsmap_get_state(mdsc->mdsmap, mds) >=
836 CEPH_MDS_STATE_ACTIVE)
837 goto out;
838 }
839
840 /* since this file/dir wasn't known to be
841 * replicated, then we want to look for the
842 * authoritative mds. */
843 mode = USE_AUTH_MDS;
844 if (frag.mds >= 0) {
845 /* choose auth mds */
846 mds = frag.mds;
847 dout("choose_mds %p %llx.%llx "
848 "frag %u mds%d (auth)\n",
849 inode, ceph_vinop(inode), frag.frag, mds);
850 if (ceph_mdsmap_get_state(mdsc->mdsmap, mds) >=
851 CEPH_MDS_STATE_ACTIVE)
852 goto out;
853 }
854 }
855 }
856
857 spin_lock(&ci->i_ceph_lock);
858 cap = NULL;
859 if (mode == USE_AUTH_MDS)
860 cap = ci->i_auth_cap;
861 if (!cap && !RB_EMPTY_ROOT(&ci->i_caps))
862 cap = rb_entry(rb_first(&ci->i_caps), struct ceph_cap, ci_node);
863 if (!cap) {
864 spin_unlock(&ci->i_ceph_lock);
865 iput(inode);
866 goto random;
867 }
868 mds = cap->session->s_mds;
869 dout("choose_mds %p %llx.%llx mds%d (%scap %p)\n",
870 inode, ceph_vinop(inode), mds,
871 cap == ci->i_auth_cap ? "auth " : "", cap);
872 spin_unlock(&ci->i_ceph_lock);
873 out:
874 iput(inode);
875 return mds;
876
877 random:
878 mds = ceph_mdsmap_get_random_mds(mdsc->mdsmap);
879 dout("choose_mds chose random mds%d\n", mds);
880 return mds;
881 }
882
883
884 /*
885 * session messages
886 */
887 static struct ceph_msg *create_session_msg(u32 op, u64 seq)
888 {
889 struct ceph_msg *msg;
890 struct ceph_mds_session_head *h;
891
892 msg = ceph_msg_new(CEPH_MSG_CLIENT_SESSION, sizeof(*h), GFP_NOFS,
893 false);
894 if (!msg) {
895 pr_err("create_session_msg ENOMEM creating msg\n");
896 return NULL;
897 }
898 h = msg->front.iov_base;
899 h->op = cpu_to_le32(op);
900 h->seq = cpu_to_le64(seq);
901
902 return msg;
903 }
904
905 /*
906 * session message, specialization for CEPH_SESSION_REQUEST_OPEN
907 * to include additional client metadata fields.
908 */
909 static struct ceph_msg *create_session_open_msg(struct ceph_mds_client *mdsc, u64 seq)
910 {
911 struct ceph_msg *msg;
912 struct ceph_mds_session_head *h;
913 int i = -1;
914 int metadata_bytes = 0;
915 int metadata_key_count = 0;
916 struct ceph_options *opt = mdsc->fsc->client->options;
917 struct ceph_mount_options *fsopt = mdsc->fsc->mount_options;
918 void *p;
919
920 const char* metadata[][2] = {
921 {"hostname", mdsc->nodename},
922 {"kernel_version", init_utsname()->release},
923 {"entity_id", opt->name ? : ""},
924 {"root", fsopt->server_path ? : "/"},
925 {NULL, NULL}
926 };
927
928 /* Calculate serialized length of metadata */
929 metadata_bytes = 4; /* map length */
930 for (i = 0; metadata[i][0]; ++i) {
931 metadata_bytes += 8 + strlen(metadata[i][0]) +
932 strlen(metadata[i][1]);
933 metadata_key_count++;
934 }
935
936 /* Allocate the message */
937 msg = ceph_msg_new(CEPH_MSG_CLIENT_SESSION, sizeof(*h) + metadata_bytes,
938 GFP_NOFS, false);
939 if (!msg) {
940 pr_err("create_session_msg ENOMEM creating msg\n");
941 return NULL;
942 }
943 h = msg->front.iov_base;
944 h->op = cpu_to_le32(CEPH_SESSION_REQUEST_OPEN);
945 h->seq = cpu_to_le64(seq);
946
947 /*
948 * Serialize client metadata into waiting buffer space, using
949 * the format that userspace expects for map<string, string>
950 *
951 * ClientSession messages with metadata are v2
952 */
953 msg->hdr.version = cpu_to_le16(2);
954 msg->hdr.compat_version = cpu_to_le16(1);
955
956 /* The write pointer, following the session_head structure */
957 p = msg->front.iov_base + sizeof(*h);
958
959 /* Number of entries in the map */
960 ceph_encode_32(&p, metadata_key_count);
961
962 /* Two length-prefixed strings for each entry in the map */
963 for (i = 0; metadata[i][0]; ++i) {
964 size_t const key_len = strlen(metadata[i][0]);
965 size_t const val_len = strlen(metadata[i][1]);
966
967 ceph_encode_32(&p, key_len);
968 memcpy(p, metadata[i][0], key_len);
969 p += key_len;
970 ceph_encode_32(&p, val_len);
971 memcpy(p, metadata[i][1], val_len);
972 p += val_len;
973 }
974
975 return msg;
976 }
977
978 /*
979 * send session open request.
980 *
981 * called under mdsc->mutex
982 */
983 static int __open_session(struct ceph_mds_client *mdsc,
984 struct ceph_mds_session *session)
985 {
986 struct ceph_msg *msg;
987 int mstate;
988 int mds = session->s_mds;
989
990 /* wait for mds to go active? */
991 mstate = ceph_mdsmap_get_state(mdsc->mdsmap, mds);
992 dout("open_session to mds%d (%s)\n", mds,
993 ceph_mds_state_name(mstate));
994 session->s_state = CEPH_MDS_SESSION_OPENING;
995 session->s_renew_requested = jiffies;
996
997 /* send connect message */
998 msg = create_session_open_msg(mdsc, session->s_seq);
999 if (!msg)
1000 return -ENOMEM;
1001 ceph_con_send(&session->s_con, msg);
1002 return 0;
1003 }
1004
1005 /*
1006 * open sessions for any export targets for the given mds
1007 *
1008 * called under mdsc->mutex
1009 */
1010 static struct ceph_mds_session *
1011 __open_export_target_session(struct ceph_mds_client *mdsc, int target)
1012 {
1013 struct ceph_mds_session *session;
1014
1015 session = __ceph_lookup_mds_session(mdsc, target);
1016 if (!session) {
1017 session = register_session(mdsc, target);
1018 if (IS_ERR(session))
1019 return session;
1020 }
1021 if (session->s_state == CEPH_MDS_SESSION_NEW ||
1022 session->s_state == CEPH_MDS_SESSION_CLOSING)
1023 __open_session(mdsc, session);
1024
1025 return session;
1026 }
1027
1028 struct ceph_mds_session *
1029 ceph_mdsc_open_export_target_session(struct ceph_mds_client *mdsc, int target)
1030 {
1031 struct ceph_mds_session *session;
1032
1033 dout("open_export_target_session to mds%d\n", target);
1034
1035 mutex_lock(&mdsc->mutex);
1036 session = __open_export_target_session(mdsc, target);
1037 mutex_unlock(&mdsc->mutex);
1038
1039 return session;
1040 }
1041
1042 static void __open_export_target_sessions(struct ceph_mds_client *mdsc,
1043 struct ceph_mds_session *session)
1044 {
1045 struct ceph_mds_info *mi;
1046 struct ceph_mds_session *ts;
1047 int i, mds = session->s_mds;
1048
1049 if (mds >= mdsc->mdsmap->m_num_mds)
1050 return;
1051
1052 mi = &mdsc->mdsmap->m_info[mds];
1053 dout("open_export_target_sessions for mds%d (%d targets)\n",
1054 session->s_mds, mi->num_export_targets);
1055
1056 for (i = 0; i < mi->num_export_targets; i++) {
1057 ts = __open_export_target_session(mdsc, mi->export_targets[i]);
1058 if (!IS_ERR(ts))
1059 ceph_put_mds_session(ts);
1060 }
1061 }
1062
1063 void ceph_mdsc_open_export_target_sessions(struct ceph_mds_client *mdsc,
1064 struct ceph_mds_session *session)
1065 {
1066 mutex_lock(&mdsc->mutex);
1067 __open_export_target_sessions(mdsc, session);
1068 mutex_unlock(&mdsc->mutex);
1069 }
1070
1071 /*
1072 * session caps
1073 */
1074
1075 static void detach_cap_releases(struct ceph_mds_session *session,
1076 struct list_head *target)
1077 {
1078 lockdep_assert_held(&session->s_cap_lock);
1079
1080 list_splice_init(&session->s_cap_releases, target);
1081 session->s_num_cap_releases = 0;
1082 dout("dispose_cap_releases mds%d\n", session->s_mds);
1083 }
1084
1085 static void dispose_cap_releases(struct ceph_mds_client *mdsc,
1086 struct list_head *dispose)
1087 {
1088 while (!list_empty(dispose)) {
1089 struct ceph_cap *cap;
1090 /* zero out the in-progress message */
1091 cap = list_first_entry(dispose, struct ceph_cap, session_caps);
1092 list_del(&cap->session_caps);
1093 ceph_put_cap(mdsc, cap);
1094 }
1095 }
1096
1097 static void cleanup_session_requests(struct ceph_mds_client *mdsc,
1098 struct ceph_mds_session *session)
1099 {
1100 struct ceph_mds_request *req;
1101 struct rb_node *p;
1102
1103 dout("cleanup_session_requests mds%d\n", session->s_mds);
1104 mutex_lock(&mdsc->mutex);
1105 while (!list_empty(&session->s_unsafe)) {
1106 req = list_first_entry(&session->s_unsafe,
1107 struct ceph_mds_request, r_unsafe_item);
1108 pr_warn_ratelimited(" dropping unsafe request %llu\n",
1109 req->r_tid);
1110 __unregister_request(mdsc, req);
1111 }
1112 /* zero r_attempts, so kick_requests() will re-send requests */
1113 p = rb_first(&mdsc->request_tree);
1114 while (p) {
1115 req = rb_entry(p, struct ceph_mds_request, r_node);
1116 p = rb_next(p);
1117 if (req->r_session &&
1118 req->r_session->s_mds == session->s_mds)
1119 req->r_attempts = 0;
1120 }
1121 mutex_unlock(&mdsc->mutex);
1122 }
1123
1124 /*
1125 * Helper to safely iterate over all caps associated with a session, with
1126 * special care taken to handle a racing __ceph_remove_cap().
1127 *
1128 * Caller must hold session s_mutex.
1129 */
1130 static int iterate_session_caps(struct ceph_mds_session *session,
1131 int (*cb)(struct inode *, struct ceph_cap *,
1132 void *), void *arg)
1133 {
1134 struct list_head *p;
1135 struct ceph_cap *cap;
1136 struct inode *inode, *last_inode = NULL;
1137 struct ceph_cap *old_cap = NULL;
1138 int ret;
1139
1140 dout("iterate_session_caps %p mds%d\n", session, session->s_mds);
1141 spin_lock(&session->s_cap_lock);
1142 p = session->s_caps.next;
1143 while (p != &session->s_caps) {
1144 cap = list_entry(p, struct ceph_cap, session_caps);
1145 inode = igrab(&cap->ci->vfs_inode);
1146 if (!inode) {
1147 p = p->next;
1148 continue;
1149 }
1150 session->s_cap_iterator = cap;
1151 spin_unlock(&session->s_cap_lock);
1152
1153 if (last_inode) {
1154 iput(last_inode);
1155 last_inode = NULL;
1156 }
1157 if (old_cap) {
1158 ceph_put_cap(session->s_mdsc, old_cap);
1159 old_cap = NULL;
1160 }
1161
1162 ret = cb(inode, cap, arg);
1163 last_inode = inode;
1164
1165 spin_lock(&session->s_cap_lock);
1166 p = p->next;
1167 if (!cap->ci) {
1168 dout("iterate_session_caps finishing cap %p removal\n",
1169 cap);
1170 BUG_ON(cap->session != session);
1171 cap->session = NULL;
1172 list_del_init(&cap->session_caps);
1173 session->s_nr_caps--;
1174 if (cap->queue_release) {
1175 list_add_tail(&cap->session_caps,
1176 &session->s_cap_releases);
1177 session->s_num_cap_releases++;
1178 } else {
1179 old_cap = cap; /* put_cap it w/o locks held */
1180 }
1181 }
1182 if (ret < 0)
1183 goto out;
1184 }
1185 ret = 0;
1186 out:
1187 session->s_cap_iterator = NULL;
1188 spin_unlock(&session->s_cap_lock);
1189
1190 iput(last_inode);
1191 if (old_cap)
1192 ceph_put_cap(session->s_mdsc, old_cap);
1193
1194 return ret;
1195 }
1196
1197 static int remove_session_caps_cb(struct inode *inode, struct ceph_cap *cap,
1198 void *arg)
1199 {
1200 struct ceph_fs_client *fsc = (struct ceph_fs_client *)arg;
1201 struct ceph_inode_info *ci = ceph_inode(inode);
1202 LIST_HEAD(to_remove);
1203 bool drop = false;
1204 bool invalidate = false;
1205
1206 dout("removing cap %p, ci is %p, inode is %p\n",
1207 cap, ci, &ci->vfs_inode);
1208 spin_lock(&ci->i_ceph_lock);
1209 __ceph_remove_cap(cap, false);
1210 if (!ci->i_auth_cap) {
1211 struct ceph_cap_flush *cf;
1212 struct ceph_mds_client *mdsc = fsc->mdsc;
1213
1214 ci->i_ceph_flags |= CEPH_I_CAP_DROPPED;
1215
1216 if (ci->i_wrbuffer_ref > 0 &&
1217 READ_ONCE(fsc->mount_state) == CEPH_MOUNT_SHUTDOWN)
1218 invalidate = true;
1219
1220 while (!list_empty(&ci->i_cap_flush_list)) {
1221 cf = list_first_entry(&ci->i_cap_flush_list,
1222 struct ceph_cap_flush, i_list);
1223 list_move(&cf->i_list, &to_remove);
1224 }
1225
1226 spin_lock(&mdsc->cap_dirty_lock);
1227
1228 list_for_each_entry(cf, &to_remove, i_list)
1229 list_del(&cf->g_list);
1230
1231 if (!list_empty(&ci->i_dirty_item)) {
1232 pr_warn_ratelimited(
1233 " dropping dirty %s state for %p %lld\n",
1234 ceph_cap_string(ci->i_dirty_caps),
1235 inode, ceph_ino(inode));
1236 ci->i_dirty_caps = 0;
1237 list_del_init(&ci->i_dirty_item);
1238 drop = true;
1239 }
1240 if (!list_empty(&ci->i_flushing_item)) {
1241 pr_warn_ratelimited(
1242 " dropping dirty+flushing %s state for %p %lld\n",
1243 ceph_cap_string(ci->i_flushing_caps),
1244 inode, ceph_ino(inode));
1245 ci->i_flushing_caps = 0;
1246 list_del_init(&ci->i_flushing_item);
1247 mdsc->num_cap_flushing--;
1248 drop = true;
1249 }
1250 spin_unlock(&mdsc->cap_dirty_lock);
1251
1252 if (atomic_read(&ci->i_filelock_ref) > 0) {
1253 /* make further file lock syscall return -EIO */
1254 ci->i_ceph_flags |= CEPH_I_ERROR_FILELOCK;
1255 pr_warn_ratelimited(" dropping file locks for %p %lld\n",
1256 inode, ceph_ino(inode));
1257 }
1258
1259 if (!ci->i_dirty_caps && ci->i_prealloc_cap_flush) {
1260 list_add(&ci->i_prealloc_cap_flush->i_list, &to_remove);
1261 ci->i_prealloc_cap_flush = NULL;
1262 }
1263 }
1264 spin_unlock(&ci->i_ceph_lock);
1265 while (!list_empty(&to_remove)) {
1266 struct ceph_cap_flush *cf;
1267 cf = list_first_entry(&to_remove,
1268 struct ceph_cap_flush, i_list);
1269 list_del(&cf->i_list);
1270 ceph_free_cap_flush(cf);
1271 }
1272
1273 wake_up_all(&ci->i_cap_wq);
1274 if (invalidate)
1275 ceph_queue_invalidate(inode);
1276 if (drop)
1277 iput(inode);
1278 return 0;
1279 }
1280
1281 /*
1282 * caller must hold session s_mutex
1283 */
1284 static void remove_session_caps(struct ceph_mds_session *session)
1285 {
1286 struct ceph_fs_client *fsc = session->s_mdsc->fsc;
1287 struct super_block *sb = fsc->sb;
1288 LIST_HEAD(dispose);
1289
1290 dout("remove_session_caps on %p\n", session);
1291 iterate_session_caps(session, remove_session_caps_cb, fsc);
1292
1293 wake_up_all(&fsc->mdsc->cap_flushing_wq);
1294
1295 spin_lock(&session->s_cap_lock);
1296 if (session->s_nr_caps > 0) {
1297 struct inode *inode;
1298 struct ceph_cap *cap, *prev = NULL;
1299 struct ceph_vino vino;
1300 /*
1301 * iterate_session_caps() skips inodes that are being
1302 * deleted, we need to wait until deletions are complete.
