]> git.proxmox.com Git - mirror_ubuntu-hirsute-kernel.git/blame - fs/ceph/mds_client.c
projects / mirror_ubuntu-hirsute-kernel.git / blame
? search:
summary | shortlog | log | commit | commitdiff | tree
blob | blame (incremental) | history | HEAD
ceph: fix a comment typo
[mirror_ubuntu-hirsute-kernel.git] / fs / ceph / mds_client.c
CommitLineData
3d14c5d2 1#include <linux/ceph/ceph_debug.h>
2f2dc053 2
496e5955 3#include <linux/fs.h>
2f2dc053 4#include <linux/wait.h>
5a0e3ad6 5#include <linux/slab.h>
54008399 6#include <linux/gfp.h>
2f2dc053 7#include <linux/sched.h>
3d14c5d2
YS		 8#include <linux/debugfs.h>
9#include <linux/seq_file.h>
dbd0c8bf 10#include <linux/utsname.h>
3e0708b9 11#include <linux/ratelimit.h>
2f2dc053 12
2f2dc053 13#include "super.h"
3d14c5d2
YS		 14#include "mds_client.h"
15
1fe60e51 16#include <linux/ceph/ceph_features.h>
3d14c5d2
YS		 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>
2f2dc053
SW		 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45 * are no longer valid.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
20cb34ae 48struct ceph_reconnect_state {
44c99757 49 int nr_caps;
20cb34ae
SW		 50 struct ceph_pagelist *pagelist;
51 bool flock;
52};
53
2f2dc053
SW		 54static void __wake_requests(struct ceph_mds_client *mdsc,
55 struct list_head *head);
56
9e32789f 57static const struct ceph_connection_operations mds_con_ops;
2f2dc053
SW		 58
59
60/*
61 * mds reply parsing
62 */
63
64/*
65 * parse individual inode info
66 */
67static int parse_reply_info_in(void **p, void *end,
14303d20 68 struct ceph_mds_reply_info_in *info,
12b4629a 69 u64 features)
2f2dc053
SW		 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
14303d20
SW		 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
2f2dc053
SW		 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;
fb01d1f8
YZ		 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
2f2dc053
SW		 103 return 0;
104bad:
105 return err;
106}
107
108/*
109 * parse a normal reply, which may contain a (dir+)dentry and/or a
110 * target inode.
111 */
112static int parse_reply_info_trace(void **p, void *end,
14303d20 113 struct ceph_mds_reply_info_parsed *info,
12b4629a 114 u64 features)
2f2dc053
SW		 115{
116 int err;
117
118 if (info->head->is_dentry) {
14303d20 119 err = parse_reply_info_in(p, end, &info->diri, features);
2f2dc053
SW		 120 if (err < 0)
121 goto out_bad;
122
123 if (unlikely(*p + sizeof(*info->dirfrag) > end))
124 goto bad;
125 info->dirfrag = *p;
126 *p += sizeof(*info->dirfrag) +
127 sizeof(u32)*le32_to_cpu(info->dirfrag->ndist);
128 if (unlikely(*p > end))
129 goto bad;
130
131 ceph_decode_32_safe(p, end, info->dname_len, bad);
132 ceph_decode_need(p, end, info->dname_len, bad);
133 info->dname = *p;
134 *p += info->dname_len;
135 info->dlease = *p;
136 *p += sizeof(*info->dlease);
137 }
138
139 if (info->head->is_target) {
14303d20 140 err = parse_reply_info_in(p, end, &info->targeti, features);
2f2dc053
SW		 141 if (err < 0)
142 goto out_bad;
143 }
144
145 if (unlikely(*p != end))
146 goto bad;
147 return 0;
148
149bad:
150 err = -EIO;
151out_bad:
152 pr_err("problem parsing mds trace %d\n", err);
153 return err;
154}
155
156/*
157 * parse readdir results
158 */
159static int parse_reply_info_dir(void **p, void *end,
14303d20 160 struct ceph_mds_reply_info_parsed *info,
12b4629a 161 u64 features)
2f2dc053
SW		 162{
163 u32 num, i = 0;
164 int err;
165
166 info->dir_dir = *p;
167 if (*p + sizeof(*info->dir_dir) > end)
168 goto bad;
169 *p += sizeof(*info->dir_dir) +
170 sizeof(u32)*le32_to_cpu(info->dir_dir->ndist);
171 if (*p > end)
172 goto bad;
173
174 ceph_decode_need(p, end, sizeof(num) + 2, bad);
c89136ea
SW		 175 num = ceph_decode_32(p);
176 info->dir_end = ceph_decode_8(p);
177 info->dir_complete = ceph_decode_8(p);
2f2dc053
SW		 178 if (num == 0)
179 goto done;
180
54008399 181 BUG_ON(!info->dir_in);
2f2dc053
SW		 182 info->dir_dname = (void *)(info->dir_in + num);
183 info->dir_dname_len = (void *)(info->dir_dname + num);
184 info->dir_dlease = (void *)(info->dir_dname_len + num);
54008399
YZ		 185 if ((unsigned long)(info->dir_dlease + num) >
186 (unsigned long)info->dir_in + info->dir_buf_size) {
187 pr_err("dir contents are larger than expected\n");
188 WARN_ON(1);
189 goto bad;
190 }
2f2dc053 191
54008399 192 info->dir_nr = num;
2f2dc053
SW		 193 while (num) {
194 /* dentry */
195 ceph_decode_need(p, end, sizeof(u32)*2, bad);
c89136ea 196 info->dir_dname_len[i] = ceph_decode_32(p);
2f2dc053
SW		 197 ceph_decode_need(p, end, info->dir_dname_len[i], bad);
198 info->dir_dname[i] = *p;
199 *p += info->dir_dname_len[i];
200 dout("parsed dir dname '%.*s'\n", info->dir_dname_len[i],
201 info->dir_dname[i]);
202 info->dir_dlease[i] = *p;
203 *p += sizeof(struct ceph_mds_reply_lease);
204
205 /* inode */
14303d20 206 err = parse_reply_info_in(p, end, &info->dir_in[i], features);
2f2dc053
SW		 207 if (err < 0)
208 goto out_bad;
209 i++;
210 num--;
211 }
212
213done:
214 if (*p != end)
215 goto bad;
216 return 0;
217
218bad:
219 err = -EIO;
220out_bad:
221 pr_err("problem parsing dir contents %d\n", err);
222 return err;
223}
224
25933abd
HS		 225/*
226 * parse fcntl F_GETLK results
227 */
228static int parse_reply_info_filelock(void **p, void *end,
14303d20 229 struct ceph_mds_reply_info_parsed *info,
12b4629a 230 u64 features)
25933abd
HS		 231{
232 if (*p + sizeof(*info->filelock_reply) > end)
233 goto bad;
234
235 info->filelock_reply = *p;
236 *p += sizeof(*info->filelock_reply);
237
238 if (unlikely(*p != end))
239 goto bad;
240 return 0;
241
242bad:
243 return -EIO;
244}
245
6e8575fa
SL		 246/*
247 * parse create results
248 */
249static int parse_reply_info_create(void **p, void *end,
250 struct ceph_mds_reply_info_parsed *info,
12b4629a 251 u64 features)
6e8575fa
SL		 252{
253 if (features & CEPH_FEATURE_REPLY_CREATE_INODE) {
254 if (*p == end) {
255 info->has_create_ino = false;
256 } else {
257 info->has_create_ino = true;
258 info->ino = ceph_decode_64(p);
259 }
260 }
261
262 if (unlikely(*p != end))
263 goto bad;
264 return 0;
265
266bad:
267 return -EIO;
268}
269
25933abd
HS		 270/*
271 * parse extra results
272 */
273static int parse_reply_info_extra(void **p, void *end,
14303d20 274 struct ceph_mds_reply_info_parsed *info,
12b4629a 275 u64 features)
25933abd
HS		 276{
277 if (info->head->op == CEPH_MDS_OP_GETFILELOCK)
14303d20 278 return parse_reply_info_filelock(p, end, info, features);
8a034497
YZ		 279 else if (info->head->op == CEPH_MDS_OP_READDIR ||
280 info->head->op == CEPH_MDS_OP_LSSNAP)
14303d20 281 return parse_reply_info_dir(p, end, info, features);
6e8575fa
SL		 282 else if (info->head->op == CEPH_MDS_OP_CREATE)
283 return parse_reply_info_create(p, end, info, features);
284 else
285 return -EIO;
25933abd
HS		 286}
287
2f2dc053
SW		 288/*
289 * parse entire mds reply
290 */
291static int parse_reply_info(struct ceph_msg *msg,
14303d20 292 struct ceph_mds_reply_info_parsed *info,
12b4629a 293 u64 features)
2f2dc053
SW		 294{
295 void *p, *end;
296 u32 len;
297 int err;
298
299 info->head = msg->front.iov_base;
300 p = msg->front.iov_base + sizeof(struct ceph_mds_reply_head);
301 end = p + msg->front.iov_len - sizeof(struct ceph_mds_reply_head);
302
303 /* trace */
304 ceph_decode_32_safe(&p, end, len, bad);
305 if (len > 0) {
32852a81 306 ceph_decode_need(&p, end, len, bad);
14303d20 307 err = parse_reply_info_trace(&p, p+len, info, features);
2f2dc053
SW		 308 if (err < 0)
309 goto out_bad;
310 }
311
25933abd 312 /* extra */
2f2dc053
SW		 313 ceph_decode_32_safe(&p, end, len, bad);
314 if (len > 0) {
32852a81 315 ceph_decode_need(&p, end, len, bad);
14303d20 316 err = parse_reply_info_extra(&p, p+len, info, features);
2f2dc053
SW		 317 if (err < 0)
318 goto out_bad;
319 }
320
321 /* snap blob */
322 ceph_decode_32_safe(&p, end, len, bad);
323 info->snapblob_len = len;
324 info->snapblob = p;
325 p += len;
326
327 if (p != end)
328 goto bad;
329 return 0;
330
331bad:
332 err = -EIO;
333out_bad:
334 pr_err("mds parse_reply err %d\n", err);
335 return err;
336}
337
338static void destroy_reply_info(struct ceph_mds_reply_info_parsed *info)
339{
54008399
YZ		 340 if (!info->dir_in)
341 return;
342 free_pages((unsigned long)info->dir_in, get_order(info->dir_buf_size));
2f2dc053
SW		 343}
344
345
346/*
347 * sessions
348 */
a687ecaf 349const char *ceph_session_state_name(int s)
2f2dc053
SW		 350{
351 switch (s) {
352 case CEPH_MDS_SESSION_NEW: return "new";
353 case CEPH_MDS_SESSION_OPENING: return "opening";
354 case CEPH_MDS_SESSION_OPEN: return "open";
355 case CEPH_MDS_SESSION_HUNG: return "hung";
356 case CEPH_MDS_SESSION_CLOSING: return "closing";
44ca18f2 357 case CEPH_MDS_SESSION_RESTARTING: return "restarting";
2f2dc053
SW		 358 case CEPH_MDS_SESSION_RECONNECTING: return "reconnecting";
359 default: return "???";
360 }
361}
362
363static struct ceph_mds_session *get_session(struct ceph_mds_session *s)
364{
365 if (atomic_inc_not_zero(&s->s_ref)) {
366 dout("mdsc get_session %p %d -> %d\n", s,
367 atomic_read(&s->s_ref)-1, atomic_read(&s->s_ref));
368 return s;
369 } else {
370 dout("mdsc get_session %p 0 -- FAIL", s);
371 return NULL;
372 }
373}
374
375void ceph_put_mds_session(struct ceph_mds_session *s)
376{
377 dout("mdsc put_session %p %d -> %d\n", s,
378 atomic_read(&s->s_ref), atomic_read(&s->s_ref)-1);
4e7a5dcd 379 if (atomic_dec_and_test(&s->s_ref)) {
6c4a1915 380 if (s->s_auth.authorizer)
27859f97
SW		 381 ceph_auth_destroy_authorizer(
382 s->s_mdsc->fsc->client->monc.auth,
383 s->s_auth.authorizer);
2f2dc053 384 kfree(s);
4e7a5dcd 385 }
2f2dc053
SW		 386}
387
388/*
389 * called under mdsc->mutex
390 */
391struct ceph_mds_session *__ceph_lookup_mds_session(struct ceph_mds_client *mdsc,
392 int mds)
393{
394 struct ceph_mds_session *session;
395
396 if (mds >= mdsc->max_sessions || mdsc->sessions[mds] == NULL)
397 return NULL;
398 session = mdsc->sessions[mds];
399 dout("lookup_mds_session %p %d\n", session,
400 atomic_read(&session->s_ref));
401 get_session(session);
402 return session;
403}
404
405static bool __have_session(struct ceph_mds_client *mdsc, int mds)
406{
407 if (mds >= mdsc->max_sessions)
408 return false;
409 return mdsc->sessions[mds];
410}
411
2600d2dd
SW		 412static int __verify_registered_session(struct ceph_mds_client *mdsc,
413 struct ceph_mds_session *s)
414{
415 if (s->s_mds >= mdsc->max_sessions ||
416 mdsc->sessions[s->s_mds] != s)
417 return -ENOENT;
418 return 0;
419}
420
2f2dc053
SW		 421/*
422 * create+register a new session for given mds.
423 * called under mdsc->mutex.