1303 * __wait_on_freeing_inode() is designed for the job,
1304 * but it is not exported, so use lookup inode function
1305 * to access it.
1306 */
1307 while (!list_empty(&session->s_caps)) {
1308 cap = list_entry(session->s_caps.next,
1309 struct ceph_cap, session_caps);
1310 if (cap == prev)
1311 break;
1312 prev = cap;
1313 vino = cap->ci->i_vino;
1314 spin_unlock(&session->s_cap_lock);
1315
1316 inode = ceph_find_inode(sb, vino);
1317 iput(inode);
1318
1319 spin_lock(&session->s_cap_lock);
1320 }
1321 }
1322
1323 // drop cap expires and unlock s_cap_lock
1324 detach_cap_releases(session, &dispose);
1325
1326 BUG_ON(session->s_nr_caps > 0);
1327 BUG_ON(!list_empty(&session->s_cap_flushing));
1328 spin_unlock(&session->s_cap_lock);
1329 dispose_cap_releases(session->s_mdsc, &dispose);
1330 }
1331
1332 /*
1333 * wake up any threads waiting on this session's caps. if the cap is
1334 * old (didn't get renewed on the client reconnect), remove it now.
1335 *
1336 * caller must hold s_mutex.
1337 */
1338 static int wake_up_session_cb(struct inode *inode, struct ceph_cap *cap,
1339 void *arg)
1340 {
1341 struct ceph_inode_info *ci = ceph_inode(inode);
1342
1343 if (arg) {
1344 spin_lock(&ci->i_ceph_lock);
1345 ci->i_wanted_max_size = 0;
1346 ci->i_requested_max_size = 0;
1347 spin_unlock(&ci->i_ceph_lock);
1348 }
1349 wake_up_all(&ci->i_cap_wq);
1350 return 0;
1351 }
1352
1353 static void wake_up_session_caps(struct ceph_mds_session *session,
1354 int reconnect)
1355 {
1356 dout("wake_up_session_caps %p mds%d\n", session, session->s_mds);
1357 iterate_session_caps(session, wake_up_session_cb,
1358 (void *)(unsigned long)reconnect);
1359 }
1360
1361 /*
1362 * Send periodic message to MDS renewing all currently held caps. The
1363 * ack will reset the expiration for all caps from this session.
1364 *
1365 * caller holds s_mutex
1366 */
1367 static int send_renew_caps(struct ceph_mds_client *mdsc,
1368 struct ceph_mds_session *session)
1369 {
1370 struct ceph_msg *msg;
1371 int state;
1372
1373 if (time_after_eq(jiffies, session->s_cap_ttl) &&
1374 time_after_eq(session->s_cap_ttl, session->s_renew_requested))
1375 pr_info("mds%d caps stale\n", session->s_mds);
1376 session->s_renew_requested = jiffies;
1377
1378 /* do not try to renew caps until a recovering mds has reconnected
1379 * with its clients. */
1380 state = ceph_mdsmap_get_state(mdsc->mdsmap, session->s_mds);
1381 if (state < CEPH_MDS_STATE_RECONNECT) {
1382 dout("send_renew_caps ignoring mds%d (%s)\n",
1383 session->s_mds, ceph_mds_state_name(state));
1384 return 0;
1385 }
1386
1387 dout("send_renew_caps to mds%d (%s)\n", session->s_mds,
1388 ceph_mds_state_name(state));
1389 msg = create_session_msg(CEPH_SESSION_REQUEST_RENEWCAPS,
1390 ++session->s_renew_seq);
1391 if (!msg)
1392 return -ENOMEM;
1393 ceph_con_send(&session->s_con, msg);
1394 return 0;
1395 }
1396
1397 static int send_flushmsg_ack(struct ceph_mds_client *mdsc,
1398 struct ceph_mds_session *session, u64 seq)
1399 {
1400 struct ceph_msg *msg;
1401
1402 dout("send_flushmsg_ack to mds%d (%s)s seq %lld\n",
1403 session->s_mds, ceph_session_state_name(session->s_state), seq);
1404 msg = create_session_msg(CEPH_SESSION_FLUSHMSG_ACK, seq);
1405 if (!msg)
1406 return -ENOMEM;
1407 ceph_con_send(&session->s_con, msg);
1408 return 0;
1409 }
1410
1411
1412 /*
1413 * Note new cap ttl, and any transition from stale -> not stale (fresh?).
1414 *
1415 * Called under session->s_mutex
1416 */
1417 static void renewed_caps(struct ceph_mds_client *mdsc,
1418 struct ceph_mds_session *session, int is_renew)
1419 {
1420 int was_stale;
1421 int wake = 0;
1422
1423 spin_lock(&session->s_cap_lock);
1424 was_stale = is_renew && time_after_eq(jiffies, session->s_cap_ttl);
1425
1426 session->s_cap_ttl = session->s_renew_requested +
1427 mdsc->mdsmap->m_session_timeout*HZ;
1428
1429 if (was_stale) {
1430 if (time_before(jiffies, session->s_cap_ttl)) {
1431 pr_info("mds%d caps renewed\n", session->s_mds);
1432 wake = 1;
1433 } else {
1434 pr_info("mds%d caps still stale\n", session->s_mds);
1435 }
1436 }
1437 dout("renewed_caps mds%d ttl now %lu, was %s, now %s\n",
1438 session->s_mds, session->s_cap_ttl, was_stale ? "stale" : "fresh",
1439 time_before(jiffies, session->s_cap_ttl) ? "stale" : "fresh");
1440 spin_unlock(&session->s_cap_lock);
1441
1442 if (wake)
1443 wake_up_session_caps(session, 0);
1444 }
1445
1446 /*
1447 * send a session close request
1448 */
1449 static int request_close_session(struct ceph_mds_client *mdsc,
1450 struct ceph_mds_session *session)
1451 {
1452 struct ceph_msg *msg;
1453
1454 dout("request_close_session mds%d state %s seq %lld\n",
1455 session->s_mds, ceph_session_state_name(session->s_state),
1456 session->s_seq);
1457 msg = create_session_msg(CEPH_SESSION_REQUEST_CLOSE, session->s_seq);
1458 if (!msg)
1459 return -ENOMEM;
1460 ceph_con_send(&session->s_con, msg);
1461 return 1;
1462 }
1463
1464 /*
1465 * Called with s_mutex held.
1466 */
1467 static int __close_session(struct ceph_mds_client *mdsc,
1468 struct ceph_mds_session *session)
1469 {
1470 if (session->s_state >= CEPH_MDS_SESSION_CLOSING)
1471 return 0;
1472 session->s_state = CEPH_MDS_SESSION_CLOSING;
1473 return request_close_session(mdsc, session);
1474 }
1475
1476 static bool drop_negative_children(struct dentry *dentry)
1477 {
1478 struct dentry *child;
1479 bool all_negative = true;
1480
1481 if (!d_is_dir(dentry))
1482 goto out;
1483
1484 spin_lock(&dentry->d_lock);
1485 list_for_each_entry(child, &dentry->d_subdirs, d_child) {
1486 if (d_really_is_positive(child)) {
1487 all_negative = false;
1488 break;
1489 }
1490 }
1491 spin_unlock(&dentry->d_lock);
1492
1493 if (all_negative)
1494 shrink_dcache_parent(dentry);
1495 out:
1496 return all_negative;
1497 }
1498
1499 /*
1500 * Trim old(er) caps.
1501 *
1502 * Because we can't cache an inode without one or more caps, we do
1503 * this indirectly: if a cap is unused, we prune its aliases, at which
1504 * point the inode will hopefully get dropped to.
1505 *
1506 * Yes, this is a bit sloppy. Our only real goal here is to respond to
1507 * memory pressure from the MDS, though, so it needn't be perfect.
1508 */
1509 static int trim_caps_cb(struct inode *inode, struct ceph_cap *cap, void *arg)
1510 {
1511 struct ceph_mds_session *session = arg;
1512 struct ceph_inode_info *ci = ceph_inode(inode);
1513 int used, wanted, oissued, mine;
1514
1515 if (session->s_trim_caps <= 0)
1516 return -1;
1517
1518 spin_lock(&ci->i_ceph_lock);
1519 mine = cap->issued | cap->implemented;
1520 used = __ceph_caps_used(ci);
1521 wanted = __ceph_caps_file_wanted(ci);
1522 oissued = __ceph_caps_issued_other(ci, cap);
1523
1524 dout("trim_caps_cb %p cap %p mine %s oissued %s used %s wanted %s\n",
1525 inode, cap, ceph_cap_string(mine), ceph_cap_string(oissued),
1526 ceph_cap_string(used), ceph_cap_string(wanted));
1527 if (cap == ci->i_auth_cap) {
1528 if (ci->i_dirty_caps || ci->i_flushing_caps ||
1529 !list_empty(&ci->i_cap_snaps))
1530 goto out;
1531 if ((used | wanted) & CEPH_CAP_ANY_WR)
1532 goto out;
1533 /* Note: it's possible that i_filelock_ref becomes non-zero
1534 * after dropping auth caps. It doesn't hurt because reply
1535 * of lock mds request will re-add auth caps. */
1536 if (atomic_read(&ci->i_filelock_ref) > 0)
1537 goto out;
1538 }
1539 /* The inode has cached pages, but it's no longer used.
1540 * we can safely drop it */
1541 if (wanted == 0 && used == CEPH_CAP_FILE_CACHE &&
1542 !(oissued & CEPH_CAP_FILE_CACHE)) {
1543 used = 0;
1544 oissued = 0;
1545 }
1546 if ((used | wanted) & ~oissued & mine)
1547 goto out; /* we need these caps */
1548
1549 if (oissued) {
1550 /* we aren't the only cap.. just remove us */
1551 __ceph_remove_cap(cap, true);
1552 session->s_trim_caps--;
1553 } else {
1554 struct dentry *dentry;
1555 /* try dropping referring dentries */
1556 spin_unlock(&ci->i_ceph_lock);
1557 dentry = d_find_any_alias(inode);
1558 if (dentry && drop_negative_children(dentry)) {
1559 int count;
1560 dput(dentry);
1561 d_prune_aliases(inode);
1562 count = atomic_read(&inode->i_count);
1563 if (count == 1)
1564 session->s_trim_caps--;
1565 dout("trim_caps_cb %p cap %p pruned, count now %d\n",
1566 inode, cap, count);
1567 } else {
1568 dput(dentry);
1569 }
1570 return 0;
1571 }
1572
1573 out:
1574 spin_unlock(&ci->i_ceph_lock);
1575 return 0;
1576 }
1577
1578 /*
1579 * Trim session cap count down to some max number.
1580 */
1581 int ceph_trim_caps(struct ceph_mds_client *mdsc,
1582 struct ceph_mds_session *session,
1583 int max_caps)
1584 {
1585 int trim_caps = session->s_nr_caps - max_caps;
1586
1587 dout("trim_caps mds%d start: %d / %d, trim %d\n",
1588 session->s_mds, session->s_nr_caps, max_caps, trim_caps);
1589 if (trim_caps > 0) {
1590 session->s_trim_caps = trim_caps;
1591 iterate_session_caps(session, trim_caps_cb, session);
1592 dout("trim_caps mds%d done: %d / %d, trimmed %d\n",
1593 session->s_mds, session->s_nr_caps, max_caps,
1594 trim_caps - session->s_trim_caps);
1595 session->s_trim_caps = 0;
1596 }
1597
1598 ceph_send_cap_releases(mdsc, session);
1599 return 0;
1600 }
1601
1602 static int check_caps_flush(struct ceph_mds_client *mdsc,
1603 u64 want_flush_tid)
1604 {
1605 int ret = 1;
1606
1607 spin_lock(&mdsc->cap_dirty_lock);
1608 if (!list_empty(&mdsc->cap_flush_list)) {
1609 struct ceph_cap_flush *cf =
1610 list_first_entry(&mdsc->cap_flush_list,
1611 struct ceph_cap_flush, g_list);
1612 if (cf->tid <= want_flush_tid) {
1613 dout("check_caps_flush still flushing tid "
1614 "%llu <= %llu\n", cf->tid, want_flush_tid);
1615 ret = 0;
1616 }
1617 }
1618 spin_unlock(&mdsc->cap_dirty_lock);
1619 return ret;
1620 }
1621
1622 /*
1623 * flush all dirty inode data to disk.
1624 *
1625 * returns true if we've flushed through want_flush_tid
1626 */
1627 static void wait_caps_flush(struct ceph_mds_client *mdsc,
1628 u64 want_flush_tid)
1629 {
1630 dout("check_caps_flush want %llu\n", want_flush_tid);
1631
1632 wait_event(mdsc->cap_flushing_wq,
1633 check_caps_flush(mdsc, want_flush_tid));
1634
1635 dout("check_caps_flush ok, flushed thru %llu\n", want_flush_tid);
1636 }
1637
1638 /*
1639 * called under s_mutex
1640 */
1641 void ceph_send_cap_releases(struct ceph_mds_client *mdsc,
1642 struct ceph_mds_session *session)
1643 {
1644 struct ceph_msg *msg = NULL;
1645 struct ceph_mds_cap_release *head;
1646 struct ceph_mds_cap_item *item;
1647 struct ceph_osd_client *osdc = &mdsc->fsc->client->osdc;
1648 struct ceph_cap *cap;
1649 LIST_HEAD(tmp_list);
1650 int num_cap_releases;
1651 __le32 barrier, *cap_barrier;
1652
1653 down_read(&osdc->lock);
1654 barrier = cpu_to_le32(osdc->epoch_barrier);
1655 up_read(&osdc->lock);
1656
1657 spin_lock(&session->s_cap_lock);
1658 again:
1659 list_splice_init(&session->s_cap_releases, &tmp_list);
1660 num_cap_releases = session->s_num_cap_releases;
1661 session->s_num_cap_releases = 0;
1662 spin_unlock(&session->s_cap_lock);
1663
1664 while (!list_empty(&tmp_list)) {
1665 if (!msg) {
1666 msg = ceph_msg_new(CEPH_MSG_CLIENT_CAPRELEASE,
1667 PAGE_SIZE, GFP_NOFS, false);
1668 if (!msg)
1669 goto out_err;
1670 head = msg->front.iov_base;
1671 head->num = cpu_to_le32(0);
1672 msg->front.iov_len = sizeof(*head);
1673
1674 msg->hdr.version = cpu_to_le16(2);
1675 msg->hdr.compat_version = cpu_to_le16(1);
1676 }
1677
1678 cap = list_first_entry(&tmp_list, struct ceph_cap,
1679 session_caps);
1680 list_del(&cap->session_caps);
1681 num_cap_releases--;
1682
1683 head = msg->front.iov_base;
1684 le32_add_cpu(&head->num, 1);
1685 item = msg->front.iov_base + msg->front.iov_len;
1686 item->ino = cpu_to_le64(cap->cap_ino);
1687 item->cap_id = cpu_to_le64(cap->cap_id);
1688 item->migrate_seq = cpu_to_le32(cap->mseq);
1689 item->seq = cpu_to_le32(cap->issue_seq);
1690 msg->front.iov_len += sizeof(*item);
1691
1692 ceph_put_cap(mdsc, cap);
1693
1694 if (le32_to_cpu(head->num) == CEPH_CAPS_PER_RELEASE) {
1695 // Append cap_barrier field
1696 cap_barrier = msg->front.iov_base + msg->front.iov_len;
1697 *cap_barrier = barrier;
1698 msg->front.iov_len += sizeof(*cap_barrier);
1699
1700 msg->hdr.front_len = cpu_to_le32(msg->front.iov_len);
1701 dout("send_cap_releases mds%d %p\n", session->s_mds, msg);
1702 ceph_con_send(&session->s_con, msg);
1703 msg = NULL;
1704 }
1705 }
1706
1707 BUG_ON(num_cap_releases != 0);
1708
1709 spin_lock(&session->s_cap_lock);
1710 if (!list_empty(&session->s_cap_releases))
1711 goto again;
1712 spin_unlock(&session->s_cap_lock);
1713
1714 if (msg) {
1715 // Append cap_barrier field
1716 cap_barrier = msg->front.iov_base + msg->front.iov_len;
1717 *cap_barrier = barrier;
1718 msg->front.iov_len += sizeof(*cap_barrier);
1719
1720 msg->hdr.front_len = cpu_to_le32(msg->front.iov_len);
1721 dout("send_cap_releases mds%d %p\n", session->s_mds, msg);
1722 ceph_con_send(&session->s_con, msg);
1723 }
1724 return;
1725 out_err:
1726 pr_err("send_cap_releases mds%d, failed to allocate message\n",
1727 session->s_mds);
1728 spin_lock(&session->s_cap_lock);
1729 list_splice(&tmp_list, &session->s_cap_releases);
1730 session->s_num_cap_releases += num_cap_releases;
1731 spin_unlock(&session->s_cap_lock);
1732 }
1733
1734 /*
1735 * requests
1736 */
1737
1738 int ceph_alloc_readdir_reply_buffer(struct ceph_mds_request *req,
1739 struct inode *dir)
1740 {
1741 struct ceph_inode_info *ci = ceph_inode(dir);
1742 struct ceph_mds_reply_info_parsed *rinfo = &req->r_reply_info;
1743 struct ceph_mount_options *opt = req->r_mdsc->fsc->mount_options;
1744 size_t size = sizeof(struct ceph_mds_reply_dir_entry);
1745 int order, num_entries;
1746
1747 spin_lock(&ci->i_ceph_lock);
1748 num_entries = ci->i_files + ci->i_subdirs;
1749 spin_unlock(&ci->i_ceph_lock);
1750 num_entries = max(num_entries, 1);
1751 num_entries = min(num_entries, opt->max_readdir);
1752
1753 order = get_order(size * num_entries);
1754 while (order >= 0) {
1755 rinfo->dir_entries = (void*)__get_free_pages(GFP_KERNEL |
1756 __GFP_NOWARN,
1757 order);
1758 if (rinfo->dir_entries)
1759 break;
1760 order--;
1761 }
1762 if (!rinfo->dir_entries)
1763 return -ENOMEM;
1764
1765 num_entries = (PAGE_SIZE << order) / size;
1766 num_entries = min(num_entries, opt->max_readdir);
1767
1768 rinfo->dir_buf_size = PAGE_SIZE << order;
1769 req->r_num_caps = num_entries + 1;
1770 req->r_args.readdir.max_entries = cpu_to_le32(num_entries);
1771 req->r_args.readdir.max_bytes = cpu_to_le32(opt->max_readdir_bytes);
1772 return 0;
1773 }
1774
1775 /*
1776 * Create an mds request.
1777 */
1778 struct ceph_mds_request *
1779 ceph_mdsc_create_request(struct ceph_mds_client *mdsc, int op, int mode)
1780 {
1781 struct ceph_mds_request *req = kzalloc(sizeof(*req), GFP_NOFS);
1782
1783 if (!req)
1784 return ERR_PTR(-ENOMEM);
1785
1786 mutex_init(&req->r_fill_mutex);
1787 req->r_mdsc = mdsc;
1788 req->r_started = jiffies;
1789 req->r_resend_mds = -1;
1790 INIT_LIST_HEAD(&req->r_unsafe_dir_item);
1791 INIT_LIST_HEAD(&req->r_unsafe_target_item);
1792 req->r_fmode = -1;
1793 kref_init(&req->r_kref);
1794 RB_CLEAR_NODE(&req->r_node);
1795 INIT_LIST_HEAD(&req->r_wait);
1796 init_completion(&req->r_completion);
1797 init_completion(&req->r_safe_completion);
1798 INIT_LIST_HEAD(&req->r_unsafe_item);
1799
1800 req->r_stamp = timespec_trunc(current_kernel_time(), mdsc->fsc->sb->s_time_gran);
1801
1802 req->r_op = op;
1803 req->r_direct_mode = mode;
1804 return req;
1805 }
1806
1807 /*
1808 * return oldest (lowest) request, tid in request tree, 0 if none.
1809 *
1810 * called under mdsc->mutex.
1811 */
1812 static struct ceph_mds_request *__get_oldest_req(struct ceph_mds_client *mdsc)
1813 {
1814 if (RB_EMPTY_ROOT(&mdsc->request_tree))
1815 return NULL;
1816 return rb_entry(rb_first(&mdsc->request_tree),
1817 struct ceph_mds_request, r_node);
1818 }
1819
1820 static inline u64 __get_oldest_tid(struct ceph_mds_client *mdsc)
1821 {
1822 return mdsc->oldest_tid;
1823 }
1824
1825 /*
1826 * Build a dentry's path. Allocate on heap; caller must kfree. Based
1827 * on build_path_from_dentry in fs/cifs/dir.c.
1828 *
1829 * If @stop_on_nosnap, generate path relative to the first non-snapped
1830 * inode.
1831 *
1832 * Encode hidden .snap dirs as a double /, i.e.
1833 * foo/.snap/bar -> foo//bar
1834 */
1835 char *ceph_mdsc_build_path(struct dentry *dentry, int *plen, u64 *base,
1836 int stop_on_nosnap)
1837 {
1838 struct dentry *temp;
1839 char *path;
1840 int len, pos;
1841 unsigned seq;
1842
1843 if (!dentry)
1844 return ERR_PTR(-EINVAL);
1845
1846 retry:
1847 len = 0;
1848 seq = read_seqbegin(&rename_lock);
1849 rcu_read_lock();
1850 for (temp = dentry; !IS_ROOT(temp);) {
1851 struct inode *inode = d_inode(temp);
1852 if (inode && ceph_snap(inode) == CEPH_SNAPDIR)
1853 len++; /* slash only */
1854 else if (stop_on_nosnap && inode &&
1855 ceph_snap(inode) == CEPH_NOSNAP)
1856 break;
1857 else
1858 len += 1 + temp->d_name.len;
1859 temp = temp->d_parent;
1860 }
1861 rcu_read_unlock();
1862 if (len)
1863 len--; /* no leading '/' */
1864
1865 path = kmalloc(len+1, GFP_NOFS);
1866 if (!path)
1867 return ERR_PTR(-ENOMEM);
1868 pos = len;
1869 path[pos] = 0; /* trailing null */
1870 rcu_read_lock();
1871 for (temp = dentry; !IS_ROOT(temp) && pos != 0; ) {
1872 struct inode *inode;
1873
1874 spin_lock(&temp->d_lock);
1875 inode = d_inode(temp);
1876 if (inode && ceph_snap(inode) == CEPH_SNAPDIR) {
1877 dout("build_path path+%d: %p SNAPDIR\n",
1878 pos, temp);
1879 } else if (stop_on_nosnap && inode &&
1880 ceph_snap(inode) == CEPH_NOSNAP) {
1881 spin_unlock(&temp->d_lock);
1882 break;
1883 } else {
1884 pos -= temp->d_name.len;
1885 if (pos < 0) {
1886 spin_unlock(&temp->d_lock);
1887 break;
1888 }
1889 strncpy(path + pos, temp->d_name.name,
1890 temp->d_name.len);
1891 }
1892 spin_unlock(&temp->d_lock);
1893 if (pos)
1894 path[--pos] = '/';
1895 temp = temp->d_parent;
1896 }
1897 rcu_read_unlock();
1898 if (pos != 0 || read_seqretry(&rename_lock, seq)) {
1899 pr_err("build_path did not end path lookup where "
1900 "expected, namelen is %d, pos is %d\n", len, pos);
1901 /* presumably this is only possible if racing with a
1902 rename of one of the parent directories (we can not
1903 lock the dentries above us to prevent this, but
1904 retrying should be harmless) */
1905 kfree(path);
1906 goto retry;
1907 }
1908
1909 *base = ceph_ino(d_inode(temp));
1910 *plen = len;
1911 dout("build_path on %p %d built %llx '%.*s'\n",
1912 dentry, d_count(dentry), *base, len, path);
1913 return path;
1914 }
1915
1916 static int build_dentry_path(struct dentry *dentry, struct inode *dir,
1917 const char **ppath, int *ppathlen, u64 *pino,
1918 int *pfreepath)
1919 {
1920 char *path;
1921
1922 rcu_read_lock();
1923 if (!dir)
1924 dir = d_inode_rcu(dentry->d_parent);
1925 if (dir && ceph_snap(dir) == CEPH_NOSNAP) {
1926 *pino = ceph_ino(dir);
1927 rcu_read_unlock();
1928 *ppath = dentry->d_name.name;
1929 *ppathlen = dentry->d_name.len;
1930 return 0;
1931 }
1932 rcu_read_unlock();
1933 path = ceph_mdsc_build_path(dentry, ppathlen, pino, 1);
1934 if (IS_ERR(path))
1935 return PTR_ERR(path);
1936 *ppath = path;
1937 *pfreepath = 1;
1938 return 0;
1939 }
1940
1941 static int build_inode_path(struct inode *inode,
1942 const char **ppath, int *ppathlen, u64 *pino,
1943 int *pfreepath)
1944 {
1945 struct dentry *dentry;
1946 char *path;
1947
1948 if (ceph_snap(inode) == CEPH_NOSNAP) {
1949 *pino = ceph_ino(inode);
1950 *ppathlen = 0;
1951 return 0;
1952 }
1953 dentry = d_find_alias(inode);
1954 path = ceph_mdsc_build_path(dentry, ppathlen, pino, 1);
1955 dput(dentry);
1956 if (IS_ERR(path))
1957 return PTR_ERR(path);
1958 *ppath = path;
1959 *pfreepath = 1;
1960 return 0;
1961 }
1962
1963 /*
1964 * request arguments may be specified via an inode *, a dentry *, or
1965 * an explicit ino+path.
1966 */
1967 static int set_request_path_attr(struct inode *rinode, struct dentry *rdentry,
1968 struct inode *rdiri, const char *rpath,
1969 u64 rino, const char **ppath, int *pathlen,
1970 u64 *ino, int *freepath)
1971 {
1972 int r = 0;
1973
1974 if (rinode) {
1975 r = build_inode_path(rinode, ppath, pathlen, ino, freepath);
1976 dout(" inode %p %llx.%llx\n", rinode, ceph_ino(rinode),
1977 ceph_snap(rinode));
1978 } else if (rdentry) {
1979 r = build_dentry_path(rdentry, rdiri, ppath, pathlen, ino,
1980 freepath);
1981 dout(" dentry %p %llx/%.*s\n", rdentry, *ino, *pathlen,
1982 *ppath);
1983 } else if (rpath || rino) {
1984 *ino = rino;
1985 *ppath = rpath;
1986 *pathlen = rpath ? strlen(rpath) : 0;
1987 dout(" path %.*s\n", *pathlen, rpath);
1988 }
1989
1990 return r;
1991 }
1992
1993 /*
1994 * called under mdsc->mutex
1995 */
1996 static struct ceph_msg *create_request_message(struct ceph_mds_client *mdsc,
1997 struct ceph_mds_request *req,
1998 int mds, bool drop_cap_releases)
1999 {
2000 struct ceph_msg *msg;
2001 struct ceph_mds_request_head *head;
2002 const char *path1 = NULL;
2003 const char *path2 = NULL;
2004 u64 ino1 = 0, ino2 = 0;
2005 int pathlen1 = 0, pathlen2 = 0;
2006 int freepath1 = 0, freepath2 = 0;
2007 int len;
2008 u16 releases;
2009 void *p, *end;
2010 int ret;
2011
2012 ret = set_request_path_attr(req->r_inode, req->r_dentry,
2013 req->r_parent, req->r_path1, req->r_ino1.ino,
2014 &path1, &pathlen1, &ino1, &freepath1);
2015 if (ret < 0) {
2016 msg = ERR_PTR(ret);
2017 goto out;
2018 }
2019
2020 ret = set_request_path_attr(NULL, req->r_old_dentry,
2021 req->r_old_dentry_dir,
2022 req->r_path2, req->r_ino2.ino,
2023 &path2, &pathlen2, &ino2, &freepath2);
2024 if (ret < 0) {
2025 msg = ERR_PTR(ret);
2026 goto out_free1;
2027 }
2028
2029 len = sizeof(*head) +
2030 pathlen1 + pathlen2 + 2*(1 + sizeof(u32) + sizeof(u64)) +
2031 sizeof(struct ceph_timespec);
2032
2033 /* calculate (max) length for cap releases */
2034 len += sizeof(struct ceph_mds_request_release) *
2035 (!!req->r_inode_drop + !!req->r_dentry_drop +
2036 !!req->r_old_inode_drop + !!req->r_old_dentry_drop);
2037 if (req->r_dentry_drop)
2038 len += req->r_dentry->d_name.len;
2039 if (req->r_old_dentry_drop)
2040 len += req->r_old_dentry->d_name.len;
2041
2042 msg = ceph_msg_new(CEPH_MSG_CLIENT_REQUEST, len, GFP_NOFS, false);
2043 if (!msg) {
2044 msg = ERR_PTR(-ENOMEM);
2045 goto out_free2;
2046 }
2047
2048 msg->hdr.version = cpu_to_le16(2);
2049 msg->hdr.tid = cpu_to_le64(req->r_tid);
2050
2051 head = msg->front.iov_base;
2052 p = msg->front.iov_base + sizeof(*head);
2053 end = msg->front.iov_base + msg->front.iov_len;
2054
2055 head->mdsmap_epoch = cpu_to_le32(mdsc->mdsmap->m_epoch);
2056 head->op = cpu_to_le32(req->r_op);
2057 head->caller_uid = cpu_to_le32(from_kuid(&init_user_ns, req->r_uid));
2058 head->caller_gid = cpu_to_le32(from_kgid(&init_user_ns, req->r_gid));
2059 head->args = req->r_args;
2060
2061 ceph_encode_filepath(&p, end, ino1, path1);
2062 ceph_encode_filepath(&p, end, ino2, path2);
2063
2064 /* make note of release offset, in case we need to replay */
2065 req->r_request_release_offset = p - msg->front.iov_base;
2066
2067 /* cap releases */
2068 releases = 0;
2069 if (req->r_inode_drop)
2070 releases += ceph_encode_inode_release(&p,
2071 req->r_inode ? req->r_inode : d_inode(req->r_dentry),
2072 mds, req->r_inode_drop, req->r_inode_unless, 0);
2073 if (req->r_dentry_drop)
2074 releases += ceph_encode_dentry_release(&p, req->r_dentry,
2075 req->r_parent, mds, req->r_dentry_drop,
2076 req->r_dentry_unless);
2077 if (req->r_old_dentry_drop)
2078 releases += ceph_encode_dentry_release(&p, req->r_old_dentry,
2079 req->r_old_dentry_dir, mds,
2080 req->r_old_dentry_drop,
2081 req->r_old_dentry_unless);
2082 if (req->r_old_inode_drop)
2083 releases += ceph_encode_inode_release(&p,
2084 d_inode(req->r_old_dentry),
2085 mds, req->r_old_inode_drop, req->r_old_inode_unless, 0);
2086
2087 if (drop_cap_releases) {
2088 releases = 0;
2089 p = msg->front.iov_base + req->r_request_release_offset;
2090 }
2091
2092 head->num_releases = cpu_to_le16(releases);
2093
2094 /* time stamp */
2095 {
2096 struct ceph_timespec ts;
2097 ceph_encode_timespec(&ts, &req->r_stamp);
2098 ceph_encode_copy(&p, &ts, sizeof(ts));
2099 }
2100
2101 BUG_ON(p > end);
2102 msg->front.iov_len = p - msg->front.iov_base;
2103 msg->hdr.front_len = cpu_to_le32(msg->front.iov_len);
2104
2105 if (req->r_pagelist) {
2106 struct ceph_pagelist *pagelist = req->r_pagelist;
2107 refcount_inc(&pagelist->refcnt);
2108 ceph_msg_data_add_pagelist(msg, pagelist);
2109 msg->hdr.data_len = cpu_to_le32(pagelist->length);
2110 } else {
2111 msg->hdr.data_len = 0;
2112 }
2113
2114 msg->hdr.data_off = cpu_to_le16(0);
2115
2116 out_free2:
2117 if (freepath2)
2118 kfree((char *)path2);
2119 out_free1:
2120 if (freepath1)
2121 kfree((char *)path1);
2122 out:
2123 return msg;
2124 }
2125
2126 /*
2127 * called under mdsc->mutex if error, under no mutex if
2128 * success.