424 */
425static struct ceph_mds_session *register_session(struct ceph_mds_client *mdsc,
426 int mds)
427{
428 struct ceph_mds_session *s;
429
c338c07c
NY		 430 if (mds >= mdsc->mdsmap->m_max_mds)
431 return ERR_PTR(-EINVAL);
432
2f2dc053 433 s = kzalloc(sizeof(*s), GFP_NOFS);
4736b009
DC		 434 if (!s)
435 return ERR_PTR(-ENOMEM);
2f2dc053
SW		 436 s->s_mdsc = mdsc;
437 s->s_mds = mds;
438 s->s_state = CEPH_MDS_SESSION_NEW;
439 s->s_ttl = 0;
440 s->s_seq = 0;
441 mutex_init(&s->s_mutex);
442
b7a9e5dd 443 ceph_con_init(&s->s_con, s, &mds_con_ops, &mdsc->fsc->client->msgr);
2f2dc053 444
d8fb02ab 445 spin_lock_init(&s->s_gen_ttl_lock);
2f2dc053 446 s->s_cap_gen = 0;
1ce208a6 447 s->s_cap_ttl = jiffies - 1;
d8fb02ab
AE		 448
449 spin_lock_init(&s->s_cap_lock);
2f2dc053
SW		 450 s->s_renew_requested = 0;
451 s->s_renew_seq = 0;
452 INIT_LIST_HEAD(&s->s_caps);
453 s->s_nr_caps = 0;
5dacf091 454 s->s_trim_caps = 0;
2f2dc053
SW		 455 atomic_set(&s->s_ref, 1);
456 INIT_LIST_HEAD(&s->s_waiting);
457 INIT_LIST_HEAD(&s->s_unsafe);
458 s->s_num_cap_releases = 0;
99a9c273 459 s->s_cap_reconnect = 0;
7c1332b8 460 s->s_cap_iterator = NULL;
2f2dc053 461 INIT_LIST_HEAD(&s->s_cap_releases);
2f2dc053
SW		 462 INIT_LIST_HEAD(&s->s_cap_flushing);
463 INIT_LIST_HEAD(&s->s_cap_snaps_flushing);
464
465 dout("register_session mds%d\n", mds);
466 if (mds >= mdsc->max_sessions) {
467 int newmax = 1 << get_count_order(mds+1);
468 struct ceph_mds_session **sa;
469
470 dout("register_session realloc to %d\n", newmax);
471 sa = kcalloc(newmax, sizeof(void *), GFP_NOFS);
472 if (sa == NULL)
42ce56e5 473 goto fail_realloc;
2f2dc053
SW		 474 if (mdsc->sessions) {
475 memcpy(sa, mdsc->sessions,
476 mdsc->max_sessions * sizeof(void *));
477 kfree(mdsc->sessions);
478 }
479 mdsc->sessions = sa;
480 mdsc->max_sessions = newmax;
481 }
482 mdsc->sessions[mds] = s;
86d8f67b 483 atomic_inc(&mdsc->num_sessions);
2f2dc053 484 atomic_inc(&s->s_ref); /* one ref to sessions[], one to caller */
42ce56e5 485
b7a9e5dd
SW		 486 ceph_con_open(&s->s_con, CEPH_ENTITY_TYPE_MDS, mds,
487 ceph_mdsmap_get_addr(mdsc->mdsmap, mds));
42ce56e5 488
2f2dc053 489 return s;
42ce56e5
SW		 490
491fail_realloc:
492 kfree(s);
493 return ERR_PTR(-ENOMEM);
2f2dc053
SW		 494}
495
496/*
497 * called under mdsc->mutex
498 */
2600d2dd 499static void __unregister_session(struct ceph_mds_client *mdsc,
42ce56e5 500 struct ceph_mds_session *s)
2f2dc053 501{
2600d2dd
SW		 502 dout("__unregister_session mds%d %p\n", s->s_mds, s);
503 BUG_ON(mdsc->sessions[s->s_mds] != s);
42ce56e5
SW		 504 mdsc->sessions[s->s_mds] = NULL;
505 ceph_con_close(&s->s_con);
506 ceph_put_mds_session(s);
86d8f67b 507 atomic_dec(&mdsc->num_sessions);
2f2dc053
SW		 508}
509
510/*
511 * drop session refs in request.
512 *
513 * should be last request ref, or hold mdsc->mutex
514 */
515static void put_request_session(struct ceph_mds_request *req)
516{
517 if (req->r_session) {
518 ceph_put_mds_session(req->r_session);
519 req->r_session = NULL;
520 }
521}
522
153c8e6b 523void ceph_mdsc_release_request(struct kref *kref)
2f2dc053 524{
153c8e6b
SW		 525 struct ceph_mds_request *req = container_of(kref,
526 struct ceph_mds_request,
527 r_kref);
54008399 528 destroy_reply_info(&req->r_reply_info);
153c8e6b
SW		 529 if (req->r_request)
530 ceph_msg_put(req->r_request);
54008399 531 if (req->r_reply)
153c8e6b 532 ceph_msg_put(req->r_reply);
153c8e6b 533 if (req->r_inode) {
41b02e1f 534 ceph_put_cap_refs(ceph_inode(req->r_inode), CEPH_CAP_PIN);
153c8e6b
SW		 535 iput(req->r_inode);
536 }
537 if (req->r_locked_dir)
41b02e1f 538 ceph_put_cap_refs(ceph_inode(req->r_locked_dir), CEPH_CAP_PIN);
e96a650a 539 iput(req->r_target_inode);
153c8e6b
SW		 540 if (req->r_dentry)
541 dput(req->r_dentry);
844d87c3
SW		 542 if (req->r_old_dentry)
543 dput(req->r_old_dentry);
544 if (req->r_old_dentry_dir) {
41b02e1f
SW		 545 /*
546 * track (and drop pins for) r_old_dentry_dir
547 * separately, since r_old_dentry's d_parent may have
548 * changed between the dir mutex being dropped and
549 * this request being freed.
550 */
551 ceph_put_cap_refs(ceph_inode(req->r_old_dentry_dir),
552 CEPH_CAP_PIN);
41b02e1f 553 iput(req->r_old_dentry_dir);
2f2dc053 554 }
153c8e6b
SW		 555 kfree(req->r_path1);
556 kfree(req->r_path2);
25e6bae3
YZ		 557 if (req->r_pagelist)
558 ceph_pagelist_release(req->r_pagelist);
153c8e6b 559 put_request_session(req);
37151668 560 ceph_unreserve_caps(req->r_mdsc, &req->r_caps_reservation);
153c8e6b 561 kfree(req);
2f2dc053
SW		 562}
563
564/*
565 * lookup session, bump ref if found.
566 *
567 * called under mdsc->mutex.
568 */
569static struct ceph_mds_request *__lookup_request(struct ceph_mds_client *mdsc,
570 u64 tid)
571{
572 struct ceph_mds_request *req;
44ca18f2
SW		 573 struct rb_node *n = mdsc->request_tree.rb_node;
574
575 while (n) {
576 req = rb_entry(n, struct ceph_mds_request, r_node);
577 if (tid < req->r_tid)
578 n = n->rb_left;
579 else if (tid > req->r_tid)
580 n = n->rb_right;
581 else {
582 ceph_mdsc_get_request(req);
583 return req;
584 }
585 }
586 return NULL;
587}
588
589static void __insert_request(struct ceph_mds_client *mdsc,
590 struct ceph_mds_request *new)
591{
592 struct rb_node **p = &mdsc->request_tree.rb_node;
593 struct rb_node *parent = NULL;
594 struct ceph_mds_request *req = NULL;
595
596 while (*p) {
597 parent = *p;
598 req = rb_entry(parent, struct ceph_mds_request, r_node);
599 if (new->r_tid < req->r_tid)
600 p = &(*p)->rb_left;
601 else if (new->r_tid > req->r_tid)
602 p = &(*p)->rb_right;
603 else
604 BUG();
605 }
606
607 rb_link_node(&new->r_node, parent, p);
608 rb_insert_color(&new->r_node, &mdsc->request_tree);
2f2dc053
SW		 609}
610
611/*
612 * Register an in-flight request, and assign a tid. Link to directory
613 * are modifying (if any).
614 *
615 * Called under mdsc->mutex.
616 */
617static void __register_request(struct ceph_mds_client *mdsc,
618 struct ceph_mds_request *req,
619 struct inode *dir)
620{
621 req->r_tid = ++mdsc->last_tid;
622 if (req->r_num_caps)
37151668
YS		 623 ceph_reserve_caps(mdsc, &req->r_caps_reservation,
624 req->r_num_caps);
2f2dc053
SW		 625 dout("__register_request %p tid %lld\n", req, req->r_tid);
626 ceph_mdsc_get_request(req);
44ca18f2 627 __insert_request(mdsc, req);
2f2dc053 628
cb4276cc
SW		 629 req->r_uid = current_fsuid();
630 req->r_gid = current_fsgid();
631
e8a7b8b1
YZ		 632 if (mdsc->oldest_tid == 0 && req->r_op != CEPH_MDS_OP_SETFILELOCK)
633 mdsc->oldest_tid = req->r_tid;
634
2f2dc053
SW		 635 if (dir) {
636 struct ceph_inode_info *ci = ceph_inode(dir);
637
3b663780 638 ihold(dir);
2f2dc053
SW		 639 spin_lock(&ci->i_unsafe_lock);
640 req->r_unsafe_dir = dir;
641 list_add_tail(&req->r_unsafe_dir_item, &ci->i_unsafe_dirops);
642 spin_unlock(&ci->i_unsafe_lock);
643 }
644}
645
646static void __unregister_request(struct ceph_mds_client *mdsc,
647 struct ceph_mds_request *req)
648{
649 dout("__unregister_request %p tid %lld\n", req, req->r_tid);
e8a7b8b1
YZ		 650
651 if (req->r_tid == mdsc->oldest_tid) {
652 struct rb_node *p = rb_next(&req->r_node);
653 mdsc->oldest_tid = 0;
654 while (p) {
655 struct ceph_mds_request *next_req =
656 rb_entry(p, struct ceph_mds_request, r_node);
657 if (next_req->r_op != CEPH_MDS_OP_SETFILELOCK) {
658 mdsc->oldest_tid = next_req->r_tid;
659 break;
660 }
661 p = rb_next(p);
662 }
663 }
664
44ca18f2 665 rb_erase(&req->r_node, &mdsc->request_tree);
80fc7314 666 RB_CLEAR_NODE(&req->r_node);
2f2dc053
SW		 667
668 if (req->r_unsafe_dir) {
669 struct ceph_inode_info *ci = ceph_inode(req->r_unsafe_dir);
670
671 spin_lock(&ci->i_unsafe_lock);
672 list_del_init(&req->r_unsafe_dir_item);
673 spin_unlock(&ci->i_unsafe_lock);
3b663780
SW		 674
675 iput(req->r_unsafe_dir);
676 req->r_unsafe_dir = NULL;
2f2dc053 677 }
94aa8ae1 678
fc55d2c9
YZ		 679 complete_all(&req->r_safe_completion);
680
94aa8ae1 681 ceph_mdsc_put_request(req);
2f2dc053
SW		 682}
683
684/*
685 * Choose mds to send request to next. If there is a hint set in the
686 * request (e.g., due to a prior forward hint from the mds), use that.
687 * Otherwise, consult frag tree and/or caps to identify the
688 * appropriate mds. If all else fails, choose randomly.
689 *
690 * Called under mdsc->mutex.
691 */
7fd7d101 692static struct dentry *get_nonsnap_parent(struct dentry *dentry)
eb6bb1c5 693{
d79698da
SW		 694 /*
695 * we don't need to worry about protecting the d_parent access
696 * here because we never renaming inside the snapped namespace
697 * except to resplice to another snapdir, and either the old or new
698 * result is a valid result.
699 */
2b0143b5 700 while (!IS_ROOT(dentry) && ceph_snap(d_inode(dentry)) != CEPH_NOSNAP)
eb6bb1c5
SW		 701 dentry = dentry->d_parent;
702 return dentry;
703}
704
2f2dc053
SW		 705static int __choose_mds(struct ceph_mds_client *mdsc,
706 struct ceph_mds_request *req)
707{
708 struct inode *inode;
709 struct ceph_inode_info *ci;
710 struct ceph_cap *cap;
711 int mode = req->r_direct_mode;
712 int mds = -1;
713 u32 hash = req->r_direct_hash;
714 bool is_hash = req->r_direct_is_hash;
715
716 /*
717 * is there a specific mds we should try? ignore hint if we have
718 * no session and the mds is not up (active or recovering).
719 */
720 if (req->r_resend_mds >= 0 &&
721 (__have_session(mdsc, req->r_resend_mds) ||
722 ceph_mdsmap_get_state(mdsc->mdsmap, req->r_resend_mds) > 0)) {
723 dout("choose_mds using resend_mds mds%d\n",
724 req->r_resend_mds);
725 return req->r_resend_mds;
726 }
727
728 if (mode == USE_RANDOM_MDS)
729 goto random;
730
731 inode = NULL;
732 if (req->r_inode) {
733 inode = req->r_inode;
734 } else if (req->r_dentry) {
d79698da
SW		 735 /* ignore race with rename; old or new d_parent is okay */
736 struct dentry *parent = req->r_dentry->d_parent;
2b0143b5 737 struct inode *dir = d_inode(parent);
eb6bb1c5 738
3d14c5d2 739 if (dir->i_sb != mdsc->fsc->sb) {
eb6bb1c5 740 /* not this fs! */
2b0143b5 741 inode = d_inode(req->r_dentry);
eb6bb1c5
SW		 742 } else if (ceph_snap(dir) != CEPH_NOSNAP) {
743 /* direct snapped/virtual snapdir requests
744 * based on parent dir inode */
d79698da 745 struct dentry *dn = get_nonsnap_parent(parent);
2b0143b5 746 inode = d_inode(dn);
eb6bb1c5 747 dout("__choose_mds using nonsnap parent %p\n", inode);
ca18bede 748 } else {
eb6bb1c5 749 /* dentry target */
2b0143b5 750 inode = d_inode(req->r_dentry);
ca18bede
YZ		 751 if (!inode || mode == USE_AUTH_MDS) {
752 /* dir + name */
753 inode = dir;
754 hash = ceph_dentry_hash(dir, req->r_dentry);
755 is_hash = true;
756 }
2f2dc053
SW		 757 }
758 }
eb6bb1c5 759
2f2dc053
SW		 760 dout("__choose_mds %p is_hash=%d (%d) mode %d\n", inode, (int)is_hash,
761 (int)hash, mode);
762 if (!inode)
763 goto random;
764 ci = ceph_inode(inode);
765
766 if (is_hash && S_ISDIR(inode->i_mode)) {
767 struct ceph_inode_frag frag;
768 int found;
769
770 ceph_choose_frag(ci, hash, &frag, &found);
771 if (found) {
772 if (mode == USE_ANY_MDS && frag.ndist > 0) {
773 u8 r;
774
775 /* choose a random replica */
776 get_random_bytes(&r, 1);
777 r %= frag.ndist;
778 mds = frag.dist[r];
779 dout("choose_mds %p %llx.%llx "
780 "frag %u mds%d (%d/%d)\n",
781 inode, ceph_vinop(inode),
d66bbd44 782 frag.frag, mds,
2f2dc053 783 (int)r, frag.ndist);
d66bbd44
SW		 784 if (ceph_mdsmap_get_state(mdsc->mdsmap, mds) >=
785 CEPH_MDS_STATE_ACTIVE)
786 return mds;
2f2dc053
SW		 787 }
788
789 /* since this file/dir wasn't known to be
790 * replicated, then we want to look for the
791 * authoritative mds. */
792 mode = USE_AUTH_MDS;
793 if (frag.mds >= 0) {
794 /* choose auth mds */
795 mds = frag.mds;
796 dout("choose_mds %p %llx.%llx "
797 "frag %u mds%d (auth)\n",
798 inode, ceph_vinop(inode), frag.frag, mds);
d66bbd44
SW		 799 if (ceph_mdsmap_get_state(mdsc->mdsmap, mds) >=
800 CEPH_MDS_STATE_ACTIVE)
801 return mds;
2f2dc053
SW		 802 }
803 }
804 }
805
be655596 806 spin_lock(&ci->i_ceph_lock);
2f2dc053
SW		 807 cap = NULL;
808 if (mode == USE_AUTH_MDS)
809 cap = ci->i_auth_cap;
810 if (!cap && !RB_EMPTY_ROOT(&ci->i_caps))
811 cap = rb_entry(rb_first(&ci->i_caps), struct ceph_cap, ci_node);
812 if (!cap) {
be655596 813 spin_unlock(&ci->i_ceph_lock);
2f2dc053
SW		 814 goto random;
815 }
816 mds = cap->session->s_mds;
817 dout("choose_mds %p %llx.%llx mds%d (%scap %p)\n",
818 inode, ceph_vinop(inode), mds,
819 cap == ci->i_auth_cap ? "auth " : "", cap);
be655596 820 spin_unlock(&ci->i_ceph_lock);
2f2dc053
SW		 821 return mds;
822
823random:
824 mds = ceph_mdsmap_get_random_mds(mdsc->mdsmap);
825 dout("choose_mds chose random mds%d\n", mds);
826 return mds;
827}
828
829
830/*
831 * session messages
832 */
833static struct ceph_msg *create_session_msg(u32 op, u64 seq)
834{
835 struct ceph_msg *msg;
836 struct ceph_mds_session_head *h;
837
b61c2763
SW		 838 msg = ceph_msg_new(CEPH_MSG_CLIENT_SESSION, sizeof(*h), GFP_NOFS,
839 false);
a79832f2 840 if (!msg) {
2f2dc053 841 pr_err("create_session_msg ENOMEM creating msg\n");
a79832f2 842 return NULL;
2f2dc053
SW		 843 }
844 h = msg->front.iov_base;
845 h->op = cpu_to_le32(op);
846 h->seq = cpu_to_le64(seq);
dbd0c8bf
JS		 847
848 return msg;
849}
850
851/*
852 * session message, specialization for CEPH_SESSION_REQUEST_OPEN
853 * to include additional client metadata fields.
854 */
855static struct ceph_msg *create_session_open_msg(struct ceph_mds_client *mdsc, u64 seq)
856{
857 struct ceph_msg *msg;
858 struct ceph_mds_session_head *h;
859 int i = -1;
860 int metadata_bytes = 0;
861 int metadata_key_count = 0;
862 struct ceph_options *opt = mdsc->fsc->client->options;
863 void *p;
864
a6a5ce4f 865 const char* metadata[][2] = {
dbd0c8bf 866 {"hostname", utsname()->nodename},
a6a5ce4f 867 {"kernel_version", utsname()->release},
dbd0c8bf
JS		 868 {"entity_id", opt->name ? opt->name : ""},
869 {NULL, NULL}
870 };
871
872 /* Calculate serialized length of metadata */
873 metadata_bytes = 4; /* map length */
874 for (i = 0; metadata[i][0] != NULL; ++i) {
875 metadata_bytes += 8 + strlen(metadata[i][0]) +
876 strlen(metadata[i][1]);
877 metadata_key_count++;
878 }
879
880 /* Allocate the message */
881 msg = ceph_msg_new(CEPH_MSG_CLIENT_SESSION, sizeof(*h) + metadata_bytes,
882 GFP_NOFS, false);
883 if (!msg) {
884 pr_err("create_session_msg ENOMEM creating msg\n");
885 return NULL;
886 }
887 h = msg->front.iov_base;
888 h->op = cpu_to_le32(CEPH_SESSION_REQUEST_OPEN);
889 h->seq = cpu_to_le64(seq);
890
891 /*
892 * Serialize client metadata into waiting buffer space, using
893 * the format that userspace expects for map<string, string>
7cfa0313
JS		 894 *
895 * ClientSession messages with metadata are v2
dbd0c8bf 896 */
7cfa0313
JS		 897 msg->hdr.version = cpu_to_le16(2);
898 msg->hdr.compat_version = cpu_to_le16(1);
dbd0c8bf
JS		 899
900 /* The write pointer, following the session_head structure */
901 p = msg->front.iov_base + sizeof(*h);
902
903 /* Number of entries in the map */
904 ceph_encode_32(&p, metadata_key_count);
905
906 /* Two length-prefixed strings for each entry in the map */
907 for (i = 0; metadata[i][0] != NULL; ++i) {
908 size_t const key_len = strlen(metadata[i][0]);
909 size_t const val_len = strlen(metadata[i][1]);
910
911 ceph_encode_32(&p, key_len);
912 memcpy(p, metadata[i][0], key_len);
913 p += key_len;
914 ceph_encode_32(&p, val_len);
915 memcpy(p, metadata[i][1], val_len);
916 p += val_len;
917 }
918
2f2dc053
SW		 919 return msg;
920}
921
922/*
923 * send session open request.
924 *
925 * called under mdsc->mutex
926 */
927static int __open_session(struct ceph_mds_client *mdsc,
928 struct ceph_mds_session *session)
929{
930 struct ceph_msg *msg;
931 int mstate;
932 int mds = session->s_mds;
2f2dc053
SW		 933
934 /* wait for mds to go active? */
935 mstate = ceph_mdsmap_get_state(mdsc->mdsmap, mds);
936 dout("open_session to mds%d (%s)\n", mds,
937 ceph_mds_state_name(mstate));
938 session->s_state = CEPH_MDS_SESSION_OPENING;
939 session->s_renew_requested = jiffies;
940
941 /* send connect message */
dbd0c8bf 942 msg = create_session_open_msg(mdsc, session->s_seq);
a79832f2
SW		 943 if (!msg)
944 return -ENOMEM;
2f2dc053 945 ceph_con_send(&session->s_con, msg);
2f2dc053
SW		 946 return 0;
947}
948
ed0552a1
SW		 949/*
950 * open sessions for any export targets for the given mds
951 *
952 * called under mdsc->mutex
953 */
5d72d13c
YZ		 954static struct ceph_mds_session *
955__open_export_target_session(struct ceph_mds_client *mdsc, int target)
956{
957 struct ceph_mds_session *session;
958
959 session = __ceph_lookup_mds_session(mdsc, target);
960 if (!session) {
961 session = register_session(mdsc, target);
962 if (IS_ERR(session))
963 return session;
964 }
965 if (session->s_state == CEPH_MDS_SESSION_NEW ||
966 session->s_state == CEPH_MDS_SESSION_CLOSING)
967 __open_session(mdsc, session);
968
969 return session;
970}
971
972struct ceph_mds_session *
973ceph_mdsc_open_export_target_session(struct ceph_mds_client *mdsc, int target)
974{
975 struct ceph_mds_session *session;
976
977 dout("open_export_target_session to mds%d\n", target);
978
979 mutex_lock(&mdsc->mutex);
980 session = __open_export_target_session(mdsc, target);
981 mutex_unlock(&mdsc->mutex);
982
983 return session;
984}
985
ed0552a1
SW		 986static void __open_export_target_sessions(struct ceph_mds_client *mdsc,
987 struct ceph_mds_session *session)
988{
989 struct ceph_mds_info *mi;
990 struct ceph_mds_session *ts;
991 int i, mds = session->s_mds;
ed0552a1
SW		 992
993 if (mds >= mdsc->mdsmap->m_max_mds)
994 return;
5d72d13c 995
ed0552a1
SW		 996 mi = &mdsc->mdsmap->m_info[mds];
997 dout("open_export_target_sessions for mds%d (%d targets)\n",
998 session->s_mds, mi->num_export_targets);
999
1000 for (i = 0; i < mi->num_export_targets; i++) {
5d72d13c
YZ		 1001 ts = __open_export_target_session(mdsc, mi->export_targets[i]);
1002 if (!IS_ERR(ts))
1003 ceph_put_mds_session(ts);
ed0552a1
SW		 1004 }
1005}
1006
154f42c2
SW		 1007void ceph_mdsc_open_export_target_sessions(struct ceph_mds_client *mdsc,
1008 struct ceph_mds_session *session)
1009{
1010 mutex_lock(&mdsc->mutex);
1011 __open_export_target_sessions(mdsc, session);
1012 mutex_unlock(&mdsc->mutex);
1013}
1014
2f2dc053
SW		 1015/*
1016 * session caps
1017 */
1018
745a8e3b
YZ		 1019/* caller holds s_cap_lock, we drop it */
1020static void cleanup_cap_releases(struct ceph_mds_client *mdsc,
1021 struct ceph_mds_session *session)
1022 __releases(session->s_cap_lock)
2f2dc053 1023{
745a8e3b
YZ		 1024 LIST_HEAD(tmp_list);
1025 list_splice_init(&session->s_cap_releases, &tmp_list);
1026 session->s_num_cap_releases = 0;
1027 spin_unlock(&session->s_cap_lock);
2f2dc053 1028
745a8e3b
YZ		 1029 dout("cleanup_cap_releases mds%d\n", session->s_mds);
1030 while (!list_empty(&tmp_list)) {
1031 struct ceph_cap *cap;
1032 /* zero out the in-progress message */
1033 cap = list_first_entry(&tmp_list,
1034 struct ceph_cap, session_caps);
1035 list_del(&cap->session_caps);
1036 ceph_put_cap(mdsc, cap);
2f2dc053 1037 }
2f2dc053
SW		 1038}
1039
1c841a96
YZ		 1040static void cleanup_session_requests(struct ceph_mds_client *mdsc,
1041 struct ceph_mds_session *session)
1042{
1043 struct ceph_mds_request *req;
1044 struct rb_node *p;
1045
1046 dout("cleanup_session_requests mds%d\n", session->s_mds);
1047 mutex_lock(&mdsc->mutex);
1048 while (!list_empty(&session->s_unsafe)) {
1049 req = list_first_entry(&session->s_unsafe,
1050 struct ceph_mds_request, r_unsafe_item);
1051 list_del_init(&req->r_unsafe_item);
3e0708b9
YZ		 1052 pr_warn_ratelimited(" dropping unsafe request %llu\n",
1053 req->r_tid);
1c841a96
YZ		 1054 __unregister_request(mdsc, req);
1055 }
1056 /* zero r_attempts, so kick_requests() will re-send requests */
1057 p = rb_first(&mdsc->request_tree);
1058 while (p) {
1059 req = rb_entry(p, struct ceph_mds_request, r_node);
1060 p = rb_next(p);
1061 if (req->r_session &&
1062 req->r_session->s_mds == session->s_mds)
1063 req->r_attempts = 0;
1064 }
1065 mutex_unlock(&mdsc->mutex);
1066}
1067
2f2dc053 1068/*
f818a736
SW		 1069 * Helper to safely iterate over all caps associated with a session, with
1070 * special care taken to handle a racing __ceph_remove_cap().
2f2dc053 1071 *
f818a736 1072 * Caller must hold session s_mutex.