2129 */
2130 static void complete_request(struct ceph_mds_client *mdsc,
2131 struct ceph_mds_request *req)
2132 {
2133 if (req->r_callback)
2134 req->r_callback(mdsc, req);
2135 else
2136 complete_all(&req->r_completion);
2137 }
2138
2139 /*
2140 * called under mdsc->mutex
2141 */
2142 static int __prepare_send_request(struct ceph_mds_client *mdsc,
2143 struct ceph_mds_request *req,
2144 int mds, bool drop_cap_releases)
2145 {
2146 struct ceph_mds_request_head *rhead;
2147 struct ceph_msg *msg;
2148 int flags = 0;
2149
2150 req->r_attempts++;
2151 if (req->r_inode) {
2152 struct ceph_cap *cap =
2153 ceph_get_cap_for_mds(ceph_inode(req->r_inode), mds);
2154
2155 if (cap)
2156 req->r_sent_on_mseq = cap->mseq;
2157 else
2158 req->r_sent_on_mseq = -1;
2159 }
2160 dout("prepare_send_request %p tid %lld %s (attempt %d)\n", req,
2161 req->r_tid, ceph_mds_op_name(req->r_op), req->r_attempts);
2162
2163 if (test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags)) {
2164 void *p;
2165 /*
2166 * Replay. Do not regenerate message (and rebuild
2167 * paths, etc.); just use the original message.
2168 * Rebuilding paths will break for renames because
2169 * d_move mangles the src name.
2170 */
2171 msg = req->r_request;
2172 rhead = msg->front.iov_base;
2173
2174 flags = le32_to_cpu(rhead->flags);
2175 flags |= CEPH_MDS_FLAG_REPLAY;
2176 rhead->flags = cpu_to_le32(flags);
2177
2178 if (req->r_target_inode)
2179 rhead->ino = cpu_to_le64(ceph_ino(req->r_target_inode));
2180
2181 rhead->num_retry = req->r_attempts - 1;
2182
2183 /* remove cap/dentry releases from message */
2184 rhead->num_releases = 0;
2185
2186 /* time stamp */
2187 p = msg->front.iov_base + req->r_request_release_offset;
2188 {
2189 struct ceph_timespec ts;
2190 ceph_encode_timespec(&ts, &req->r_stamp);
2191 ceph_encode_copy(&p, &ts, sizeof(ts));
2192 }
2193
2194 msg->front.iov_len = p - msg->front.iov_base;
2195 msg->hdr.front_len = cpu_to_le32(msg->front.iov_len);
2196 return 0;
2197 }
2198
2199 if (req->r_request) {
2200 ceph_msg_put(req->r_request);
2201 req->r_request = NULL;
2202 }
2203 msg = create_request_message(mdsc, req, mds, drop_cap_releases);
2204 if (IS_ERR(msg)) {
2205 req->r_err = PTR_ERR(msg);
2206 return PTR_ERR(msg);
2207 }
2208 req->r_request = msg;
2209
2210 rhead = msg->front.iov_base;
2211 rhead->oldest_client_tid = cpu_to_le64(__get_oldest_tid(mdsc));
2212 if (test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags))
2213 flags |= CEPH_MDS_FLAG_REPLAY;
2214 if (req->r_parent)
2215 flags |= CEPH_MDS_FLAG_WANT_DENTRY;
2216 rhead->flags = cpu_to_le32(flags);
2217 rhead->num_fwd = req->r_num_fwd;
2218 rhead->num_retry = req->r_attempts - 1;
2219 rhead->ino = 0;
2220
2221 dout(" r_parent = %p\n", req->r_parent);
2222 return 0;
2223 }
2224
2225 /*
2226 * send request, or put it on the appropriate wait list.
2227 */
2228 static int __do_request(struct ceph_mds_client *mdsc,
2229 struct ceph_mds_request *req)
2230 {
2231 struct ceph_mds_session *session = NULL;
2232 int mds = -1;
2233 int err = 0;
2234
2235 if (req->r_err || test_bit(CEPH_MDS_R_GOT_RESULT, &req->r_req_flags)) {
2236 if (test_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags))
2237 __unregister_request(mdsc, req);
2238 goto out;
2239 }
2240
2241 if (req->r_timeout &&
2242 time_after_eq(jiffies, req->r_started + req->r_timeout)) {
2243 dout("do_request timed out\n");
2244 err = -EIO;
2245 goto finish;
2246 }
2247 if (READ_ONCE(mdsc->fsc->mount_state) == CEPH_MOUNT_SHUTDOWN) {
2248 dout("do_request forced umount\n");
2249 err = -EIO;
2250 goto finish;
2251 }
2252 if (READ_ONCE(mdsc->fsc->mount_state) == CEPH_MOUNT_MOUNTING) {
2253 if (mdsc->mdsmap_err) {
2254 err = mdsc->mdsmap_err;
2255 dout("do_request mdsmap err %d\n", err);
2256 goto finish;
2257 }
2258 if (mdsc->mdsmap->m_epoch == 0) {
2259 dout("do_request no mdsmap, waiting for map\n");
2260 list_add(&req->r_wait, &mdsc->waiting_for_map);
2261 goto finish;
2262 }
2263 if (!(mdsc->fsc->mount_options->flags &
2264 CEPH_MOUNT_OPT_MOUNTWAIT) &&
2265 !ceph_mdsmap_is_cluster_available(mdsc->mdsmap)) {
2266 err = -ENOENT;
2267 pr_info("probably no mds server is up\n");
2268 goto finish;
2269 }
2270 }
2271
2272 put_request_session(req);
2273
2274 mds = __choose_mds(mdsc, req);
2275 if (mds < 0 ||
2276 ceph_mdsmap_get_state(mdsc->mdsmap, mds) < CEPH_MDS_STATE_ACTIVE) {
2277 dout("do_request no mds or not active, waiting for map\n");
2278 list_add(&req->r_wait, &mdsc->waiting_for_map);
2279 goto out;
2280 }
2281
2282 /* get, open session */
2283 session = __ceph_lookup_mds_session(mdsc, mds);
2284 if (!session) {
2285 session = register_session(mdsc, mds);
2286 if (IS_ERR(session)) {
2287 err = PTR_ERR(session);
2288 goto finish;
2289 }
2290 }
2291 req->r_session = get_session(session);
2292
2293 dout("do_request mds%d session %p state %s\n", mds, session,
2294 ceph_session_state_name(session->s_state));
2295 if (session->s_state != CEPH_MDS_SESSION_OPEN &&
2296 session->s_state != CEPH_MDS_SESSION_HUNG) {
2297 if (session->s_state == CEPH_MDS_SESSION_REJECTED) {
2298 err = -EACCES;
2299 goto out_session;
2300 }
2301 if (session->s_state == CEPH_MDS_SESSION_NEW ||
2302 session->s_state == CEPH_MDS_SESSION_CLOSING)
2303 __open_session(mdsc, session);
2304 list_add(&req->r_wait, &session->s_waiting);
2305 goto out_session;
2306 }
2307
2308 /* send request */
2309 req->r_resend_mds = -1; /* forget any previous mds hint */
2310
2311 if (req->r_request_started == 0) /* note request start time */
2312 req->r_request_started = jiffies;
2313
2314 err = __prepare_send_request(mdsc, req, mds, false);
2315 if (!err) {
2316 ceph_msg_get(req->r_request);
2317 ceph_con_send(&session->s_con, req->r_request);
2318 }
2319
2320 out_session:
2321 ceph_put_mds_session(session);
2322 finish:
2323 if (err) {
2324 dout("__do_request early error %d\n", err);
2325 req->r_err = err;
2326 complete_request(mdsc, req);
2327 __unregister_request(mdsc, req);
2328 }
2329 out:
2330 return err;
2331 }
2332
2333 /*
2334 * called under mdsc->mutex
2335 */
2336 static void __wake_requests(struct ceph_mds_client *mdsc,
2337 struct list_head *head)
2338 {
2339 struct ceph_mds_request *req;
2340 LIST_HEAD(tmp_list);
2341
2342 list_splice_init(head, &tmp_list);
2343
2344 while (!list_empty(&tmp_list)) {
2345 req = list_entry(tmp_list.next,
2346 struct ceph_mds_request, r_wait);
2347 list_del_init(&req->r_wait);
2348 dout(" wake request %p tid %llu\n", req, req->r_tid);
2349 __do_request(mdsc, req);
2350 }
2351 }
2352
2353 /*
2354 * Wake up threads with requests pending for @mds, so that they can
2355 * resubmit their requests to a possibly different mds.
2356 */
2357 static void kick_requests(struct ceph_mds_client *mdsc, int mds)
2358 {
2359 struct ceph_mds_request *req;
2360 struct rb_node *p = rb_first(&mdsc->request_tree);
2361
2362 dout("kick_requests mds%d\n", mds);
2363 while (p) {
2364 req = rb_entry(p, struct ceph_mds_request, r_node);
2365 p = rb_next(p);
2366 if (test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags))
2367 continue;
2368 if (req->r_attempts > 0)
2369 continue; /* only new requests */
2370 if (req->r_session &&
2371 req->r_session->s_mds == mds) {
2372 dout(" kicking tid %llu\n", req->r_tid);
2373 list_del_init(&req->r_wait);
2374 __do_request(mdsc, req);
2375 }
2376 }
2377 }
2378
2379 void ceph_mdsc_submit_request(struct ceph_mds_client *mdsc,
2380 struct ceph_mds_request *req)
2381 {
2382 dout("submit_request on %p\n", req);
2383 mutex_lock(&mdsc->mutex);
2384 __register_request(mdsc, req, NULL);
2385 __do_request(mdsc, req);
2386 mutex_unlock(&mdsc->mutex);
2387 }
2388
2389 /*
2390 * Synchrously perform an mds request. Take care of all of the
2391 * session setup, forwarding, retry details.
2392 */
2393 int ceph_mdsc_do_request(struct ceph_mds_client *mdsc,
2394 struct inode *dir,
2395 struct ceph_mds_request *req)
2396 {
2397 int err;
2398
2399 dout("do_request on %p\n", req);
2400
2401 /* take CAP_PIN refs for r_inode, r_parent, r_old_dentry */
2402 if (req->r_inode)
2403 ceph_get_cap_refs(ceph_inode(req->r_inode), CEPH_CAP_PIN);
2404 if (req->r_parent)
2405 ceph_get_cap_refs(ceph_inode(req->r_parent), CEPH_CAP_PIN);
2406 if (req->r_old_dentry_dir)
2407 ceph_get_cap_refs(ceph_inode(req->r_old_dentry_dir),
2408 CEPH_CAP_PIN);
2409
2410 /* issue */
2411 mutex_lock(&mdsc->mutex);
2412 __register_request(mdsc, req, dir);
2413 __do_request(mdsc, req);
2414
2415 if (req->r_err) {
2416 err = req->r_err;
2417 goto out;
2418 }
2419
2420 /* wait */
2421 mutex_unlock(&mdsc->mutex);
2422 dout("do_request waiting\n");
2423 if (!req->r_timeout && req->r_wait_for_completion) {
2424 err = req->r_wait_for_completion(mdsc, req);
2425 } else {
2426 long timeleft = wait_for_completion_killable_timeout(
2427 &req->r_completion,
2428 ceph_timeout_jiffies(req->r_timeout));
2429 if (timeleft > 0)
2430 err = 0;
2431 else if (!timeleft)
2432 err = -EIO; /* timed out */
2433 else
2434 err = timeleft; /* killed */
2435 }
2436 dout("do_request waited, got %d\n", err);
2437 mutex_lock(&mdsc->mutex);
2438
2439 /* only abort if we didn't race with a real reply */
2440 if (test_bit(CEPH_MDS_R_GOT_RESULT, &req->r_req_flags)) {
2441 err = le32_to_cpu(req->r_reply_info.head->result);
2442 } else if (err < 0) {
2443 dout("aborted request %lld with %d\n", req->r_tid, err);
2444
2445 /*
2446 * ensure we aren't running concurrently with
2447 * ceph_fill_trace or ceph_readdir_prepopulate, which
2448 * rely on locks (dir mutex) held by our caller.
2449 */
2450 mutex_lock(&req->r_fill_mutex);
2451 req->r_err = err;
2452 set_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags);
2453 mutex_unlock(&req->r_fill_mutex);
2454
2455 if (req->r_parent &&
2456 (req->r_op & CEPH_MDS_OP_WRITE))
2457 ceph_invalidate_dir_request(req);
2458 } else {
2459 err = req->r_err;
2460 }
2461
2462 out:
2463 mutex_unlock(&mdsc->mutex);
2464 dout("do_request %p done, result %d\n", req, err);
2465 return err;
2466 }
2467
2468 /*
2469 * Invalidate dir's completeness, dentry lease state on an aborted MDS
2470 * namespace request.
2471 */
2472 void ceph_invalidate_dir_request(struct ceph_mds_request *req)
2473 {
2474 struct inode *dir = req->r_parent;
2475 struct inode *old_dir = req->r_old_dentry_dir;
2476
2477 dout("invalidate_dir_request %p %p (complete, lease(s))\n", dir, old_dir);
2478
2479 ceph_dir_clear_complete(dir);
2480 if (old_dir)
2481 ceph_dir_clear_complete(old_dir);
2482 if (req->r_dentry)
2483 ceph_invalidate_dentry_lease(req->r_dentry);
2484 if (req->r_old_dentry)
2485 ceph_invalidate_dentry_lease(req->r_old_dentry);
2486 }
2487
2488 /*
2489 * Handle mds reply.
2490 *
2491 * We take the session mutex and parse and process the reply immediately.
2492 * This preserves the logical ordering of replies, capabilities, etc., sent
2493 * by the MDS as they are applied to our local cache.
2494 */
2495 static void handle_reply(struct ceph_mds_session *session, struct ceph_msg *msg)
2496 {
2497 struct ceph_mds_client *mdsc = session->s_mdsc;
2498 struct ceph_mds_request *req;
2499 struct ceph_mds_reply_head *head = msg->front.iov_base;
2500 struct ceph_mds_reply_info_parsed *rinfo; /* parsed reply info */
2501 struct ceph_snap_realm *realm;
2502 u64 tid;
2503 int err, result;
2504 int mds = session->s_mds;
2505
2506 if (msg->front.iov_len < sizeof(*head)) {
2507 pr_err("mdsc_handle_reply got corrupt (short) reply\n");
2508 ceph_msg_dump(msg);
2509 return;
2510 }
2511
2512 /* get request, session */
2513 tid = le64_to_cpu(msg->hdr.tid);
2514 mutex_lock(&mdsc->mutex);
2515 req = lookup_get_request(mdsc, tid);
2516 if (!req) {
2517 dout("handle_reply on unknown tid %llu\n", tid);
2518 mutex_unlock(&mdsc->mutex);
2519 return;
2520 }
2521 dout("handle_reply %p\n", req);
2522
2523 /* correct session? */
2524 if (req->r_session != session) {
2525 pr_err("mdsc_handle_reply got %llu on session mds%d"
2526 " not mds%d\n", tid, session->s_mds,
2527 req->r_session ? req->r_session->s_mds : -1);
2528 mutex_unlock(&mdsc->mutex);
2529 goto out;
2530 }
2531
2532 /* dup? */
2533 if ((test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags) && !head->safe) ||
2534 (test_bit(CEPH_MDS_R_GOT_SAFE, &req->r_req_flags) && head->safe)) {
2535 pr_warn("got a dup %s reply on %llu from mds%d\n",
2536 head->safe ? "safe" : "unsafe", tid, mds);
2537 mutex_unlock(&mdsc->mutex);
2538 goto out;
2539 }
2540 if (test_bit(CEPH_MDS_R_GOT_SAFE, &req->r_req_flags)) {
2541 pr_warn("got unsafe after safe on %llu from mds%d\n",
2542 tid, mds);
2543 mutex_unlock(&mdsc->mutex);
2544 goto out;
2545 }
2546
2547 result = le32_to_cpu(head->result);
2548
2549 /*
2550 * Handle an ESTALE
2551 * if we're not talking to the authority, send to them
2552 * if the authority has changed while we weren't looking,
2553 * send to new authority
2554 * Otherwise we just have to return an ESTALE
2555 */
2556 if (result == -ESTALE) {
2557 dout("got ESTALE on request %llu\n", req->r_tid);
2558 req->r_resend_mds = -1;
2559 if (req->r_direct_mode != USE_AUTH_MDS) {
2560 dout("not using auth, setting for that now\n");
2561 req->r_direct_mode = USE_AUTH_MDS;
2562 __do_request(mdsc, req);
2563 mutex_unlock(&mdsc->mutex);
2564 goto out;
2565 } else {
2566 int mds = __choose_mds(mdsc, req);
2567 if (mds >= 0 && mds != req->r_session->s_mds) {
2568 dout("but auth changed, so resending\n");
2569 __do_request(mdsc, req);
2570 mutex_unlock(&mdsc->mutex);
2571 goto out;
2572 }
2573 }
2574 dout("have to return ESTALE on request %llu\n", req->r_tid);
2575 }
2576
2577
2578 if (head->safe) {
2579 set_bit(CEPH_MDS_R_GOT_SAFE, &req->r_req_flags);
2580 __unregister_request(mdsc, req);
2581
2582 if (test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags)) {
2583 /*
2584 * We already handled the unsafe response, now do the
2585 * cleanup. No need to examine the response; the MDS
2586 * doesn't include any result info in the safe
2587 * response. And even if it did, there is nothing
2588 * useful we could do with a revised return value.