2f2dc053
SW		 1073 */
1074static int iterate_session_caps(struct ceph_mds_session *session,
1075 int (*cb)(struct inode *, struct ceph_cap *,
1076 void *), void *arg)
1077{
7c1332b8
SW		 1078 struct list_head *p;
1079 struct ceph_cap *cap;
1080 struct inode *inode, *last_inode = NULL;
1081 struct ceph_cap *old_cap = NULL;
2f2dc053
SW		 1082 int ret;
1083
1084 dout("iterate_session_caps %p mds%d\n", session, session->s_mds);
1085 spin_lock(&session->s_cap_lock);
7c1332b8
SW		 1086 p = session->s_caps.next;
1087 while (p != &session->s_caps) {
1088 cap = list_entry(p, struct ceph_cap, session_caps);
2f2dc053 1089 inode = igrab(&cap->ci->vfs_inode);
7c1332b8
SW		 1090 if (!inode) {
1091 p = p->next;
2f2dc053 1092 continue;
7c1332b8
SW		 1093 }
1094 session->s_cap_iterator = cap;
2f2dc053 1095 spin_unlock(&session->s_cap_lock);
7c1332b8
SW		 1096
1097 if (last_inode) {
1098 iput(last_inode);
1099 last_inode = NULL;
1100 }
1101 if (old_cap) {
37151668 1102 ceph_put_cap(session->s_mdsc, old_cap);
7c1332b8
SW		 1103 old_cap = NULL;
1104 }
1105
2f2dc053 1106 ret = cb(inode, cap, arg);
7c1332b8
SW		 1107 last_inode = inode;
1108
2f2dc053 1109 spin_lock(&session->s_cap_lock);
7c1332b8
SW		 1110 p = p->next;
1111 if (cap->ci == NULL) {
1112 dout("iterate_session_caps finishing cap %p removal\n",
1113 cap);
1114 BUG_ON(cap->session != session);
745a8e3b 1115 cap->session = NULL;
7c1332b8
SW		 1116 list_del_init(&cap->session_caps);
1117 session->s_nr_caps--;
745a8e3b
YZ		 1118 if (cap->queue_release) {
1119 list_add_tail(&cap->session_caps,
1120 &session->s_cap_releases);
1121 session->s_num_cap_releases++;
1122 } else {
1123 old_cap = cap; /* put_cap it w/o locks held */
1124 }
7c1332b8 1125 }
5dacf091
SW		 1126 if (ret < 0)
1127 goto out;
2f2dc053 1128 }
5dacf091
SW		 1129 ret = 0;
1130out:
7c1332b8 1131 session->s_cap_iterator = NULL;
2f2dc053 1132 spin_unlock(&session->s_cap_lock);
7c1332b8 1133
e96a650a 1134 iput(last_inode);
7c1332b8 1135 if (old_cap)
37151668 1136 ceph_put_cap(session->s_mdsc, old_cap);
7c1332b8 1137
5dacf091 1138 return ret;
2f2dc053
SW		 1139}
1140
1141static int remove_session_caps_cb(struct inode *inode, struct ceph_cap *cap,
6c99f254 1142 void *arg)
2f2dc053
SW		 1143{
1144 struct ceph_inode_info *ci = ceph_inode(inode);
553adfd9 1145 LIST_HEAD(to_remove);
6c99f254
SW		 1146 int drop = 0;
1147
2f2dc053
SW		 1148 dout("removing cap %p, ci is %p, inode is %p\n",
1149 cap, ci, &ci->vfs_inode);
be655596 1150 spin_lock(&ci->i_ceph_lock);
a096b09a 1151 __ceph_remove_cap(cap, false);
571ade33 1152 if (!ci->i_auth_cap) {
553adfd9 1153 struct ceph_cap_flush *cf;
6c99f254 1154 struct ceph_mds_client *mdsc =
3d14c5d2 1155 ceph_sb_to_client(inode->i_sb)->mdsc;
6c99f254 1156
553adfd9
YZ		 1157 while (true) {
1158 struct rb_node *n = rb_first(&ci->i_cap_flush_tree);
1159 if (!n)
1160 break;
1161 cf = rb_entry(n, struct ceph_cap_flush, i_node);
1162 rb_erase(&cf->i_node, &ci->i_cap_flush_tree);
1163 list_add(&cf->list, &to_remove);
1164 }
1165
6c99f254 1166 spin_lock(&mdsc->cap_dirty_lock);
8310b089
YZ		 1167
1168 list_for_each_entry(cf, &to_remove, list)
1169 rb_erase(&cf->g_node, &mdsc->cap_flush_tree);
1170
6c99f254 1171 if (!list_empty(&ci->i_dirty_item)) {
3e0708b9
YZ		 1172 pr_warn_ratelimited(
1173 " dropping dirty %s state for %p %lld\n",
6c99f254
SW		 1174 ceph_cap_string(ci->i_dirty_caps),
1175 inode, ceph_ino(inode));
1176 ci->i_dirty_caps = 0;
1177 list_del_init(&ci->i_dirty_item);
1178 drop = 1;
1179 }
1180 if (!list_empty(&ci->i_flushing_item)) {
3e0708b9
YZ		 1181 pr_warn_ratelimited(
1182 " dropping dirty+flushing %s state for %p %lld\n",
6c99f254
SW		 1183 ceph_cap_string(ci->i_flushing_caps),
1184 inode, ceph_ino(inode));
1185 ci->i_flushing_caps = 0;
1186 list_del_init(&ci->i_flushing_item);
1187 mdsc->num_cap_flushing--;
1188 drop = 1;
1189 }
6c99f254 1190 spin_unlock(&mdsc->cap_dirty_lock);
553adfd9 1191
f66fd9f0
YZ		 1192 if (!ci->i_dirty_caps && ci->i_prealloc_cap_flush) {
1193 list_add(&ci->i_prealloc_cap_flush->list, &to_remove);
1194 ci->i_prealloc_cap_flush = NULL;
1195 }
6c99f254 1196 }
be655596 1197 spin_unlock(&ci->i_ceph_lock);
553adfd9
YZ		 1198 while (!list_empty(&to_remove)) {
1199 struct ceph_cap_flush *cf;
1200 cf = list_first_entry(&to_remove,
1201 struct ceph_cap_flush, list);
1202 list_del(&cf->list);
f66fd9f0 1203 ceph_free_cap_flush(cf);
553adfd9 1204 }
6c99f254
SW		 1205 while (drop--)
1206 iput(inode);
2f2dc053
SW		 1207 return 0;
1208}
1209
1210/*
1211 * caller must hold session s_mutex
1212 */
1213static void remove_session_caps(struct ceph_mds_session *session)
1214{
1215 dout("remove_session_caps on %p\n", session);
1216 iterate_session_caps(session, remove_session_caps_cb, NULL);
6f60f889
YZ		 1217
1218 spin_lock(&session->s_cap_lock);
1219 if (session->s_nr_caps > 0) {
1220 struct super_block *sb = session->s_mdsc->fsc->sb;
1221 struct inode *inode;
1222 struct ceph_cap *cap, *prev = NULL;
1223 struct ceph_vino vino;
1224 /*
1225 * iterate_session_caps() skips inodes that are being
1226 * deleted, we need to wait until deletions are complete.
1227 * __wait_on_freeing_inode() is designed for the job,
1228 * but it is not exported, so use lookup inode function
1229 * to access it.
1230 */
1231 while (!list_empty(&session->s_caps)) {
1232 cap = list_entry(session->s_caps.next,
1233 struct ceph_cap, session_caps);
1234 if (cap == prev)
1235 break;
1236 prev = cap;
1237 vino = cap->ci->i_vino;
1238 spin_unlock(&session->s_cap_lock);
1239
ed284c49 1240 inode = ceph_find_inode(sb, vino);
6f60f889
YZ		 1241 iput(inode);
1242
1243 spin_lock(&session->s_cap_lock);
1244 }
1245 }
745a8e3b
YZ		 1246
1247 // drop cap expires and unlock s_cap_lock
1248 cleanup_cap_releases(session->s_mdsc, session);
6f60f889 1249
2f2dc053 1250 BUG_ON(session->s_nr_caps > 0);
6c99f254 1251 BUG_ON(!list_empty(&session->s_cap_flushing));
2f2dc053
SW		 1252}
1253
1254/*
1255 * wake up any threads waiting on this session's caps. if the cap is
1256 * old (didn't get renewed on the client reconnect), remove it now.
1257 *
1258 * caller must hold s_mutex.
1259 */
1260static int wake_up_session_cb(struct inode *inode, struct ceph_cap *cap,
1261 void *arg)
1262{
0dc2570f
SW		 1263 struct ceph_inode_info *ci = ceph_inode(inode);
1264
03066f23 1265 wake_up_all(&ci->i_cap_wq);
0dc2570f 1266 if (arg) {
be655596 1267 spin_lock(&ci->i_ceph_lock);
0dc2570f
SW		 1268 ci->i_wanted_max_size = 0;
1269 ci->i_requested_max_size = 0;
be655596 1270 spin_unlock(&ci->i_ceph_lock);
0dc2570f 1271 }
2f2dc053
SW		 1272 return 0;
1273}
1274
0dc2570f
SW		 1275static void wake_up_session_caps(struct ceph_mds_session *session,
1276 int reconnect)
2f2dc053
SW		 1277{
1278 dout("wake_up_session_caps %p mds%d\n", session, session->s_mds);
0dc2570f
SW		 1279 iterate_session_caps(session, wake_up_session_cb,
1280 (void *)(unsigned long)reconnect);
2f2dc053
SW		 1281}
1282
1283/*
1284 * Send periodic message to MDS renewing all currently held caps. The
1285 * ack will reset the expiration for all caps from this session.
1286 *
1287 * caller holds s_mutex
1288 */
1289static int send_renew_caps(struct ceph_mds_client *mdsc,
1290 struct ceph_mds_session *session)
1291{
1292 struct ceph_msg *msg;
1293 int state;
1294
1295 if (time_after_eq(jiffies, session->s_cap_ttl) &&
1296 time_after_eq(session->s_cap_ttl, session->s_renew_requested))
1297 pr_info("mds%d caps stale\n", session->s_mds);
e4cb4cb8 1298 session->s_renew_requested = jiffies;
2f2dc053
SW		 1299
1300 /* do not try to renew caps until a recovering mds has reconnected
1301 * with its clients. */
1302 state = ceph_mdsmap_get_state(mdsc->mdsmap, session->s_mds);
1303 if (state < CEPH_MDS_STATE_RECONNECT) {
1304 dout("send_renew_caps ignoring mds%d (%s)\n",
1305 session->s_mds, ceph_mds_state_name(state));
1306 return 0;
1307 }
1308
1309 dout("send_renew_caps to mds%d (%s)\n", session->s_mds,
1310 ceph_mds_state_name(state));
2f2dc053
SW		 1311 msg = create_session_msg(CEPH_SESSION_REQUEST_RENEWCAPS,
1312 ++session->s_renew_seq);
a79832f2
SW		 1313 if (!msg)
1314 return -ENOMEM;
2f2dc053
SW		 1315 ceph_con_send(&session->s_con, msg);
1316 return 0;
1317}
1318
186e4f7a
YZ		 1319static int send_flushmsg_ack(struct ceph_mds_client *mdsc,
1320 struct ceph_mds_session *session, u64 seq)
1321{
1322 struct ceph_msg *msg;
1323
1324 dout("send_flushmsg_ack to mds%d (%s)s seq %lld\n",
a687ecaf 1325 session->s_mds, ceph_session_state_name(session->s_state), seq);
186e4f7a
YZ		 1326 msg = create_session_msg(CEPH_SESSION_FLUSHMSG_ACK, seq);
1327 if (!msg)
1328 return -ENOMEM;
1329 ceph_con_send(&session->s_con, msg);
1330 return 0;
1331}
1332
1333
2f2dc053
SW		 1334/*
1335 * Note new cap ttl, and any transition from stale -> not stale (fresh?).
0dc2570f
SW		 1336 *
1337 * Called under session->s_mutex
2f2dc053
SW		 1338 */
1339static void renewed_caps(struct ceph_mds_client *mdsc,
1340 struct ceph_mds_session *session, int is_renew)
1341{
1342 int was_stale;
1343 int wake = 0;
1344
1345 spin_lock(&session->s_cap_lock);
1ce208a6 1346 was_stale = is_renew && time_after_eq(jiffies, session->s_cap_ttl);
2f2dc053
SW		 1347
1348 session->s_cap_ttl = session->s_renew_requested +
1349 mdsc->mdsmap->m_session_timeout*HZ;
1350
1351 if (was_stale) {
1352 if (time_before(jiffies, session->s_cap_ttl)) {
1353 pr_info("mds%d caps renewed\n", session->s_mds);
1354 wake = 1;
1355 } else {
1356 pr_info("mds%d caps still stale\n", session->s_mds);
1357 }
1358 }
1359 dout("renewed_caps mds%d ttl now %lu, was %s, now %s\n",
1360 session->s_mds, session->s_cap_ttl, was_stale ? "stale" : "fresh",
1361 time_before(jiffies, session->s_cap_ttl) ? "stale" : "fresh");
1362 spin_unlock(&session->s_cap_lock);
1363
1364 if (wake)
0dc2570f 1365 wake_up_session_caps(session, 0);
2f2dc053
SW		 1366}
1367
1368/*
1369 * send a session close request
1370 */
1371static int request_close_session(struct ceph_mds_client *mdsc,
1372 struct ceph_mds_session *session)
1373{
1374 struct ceph_msg *msg;
2f2dc053
SW		 1375
1376 dout("request_close_session mds%d state %s seq %lld\n",
a687ecaf 1377 session->s_mds, ceph_session_state_name(session->s_state),
2f2dc053
SW		 1378 session->s_seq);
1379 msg = create_session_msg(CEPH_SESSION_REQUEST_CLOSE, session->s_seq);
a79832f2
SW		 1380 if (!msg)
1381 return -ENOMEM;
1382 ceph_con_send(&session->s_con, msg);
1383 return 0;
2f2dc053
SW		 1384}
1385
1386/*
1387 * Called with s_mutex held.
1388 */
1389static int __close_session(struct ceph_mds_client *mdsc,
1390 struct ceph_mds_session *session)
1391{
1392 if (session->s_state >= CEPH_MDS_SESSION_CLOSING)
1393 return 0;
1394 session->s_state = CEPH_MDS_SESSION_CLOSING;
1395 return request_close_session(mdsc, session);
1396}
1397
1398/*
1399 * Trim old(er) caps.
1400 *
1401 * Because we can't cache an inode without one or more caps, we do
1402 * this indirectly: if a cap is unused, we prune its aliases, at which
1403 * point the inode will hopefully get dropped to.
1404 *
1405 * Yes, this is a bit sloppy. Our only real goal here is to respond to
1406 * memory pressure from the MDS, though, so it needn't be perfect.
1407 */
1408static int trim_caps_cb(struct inode *inode, struct ceph_cap *cap, void *arg)
1409{
1410 struct ceph_mds_session *session = arg;
1411 struct ceph_inode_info *ci = ceph_inode(inode);
979abfdd 1412 int used, wanted, oissued, mine;
2f2dc053
SW		 1413
1414 if (session->s_trim_caps <= 0)
1415 return -1;
1416
be655596 1417 spin_lock(&ci->i_ceph_lock);
2f2dc053
SW		 1418 mine = cap->issued | cap->implemented;
1419 used = __ceph_caps_used(ci);
979abfdd 1420 wanted = __ceph_caps_file_wanted(ci);
2f2dc053
SW		 1421 oissued = __ceph_caps_issued_other(ci, cap);
1422
979abfdd 1423 dout("trim_caps_cb %p cap %p mine %s oissued %s used %s wanted %s\n",
2f2dc053 1424 inode, cap, ceph_cap_string(mine), ceph_cap_string(oissued),
979abfdd
YZ		 1425 ceph_cap_string(used), ceph_cap_string(wanted));
1426 if (cap == ci->i_auth_cap) {
622f3e25
YZ		 1427 if (ci->i_dirty_caps || ci->i_flushing_caps ||
1428 !list_empty(&ci->i_cap_snaps))
979abfdd
YZ		 1429 goto out;
1430 if ((used | wanted) & CEPH_CAP_ANY_WR)
1431 goto out;
1432 }
1433 if ((used | wanted) & ~oissued & mine)
2f2dc053
SW		 1434 goto out; /* we need these caps */
1435
1436 session->s_trim_caps--;
1437 if (oissued) {
1438 /* we aren't the only cap.. just remove us */
a096b09a 1439 __ceph_remove_cap(cap, true);
2f2dc053
SW		 1440 } else {
1441 /* try to drop referring dentries */
be655596 1442 spin_unlock(&ci->i_ceph_lock);
2f2dc053
SW		 1443 d_prune_aliases(inode);
1444 dout("trim_caps_cb %p cap %p pruned, count now %d\n",
1445 inode, cap, atomic_read(&inode->i_count));
1446 return 0;
1447 }
1448
1449out:
be655596 1450 spin_unlock(&ci->i_ceph_lock);
2f2dc053
SW		 1451 return 0;
1452}
1453
1454/*
1455 * Trim session cap count down to some max number.
1456 */
1457static int trim_caps(struct ceph_mds_client *mdsc,
1458 struct ceph_mds_session *session,
1459 int max_caps)
1460{
1461 int trim_caps = session->s_nr_caps - max_caps;
1462
1463 dout("trim_caps mds%d start: %d / %d, trim %d\n",
1464 session->s_mds, session->s_nr_caps, max_caps, trim_caps);
1465 if (trim_caps > 0) {
1466 session->s_trim_caps = trim_caps;
1467 iterate_session_caps(session, trim_caps_cb, session);
1468 dout("trim_caps mds%d done: %d / %d, trimmed %d\n",
1469 session->s_mds, session->s_nr_caps, max_caps,
1470 trim_caps - session->s_trim_caps);
5dacf091 1471 session->s_trim_caps = 0;
2f2dc053 1472 }
a56371d9 1473
a56371d9 1474 ceph_send_cap_releases(mdsc, session);
2f2dc053
SW		 1475 return 0;
1476}
1477
8310b089
YZ		 1478static int check_capsnap_flush(struct ceph_inode_info *ci,
1479 u64 want_snap_seq)
d3383a8e 1480{
8310b089 1481 int ret = 1;
d3383a8e 1482 spin_lock(&ci->i_ceph_lock);
affbc19a
YZ		 1483 if (want_snap_seq > 0 && !list_empty(&ci->i_cap_snaps)) {
1484 struct ceph_cap_snap *capsnap =
1485 list_first_entry(&ci->i_cap_snaps,
1486 struct ceph_cap_snap, ci_item);
8310b089 1487 ret = capsnap->follows >= want_snap_seq;
affbc19a 1488 }
d3383a8e 1489 spin_unlock(&ci->i_ceph_lock);
8310b089
YZ		 1490 return ret;
1491}
1492
1493static int check_caps_flush(struct ceph_mds_client *mdsc,
1494 u64 want_flush_tid)
1495{
1496 struct rb_node *n;
1497 struct ceph_cap_flush *cf;
1498 int ret = 1;
1499
1500 spin_lock(&mdsc->cap_dirty_lock);
1501 n = rb_first(&mdsc->cap_flush_tree);
1502 cf = n ? rb_entry(n, struct ceph_cap_flush, g_node) : NULL;
1503 if (cf && cf->tid <= want_flush_tid) {
1504 dout("check_caps_flush still flushing tid %llu <= %llu\n",
1505 cf->tid, want_flush_tid);
1506 ret = 0;
1507 }
1508 spin_unlock(&mdsc->cap_dirty_lock);
1509 return ret;
d3383a8e
YZ		 1510}
1511
2f2dc053
SW		 1512/*
1513 * flush all dirty inode data to disk.
1514 *
8310b089 1515 * returns true if we've flushed through want_flush_tid
2f2dc053 1516 */
affbc19a 1517static void wait_caps_flush(struct ceph_mds_client *mdsc,
8310b089 1518 u64 want_flush_tid, u64 want_snap_seq)
2f2dc053 1519{
d3383a8e 1520 int mds;
2f2dc053 1521
8310b089
YZ		 1522 dout("check_caps_flush want %llu snap want %llu\n",
1523 want_flush_tid, want_snap_seq);
2f2dc053 1524 mutex_lock(&mdsc->mutex);
affbc19a 1525 for (mds = 0; mds < mdsc->max_sessions; ) {
2f2dc053 1526 struct ceph_mds_session *session = mdsc->sessions[mds];
8310b089 1527 struct inode *inode = NULL;
2f2dc053 1528
affbc19a
YZ		 1529 if (!session) {
1530 mds++;
2f2dc053 1531 continue;
affbc19a 1532 }
2f2dc053
SW		 1533 get_session(session);
1534 mutex_unlock(&mdsc->mutex);
1535
1536 mutex_lock(&session->s_mutex);
affbc19a
YZ		 1537 if (!list_empty(&session->s_cap_snaps_flushing)) {
1538 struct ceph_cap_snap *capsnap =
1539 list_first_entry(&session->s_cap_snaps_flushing,
1540 struct ceph_cap_snap,
1541 flushing_item);
1542 struct ceph_inode_info *ci = capsnap->ci;
8310b089 1543 if (!check_capsnap_flush(ci, want_snap_seq)) {
affbc19a
YZ		 1544 dout("check_cap_flush still flushing snap %p "
1545 "follows %lld <= %lld to mds%d\n",
1546 &ci->vfs_inode, capsnap->follows,
1547 want_snap_seq, mds);
8310b089 1548 inode = igrab(&ci->vfs_inode);
2f2dc053 1549 }
2f2dc053
SW		 1550 }
1551 mutex_unlock(&session->s_mutex);
1552 ceph_put_mds_session(session);
1553
8310b089 1554 if (inode) {
affbc19a 1555 wait_event(mdsc->cap_flushing_wq,
8310b089
YZ		 1556 check_capsnap_flush(ceph_inode(inode),
1557 want_snap_seq));
1558 iput(inode);
1559 } else {
affbc19a 1560 mds++;
8310b089 1561 }
affbc19a 1562
2f2dc053
SW		 1563 mutex_lock(&mdsc->mutex);
1564 }
2f2dc053 1565 mutex_unlock(&mdsc->mutex);
8310b089
YZ		 1566
1567 wait_event(mdsc->cap_flushing_wq,
1568 check_caps_flush(mdsc, want_flush_tid));
1569
1570 dout("check_caps_flush ok, flushed thru %llu\n", want_flush_tid);
2f2dc053
SW		 1571}
1572
1573/*
1574 * called under s_mutex
1575 */
3d7ded4d
SW		 1576void ceph_send_cap_releases(struct ceph_mds_client *mdsc,
1577 struct ceph_mds_session *session)
2f2dc053 1578{
745a8e3b
YZ		 1579 struct ceph_msg *msg = NULL;
1580 struct ceph_mds_cap_release *head;
1581 struct ceph_mds_cap_item *item;
1582 struct ceph_cap *cap;
1583 LIST_HEAD(tmp_list);
1584 int num_cap_releases;
2f2dc053 1585
0f8605f2 1586 spin_lock(&session->s_cap_lock);
745a8e3b
YZ		 1587again:
1588 list_splice_init(&session->s_cap_releases, &tmp_list);
1589 num_cap_releases = session->s_num_cap_releases;
1590 session->s_num_cap_releases = 0;
2f2dc053 1591 spin_unlock(&session->s_cap_lock);
e01a5946 1592
745a8e3b
YZ		 1593 while (!list_empty(&tmp_list)) {
1594 if (!msg) {
1595 msg = ceph_msg_new(CEPH_MSG_CLIENT_CAPRELEASE,
1596 PAGE_CACHE_SIZE, GFP_NOFS, false);
1597 if (!msg)
1598 goto out_err;
1599 head = msg->front.iov_base;
1600 head->num = cpu_to_le32(0);
1601 msg->front.iov_len = sizeof(*head);
1602 }
1603 cap = list_first_entry(&tmp_list, struct ceph_cap,
1604 session_caps);
1605 list_del(&cap->session_caps);
1606 num_cap_releases--;
e01a5946 1607
00bd8edb 1608 head = msg->front.iov_base;
745a8e3b
YZ		 1609 le32_add_cpu(&head->num, 1);
1610 item = msg->front.iov_base + msg->front.iov_len;
1611 item->ino = cpu_to_le64(cap->cap_ino);
1612 item->cap_id = cpu_to_le64(cap->cap_id);
1613 item->migrate_seq = cpu_to_le32(cap->mseq);
1614 item->seq = cpu_to_le32(cap->issue_seq);
1615 msg->front.iov_len += sizeof(*item);
1616
1617 ceph_put_cap(mdsc, cap);
1618
1619 if (le32_to_cpu(head->num) == CEPH_CAPS_PER_RELEASE) {
1620 msg->hdr.front_len = cpu_to_le32(msg->front.iov_len);
1621 dout("send_cap_releases mds%d %p\n", session->s_mds, msg);
1622 ceph_con_send(&session->s_con, msg);
1623 msg = NULL;
1624 }
00bd8edb 1625 }
e01a5946 1626
745a8e3b 1627 BUG_ON(num_cap_releases != 0);
e01a5946 1628
745a8e3b
YZ		 1629 spin_lock(&session->s_cap_lock);
1630 if (!list_empty(&session->s_cap_releases))
1631 goto again;
1632 spin_unlock(&session->s_cap_lock);
1633
1634 if (msg) {
1635 msg->hdr.front_len = cpu_to_le32(msg->front.iov_len);
1636 dout("send_cap_releases mds%d %p\n", session->s_mds, msg);
1637 ceph_con_send(&session->s_con, msg);
e01a5946 1638 }
745a8e3b
YZ		 1639 return;
1640out_err:
1641 pr_err("send_cap_releases mds%d, failed to allocate message\n",
1642 session->s_mds);
1643 spin_lock(&session->s_cap_lock);
1644 list_splice(&tmp_list, &session->s_cap_releases);
1645 session->s_num_cap_releases += num_cap_releases;
1646 spin_unlock(&session->s_cap_lock);
e01a5946
SW		 1647}
1648
2f2dc053
SW		 1649/*
1650 * requests
1651 */
1652
54008399
YZ		 1653int ceph_alloc_readdir_reply_buffer(struct ceph_mds_request *req,
1654 struct inode *dir)
1655{
1656 struct ceph_inode_info *ci = ceph_inode(dir);
1657 struct ceph_mds_reply_info_parsed *rinfo = &req->r_reply_info;
1658 struct ceph_mount_options *opt = req->r_mdsc->fsc->mount_options;
1659 size_t size = sizeof(*rinfo->dir_in) + sizeof(*rinfo->dir_dname_len) +
1660 sizeof(*rinfo->dir_dname) + sizeof(*rinfo->dir_dlease);
1661 int order, num_entries;
1662
1663 spin_lock(&ci->i_ceph_lock);
1664 num_entries = ci->i_files + ci->i_subdirs;
1665 spin_unlock(&ci->i_ceph_lock);
1666 num_entries = max(num_entries, 1);
1667 num_entries = min(num_entries, opt->max_readdir);
1668
1669 order = get_order(size * num_entries);
1670 while (order >= 0) {
687265e5
YZ		 1671 rinfo->dir_in = (void*)__get_free_pages(GFP_KERNEL |
1672 __GFP_NOWARN,
54008399
YZ		 1673 order);
1674 if (rinfo->dir_in)
1675 break;
1676 order--;
1677 }
1678 if (!rinfo->dir_in)
1679 return -ENOMEM;
1680
1681 num_entries = (PAGE_SIZE << order) / size;
1682 num_entries = min(num_entries, opt->max_readdir);
1683
1684 rinfo->dir_buf_size = PAGE_SIZE << order;
1685 req->r_num_caps = num_entries + 1;
1686 req->r_args.readdir.max_entries = cpu_to_le32(num_entries);
1687 req->r_args.readdir.max_bytes = cpu_to_le32(opt->max_readdir_bytes);
1688 return 0;
1689}
1690
2f2dc053
SW		 1691/*
1692 * Create an mds request.