2589 */
2590 dout("got safe reply %llu, mds%d\n", tid, mds);
2591
2592 /* last unsafe request during umount? */
2593 if (mdsc->stopping && !__get_oldest_req(mdsc))
2594 complete_all(&mdsc->safe_umount_waiters);
2595 mutex_unlock(&mdsc->mutex);
2596 goto out;
2597 }
2598 } else {
2599 set_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags);
2600 list_add_tail(&req->r_unsafe_item, &req->r_session->s_unsafe);
2601 if (req->r_unsafe_dir) {
2602 struct ceph_inode_info *ci =
2603 ceph_inode(req->r_unsafe_dir);
2604 spin_lock(&ci->i_unsafe_lock);
2605 list_add_tail(&req->r_unsafe_dir_item,
2606 &ci->i_unsafe_dirops);
2607 spin_unlock(&ci->i_unsafe_lock);
2608 }
2609 }
2610
2611 dout("handle_reply tid %lld result %d\n", tid, result);
2612 rinfo = &req->r_reply_info;
2613 err = parse_reply_info(msg, rinfo, session->s_con.peer_features);
2614 mutex_unlock(&mdsc->mutex);
2615
2616 mutex_lock(&session->s_mutex);
2617 if (err < 0) {
2618 pr_err("mdsc_handle_reply got corrupt reply mds%d(tid:%lld)\n", mds, tid);
2619 ceph_msg_dump(msg);
2620 goto out_err;
2621 }
2622
2623 /* snap trace */
2624 realm = NULL;
2625 if (rinfo->snapblob_len) {
2626 down_write(&mdsc->snap_rwsem);
2627 ceph_update_snap_trace(mdsc, rinfo->snapblob,
2628 rinfo->snapblob + rinfo->snapblob_len,
2629 le32_to_cpu(head->op) == CEPH_MDS_OP_RMSNAP,
2630 &realm);
2631 downgrade_write(&mdsc->snap_rwsem);
2632 } else {
2633 down_read(&mdsc->snap_rwsem);
2634 }
2635
2636 /* insert trace into our cache */
2637 mutex_lock(&req->r_fill_mutex);
2638 current->journal_info = req;
2639 err = ceph_fill_trace(mdsc->fsc->sb, req);
2640 if (err == 0) {
2641 if (result == 0 && (req->r_op == CEPH_MDS_OP_READDIR ||
2642 req->r_op == CEPH_MDS_OP_LSSNAP))
2643 ceph_readdir_prepopulate(req, req->r_session);
2644 ceph_unreserve_caps(mdsc, &req->r_caps_reservation);
2645 }
2646 current->journal_info = NULL;
2647 mutex_unlock(&req->r_fill_mutex);
2648
2649 up_read(&mdsc->snap_rwsem);
2650 if (realm)
2651 ceph_put_snap_realm(mdsc, realm);
2652
2653 if (err == 0 && req->r_target_inode &&
2654 test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags)) {
2655 struct ceph_inode_info *ci = ceph_inode(req->r_target_inode);
2656 spin_lock(&ci->i_unsafe_lock);
2657 list_add_tail(&req->r_unsafe_target_item, &ci->i_unsafe_iops);
2658 spin_unlock(&ci->i_unsafe_lock);
2659 }
2660 out_err:
2661 mutex_lock(&mdsc->mutex);
2662 if (!test_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags)) {
2663 if (err) {
2664 req->r_err = err;
2665 } else {
2666 req->r_reply = ceph_msg_get(msg);
2667 set_bit(CEPH_MDS_R_GOT_RESULT, &req->r_req_flags);
2668 }
2669 } else {
2670 dout("reply arrived after request %lld was aborted\n", tid);
2671 }
2672 mutex_unlock(&mdsc->mutex);
2673
2674 mutex_unlock(&session->s_mutex);
2675
2676 /* kick calling process */
2677 complete_request(mdsc, req);
2678 out:
2679 ceph_mdsc_put_request(req);
2680 return;
2681 }
2682
2683
2684
2685 /*
2686 * handle mds notification that our request has been forwarded.
2687 */
2688 static void handle_forward(struct ceph_mds_client *mdsc,
2689 struct ceph_mds_session *session,
2690 struct ceph_msg *msg)
2691 {
2692 struct ceph_mds_request *req;
2693 u64 tid = le64_to_cpu(msg->hdr.tid);
2694 u32 next_mds;
2695 u32 fwd_seq;
2696 int err = -EINVAL;
2697 void *p = msg->front.iov_base;
2698 void *end = p + msg->front.iov_len;
2699
2700 ceph_decode_need(&p, end, 2*sizeof(u32), bad);
2701 next_mds = ceph_decode_32(&p);
2702 fwd_seq = ceph_decode_32(&p);
2703
2704 mutex_lock(&mdsc->mutex);
2705 req = lookup_get_request(mdsc, tid);
2706 if (!req) {
2707 dout("forward tid %llu to mds%d - req dne\n", tid, next_mds);
2708 goto out; /* dup reply? */
2709 }
2710
2711 if (test_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags)) {
2712 dout("forward tid %llu aborted, unregistering\n", tid);
2713 __unregister_request(mdsc, req);
2714 } else if (fwd_seq <= req->r_num_fwd) {
2715 dout("forward tid %llu to mds%d - old seq %d <= %d\n",
2716 tid, next_mds, req->r_num_fwd, fwd_seq);
2717 } else {
2718 /* resend. forward race not possible; mds would drop */
2719 dout("forward tid %llu to mds%d (we resend)\n", tid, next_mds);
2720 BUG_ON(req->r_err);
2721 BUG_ON(test_bit(CEPH_MDS_R_GOT_RESULT, &req->r_req_flags));
2722 req->r_attempts = 0;
2723 req->r_num_fwd = fwd_seq;
2724 req->r_resend_mds = next_mds;
2725 put_request_session(req);
2726 __do_request(mdsc, req);
2727 }
2728 ceph_mdsc_put_request(req);
2729 out:
2730 mutex_unlock(&mdsc->mutex);
2731 return;
2732
2733 bad:
2734 pr_err("mdsc_handle_forward decode error err=%d\n", err);
2735 }
2736
2737 /*
2738 * handle a mds session control message
2739 */
2740 static void handle_session(struct ceph_mds_session *session,
2741 struct ceph_msg *msg)
2742 {
2743 struct ceph_mds_client *mdsc = session->s_mdsc;
2744 u32 op;
2745 u64 seq;
2746 int mds = session->s_mds;
2747 struct ceph_mds_session_head *h = msg->front.iov_base;
2748 int wake = 0;
2749
2750 /* decode */
2751 if (msg->front.iov_len != sizeof(*h))
2752 goto bad;
2753 op = le32_to_cpu(h->op);
2754 seq = le64_to_cpu(h->seq);
2755
2756 mutex_lock(&mdsc->mutex);
2757 if (op == CEPH_SESSION_CLOSE) {
2758 get_session(session);
2759 __unregister_session(mdsc, session);
2760 }
2761 /* FIXME: this ttl calculation is generous */
2762 session->s_ttl = jiffies + HZ*mdsc->mdsmap->m_session_autoclose;
2763 mutex_unlock(&mdsc->mutex);
2764
2765 mutex_lock(&session->s_mutex);
2766
2767 dout("handle_session mds%d %s %p state %s seq %llu\n",
2768 mds, ceph_session_op_name(op), session,
2769 ceph_session_state_name(session->s_state), seq);
2770
2771 if (session->s_state == CEPH_MDS_SESSION_HUNG) {
2772 session->s_state = CEPH_MDS_SESSION_OPEN;
2773 pr_info("mds%d came back\n", session->s_mds);
2774 }
2775
2776 switch (op) {
2777 case CEPH_SESSION_OPEN:
2778 if (session->s_state == CEPH_MDS_SESSION_RECONNECTING)
2779 pr_info("mds%d reconnect success\n", session->s_mds);
2780 session->s_state = CEPH_MDS_SESSION_OPEN;
2781 renewed_caps(mdsc, session, 0);
2782 wake = 1;
2783 if (mdsc->stopping)
2784 __close_session(mdsc, session);
2785 break;
2786
2787 case CEPH_SESSION_RENEWCAPS:
2788 if (session->s_renew_seq == seq)
2789 renewed_caps(mdsc, session, 1);
2790 break;
2791
2792 case CEPH_SESSION_CLOSE:
2793 if (session->s_state == CEPH_MDS_SESSION_RECONNECTING)
2794 pr_info("mds%d reconnect denied\n", session->s_mds);
2795 cleanup_session_requests(mdsc, session);
2796 remove_session_caps(session);
2797 wake = 2; /* for good measure */
2798 wake_up_all(&mdsc->session_close_wq);
2799 break;
2800
2801 case CEPH_SESSION_STALE:
2802 pr_info("mds%d caps went stale, renewing\n",
2803 session->s_mds);
2804 spin_lock(&session->s_gen_ttl_lock);
2805 session->s_cap_gen++;
2806 session->s_cap_ttl = jiffies - 1;
2807 spin_unlock(&session->s_gen_ttl_lock);
2808 send_renew_caps(mdsc, session);
2809 break;
2810
2811 case CEPH_SESSION_RECALL_STATE:
2812 ceph_trim_caps(mdsc, session, le32_to_cpu(h->max_caps));
2813 break;
2814
2815 case CEPH_SESSION_FLUSHMSG:
2816 send_flushmsg_ack(mdsc, session, seq);
2817 break;
2818
2819 case CEPH_SESSION_FORCE_RO:
2820 dout("force_session_readonly %p\n", session);
2821 spin_lock(&session->s_cap_lock);
2822 session->s_readonly = true;
2823 spin_unlock(&session->s_cap_lock);
2824 wake_up_session_caps(session, 0);
2825 break;
2826
2827 case CEPH_SESSION_REJECT:
2828 WARN_ON(session->s_state != CEPH_MDS_SESSION_OPENING);
2829 pr_info("mds%d rejected session\n", session->s_mds);
2830 session->s_state = CEPH_MDS_SESSION_REJECTED;
2831 cleanup_session_requests(mdsc, session);
2832 remove_session_caps(session);
2833 wake = 2; /* for good measure */
2834 break;
2835
2836 default:
2837 pr_err("mdsc_handle_session bad op %d mds%d\n", op, mds);
2838 WARN_ON(1);
2839 }
2840
2841 mutex_unlock(&session->s_mutex);
2842 if (wake) {
2843 mutex_lock(&mdsc->mutex);
2844 __wake_requests(mdsc, &session->s_waiting);
2845 if (wake == 2)
2846 kick_requests(mdsc, mds);
2847 mutex_unlock(&mdsc->mutex);
2848 }
2849 if (op == CEPH_SESSION_CLOSE)
2850 ceph_put_mds_session(session);
2851 return;
2852
2853 bad:
2854 pr_err("mdsc_handle_session corrupt message mds%d len %d\n", mds,
2855 (int)msg->front.iov_len);
2856 ceph_msg_dump(msg);
2857 return;
2858 }
2859
2860
2861 /*
2862 * called under session->mutex.
2863 */
2864 static void replay_unsafe_requests(struct ceph_mds_client *mdsc,
2865 struct ceph_mds_session *session)
2866 {
2867 struct ceph_mds_request *req, *nreq;
2868 struct rb_node *p;
2869 int err;
2870
2871 dout("replay_unsafe_requests mds%d\n", session->s_mds);
2872
2873 mutex_lock(&mdsc->mutex);
2874 list_for_each_entry_safe(req, nreq, &session->s_unsafe, r_unsafe_item) {
2875 err = __prepare_send_request(mdsc, req, session->s_mds, true);
2876 if (!err) {
2877 ceph_msg_get(req->r_request);
2878 ceph_con_send(&session->s_con, req->r_request);
2879 }
2880 }
2881
2882 /*
2883 * also re-send old requests when MDS enters reconnect stage. So that MDS
2884 * can process completed request in clientreplay stage.
2885 */
2886 p = rb_first(&mdsc->request_tree);
2887 while (p) {
2888 req = rb_entry(p, struct ceph_mds_request, r_node);
2889 p = rb_next(p);
2890 if (test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags))
2891 continue;
2892 if (req->r_attempts == 0)
2893 continue; /* only old requests */
2894 if (req->r_session &&
2895 req->r_session->s_mds == session->s_mds) {
2896 err = __prepare_send_request(mdsc, req,
2897 session->s_mds, true);
2898 if (!err) {
2899 ceph_msg_get(req->r_request);
2900 ceph_con_send(&session->s_con, req->r_request);
2901 }
2902 }
2903 }
2904 mutex_unlock(&mdsc->mutex);
2905 }
2906
2907 /*
2908 * Encode information about a cap for a reconnect with the MDS.