1693 */
1694struct ceph_mds_request *
1695ceph_mdsc_create_request(struct ceph_mds_client *mdsc, int op, int mode)
1696{
1697 struct ceph_mds_request *req = kzalloc(sizeof(*req), GFP_NOFS);
1698
1699 if (!req)
1700 return ERR_PTR(-ENOMEM);
1701
b4556396 1702 mutex_init(&req->r_fill_mutex);
37151668 1703 req->r_mdsc = mdsc;
2f2dc053
SW		 1704 req->r_started = jiffies;
1705 req->r_resend_mds = -1;
1706 INIT_LIST_HEAD(&req->r_unsafe_dir_item);
1707 req->r_fmode = -1;
153c8e6b 1708 kref_init(&req->r_kref);
2f2dc053
SW		 1709 INIT_LIST_HEAD(&req->r_wait);
1710 init_completion(&req->r_completion);
1711 init_completion(&req->r_safe_completion);
1712 INIT_LIST_HEAD(&req->r_unsafe_item);
1713
b8e69066
SW		 1714 req->r_stamp = CURRENT_TIME;
1715
2f2dc053
SW		 1716 req->r_op = op;
1717 req->r_direct_mode = mode;
1718 return req;
1719}
1720
1721/*
44ca18f2 1722 * return oldest (lowest) request, tid in request tree, 0 if none.
2f2dc053
SW		 1723 *
1724 * called under mdsc->mutex.
1725 */
44ca18f2
SW		 1726static struct ceph_mds_request *__get_oldest_req(struct ceph_mds_client *mdsc)
1727{
1728 if (RB_EMPTY_ROOT(&mdsc->request_tree))
1729 return NULL;
1730 return rb_entry(rb_first(&mdsc->request_tree),
1731 struct ceph_mds_request, r_node);
1732}
1733
e8a7b8b1 1734static inline u64 __get_oldest_tid(struct ceph_mds_client *mdsc)
2f2dc053 1735{
e8a7b8b1 1736 return mdsc->oldest_tid;
2f2dc053
SW		 1737}
1738
1739/*
1740 * Build a dentry's path. Allocate on heap; caller must kfree. Based
1741 * on build_path_from_dentry in fs/cifs/dir.c.
1742 *
1743 * If @stop_on_nosnap, generate path relative to the first non-snapped
1744 * inode.
1745 *
1746 * Encode hidden .snap dirs as a double /, i.e.
1747 * foo/.snap/bar -> foo//bar
1748 */
1749char *ceph_mdsc_build_path(struct dentry *dentry, int *plen, u64 *base,
1750 int stop_on_nosnap)
1751{
1752 struct dentry *temp;
1753 char *path;
1754 int len, pos;
1b71fe2e 1755 unsigned seq;
2f2dc053
SW		 1756
1757 if (dentry == NULL)
1758 return ERR_PTR(-EINVAL);
1759
1760retry:
1761 len = 0;
1b71fe2e
AV		 1762 seq = read_seqbegin(&rename_lock);
1763 rcu_read_lock();
2f2dc053 1764 for (temp = dentry; !IS_ROOT(temp);) {
2b0143b5 1765 struct inode *inode = d_inode(temp);
2f2dc053
SW		 1766 if (inode && ceph_snap(inode) == CEPH_SNAPDIR)
1767 len++; /* slash only */
1768 else if (stop_on_nosnap && inode &&
1769 ceph_snap(inode) == CEPH_NOSNAP)
1770 break;
1771 else
1772 len += 1 + temp->d_name.len;
1773 temp = temp->d_parent;
2f2dc053 1774 }
1b71fe2e 1775 rcu_read_unlock();
2f2dc053
SW		 1776 if (len)
1777 len--; /* no leading '/' */
1778
1779 path = kmalloc(len+1, GFP_NOFS);
1780 if (path == NULL)
1781 return ERR_PTR(-ENOMEM);
1782 pos = len;
1783 path[pos] = 0; /* trailing null */
1b71fe2e 1784 rcu_read_lock();
2f2dc053 1785 for (temp = dentry; !IS_ROOT(temp) && pos != 0; ) {
1b71fe2e 1786 struct inode *inode;
2f2dc053 1787
1b71fe2e 1788 spin_lock(&temp->d_lock);
2b0143b5 1789 inode = d_inode(temp);
2f2dc053 1790 if (inode && ceph_snap(inode) == CEPH_SNAPDIR) {
104648ad 1791 dout("build_path path+%d: %p SNAPDIR\n",
2f2dc053
SW		 1792 pos, temp);
1793 } else if (stop_on_nosnap && inode &&
1794 ceph_snap(inode) == CEPH_NOSNAP) {
9d5a09e6 1795 spin_unlock(&temp->d_lock);
2f2dc053
SW		 1796 break;
1797 } else {
1798 pos -= temp->d_name.len;
1b71fe2e
AV		 1799 if (pos < 0) {
1800 spin_unlock(&temp->d_lock);
2f2dc053 1801 break;
1b71fe2e 1802 }
2f2dc053
SW		 1803 strncpy(path + pos, temp->d_name.name,
1804 temp->d_name.len);
2f2dc053 1805 }
1b71fe2e 1806 spin_unlock(&temp->d_lock);
2f2dc053
SW		 1807 if (pos)
1808 path[--pos] = '/';
1809 temp = temp->d_parent;
2f2dc053 1810 }
1b71fe2e
AV		 1811 rcu_read_unlock();
1812 if (pos != 0 || read_seqretry(&rename_lock, seq)) {
104648ad 1813 pr_err("build_path did not end path lookup where "
2f2dc053
SW		 1814 "expected, namelen is %d, pos is %d\n", len, pos);
1815 /* presumably this is only possible if racing with a
1816 rename of one of the parent directories (we can not
1817 lock the dentries above us to prevent this, but
1818 retrying should be harmless) */
1819 kfree(path);
1820 goto retry;
1821 }
1822
2b0143b5 1823 *base = ceph_ino(d_inode(temp));
2f2dc053 1824 *plen = len;
104648ad 1825 dout("build_path on %p %d built %llx '%.*s'\n",
84d08fa8 1826 dentry, d_count(dentry), *base, len, path);
2f2dc053
SW		 1827 return path;
1828}
1829
1830static int build_dentry_path(struct dentry *dentry,
1831 const char **ppath, int *ppathlen, u64 *pino,
1832 int *pfreepath)
1833{
1834 char *path;
1835
2b0143b5
DH		 1836 if (ceph_snap(d_inode(dentry->d_parent)) == CEPH_NOSNAP) {
1837 *pino = ceph_ino(d_inode(dentry->d_parent));
2f2dc053
SW		 1838 *ppath = dentry->d_name.name;
1839 *ppathlen = dentry->d_name.len;
1840 return 0;
1841 }
1842 path = ceph_mdsc_build_path(dentry, ppathlen, pino, 1);
1843 if (IS_ERR(path))
1844 return PTR_ERR(path);
1845 *ppath = path;
1846 *pfreepath = 1;
1847 return 0;
1848}
1849
1850static int build_inode_path(struct inode *inode,
1851 const char **ppath, int *ppathlen, u64 *pino,
1852 int *pfreepath)
1853{
1854 struct dentry *dentry;
1855 char *path;
1856
1857 if (ceph_snap(inode) == CEPH_NOSNAP) {
1858 *pino = ceph_ino(inode);
1859 *ppathlen = 0;
1860 return 0;
1861 }
1862 dentry = d_find_alias(inode);
1863 path = ceph_mdsc_build_path(dentry, ppathlen, pino, 1);
1864 dput(dentry);
1865 if (IS_ERR(path))
1866 return PTR_ERR(path);
1867 *ppath = path;
1868 *pfreepath = 1;
1869 return 0;
1870}
1871
1872/*
1873 * request arguments may be specified via an inode *, a dentry *, or
1874 * an explicit ino+path.
1875 */
1876static int set_request_path_attr(struct inode *rinode, struct dentry *rdentry,
1877 const char *rpath, u64 rino,
1878 const char **ppath, int *pathlen,
1879 u64 *ino, int *freepath)
1880{
1881 int r = 0;
1882
1883 if (rinode) {
1884 r = build_inode_path(rinode, ppath, pathlen, ino, freepath);
1885 dout(" inode %p %llx.%llx\n", rinode, ceph_ino(rinode),
1886 ceph_snap(rinode));
1887 } else if (rdentry) {
1888 r = build_dentry_path(rdentry, ppath, pathlen, ino, freepath);
1889 dout(" dentry %p %llx/%.*s\n", rdentry, *ino, *pathlen,
1890 *ppath);
795858db 1891 } else if (rpath || rino) {
2f2dc053
SW		 1892 *ino = rino;
1893 *ppath = rpath;
b000056a 1894 *pathlen = rpath ? strlen(rpath) : 0;
2f2dc053
SW		 1895 dout(" path %.*s\n", *pathlen, rpath);
1896 }
1897
1898 return r;
1899}
1900
1901/*
1902 * called under mdsc->mutex
1903 */
1904static struct ceph_msg *create_request_message(struct ceph_mds_client *mdsc,
1905 struct ceph_mds_request *req,
6e6f0923 1906 int mds, bool drop_cap_releases)
2f2dc053
SW		 1907{
1908 struct ceph_msg *msg;
1909 struct ceph_mds_request_head *head;
1910 const char *path1 = NULL;
1911 const char *path2 = NULL;
1912 u64 ino1 = 0, ino2 = 0;
1913 int pathlen1 = 0, pathlen2 = 0;
1914 int freepath1 = 0, freepath2 = 0;
1915 int len;
1916 u16 releases;
1917 void *p, *end;
1918 int ret;
1919
1920 ret = set_request_path_attr(req->r_inode, req->r_dentry,
1921 req->r_path1, req->r_ino1.ino,
1922 &path1, &pathlen1, &ino1, &freepath1);
1923 if (ret < 0) {
1924 msg = ERR_PTR(ret);
1925 goto out;
1926 }
1927
1928 ret = set_request_path_attr(NULL, req->r_old_dentry,
1929 req->r_path2, req->r_ino2.ino,
1930 &path2, &pathlen2, &ino2, &freepath2);
1931 if (ret < 0) {
1932 msg = ERR_PTR(ret);
1933 goto out_free1;
1934 }
1935
1936 len = sizeof(*head) +
b8e69066
SW		 1937 pathlen1 + pathlen2 + 2*(1 + sizeof(u32) + sizeof(u64)) +
1938 sizeof(struct timespec);
2f2dc053
SW		 1939
1940 /* calculate (max) length for cap releases */
1941 len += sizeof(struct ceph_mds_request_release) *
1942 (!!req->r_inode_drop + !!req->r_dentry_drop +
1943 !!req->r_old_inode_drop + !!req->r_old_dentry_drop);
1944 if (req->r_dentry_drop)
1945 len += req->r_dentry->d_name.len;
1946 if (req->r_old_dentry_drop)
1947 len += req->r_old_dentry->d_name.len;
1948
b61c2763 1949 msg = ceph_msg_new(CEPH_MSG_CLIENT_REQUEST, len, GFP_NOFS, false);
a79832f2
SW		 1950 if (!msg) {
1951 msg = ERR_PTR(-ENOMEM);
2f2dc053 1952 goto out_free2;
a79832f2 1953 }
2f2dc053 1954
7cfa0313 1955 msg->hdr.version = cpu_to_le16(2);
6df058c0
SW		 1956 msg->hdr.tid = cpu_to_le64(req->r_tid);
1957
2f2dc053
SW		 1958 head = msg->front.iov_base;
1959 p = msg->front.iov_base + sizeof(*head);
1960 end = msg->front.iov_base + msg->front.iov_len;
1961
1962 head->mdsmap_epoch = cpu_to_le32(mdsc->mdsmap->m_epoch);
1963 head->op = cpu_to_le32(req->r_op);
ff3d0046
EB		 1964 head->caller_uid = cpu_to_le32(from_kuid(&init_user_ns, req->r_uid));
1965 head->caller_gid = cpu_to_le32(from_kgid(&init_user_ns, req->r_gid));
2f2dc053
SW		 1966 head->args = req->r_args;
1967
1968 ceph_encode_filepath(&p, end, ino1, path1);
1969 ceph_encode_filepath(&p, end, ino2, path2);
1970
e979cf50
SW		 1971 /* make note of release offset, in case we need to replay */
1972 req->r_request_release_offset = p - msg->front.iov_base;
1973
2f2dc053
SW		 1974 /* cap releases */
1975 releases = 0;
1976 if (req->r_inode_drop)
1977 releases += ceph_encode_inode_release(&p,
2b0143b5 1978 req->r_inode ? req->r_inode : d_inode(req->r_dentry),
2f2dc053
SW		 1979 mds, req->r_inode_drop, req->r_inode_unless, 0);
1980 if (req->r_dentry_drop)
1981 releases += ceph_encode_dentry_release(&p, req->r_dentry,
1982 mds, req->r_dentry_drop, req->r_dentry_unless);
1983 if (req->r_old_dentry_drop)
1984 releases += ceph_encode_dentry_release(&p, req->r_old_dentry,
1985 mds, req->r_old_dentry_drop, req->r_old_dentry_unless);
1986 if (req->r_old_inode_drop)
1987 releases += ceph_encode_inode_release(&p,
2b0143b5 1988 d_inode(req->r_old_dentry),
2f2dc053 1989 mds, req->r_old_inode_drop, req->r_old_inode_unless, 0);
6e6f0923
YZ		 1990
1991 if (drop_cap_releases) {
1992 releases = 0;
1993 p = msg->front.iov_base + req->r_request_release_offset;
1994 }
1995
2f2dc053
SW		 1996 head->num_releases = cpu_to_le16(releases);
1997
b8e69066 1998 /* time stamp */
1f041a89
YZ		 1999 {
2000 struct ceph_timespec ts;
2001 ceph_encode_timespec(&ts, &req->r_stamp);
2002 ceph_encode_copy(&p, &ts, sizeof(ts));
2003 }
b8e69066 2004
2f2dc053
SW		 2005 BUG_ON(p > end);
2006 msg->front.iov_len = p - msg->front.iov_base;
2007 msg->hdr.front_len = cpu_to_le32(msg->front.iov_len);
2008
25e6bae3
YZ		 2009 if (req->r_pagelist) {
2010 struct ceph_pagelist *pagelist = req->r_pagelist;
2011 atomic_inc(&pagelist->refcnt);
2012 ceph_msg_data_add_pagelist(msg, pagelist);
2013 msg->hdr.data_len = cpu_to_le32(pagelist->length);
2014 } else {
2015 msg->hdr.data_len = 0;
ebf18f47 2016 }
02afca6c 2017
2f2dc053
SW		 2018 msg->hdr.data_off = cpu_to_le16(0);
2019
2020out_free2:
2021 if (freepath2)
2022 kfree((char *)path2);
2023out_free1:
2024 if (freepath1)
2025 kfree((char *)path1);
2026out:
2027 return msg;
2028}
2029
2030/*
2031 * called under mdsc->mutex if error, under no mutex if
2032 * success.
2033 */
2034static void complete_request(struct ceph_mds_client *mdsc,
2035 struct ceph_mds_request *req)
2036{
2037 if (req->r_callback)
2038 req->r_callback(mdsc, req);
2039 else
03066f23 2040 complete_all(&req->r_completion);
2f2dc053
SW		 2041}
2042
2043/*
2044 * called under mdsc->mutex
2045 */
2046static int __prepare_send_request(struct ceph_mds_client *mdsc,
2047 struct ceph_mds_request *req,
6e6f0923 2048 int mds, bool drop_cap_releases)
2f2dc053
SW		 2049{
2050 struct ceph_mds_request_head *rhead;
2051 struct ceph_msg *msg;
2052 int flags = 0;
2053
2f2dc053 2054 req->r_attempts++;
e55b71f8
GF		 2055 if (req->r_inode) {
2056 struct ceph_cap *cap =
2057 ceph_get_cap_for_mds(ceph_inode(req->r_inode), mds);
2058
2059 if (cap)
2060 req->r_sent_on_mseq = cap->mseq;
2061 else
2062 req->r_sent_on_mseq = -1;
2063 }
2f2dc053
SW		 2064 dout("prepare_send_request %p tid %lld %s (attempt %d)\n", req,
2065 req->r_tid, ceph_mds_op_name(req->r_op), req->r_attempts);
2066
01a92f17 2067 if (req->r_got_unsafe) {
c5c9a0bf 2068 void *p;
01a92f17
SW		 2069 /*
2070 * Replay. Do not regenerate message (and rebuild
2071 * paths, etc.); just use the original message.
2072 * Rebuilding paths will break for renames because
2073 * d_move mangles the src name.
2074 */
2075 msg = req->r_request;
2076 rhead = msg->front.iov_base;
2077
2078 flags = le32_to_cpu(rhead->flags);
2079 flags |= CEPH_MDS_FLAG_REPLAY;
2080 rhead->flags = cpu_to_le32(flags);
2081
2082 if (req->r_target_inode)
2083 rhead->ino = cpu_to_le64(ceph_ino(req->r_target_inode));
2084
2085 rhead->num_retry = req->r_attempts - 1;
e979cf50
SW		 2086
2087 /* remove cap/dentry releases from message */
2088 rhead->num_releases = 0;
c5c9a0bf
YZ		 2089
2090 /* time stamp */
2091 p = msg->front.iov_base + req->r_request_release_offset;
1f041a89
YZ		 2092 {
2093 struct ceph_timespec ts;
2094 ceph_encode_timespec(&ts, &req->r_stamp);
2095 ceph_encode_copy(&p, &ts, sizeof(ts));
2096 }
c5c9a0bf
YZ		 2097
2098 msg->front.iov_len = p - msg->front.iov_base;
2099 msg->hdr.front_len = cpu_to_le32(msg->front.iov_len);
01a92f17
SW		 2100 return 0;
2101 }
2102
2f2dc053
SW		 2103 if (req->r_request) {
2104 ceph_msg_put(req->r_request);
2105 req->r_request = NULL;
2106 }
6e6f0923 2107 msg = create_request_message(mdsc, req, mds, drop_cap_releases);
2f2dc053 2108 if (IS_ERR(msg)) {
e1518c7c 2109 req->r_err = PTR_ERR(msg);
a79832f2 2110 return PTR_ERR(msg);
2f2dc053
SW		 2111 }
2112 req->r_request = msg;
2113
2114 rhead = msg->front.iov_base;
2f2dc053
SW		 2115 rhead->oldest_client_tid = cpu_to_le64(__get_oldest_tid(mdsc));
2116 if (req->r_got_unsafe)
2117 flags |= CEPH_MDS_FLAG_REPLAY;
2118 if (req->r_locked_dir)
2119 flags |= CEPH_MDS_FLAG_WANT_DENTRY;
2120 rhead->flags = cpu_to_le32(flags);
2121 rhead->num_fwd = req->r_num_fwd;
2122 rhead->num_retry = req->r_attempts - 1;
01a92f17 2123 rhead->ino = 0;
2f2dc053
SW		 2124
2125 dout(" r_locked_dir = %p\n", req->r_locked_dir);
2f2dc053
SW		 2126 return 0;
2127}
2128
2129/*
2130 * send request, or put it on the appropriate wait list.
2131 */
2132static int __do_request(struct ceph_mds_client *mdsc,
2133 struct ceph_mds_request *req)
2134{
2135 struct ceph_mds_session *session = NULL;
2136 int mds = -1;
48fec5d0 2137 int err = 0;
2f2dc053 2138
eb1b8af3
YZ		 2139 if (req->r_err || req->r_got_result) {
2140 if (req->r_aborted)
2141 __unregister_request(mdsc, req);
2f2dc053 2142 goto out;
eb1b8af3 2143 }
2f2dc053
SW		 2144
2145 if (req->r_timeout &&
2146 time_after_eq(jiffies, req->r_started + req->r_timeout)) {
2147 dout("do_request timed out\n");
2148 err = -EIO;
2149 goto finish;
2150 }
48fec5d0
YZ		 2151 if (ACCESS_ONCE(mdsc->fsc->mount_state) == CEPH_MOUNT_SHUTDOWN) {
2152 dout("do_request forced umount\n");
2153 err = -EIO;
2154 goto finish;
2155 }
2f2dc053 2156
dc69e2e9
SW		 2157 put_request_session(req);
2158
2f2dc053
SW		 2159 mds = __choose_mds(mdsc, req);
2160 if (mds < 0 ||
2161 ceph_mdsmap_get_state(mdsc->mdsmap, mds) < CEPH_MDS_STATE_ACTIVE) {
2162 dout("do_request no mds or not active, waiting for map\n");
2163 list_add(&req->r_wait, &mdsc->waiting_for_map);
2164 goto out;
2165 }
2166
2167 /* get, open session */
2168 session = __ceph_lookup_mds_session(mdsc, mds);
9c423956 2169 if (!session) {
2f2dc053 2170 session = register_session(mdsc, mds);
9c423956
SW		 2171 if (IS_ERR(session)) {
2172 err = PTR_ERR(session);
2173 goto finish;
2174 }
2175 }
dc69e2e9
SW		 2176 req->r_session = get_session(session);
2177
2f2dc053 2178 dout("do_request mds%d session %p state %s\n", mds, session,
a687ecaf 2179 ceph_session_state_name(session->s_state));
2f2dc053
SW		 2180 if (session->s_state != CEPH_MDS_SESSION_OPEN &&
2181 session->s_state != CEPH_MDS_SESSION_HUNG) {
2182 if (session->s_state == CEPH_MDS_SESSION_NEW ||
2183 session->s_state == CEPH_MDS_SESSION_CLOSING)
2184 __open_session(mdsc, session);
2185 list_add(&req->r_wait, &session->s_waiting);
2186 goto out_session;
2187 }
2188
2189 /* send request */
2f2dc053
SW		 2190 req->r_resend_mds = -1; /* forget any previous mds hint */
2191
2192 if (req->r_request_started == 0) /* note request start time */
2193 req->r_request_started = jiffies;
2194
6e6f0923 2195 err = __prepare_send_request(mdsc, req, mds, false);
2f2dc053
SW		 2196 if (!err) {
2197 ceph_msg_get(req->r_request);
2198 ceph_con_send(&session->s_con, req->r_request);
2199 }
2200
2201out_session:
2202 ceph_put_mds_session(session);
48fec5d0
YZ		 2203finish:
2204 if (err) {
2205 dout("__do_request early error %d\n", err);
2206 req->r_err = err;
2207 complete_request(mdsc, req);
2208 __unregister_request(mdsc, req);
2209 }
2f2dc053
SW		 2210out:
2211 return err;
2f2dc053
SW		 2212}
2213
2214/*
2215 * called under mdsc->mutex
2216 */
2217static void __wake_requests(struct ceph_mds_client *mdsc,
2218 struct list_head *head)
2219{
ed75ec2c
YZ		 2220 struct ceph_mds_request *req;
2221 LIST_HEAD(tmp_list);
2222
2223 list_splice_init(head, &tmp_list);
2f2dc053 2224
ed75ec2c
YZ		 2225 while (!list_empty(&tmp_list)) {
2226 req = list_entry(tmp_list.next,
2227 struct ceph_mds_request, r_wait);
2f2dc053 2228 list_del_init(&req->r_wait);
7971bd92 2229 dout(" wake request %p tid %llu\n", req, req->r_tid);
2f2dc053
SW		 2230 __do_request(mdsc, req);
2231 }
2232}
2233
2234/*
2235 * Wake up threads with requests pending for @mds, so that they can
29790f26 2236 * resubmit their requests to a possibly different mds.
2f2dc053 2237 */
29790f26 2238static void kick_requests(struct ceph_mds_client *mdsc, int mds)
2f2dc053 2239{
44ca18f2 2240 struct ceph_mds_request *req;
282c1052 2241 struct rb_node *p = rb_first(&mdsc->request_tree);
2f2dc053
SW		 2242
2243 dout("kick_requests mds%d\n", mds);
282c1052 2244 while (p) {
44ca18f2 2245 req = rb_entry(p, struct ceph_mds_request, r_node);
282c1052 2246 p = rb_next(p);
44ca18f2
SW		 2247 if (req->r_got_unsafe)
2248 continue;
3de22be6
YZ		 2249 if (req->r_attempts > 0)
2250 continue; /* only new requests */
44ca18f2
SW		 2251 if (req->r_session &&
2252 req->r_session->s_mds == mds) {
2253 dout(" kicking tid %llu\n", req->r_tid);
03974e81 2254 list_del_init(&req->r_wait);
44ca18f2 2255 __do_request(mdsc, req);
2f2dc053
SW		 2256 }
2257 }
2258}
2259
2260void ceph_mdsc_submit_request(struct ceph_mds_client *mdsc,
2261 struct ceph_mds_request *req)
2262{
2263 dout("submit_request on %p\n", req);
2264 mutex_lock(&mdsc->mutex);
2265 __register_request(mdsc, req, NULL);
2266 __do_request(mdsc, req);
2267 mutex_unlock(&mdsc->mutex);
2268}
2269
2270/*
2271 * Synchrously perform an mds request. Take care of all of the
2272 * session setup, forwarding, retry details.