2909 */
2910 static int encode_caps_cb(struct inode *inode, struct ceph_cap *cap,
2911 void *arg)
2912 {
2913 union {
2914 struct ceph_mds_cap_reconnect v2;
2915 struct ceph_mds_cap_reconnect_v1 v1;
2916 } rec;
2917 struct ceph_inode_info *ci = cap->ci;
2918 struct ceph_reconnect_state *recon_state = arg;
2919 struct ceph_pagelist *pagelist = recon_state->pagelist;
2920 char *path;
2921 int pathlen, err;
2922 u64 pathbase;
2923 u64 snap_follows;
2924 struct dentry *dentry;
2925
2926 dout(" adding %p ino %llx.%llx cap %p %lld %s\n",
2927 inode, ceph_vinop(inode), cap, cap->cap_id,
2928 ceph_cap_string(cap->issued));
2929 err = ceph_pagelist_encode_64(pagelist, ceph_ino(inode));
2930 if (err)
2931 return err;
2932
2933 dentry = d_find_alias(inode);
2934 if (dentry) {
2935 path = ceph_mdsc_build_path(dentry, &pathlen, &pathbase, 0);
2936 if (IS_ERR(path)) {
2937 err = PTR_ERR(path);
2938 goto out_dput;
2939 }
2940 } else {
2941 path = NULL;
2942 pathlen = 0;
2943 pathbase = 0;
2944 }
2945
2946 spin_lock(&ci->i_ceph_lock);
2947 cap->seq = 0; /* reset cap seq */
2948 cap->issue_seq = 0; /* and issue_seq */
2949 cap->mseq = 0; /* and migrate_seq */
2950 cap->cap_gen = cap->session->s_cap_gen;
2951
2952 if (recon_state->msg_version >= 2) {
2953 rec.v2.cap_id = cpu_to_le64(cap->cap_id);
2954 rec.v2.wanted = cpu_to_le32(__ceph_caps_wanted(ci));
2955 rec.v2.issued = cpu_to_le32(cap->issued);
2956 rec.v2.snaprealm = cpu_to_le64(ci->i_snap_realm->ino);
2957 rec.v2.pathbase = cpu_to_le64(pathbase);
2958 rec.v2.flock_len = (__force __le32)
2959 ((ci->i_ceph_flags & CEPH_I_ERROR_FILELOCK) ? 0 : 1);
2960 } else {
2961 struct timespec ts;
2962 rec.v1.cap_id = cpu_to_le64(cap->cap_id);
2963 rec.v1.wanted = cpu_to_le32(__ceph_caps_wanted(ci));
2964 rec.v1.issued = cpu_to_le32(cap->issued);
2965 rec.v1.size = cpu_to_le64(inode->i_size);
2966 ts = timespec64_to_timespec(inode->i_mtime);
2967 ceph_encode_timespec(&rec.v1.mtime, &ts);
2968 ts = timespec64_to_timespec(inode->i_atime);
2969 ceph_encode_timespec(&rec.v1.atime, &ts);
2970 rec.v1.snaprealm = cpu_to_le64(ci->i_snap_realm->ino);
2971 rec.v1.pathbase = cpu_to_le64(pathbase);
2972 }
2973
2974 if (list_empty(&ci->i_cap_snaps)) {
2975 snap_follows = ci->i_head_snapc ? ci->i_head_snapc->seq : 0;
2976 } else {
2977 struct ceph_cap_snap *capsnap =
2978 list_first_entry(&ci->i_cap_snaps,
2979 struct ceph_cap_snap, ci_item);
2980 snap_follows = capsnap->follows;
2981 }
2982 spin_unlock(&ci->i_ceph_lock);
2983
2984 if (recon_state->msg_version >= 2) {
2985 int num_fcntl_locks, num_flock_locks;
2986 struct ceph_filelock *flocks = NULL;
2987 size_t struct_len, total_len = 0;
2988 u8 struct_v = 0;
2989
2990 encode_again:
2991 if (rec.v2.flock_len) {
2992 ceph_count_locks(inode, &num_fcntl_locks, &num_flock_locks);
2993 } else {
2994 num_fcntl_locks = 0;
2995 num_flock_locks = 0;
2996 }
2997 if (num_fcntl_locks + num_flock_locks > 0) {
2998 flocks = kmalloc_array(num_fcntl_locks + num_flock_locks,
2999 sizeof(struct ceph_filelock),
3000 GFP_NOFS);
3001 if (!flocks) {
3002 err = -ENOMEM;
3003 goto out_free;
3004 }
3005 err = ceph_encode_locks_to_buffer(inode, flocks,
3006 num_fcntl_locks,
3007 num_flock_locks);
3008 if (err) {
3009 kfree(flocks);
3010 flocks = NULL;
3011 if (err == -ENOSPC)
3012 goto encode_again;
3013 goto out_free;
3014 }
3015 } else {
3016 kfree(flocks);
3017 flocks = NULL;
3018 }
3019
3020 if (recon_state->msg_version >= 3) {
3021 /* version, compat_version and struct_len */
3022 total_len = 2 * sizeof(u8) + sizeof(u32);
3023 struct_v = 2;
3024 }
3025 /*
3026 * number of encoded locks is stable, so copy to pagelist
3027 */
3028 struct_len = 2 * sizeof(u32) +
3029 (num_fcntl_locks + num_flock_locks) *
3030 sizeof(struct ceph_filelock);
3031 rec.v2.flock_len = cpu_to_le32(struct_len);
3032
3033 struct_len += sizeof(rec.v2);
3034 struct_len += sizeof(u32) + pathlen;
3035
3036 if (struct_v >= 2)
3037 struct_len += sizeof(u64); /* snap_follows */
3038
3039 total_len += struct_len;
3040 err = ceph_pagelist_reserve(pagelist, total_len);
3041
3042 if (!err) {
3043 if (recon_state->msg_version >= 3) {
3044 ceph_pagelist_encode_8(pagelist, struct_v);
3045 ceph_pagelist_encode_8(pagelist, 1);
3046 ceph_pagelist_encode_32(pagelist, struct_len);
3047 }
3048 ceph_pagelist_encode_string(pagelist, path, pathlen);
3049 ceph_pagelist_append(pagelist, &rec, sizeof(rec.v2));
3050 ceph_locks_to_pagelist(flocks, pagelist,
3051 num_fcntl_locks,
3052 num_flock_locks);
3053 if (struct_v >= 2)
3054 ceph_pagelist_encode_64(pagelist, snap_follows);
3055 }
3056 kfree(flocks);
3057 } else {
3058 size_t size = sizeof(u32) + pathlen + sizeof(rec.v1);
3059 err = ceph_pagelist_reserve(pagelist, size);
3060 if (!err) {
3061 ceph_pagelist_encode_string(pagelist, path, pathlen);
3062 ceph_pagelist_append(pagelist, &rec, sizeof(rec.v1));
3063 }
3064 }
3065
3066 recon_state->nr_caps++;
3067 out_free:
3068 kfree(path);
3069 out_dput:
3070 dput(dentry);
3071 return err;
3072 }
3073
3074
3075 /*
3076 * If an MDS fails and recovers, clients need to reconnect in order to
3077 * reestablish shared state. This includes all caps issued through
3078 * this session _and_ the snap_realm hierarchy. Because it's not
3079 * clear which snap realms the mds cares about, we send everything we
3080 * know about.. that ensures we'll then get any new info the
3081 * recovering MDS might have.
3082 *
3083 * This is a relatively heavyweight operation, but it's rare.
3084 *
3085 * called with mdsc->mutex held.
3086 */
3087 static void send_mds_reconnect(struct ceph_mds_client *mdsc,
3088 struct ceph_mds_session *session)
3089 {
3090 struct ceph_msg *reply;
3091 struct rb_node *p;
3092 int mds = session->s_mds;
3093 int err = -ENOMEM;
3094 int s_nr_caps;
3095 struct ceph_pagelist *pagelist;
3096 struct ceph_reconnect_state recon_state;
3097 LIST_HEAD(dispose);
3098
3099 pr_info("mds%d reconnect start\n", mds);
3100
3101 pagelist = kmalloc(sizeof(*pagelist), GFP_NOFS);
3102 if (!pagelist)
3103 goto fail_nopagelist;
3104 ceph_pagelist_init(pagelist);
3105
3106 reply = ceph_msg_new(CEPH_MSG_CLIENT_RECONNECT, 0, GFP_NOFS, false);
3107 if (!reply)
3108 goto fail_nomsg;
3109
3110 mutex_lock(&session->s_mutex);
3111 session->s_state = CEPH_MDS_SESSION_RECONNECTING;
3112 session->s_seq = 0;
3113
3114 dout("session %p state %s\n", session,
3115 ceph_session_state_name(session->s_state));
3116
3117 spin_lock(&session->s_gen_ttl_lock);
3118 session->s_cap_gen++;
3119 spin_unlock(&session->s_gen_ttl_lock);
3120
3121 spin_lock(&session->s_cap_lock);
3122 /* don't know if session is readonly */
3123 session->s_readonly = 0;
3124 /*
3125 * notify __ceph_remove_cap() that we are composing cap reconnect.
3126 * If a cap get released before being added to the cap reconnect,
3127 * __ceph_remove_cap() should skip queuing cap release.
3128 */
3129 session->s_cap_reconnect = 1;
3130 /* drop old cap expires; we're about to reestablish that state */
3131 detach_cap_releases(session, &dispose);
3132 spin_unlock(&session->s_cap_lock);
3133 dispose_cap_releases(mdsc, &dispose);
3134
3135 /* trim unused caps to reduce MDS's cache rejoin time */
3136 if (mdsc->fsc->sb->s_root)
3137 shrink_dcache_parent(mdsc->fsc->sb->s_root);
3138
3139 ceph_con_close(&session->s_con);
3140 ceph_con_open(&session->s_con,
3141 CEPH_ENTITY_TYPE_MDS, mds,
3142 ceph_mdsmap_get_addr(mdsc->mdsmap, mds));
3143
3144 /* replay unsafe requests */
3145 replay_unsafe_requests(mdsc, session);
3146
3147 down_read(&mdsc->snap_rwsem);
3148
3149 /* traverse this session's caps */
3150 s_nr_caps = session->s_nr_caps;
3151 err = ceph_pagelist_encode_32(pagelist, s_nr_caps);
3152 if (err)
3153 goto fail;
3154
3155 recon_state.nr_caps = 0;
3156 recon_state.pagelist = pagelist;
3157 if (session->s_con.peer_features & CEPH_FEATURE_MDSENC)
3158 recon_state.msg_version = 3;
3159 else if (session->s_con.peer_features & CEPH_FEATURE_FLOCK)
3160 recon_state.msg_version = 2;
3161 else
3162 recon_state.msg_version = 1;
3163 err = iterate_session_caps(session, encode_caps_cb, &recon_state);
3164 if (err < 0)
3165 goto fail;
3166
3167 spin_lock(&session->s_cap_lock);
3168 session->s_cap_reconnect = 0;
3169 spin_unlock(&session->s_cap_lock);
3170
3171 /*
3172 * snaprealms. we provide mds with the ino, seq (version), and
3173 * parent for all of our realms. If the mds has any newer info,
3174 * it will tell us.
3175 */
3176 for (p = rb_first(&mdsc->snap_realms); p; p = rb_next(p)) {
3177 struct ceph_snap_realm *realm =
3178 rb_entry(p, struct ceph_snap_realm, node);
3179 struct ceph_mds_snaprealm_reconnect sr_rec;
3180
3181 dout(" adding snap realm %llx seq %lld parent %llx\n",
3182 realm->ino, realm->seq, realm->parent_ino);
3183 sr_rec.ino = cpu_to_le64(realm->ino);
3184 sr_rec.seq = cpu_to_le64(realm->seq);
3185 sr_rec.parent = cpu_to_le64(realm->parent_ino);
3186 err = ceph_pagelist_append(pagelist, &sr_rec, sizeof(sr_rec));
3187 if (err)
3188 goto fail;
3189 }
3190
3191 reply->hdr.version = cpu_to_le16(recon_state.msg_version);
3192
3193 /* raced with cap release? */
3194 if (s_nr_caps != recon_state.nr_caps) {
3195 struct page *page = list_first_entry(&pagelist->head,
3196 struct page, lru);
3197 __le32 *addr = kmap_atomic(page);
3198 *addr = cpu_to_le32(recon_state.nr_caps);
3199 kunmap_atomic(addr);
3200 }
3201
3202 reply->hdr.data_len = cpu_to_le32(pagelist->length);
3203 ceph_msg_data_add_pagelist(reply, pagelist);
3204
3205 ceph_early_kick_flushing_caps(mdsc, session);
3206
3207 ceph_con_send(&session->s_con, reply);
3208
3209 mutex_unlock(&session->s_mutex);
3210
3211 mutex_lock(&mdsc->mutex);
3212 __wake_requests(mdsc, &session->s_waiting);
3213 mutex_unlock(&mdsc->mutex);
3214
3215 up_read(&mdsc->snap_rwsem);
3216 return;
3217
3218 fail:
3219 ceph_msg_put(reply);
3220 up_read(&mdsc->snap_rwsem);
3221 mutex_unlock(&session->s_mutex);
3222 fail_nomsg:
3223 ceph_pagelist_release(pagelist);
3224 fail_nopagelist:
3225 pr_err("error %d preparing reconnect for mds%d\n", err, mds);
3226 return;
3227 }
3228
3229
3230 /*
3231 * compare old and new mdsmaps, kicking requests
3232 * and closing out old connections as necessary
3233 *
3234 * called under mdsc->mutex.
3235 */
3236 static void check_new_map(struct ceph_mds_client *mdsc,
3237 struct ceph_mdsmap *newmap,
3238 struct ceph_mdsmap *oldmap)
3239 {
3240 int i;
3241 int oldstate, newstate;
3242 struct ceph_mds_session *s;
3243
3244 dout("check_new_map new %u old %u\n",
3245 newmap->m_epoch, oldmap->m_epoch);
3246
3247 for (i = 0; i < oldmap->m_num_mds && i < mdsc->max_sessions; i++) {
3248 if (!mdsc->sessions[i])
3249 continue;
3250 s = mdsc->sessions[i];
3251 oldstate = ceph_mdsmap_get_state(oldmap, i);
3252 newstate = ceph_mdsmap_get_state(newmap, i);
3253
3254 dout("check_new_map mds%d state %s%s -> %s%s (session %s)\n",
3255 i, ceph_mds_state_name(oldstate),
3256 ceph_mdsmap_is_laggy(oldmap, i) ? " (laggy)" : "",
3257 ceph_mds_state_name(newstate),
3258 ceph_mdsmap_is_laggy(newmap, i) ? " (laggy)" : "",
3259 ceph_session_state_name(s->s_state));
3260
3261 if (i >= newmap->m_num_mds ||
3262 memcmp(ceph_mdsmap_get_addr(oldmap, i),
3263 ceph_mdsmap_get_addr(newmap, i),
3264 sizeof(struct ceph_entity_addr))) {
3265 if (s->s_state == CEPH_MDS_SESSION_OPENING) {
3266 /* the session never opened, just close it
3267 * out now */
3268 get_session(s);
3269 __unregister_session(mdsc, s);
3270 __wake_requests(mdsc, &s->s_waiting);
3271 ceph_put_mds_session(s);
3272 } else if (i >= newmap->m_num_mds) {
3273 /* force close session for stopped mds */
3274 get_session(s);
3275 __unregister_session(mdsc, s);
3276 __wake_requests(mdsc, &s->s_waiting);
3277 kick_requests(mdsc, i);
3278 mutex_unlock(&mdsc->mutex);
3279
3280 mutex_lock(&s->s_mutex);
3281 cleanup_session_requests(mdsc, s);
3282 remove_session_caps(s);
3283 mutex_unlock(&s->s_mutex);
3284
3285 ceph_put_mds_session(s);
3286
3287 mutex_lock(&mdsc->mutex);
3288 } else {
3289 /* just close it */
3290 mutex_unlock(&mdsc->mutex);
3291 mutex_lock(&s->s_mutex);
3292 mutex_lock(&mdsc->mutex);
3293 ceph_con_close(&s->s_con);
3294 mutex_unlock(&s->s_mutex);
3295 s->s_state = CEPH_MDS_SESSION_RESTARTING;
3296 }
3297 } else if (oldstate == newstate) {
3298 continue; /* nothing new with this mds */
3299 }
3300
3301 /*
3302 * send reconnect?
3303 */
3304 if (s->s_state == CEPH_MDS_SESSION_RESTARTING &&
3305 newstate >= CEPH_MDS_STATE_RECONNECT) {
3306 mutex_unlock(&mdsc->mutex);
3307 send_mds_reconnect(mdsc, s);
3308 mutex_lock(&mdsc->mutex);
3309 }
3310
3311 /*
3312 * kick request on any mds that has gone active.
3313 */
3314 if (oldstate < CEPH_MDS_STATE_ACTIVE &&
3315 newstate >= CEPH_MDS_STATE_ACTIVE) {
3316 if (oldstate != CEPH_MDS_STATE_CREATING &&
3317 oldstate != CEPH_MDS_STATE_STARTING)
3318 pr_info("mds%d recovery completed\n", s->s_mds);
3319 kick_requests(mdsc, i);
3320 ceph_kick_flushing_caps(mdsc, s);
3321 wake_up_session_caps(s, 1);
3322 }
3323 }
3324
3325 for (i = 0; i < newmap->m_num_mds && i < mdsc->max_sessions; i++) {
3326 s = mdsc->sessions[i];
3327 if (!s)
3328 continue;
3329 if (!ceph_mdsmap_is_laggy(newmap, i))
3330 continue;
3331 if (s->s_state == CEPH_MDS_SESSION_OPEN ||
3332 s->s_state == CEPH_MDS_SESSION_HUNG ||
3333 s->s_state == CEPH_MDS_SESSION_CLOSING) {
3334 dout(" connecting to export targets of laggy mds%d\n",
3335 i);
3336 __open_export_target_sessions(mdsc, s);
3337 }
3338 }
3339 }
3340
3341
3342
3343 /*
3344 * leases
3345 */
3346
3347 /*
3348 * caller must hold session s_mutex, dentry->d_lock
3349 */
3350 void __ceph_mdsc_drop_dentry_lease(struct dentry *dentry)
3351 {
3352 struct ceph_dentry_info *di = ceph_dentry(dentry);
3353
3354 ceph_put_mds_session(di->lease_session);
3355 di->lease_session = NULL;
3356 }
3357
3358 static void handle_lease(struct ceph_mds_client *mdsc,
3359 struct ceph_mds_session *session,
3360 struct ceph_msg *msg)
3361 {
3362 struct super_block *sb = mdsc->fsc->sb;
3363 struct inode *inode;
3364 struct dentry *parent, *dentry;
3365 struct ceph_dentry_info *di;
3366 int mds = session->s_mds;
3367 struct ceph_mds_lease *h = msg->front.iov_base;
3368 u32 seq;
3369 struct ceph_vino vino;
3370 struct qstr dname;
3371 int release = 0;
3372
3373 dout("handle_lease from mds%d\n", mds);
3374
3375 /* decode */
3376 if (msg->front.iov_len < sizeof(*h) + sizeof(u32))
3377 goto bad;
3378 vino.ino = le64_to_cpu(h->ino);
3379 vino.snap = CEPH_NOSNAP;
3380 seq = le32_to_cpu(h->seq);
3381 dname.name = (void *)h + sizeof(*h) + sizeof(u32);
3382 dname.len = msg->front.iov_len - sizeof(*h) - sizeof(u32);
3383 if (dname.len != get_unaligned_le32(h+1))
3384 goto bad;
3385
3386 /* lookup inode */
3387 inode = ceph_find_inode(sb, vino);
3388 dout("handle_lease %s, ino %llx %p %.*s\n",
3389 ceph_lease_op_name(h->action), vino.ino, inode,
3390 dname.len, dname.name);
3391
3392 mutex_lock(&session->s_mutex);
3393 session->s_seq++;
3394
3395 if (!inode) {
3396 dout("handle_lease no inode %llx\n", vino.ino);
3397 goto release;
3398 }
3399
3400 /* dentry */
3401 parent = d_find_alias(inode);
3402 if (!parent) {
3403 dout("no parent dentry on inode %p\n", inode);
3404 WARN_ON(1);
3405 goto release; /* hrm... */
3406 }
3407 dname.hash = full_name_hash(parent, dname.name, dname.len);
3408 dentry = d_lookup(parent, &dname);
3409 dput(parent);
3410 if (!dentry)
3411 goto release;
3412
3413 spin_lock(&dentry->d_lock);
3414 di = ceph_dentry(dentry);
3415 switch (h->action) {
3416 case CEPH_MDS_LEASE_REVOKE:
3417 if (di->lease_session == session) {
3418 if (ceph_seq_cmp(di->lease_seq, seq) > 0)
3419 h->seq = cpu_to_le32(di->lease_seq);
3420 __ceph_mdsc_drop_dentry_lease(dentry);
3421 }
3422 release = 1;
3423 break;
3424
3425 case CEPH_MDS_LEASE_RENEW:
3426 if (di->lease_session == session &&
3427 di->lease_gen == session->s_cap_gen &&
3428 di->lease_renew_from &&
3429 di->lease_renew_after == 0) {
3430 unsigned long duration =
3431 msecs_to_jiffies(le32_to_cpu(h->duration_ms));
3432
3433 di->lease_seq = seq;
3434 di->time = di->lease_renew_from + duration;
3435 di->lease_renew_after = di->lease_renew_from +
3436 (duration >> 1);
3437 di->lease_renew_from = 0;
3438 }
3439 break;
3440 }
3441 spin_unlock(&dentry->d_lock);
3442 dput(dentry);
3443
3444 if (!release)
3445 goto out;
3446
3447 release:
3448 /* let's just reuse the same message */
3449 h->action = CEPH_MDS_LEASE_REVOKE_ACK;
3450 ceph_msg_get(msg);
3451 ceph_con_send(&session->s_con, msg);
3452
3453 out:
3454 iput(inode);
3455 mutex_unlock(&session->s_mutex);
3456 return;
3457
3458 bad:
3459 pr_err("corrupt lease message\n");
3460 ceph_msg_dump(msg);
3461 }
3462
3463 void ceph_mdsc_lease_send_msg(struct ceph_mds_session *session,
3464 struct inode *inode,
3465 struct dentry *dentry, char action,
3466 u32 seq)
3467 {
3468 struct ceph_msg *msg;
3469 struct ceph_mds_lease *lease;
3470 int len = sizeof(*lease) + sizeof(u32);
3471 int dnamelen = 0;
3472
3473 dout("lease_send_msg inode %p dentry %p %s to mds%d\n",
3474 inode, dentry, ceph_lease_op_name(action), session->s_mds);
3475 dnamelen = dentry->d_name.len;
3476 len += dnamelen;
3477
3478 msg = ceph_msg_new(CEPH_MSG_CLIENT_LEASE, len, GFP_NOFS, false);
3479 if (!msg)
3480 return;
3481 lease = msg->front.iov_base;
3482 lease->action = action;
3483 lease->ino = cpu_to_le64(ceph_vino(inode).ino);
3484 lease->first = lease->last = cpu_to_le64(ceph_vino(inode).snap);
3485 lease->seq = cpu_to_le32(seq);
3486 put_unaligned_le32(dnamelen, lease + 1);
3487 memcpy((void *)(lease + 1) + 4, dentry->d_name.name, dnamelen);
3488
3489 /*
3490 * if this is a preemptive lease RELEASE, no need to
3491 * flush request stream, since the actual request will
3492 * soon follow.
3493 */
3494 msg->more_to_follow = (action == CEPH_MDS_LEASE_RELEASE);
3495
3496 ceph_con_send(&session->s_con, msg);
3497 }
3498
3499 /*
3500 * lock unlock sessions, to wait ongoing session activities
3501 */
3502 static void lock_unlock_sessions(struct ceph_mds_client *mdsc)
3503 {
3504 int i;
3505
3506 mutex_lock(&mdsc->mutex);
3507 for (i = 0; i < mdsc->max_sessions; i++) {
3508 struct ceph_mds_session *s = __ceph_lookup_mds_session(mdsc, i);
3509 if (!s)
3510 continue;
3511 mutex_unlock(&mdsc->mutex);
3512 mutex_lock(&s->s_mutex);
3513 mutex_unlock(&s->s_mutex);
3514 ceph_put_mds_session(s);
3515 mutex_lock(&mdsc->mutex);
3516 }
3517 mutex_unlock(&mdsc->mutex);
3518 }
3519
3520
3521
3522 /*
3523 * delayed work -- periodically trim expired leases, renew caps with mds
3524 */
3525 static void schedule_delayed(struct ceph_mds_client *mdsc)
3526 {
3527 int delay = 5;
3528 unsigned hz = round_jiffies_relative(HZ * delay);
3529 schedule_delayed_work(&mdsc->delayed_work, hz);
3530 }
3531
3532 static void delayed_work(struct work_struct *work)
3533 {
3534 int i;
3535 struct ceph_mds_client *mdsc =
3536 container_of(work, struct ceph_mds_client, delayed_work.work);
3537 int renew_interval;
3538 int renew_caps;
3539
3540 dout("mdsc delayed_work\n");
3541 ceph_check_delayed_caps(mdsc);
3542
3543 mutex_lock(&mdsc->mutex);
3544 renew_interval = mdsc->mdsmap->m_session_timeout >> 2;
3545 renew_caps = time_after_eq(jiffies, HZ*renew_interval +
3546 mdsc->last_renew_caps);
3547 if (renew_caps)
3548 mdsc->last_renew_caps = jiffies;
3549
3550 for (i = 0; i < mdsc->max_sessions; i++) {
3551 struct ceph_mds_session *s = __ceph_lookup_mds_session(mdsc, i);
3552 if (!s)
3553 continue;
3554 if (s->s_state == CEPH_MDS_SESSION_CLOSING) {
3555 dout("resending session close request for mds%d\n",
3556 s->s_mds);
3557 request_close_session(mdsc, s);
3558 ceph_put_mds_session(s);
3559 continue;
3560 }
3561 if (s->s_ttl && time_after(jiffies, s->s_ttl)) {
3562 if (s->s_state == CEPH_MDS_SESSION_OPEN) {
3563 s->s_state = CEPH_MDS_SESSION_HUNG;
3564 pr_info("mds%d hung\n", s->s_mds);
3565 }
3566 }
3567 if (s->s_state < CEPH_MDS_SESSION_OPEN) {
3568 /* this mds is failed or recovering, just wait */
3569 ceph_put_mds_session(s);
3570 continue;
3571 }
3572 mutex_unlock(&mdsc->mutex);
3573
3574 mutex_lock(&s->s_mutex);
3575 if (renew_caps)
3576 send_renew_caps(mdsc, s);
3577 else
3578 ceph_con_keepalive(&s->s_con);
3579 if (s->s_state == CEPH_MDS_SESSION_OPEN ||
3580 s->s_state == CEPH_MDS_SESSION_HUNG)
3581 ceph_send_cap_releases(mdsc, s);
3582 mutex_unlock(&s->s_mutex);
3583 ceph_put_mds_session(s);
3584
3585 mutex_lock(&mdsc->mutex);
3586 }
3587 mutex_unlock(&mdsc->mutex);
3588
3589 schedule_delayed(mdsc);
3590 }
3591
3592 int ceph_mdsc_init(struct ceph_fs_client *fsc)
3593
3594 {
3595 struct ceph_mds_client *mdsc;
3596
3597 mdsc = kzalloc(sizeof(struct ceph_mds_client), GFP_NOFS);
3598 if (!mdsc)
3599 return -ENOMEM;
3600 mdsc->fsc = fsc;
3601 mutex_init(&mdsc->mutex);
3602 mdsc->mdsmap = kzalloc(sizeof(*mdsc->mdsmap), GFP_NOFS);
3603 if (!mdsc->mdsmap) {
3604 kfree(mdsc);
3605 return -ENOMEM;
3606 }
3607
3608 fsc->mdsc = mdsc;
3609 init_completion(&mdsc->safe_umount_waiters);
3610 init_waitqueue_head(&mdsc->session_close_wq);
3611 INIT_LIST_HEAD(&mdsc->waiting_for_map);
3612 mdsc->sessions = NULL;
3613 atomic_set(&mdsc->num_sessions, 0);
3614 mdsc->max_sessions = 0;
3615 mdsc->stopping = 0;
3616 atomic64_set(&mdsc->quotarealms_count, 0);
3617 mdsc->last_snap_seq = 0;
3618 init_rwsem(&mdsc->snap_rwsem);
3619 mdsc->snap_realms = RB_ROOT;
3620 INIT_LIST_HEAD(&mdsc->snap_empty);
3621 spin_lock_init(&mdsc->snap_empty_lock);
3622 mdsc->last_tid = 0;
3623 mdsc->oldest_tid = 0;
3624 mdsc->request_tree = RB_ROOT;
3625 INIT_DELAYED_WORK(&mdsc->delayed_work, delayed_work);
3626 mdsc->last_renew_caps = jiffies;
3627 INIT_LIST_HEAD(&mdsc->cap_delay_list);
3628 spin_lock_init(&mdsc->cap_delay_lock);
3629 INIT_LIST_HEAD(&mdsc->snap_flush_list);
3630 spin_lock_init(&mdsc->snap_flush_lock);
3631 mdsc->last_cap_flush_tid = 1;
3632 INIT_LIST_HEAD(&mdsc->cap_flush_list);
3633 INIT_LIST_HEAD(&mdsc->cap_dirty);
3634 INIT_LIST_HEAD(&mdsc->cap_dirty_migrating);
3635 mdsc->num_cap_flushing = 0;
3636 spin_lock_init(&mdsc->cap_dirty_lock);
3637 init_waitqueue_head(&mdsc->cap_flushing_wq);
3638 spin_lock_init(&mdsc->dentry_lru_lock);
3639 INIT_LIST_HEAD(&mdsc->dentry_lru);
3640
3641 ceph_caps_init(mdsc);
3642 ceph_adjust_min_caps(mdsc, fsc->min_caps);
3643
3644 init_rwsem(&mdsc->pool_perm_rwsem);
3645 mdsc->pool_perm_tree = RB_ROOT;
3646
3647 strscpy(mdsc->nodename, utsname()->nodename,
3648 sizeof(mdsc->nodename));
3649 return 0;
3650 }
3651
3652 /*
3653 * Wait for safe replies on open mds requests. If we time out, drop
3654 * all requests from the tree to avoid dangling dentry refs.
3655 */
3656 static void wait_requests(struct ceph_mds_client *mdsc)
3657 {
3658 struct ceph_options *opts = mdsc->fsc->client->options;
3659 struct ceph_mds_request *req;
3660
3661 mutex_lock(&mdsc->mutex);
3662 if (__get_oldest_req(mdsc)) {
3663 mutex_unlock(&mdsc->mutex);
3664
3665 dout("wait_requests waiting for requests\n");
3666 wait_for_completion_timeout(&mdsc->safe_umount_waiters,
3667 ceph_timeout_jiffies(opts->mount_timeout));
3668
3669 /* tear down remaining requests */
3670 mutex_lock(&mdsc->mutex);
3671 while ((req = __get_oldest_req(mdsc))) {
3672 dout("wait_requests timed out on tid %llu\n",
3673 req->r_tid);
3674 __unregister_request(mdsc, req);
3675 }
3676 }
3677 mutex_unlock(&mdsc->mutex);
3678 dout("wait_requests done\n");
3679 }
3680
3681 /*
3682 * called before mount is ro, and before dentries are torn down.
3683 * (hmm, does this still race with new lookups?)
3684 */
3685 void ceph_mdsc_pre_umount(struct ceph_mds_client *mdsc)
3686 {
3687 dout("pre_umount\n");
3688 mdsc->stopping = 1;
3689
3690 lock_unlock_sessions(mdsc);
3691 ceph_flush_dirty_caps(mdsc);
3692 wait_requests(mdsc);
3693
3694 /*
3695 * wait for reply handlers to drop their request refs and
3696 * their inode/dcache refs
3697 */
3698 ceph_msgr_flush();
3699 }
3700
3701 /*
3702 * wait for all write mds requests to flush.
3703 */
3704 static void wait_unsafe_requests(struct ceph_mds_client *mdsc, u64 want_tid)
3705 {
3706 struct ceph_mds_request *req = NULL, *nextreq;
3707 struct rb_node *n;
3708
3709 mutex_lock(&mdsc->mutex);
3710 dout("wait_unsafe_requests want %lld\n", want_tid);
3711 restart:
3712 req = __get_oldest_req(mdsc);
3713 while (req && req->r_tid <= want_tid) {
3714 /* find next request */
3715 n = rb_next(&req->r_node);
3716 if (n)
3717 nextreq = rb_entry(n, struct ceph_mds_request, r_node);
3718 else
3719 nextreq = NULL;
3720 if (req->r_op != CEPH_MDS_OP_SETFILELOCK &&
3721 (req->r_op & CEPH_MDS_OP_WRITE)) {
3722 /* write op */
3723 ceph_mdsc_get_request(req);
3724 if (nextreq)
3725 ceph_mdsc_get_request(nextreq);
3726 mutex_unlock(&mdsc->mutex);
3727 dout("wait_unsafe_requests wait on %llu (want %llu)\n",
3728 req->r_tid, want_tid);
3729 wait_for_completion(&req->r_safe_completion);
3730 mutex_lock(&mdsc->mutex);
3731 ceph_mdsc_put_request(req);
3732 if (!nextreq)
3733 break; /* next dne before, so we're done! */
3734 if (RB_EMPTY_NODE(&nextreq->r_node)) {
3735 /* next request was removed from tree */
3736 ceph_mdsc_put_request(nextreq);
3737 goto restart;
3738 }
3739 ceph_mdsc_put_request(nextreq); /* won't go away */
3740 }
3741 req = nextreq;
3742 }
3743 mutex_unlock(&mdsc->mutex);
3744 dout("wait_unsafe_requests done\n");
3745 }
3746
3747 void ceph_mdsc_sync(struct ceph_mds_client *mdsc)
3748 {
3749 u64 want_tid, want_flush;
3750
3751 if (READ_ONCE(mdsc->fsc->mount_state) == CEPH_MOUNT_SHUTDOWN)
3752 return;
3753
3754 dout("sync\n");
3755 mutex_lock(&mdsc->mutex);
3756 want_tid = mdsc->last_tid;
3757 mutex_unlock(&mdsc->mutex);
3758
3759 ceph_flush_dirty_caps(mdsc);
3760 spin_lock(&mdsc->cap_dirty_lock);
3761 want_flush = mdsc->last_cap_flush_tid;
3762 if (!list_empty(&mdsc->cap_flush_list)) {
3763 struct ceph_cap_flush *cf =
3764 list_last_entry(&mdsc->cap_flush_list,
3765 struct ceph_cap_flush, g_list);
3766 cf->wake = true;
3767 }
3768 spin_unlock(&mdsc->cap_dirty_lock);
3769
3770 dout("sync want tid %lld flush_seq %lld\n",
3771 want_tid, want_flush);
3772
3773 wait_unsafe_requests(mdsc, want_tid);
3774 wait_caps_flush(mdsc, want_flush);
3775 }
3776
3777 /*
3778 * true if all sessions are closed, or we force unmount
3779 */
3780 static bool done_closing_sessions(struct ceph_mds_client *mdsc, int skipped)
3781 {
3782 if (READ_ONCE(mdsc->fsc->mount_state) == CEPH_MOUNT_SHUTDOWN)
3783 return true;
3784 return atomic_read(&mdsc->num_sessions) <= skipped;
3785 }
3786
3787 /*
3788 * called after sb is ro.
3789 */
3790 void ceph_mdsc_close_sessions(struct ceph_mds_client *mdsc)
3791 {
3792 struct ceph_options *opts = mdsc->fsc->client->options;
3793 struct ceph_mds_session *session;
3794 int i;
3795 int skipped = 0;
3796
3797 dout("close_sessions\n");
3798
3799 /* close sessions */
3800 mutex_lock(&mdsc->mutex);
3801 for (i = 0; i < mdsc->max_sessions; i++) {
3802 session = __ceph_lookup_mds_session(mdsc, i);
3803 if (!session)
3804 continue;
3805 mutex_unlock(&mdsc->mutex);
3806 mutex_lock(&session->s_mutex);
3807 if (__close_session(mdsc, session) <= 0)
3808 skipped++;
3809 mutex_unlock(&session->s_mutex);
3810 ceph_put_mds_session(session);
3811 mutex_lock(&mdsc->mutex);
3812 }
3813 mutex_unlock(&mdsc->mutex);
3814
3815 dout("waiting for sessions to close\n");
3816 wait_event_timeout(mdsc->session_close_wq,
3817 done_closing_sessions(mdsc, skipped),
3818 ceph_timeout_jiffies(opts->mount_timeout));
3819
3820 /* tear down remaining sessions */
3821 mutex_lock(&mdsc->mutex);
3822 for (i = 0; i < mdsc->max_sessions; i++) {
3823 if (mdsc->sessions[i]) {
3824 session = get_session(mdsc->sessions[i]);
3825 __unregister_session(mdsc, session);
3826 mutex_unlock(&mdsc->mutex);
3827 mutex_lock(&session->s_mutex);
3828 remove_session_caps(session);
3829 mutex_unlock(&session->s_mutex);
3830 ceph_put_mds_session(session);
3831 mutex_lock(&mdsc->mutex);
3832 }
3833 }
3834 WARN_ON(!list_empty(&mdsc->cap_delay_list));
3835 mutex_unlock(&mdsc->mutex);
3836
3837 ceph_cleanup_empty_realms(mdsc);
3838
3839 cancel_delayed_work_sync(&mdsc->delayed_work); /* cancel timer */
3840
3841 dout("stopped\n");
3842 }
3843
3844 void ceph_mdsc_force_umount(struct ceph_mds_client *mdsc)
3845 {
3846 struct ceph_mds_session *session;
3847 int mds;
3848
3849 dout("force umount\n");
3850
3851 mutex_lock(&mdsc->mutex);
3852 for (mds = 0; mds < mdsc->max_sessions; mds++) {
3853 session = __ceph_lookup_mds_session(mdsc, mds);
3854 if (!session)
3855 continue;
3856 mutex_unlock(&mdsc->mutex);
3857 mutex_lock(&session->s_mutex);
3858 __close_session(mdsc, session);
3859 if (session->s_state == CEPH_MDS_SESSION_CLOSING) {
3860 cleanup_session_requests(mdsc, session);
3861 remove_session_caps(session);
3862 }
3863 mutex_unlock(&session->s_mutex);
3864 ceph_put_mds_session(session);
3865 mutex_lock(&mdsc->mutex);
3866 kick_requests(mdsc, mds);
3867 }
3868 __wake_requests(mdsc, &mdsc->waiting_for_map);
3869 mutex_unlock(&mdsc->mutex);
3870 }
3871
3872 static void ceph_mdsc_stop(struct ceph_mds_client *mdsc)
3873 {
3874 dout("stop\n");
3875 cancel_delayed_work_sync(&mdsc->delayed_work); /* cancel timer */
3876 if (mdsc->mdsmap)
3877 ceph_mdsmap_destroy(mdsc->mdsmap);
3878 kfree(mdsc->sessions);
3879 ceph_caps_finalize(mdsc);
3880 ceph_pool_perm_destroy(mdsc);
3881 }
3882
3883 void ceph_mdsc_destroy(struct ceph_fs_client *fsc)
3884 {
3885 struct ceph_mds_client *mdsc = fsc->mdsc;
3886 dout("mdsc_destroy %p\n", mdsc);
3887
3888 if (!mdsc)
3889 return;
3890
3891 /* flush out any connection work with references to us */
3892 ceph_msgr_flush();
3893
3894 ceph_mdsc_stop(mdsc);
3895
3896 fsc->mdsc = NULL;
3897 kfree(mdsc);
3898 dout("mdsc_destroy %p done\n", mdsc);
3899 }
3900
3901 void ceph_mdsc_handle_fsmap(struct ceph_mds_client *mdsc, struct ceph_msg *msg)
3902 {
3903 struct ceph_fs_client *fsc = mdsc->fsc;
3904 const char *mds_namespace = fsc->mount_options->mds_namespace;
3905 void *p = msg->front.iov_base;
3906 void *end = p + msg->front.iov_len;
3907 u32 epoch;
3908 u32 map_len;
3909 u32 num_fs;
3910 u32 mount_fscid = (u32)-1;
3911 u8 struct_v, struct_cv;
3912 int err = -EINVAL;
3913
3914 ceph_decode_need(&p, end, sizeof(u32), bad);
3915 epoch = ceph_decode_32(&p);
3916
3917 dout("handle_fsmap epoch %u\n", epoch);
3918
3919 ceph_decode_need(&p, end, 2 + sizeof(u32), bad);
3920 struct_v = ceph_decode_8(&p);
3921 struct_cv = ceph_decode_8(&p);
3922 map_len = ceph_decode_32(&p);
3923
3924 ceph_decode_need(&p, end, sizeof(u32) * 3, bad);
3925 p += sizeof(u32) * 2; /* skip epoch and legacy_client_fscid */
3926
3927 num_fs = ceph_decode_32(&p);
3928 while (num_fs-- > 0) {
3929 void *info_p, *info_end;
3930 u32 info_len;
3931 u8 info_v, info_cv;
3932 u32 fscid, namelen;
3933
3934 ceph_decode_need(&p, end, 2 + sizeof(u32), bad);
3935 info_v = ceph_decode_8(&p);
3936 info_cv = ceph_decode_8(&p);
3937 info_len = ceph_decode_32(&p);
3938 ceph_decode_need(&p, end, info_len, bad);
3939 info_p = p;
3940 info_end = p + info_len;
3941 p = info_end;
3942
3943 ceph_decode_need(&info_p, info_end, sizeof(u32) * 2, bad);
3944 fscid = ceph_decode_32(&info_p);
3945 namelen = ceph_decode_32(&info_p);
3946 ceph_decode_need(&info_p, info_end, namelen, bad);
3947
3948 if (mds_namespace &&
3949 strlen(mds_namespace) == namelen &&
3950 !strncmp(mds_namespace, (char *)info_p, namelen)) {
3951 mount_fscid = fscid;
3952 break;
3953 }
3954 }
3955
3956 ceph_monc_got_map(&fsc->client->monc, CEPH_SUB_FSMAP, epoch);
3957 if (mount_fscid != (u32)-1) {
3958 fsc->client->monc.fs_cluster_id = mount_fscid;
3959 ceph_monc_want_map(&fsc->client->monc, CEPH_SUB_MDSMAP,
3960 0, true);
3961 ceph_monc_renew_subs(&fsc->client->monc);
3962 } else {
3963 err = -ENOENT;
3964 goto err_out;
3965 }
3966 return;
3967
3968 bad:
3969 pr_err("error decoding fsmap\n");
3970 err_out:
3971 mutex_lock(&mdsc->mutex);
3972 mdsc->mdsmap_err = err;
3973 __wake_requests(mdsc, &mdsc->waiting_for_map);
3974 mutex_unlock(&mdsc->mutex);
3975 }
3976
3977 /*
3978 * handle mds map update.
3979 */
3980 void ceph_mdsc_handle_mdsmap(struct ceph_mds_client *mdsc, struct ceph_msg *msg)
3981 {
3982 u32 epoch;
3983 u32 maplen;
3984 void *p = msg->front.iov_base;
3985 void *end = p + msg->front.iov_len;
3986 struct ceph_mdsmap *newmap, *oldmap;
3987 struct ceph_fsid fsid;
3988 int err = -EINVAL;
3989
3990 ceph_decode_need(&p, end, sizeof(fsid)+2*sizeof(u32), bad);
3991 ceph_decode_copy(&p, &fsid, sizeof(fsid));
3992 if (ceph_check_fsid(mdsc->fsc->client, &fsid) < 0)
3993 return;
3994 epoch = ceph_decode_32(&p);
3995 maplen = ceph_decode_32(&p);
3996 dout("handle_map epoch %u len %d\n", epoch, (int)maplen);
3997
3998 /* do we need it? */
3999 mutex_lock(&mdsc->mutex);
4000 if (mdsc->mdsmap && epoch <= mdsc->mdsmap->m_epoch) {
4001 dout("handle_map epoch %u <= our %u\n",
4002 epoch, mdsc->mdsmap->m_epoch);
4003 mutex_unlock(&mdsc->mutex);
4004 return;
4005 }
4006
4007 newmap = ceph_mdsmap_decode(&p, end);
4008 if (IS_ERR(newmap)) {
4009 err = PTR_ERR(newmap);
4010 goto bad_unlock;
4011 }
4012
4013 /* swap into place */
4014 if (mdsc->mdsmap) {
4015 oldmap = mdsc->mdsmap;
4016 mdsc->mdsmap = newmap;
4017 check_new_map(mdsc, newmap, oldmap);
4018 ceph_mdsmap_destroy(oldmap);
4019 } else {
4020 mdsc->mdsmap = newmap; /* first mds map */
4021 }
4022 mdsc->fsc->sb->s_maxbytes = mdsc->mdsmap->m_max_file_size;
4023
4024 __wake_requests(mdsc, &mdsc->waiting_for_map);
4025 ceph_monc_got_map(&mdsc->fsc->client->monc, CEPH_SUB_MDSMAP,
4026 mdsc->mdsmap->m_epoch);
4027
4028 mutex_unlock(&mdsc->mutex);
4029 schedule_delayed(mdsc);
4030 return;
4031
4032 bad_unlock:
4033 mutex_unlock(&mdsc->mutex);
4034 bad:
4035 pr_err("error decoding mdsmap %d\n", err);
4036 return;
4037 }
4038
4039 static struct ceph_connection *con_get(struct ceph_connection *con)
4040 {
4041 struct ceph_mds_session *s = con->private;
4042
4043 if (get_session(s)) {
4044 dout("mdsc con_get %p ok (%d)\n", s, refcount_read(&s->s_ref));
4045 return con;
4046 }
4047 dout("mdsc con_get %p FAIL\n", s);
4048 return NULL;
4049 }
4050
4051 static void con_put(struct ceph_connection *con)
4052 {
4053 struct ceph_mds_session *s = con->private;
4054
4055 dout("mdsc con_put %p (%d)\n", s, refcount_read(&s->s_ref) - 1);
4056 ceph_put_mds_session(s);
4057 }
4058
4059 /*
4060 * if the client is unresponsive for long enough, the mds will kill
4061 * the session entirely.
4062 */
4063 static void peer_reset(struct ceph_connection *con)
4064 {
4065 struct ceph_mds_session *s = con->private;
4066 struct ceph_mds_client *mdsc = s->s_mdsc;
4067
4068 pr_warn("mds%d closed our session\n", s->s_mds);
4069 send_mds_reconnect(mdsc, s);
4070 }
4071
4072 static void dispatch(struct ceph_connection *con, struct ceph_msg *msg)
4073 {
4074 struct ceph_mds_session *s = con->private;
4075 struct ceph_mds_client *mdsc = s->s_mdsc;
4076 int type = le16_to_cpu(msg->hdr.type);
4077
4078 mutex_lock(&mdsc->mutex);
4079 if (__verify_registered_session(mdsc, s) < 0) {
4080 mutex_unlock(&mdsc->mutex);
4081 goto out;
4082 }
4083 mutex_unlock(&mdsc->mutex);
4084
4085 switch (type) {
4086 case CEPH_MSG_MDS_MAP:
4087 ceph_mdsc_handle_mdsmap(mdsc, msg);
4088 break;
4089 case CEPH_MSG_FS_MAP_USER:
4090 ceph_mdsc_handle_fsmap(mdsc, msg);
4091 break;
4092 case CEPH_MSG_CLIENT_SESSION:
4093 handle_session(s, msg);
4094 break;
4095 case CEPH_MSG_CLIENT_REPLY:
4096 handle_reply(s, msg);
4097 break;
4098 case CEPH_MSG_CLIENT_REQUEST_FORWARD:
4099 handle_forward(mdsc, s, msg);
4100 break;
4101 case CEPH_MSG_CLIENT_CAPS:
4102 ceph_handle_caps(s, msg);
4103 break;
4104 case CEPH_MSG_CLIENT_SNAP:
4105 ceph_handle_snap(mdsc, s, msg);
4106 break;
4107 case CEPH_MSG_CLIENT_LEASE:
4108 handle_lease(mdsc, s, msg);
4109 break;
4110 case CEPH_MSG_CLIENT_QUOTA:
4111 ceph_handle_quota(mdsc, s, msg);
4112 break;
4113
4114 default:
4115 pr_err("received unknown message type %d %s\n", type,
4116 ceph_msg_type_name(type));
4117 }
4118 out:
4119 ceph_msg_put(msg);
4120 }
4121
4122 /*
4123 * authentication
4124 */
4125
4126 /*
4127 * Note: returned pointer is the address of a structure that's
4128 * managed separately. Caller must *not* attempt to free it.
4129 */
4130 static struct ceph_auth_handshake *get_authorizer(struct ceph_connection *con,
4131 int *proto, int force_new)
4132 {
4133 struct ceph_mds_session *s = con->private;
4134 struct ceph_mds_client *mdsc = s->s_mdsc;
4135 struct ceph_auth_client *ac = mdsc->fsc->client->monc.auth;
4136 struct ceph_auth_handshake *auth = &s->s_auth;
4137
4138 if (force_new && auth->authorizer) {
4139 ceph_auth_destroy_authorizer(auth->authorizer);
4140 auth->authorizer = NULL;
4141 }
4142 if (!auth->authorizer) {
4143 int ret = ceph_auth_create_authorizer(ac, CEPH_ENTITY_TYPE_MDS,
4144 auth);
4145 if (ret)
4146 return ERR_PTR(ret);
4147 } else {
4148 int ret = ceph_auth_update_authorizer(ac, CEPH_ENTITY_TYPE_MDS,
4149 auth);
4150 if (ret)
4151 return ERR_PTR(ret);
4152 }
4153 *proto = ac->protocol;
4154
4155 return auth;
4156 }
4157
4158
4159 static int verify_authorizer_reply(struct ceph_connection *con)
4160 {
4161 struct ceph_mds_session *s = con->private;
4162 struct ceph_mds_client *mdsc = s->s_mdsc;
4163 struct ceph_auth_client *ac = mdsc->fsc->client->monc.auth;
4164
4165 return ceph_auth_verify_authorizer_reply(ac, s->s_auth.authorizer);
4166 }
4167
4168 static int invalidate_authorizer(struct ceph_connection *con)
4169 {
4170 struct ceph_mds_session *s = con->private;
4171 struct ceph_mds_client *mdsc = s->s_mdsc;
4172 struct ceph_auth_client *ac = mdsc->fsc->client->monc.auth;
4173
4174 ceph_auth_invalidate_authorizer(ac, CEPH_ENTITY_TYPE_MDS);
4175
4176 return ceph_monc_validate_auth(&mdsc->fsc->client->monc);
4177 }
4178
4179 static struct ceph_msg *mds_alloc_msg(struct ceph_connection *con,
4180 struct ceph_msg_header *hdr, int *skip)
4181 {
4182 struct ceph_msg *msg;
4183 int type = (int) le16_to_cpu(hdr->type);
4184 int front_len = (int) le32_to_cpu(hdr->front_len);
4185
4186 if (con->in_msg)
4187 return con->in_msg;
4188
4189 *skip = 0;
4190 msg = ceph_msg_new(type, front_len, GFP_NOFS, false);
4191 if (!msg) {
4192 pr_err("unable to allocate msg type %d len %d\n",
4193 type, front_len);
4194 return NULL;
4195 }
4196
4197 return msg;
4198 }
4199
4200 static int mds_sign_message(struct ceph_msg *msg)
4201 {
4202 struct ceph_mds_session *s = msg->con->private;
4203 struct ceph_auth_handshake *auth = &s->s_auth;
4204
4205 return ceph_auth_sign_message(auth, msg);
4206 }
4207
4208 static int mds_check_message_signature(struct ceph_msg *msg)
4209 {
4210 struct ceph_mds_session *s = msg->con->private;
4211 struct ceph_auth_handshake *auth = &s->s_auth;
4212
4213 return ceph_auth_check_message_signature(auth, msg);
4214 }
4215
4216 static const struct ceph_connection_operations mds_con_ops = {
4217 .get = con_get,
4218 .put = con_put,
4219 .dispatch = dispatch,
4220 .get_authorizer = get_authorizer,
4221 .verify_authorizer_reply = verify_authorizer_reply,
4222 .invalidate_authorizer = invalidate_authorizer,
4223 .peer_reset = peer_reset,
4224 .alloc_msg = mds_alloc_msg,
4225 .sign_message = mds_sign_message,
4226 .check_message_signature = mds_check_message_signature,
4227 };
4228
4229 /* eof */
Ubuntu Eoan Linux kernel mirror