2273 */
2274int ceph_mdsc_do_request(struct ceph_mds_client *mdsc,
2275 struct inode *dir,
2276 struct ceph_mds_request *req)
2277{
2278 int err;
2279
2280 dout("do_request on %p\n", req);
2281
2282 /* take CAP_PIN refs for r_inode, r_locked_dir, r_old_dentry */
2283 if (req->r_inode)
2284 ceph_get_cap_refs(ceph_inode(req->r_inode), CEPH_CAP_PIN);
2285 if (req->r_locked_dir)
2286 ceph_get_cap_refs(ceph_inode(req->r_locked_dir), CEPH_CAP_PIN);
844d87c3 2287 if (req->r_old_dentry_dir)
41b02e1f
SW		 2288 ceph_get_cap_refs(ceph_inode(req->r_old_dentry_dir),
2289 CEPH_CAP_PIN);
2f2dc053
SW		 2290
2291 /* issue */
2292 mutex_lock(&mdsc->mutex);
2293 __register_request(mdsc, req, dir);
2294 __do_request(mdsc, req);
2295
e1518c7c
SW		 2296 if (req->r_err) {
2297 err = req->r_err;
e1518c7c 2298 goto out;
2f2dc053
SW		 2299 }
2300
e1518c7c
SW		 2301 /* wait */
2302 mutex_unlock(&mdsc->mutex);
2303 dout("do_request waiting\n");
5be73034 2304 if (!req->r_timeout && req->r_wait_for_completion) {
9280be24 2305 err = req->r_wait_for_completion(mdsc, req);
e1518c7c 2306 } else {
5be73034
ID		 2307 long timeleft = wait_for_completion_killable_timeout(
2308 &req->r_completion,
2309 ceph_timeout_jiffies(req->r_timeout));
2310 if (timeleft > 0)
2311 err = 0;
2312 else if (!timeleft)
2313 err = -EIO; /* timed out */
2314 else
2315 err = timeleft; /* killed */
e1518c7c
SW		 2316 }
2317 dout("do_request waited, got %d\n", err);
2318 mutex_lock(&mdsc->mutex);
5b1daecd 2319
e1518c7c
SW		 2320 /* only abort if we didn't race with a real reply */
2321 if (req->r_got_result) {
2322 err = le32_to_cpu(req->r_reply_info.head->result);
2323 } else if (err < 0) {
2324 dout("aborted request %lld with %d\n", req->r_tid, err);
b4556396
SW		 2325
2326 /*
2327 * ensure we aren't running concurrently with
2328 * ceph_fill_trace or ceph_readdir_prepopulate, which
2329 * rely on locks (dir mutex) held by our caller.
2330 */
2331 mutex_lock(&req->r_fill_mutex);
e1518c7c
SW		 2332 req->r_err = err;
2333 req->r_aborted = true;
b4556396 2334 mutex_unlock(&req->r_fill_mutex);
5b1daecd 2335
e1518c7c 2336 if (req->r_locked_dir &&
167c9e35
SW		 2337 (req->r_op & CEPH_MDS_OP_WRITE))
2338 ceph_invalidate_dir_request(req);
2f2dc053 2339 } else {
e1518c7c 2340 err = req->r_err;
2f2dc053 2341 }
2f2dc053 2342
e1518c7c
SW		 2343out:
2344 mutex_unlock(&mdsc->mutex);
2f2dc053
SW		 2345 dout("do_request %p done, result %d\n", req, err);
2346 return err;
2347}
2348
167c9e35 2349/*
2f276c51 2350 * Invalidate dir's completeness, dentry lease state on an aborted MDS
167c9e35
SW		 2351 * namespace request.
2352 */
2353void ceph_invalidate_dir_request(struct ceph_mds_request *req)
2354{
2355 struct inode *inode = req->r_locked_dir;
167c9e35 2356
2f276c51 2357 dout("invalidate_dir_request %p (complete, lease(s))\n", inode);
167c9e35 2358
2f276c51 2359 ceph_dir_clear_complete(inode);
167c9e35
SW		 2360 if (req->r_dentry)
2361 ceph_invalidate_dentry_lease(req->r_dentry);
2362 if (req->r_old_dentry)
2363 ceph_invalidate_dentry_lease(req->r_old_dentry);
2364}
2365
2f2dc053
SW		 2366/*
2367 * Handle mds reply.
2368 *
2369 * We take the session mutex and parse and process the reply immediately.
2370 * This preserves the logical ordering of replies, capabilities, etc., sent
2371 * by the MDS as they are applied to our local cache.
2372 */
2373static void handle_reply(struct ceph_mds_session *session, struct ceph_msg *msg)
2374{
2375 struct ceph_mds_client *mdsc = session->s_mdsc;
2376 struct ceph_mds_request *req;
2377 struct ceph_mds_reply_head *head = msg->front.iov_base;
2378 struct ceph_mds_reply_info_parsed *rinfo; /* parsed reply info */
982d6011 2379 struct ceph_snap_realm *realm;
2f2dc053
SW		 2380 u64 tid;
2381 int err, result;
2600d2dd 2382 int mds = session->s_mds;
2f2dc053 2383
2f2dc053
SW		 2384 if (msg->front.iov_len < sizeof(*head)) {
2385 pr_err("mdsc_handle_reply got corrupt (short) reply\n");
9ec7cab1 2386 ceph_msg_dump(msg);
2f2dc053
SW		 2387 return;
2388 }
2389
2390 /* get request, session */
6df058c0 2391 tid = le64_to_cpu(msg->hdr.tid);
2f2dc053
SW		 2392 mutex_lock(&mdsc->mutex);
2393 req = __lookup_request(mdsc, tid);
2394 if (!req) {
2395 dout("handle_reply on unknown tid %llu\n", tid);
2396 mutex_unlock(&mdsc->mutex);
2397 return;
2398 }
2399 dout("handle_reply %p\n", req);
2f2dc053
SW		 2400
2401 /* correct session? */
d96d6049 2402 if (req->r_session != session) {
2f2dc053
SW		 2403 pr_err("mdsc_handle_reply got %llu on session mds%d"
2404 " not mds%d\n", tid, session->s_mds,
2405 req->r_session ? req->r_session->s_mds : -1);
2406 mutex_unlock(&mdsc->mutex);
2407 goto out;
2408 }
2409
2410 /* dup? */
2411 if ((req->r_got_unsafe && !head->safe) ||
2412 (req->r_got_safe && head->safe)) {
f3ae1b97 2413 pr_warn("got a dup %s reply on %llu from mds%d\n",
2f2dc053
SW		 2414 head->safe ? "safe" : "unsafe", tid, mds);
2415 mutex_unlock(&mdsc->mutex);
2416 goto out;
2417 }
1550d34e 2418 if (req->r_got_safe) {
f3ae1b97 2419 pr_warn("got unsafe after safe on %llu from mds%d\n",
85792d0d
SW		 2420 tid, mds);
2421 mutex_unlock(&mdsc->mutex);
2422 goto out;
2423 }
2f2dc053
SW		 2424
2425 result = le32_to_cpu(head->result);
2426
2427 /*
e55b71f8
GF		 2428 * Handle an ESTALE
2429 * if we're not talking to the authority, send to them
2430 * if the authority has changed while we weren't looking,
2431 * send to new authority
2432 * Otherwise we just have to return an ESTALE
2f2dc053
SW		 2433 */
2434 if (result == -ESTALE) {
e55b71f8 2435 dout("got ESTALE on request %llu", req->r_tid);
51da8e8c 2436 req->r_resend_mds = -1;
ca18bede 2437 if (req->r_direct_mode != USE_AUTH_MDS) {
e55b71f8
GF		 2438 dout("not using auth, setting for that now");
2439 req->r_direct_mode = USE_AUTH_MDS;
2f2dc053
SW		 2440 __do_request(mdsc, req);
2441 mutex_unlock(&mdsc->mutex);
2442 goto out;
e55b71f8 2443 } else {
ca18bede
YZ		 2444 int mds = __choose_mds(mdsc, req);
2445 if (mds >= 0 && mds != req->r_session->s_mds) {
2446 dout("but auth changed, so resending");
e55b71f8
GF		 2447 __do_request(mdsc, req);
2448 mutex_unlock(&mdsc->mutex);
2449 goto out;
2450 }
2f2dc053 2451 }
e55b71f8 2452 dout("have to return ESTALE on request %llu", req->r_tid);
2f2dc053
SW		 2453 }
2454
e55b71f8 2455
2f2dc053
SW		 2456 if (head->safe) {
2457 req->r_got_safe = true;
2458 __unregister_request(mdsc, req);
2f2dc053
SW		 2459
2460 if (req->r_got_unsafe) {
2461 /*
2462 * We already handled the unsafe response, now do the
2463 * cleanup. No need to examine the response; the MDS
2464 * doesn't include any result info in the safe
2465 * response. And even if it did, there is nothing
2466 * useful we could do with a revised return value.
2467 */
2468 dout("got safe reply %llu, mds%d\n", tid, mds);
2469 list_del_init(&req->r_unsafe_item);
2470
2471 /* last unsafe request during umount? */
44ca18f2 2472 if (mdsc->stopping && !__get_oldest_req(mdsc))
03066f23 2473 complete_all(&mdsc->safe_umount_waiters);
2f2dc053
SW		 2474 mutex_unlock(&mdsc->mutex);
2475 goto out;
2476 }
e1518c7c 2477 } else {
2f2dc053
SW		 2478 req->r_got_unsafe = true;
2479 list_add_tail(&req->r_unsafe_item, &req->r_session->s_unsafe);
2480 }
2481
2482 dout("handle_reply tid %lld result %d\n", tid, result);
2483 rinfo = &req->r_reply_info;
14303d20 2484 err = parse_reply_info(msg, rinfo, session->s_con.peer_features);
2f2dc053
SW		 2485 mutex_unlock(&mdsc->mutex);
2486
2487 mutex_lock(&session->s_mutex);
2488 if (err < 0) {
25933abd 2489 pr_err("mdsc_handle_reply got corrupt reply mds%d(tid:%lld)\n", mds, tid);
9ec7cab1 2490 ceph_msg_dump(msg);
2f2dc053
SW		 2491 goto out_err;
2492 }
2493
2494 /* snap trace */
982d6011 2495 realm = NULL;
2f2dc053
SW		 2496 if (rinfo->snapblob_len) {
2497 down_write(&mdsc->snap_rwsem);
2498 ceph_update_snap_trace(mdsc, rinfo->snapblob,
982d6011
YZ		 2499 rinfo->snapblob + rinfo->snapblob_len,
2500 le32_to_cpu(head->op) == CEPH_MDS_OP_RMSNAP,
2501 &realm);
2f2dc053
SW		 2502 downgrade_write(&mdsc->snap_rwsem);
2503 } else {
2504 down_read(&mdsc->snap_rwsem);
2505 }
2506
2507 /* insert trace into our cache */
b4556396 2508 mutex_lock(&req->r_fill_mutex);
3d14c5d2 2509 err = ceph_fill_trace(mdsc->fsc->sb, req, req->r_session);
2f2dc053 2510 if (err == 0) {
6e8575fa 2511 if (result == 0 && (req->r_op == CEPH_MDS_OP_READDIR ||
81c6aea5 2512 req->r_op == CEPH_MDS_OP_LSSNAP))
2f2dc053 2513 ceph_readdir_prepopulate(req, req->r_session);
37151668 2514 ceph_unreserve_caps(mdsc, &req->r_caps_reservation);
2f2dc053 2515 }
b4556396 2516 mutex_unlock(&req->r_fill_mutex);
2f2dc053
SW		 2517
2518 up_read(&mdsc->snap_rwsem);
982d6011
YZ		 2519 if (realm)
2520 ceph_put_snap_realm(mdsc, realm);
2f2dc053 2521out_err:
e1518c7c
SW		 2522 mutex_lock(&mdsc->mutex);
2523 if (!req->r_aborted) {
2524 if (err) {
2525 req->r_err = err;
2526 } else {
5fdb1389 2527 req->r_reply = ceph_msg_get(msg);
e1518c7c
SW		 2528 req->r_got_result = true;
2529 }
2f2dc053 2530 } else {
e1518c7c 2531 dout("reply arrived after request %lld was aborted\n", tid);
2f2dc053 2532 }
e1518c7c 2533 mutex_unlock(&mdsc->mutex);
2f2dc053 2534
2f2dc053
SW		 2535 mutex_unlock(&session->s_mutex);
2536
2537 /* kick calling process */
2538 complete_request(mdsc, req);
2539out:
2540 ceph_mdsc_put_request(req);
2541 return;
2542}
2543
2544
2545
2546/*
2547 * handle mds notification that our request has been forwarded.
2548 */
2600d2dd
SW		 2549static void handle_forward(struct ceph_mds_client *mdsc,
2550 struct ceph_mds_session *session,
2551 struct ceph_msg *msg)
2f2dc053
SW		 2552{
2553 struct ceph_mds_request *req;
a1ea787c 2554 u64 tid = le64_to_cpu(msg->hdr.tid);
2f2dc053
SW		 2555 u32 next_mds;
2556 u32 fwd_seq;
2f2dc053
SW		 2557 int err = -EINVAL;
2558 void *p = msg->front.iov_base;
2559 void *end = p + msg->front.iov_len;
2f2dc053 2560
a1ea787c 2561 ceph_decode_need(&p, end, 2*sizeof(u32), bad);
c89136ea
SW		 2562 next_mds = ceph_decode_32(&p);
2563 fwd_seq = ceph_decode_32(&p);
2f2dc053
SW		 2564
2565 mutex_lock(&mdsc->mutex);
2566 req = __lookup_request(mdsc, tid);
2567 if (!req) {
2a8e5e36 2568 dout("forward tid %llu to mds%d - req dne\n", tid, next_mds);
2f2dc053
SW		 2569 goto out; /* dup reply? */
2570 }
2571
2a8e5e36
SW		 2572 if (req->r_aborted) {
2573 dout("forward tid %llu aborted, unregistering\n", tid);
2574 __unregister_request(mdsc, req);
2575 } else if (fwd_seq <= req->r_num_fwd) {
2576 dout("forward tid %llu to mds%d - old seq %d <= %d\n",
2f2dc053
SW		 2577 tid, next_mds, req->r_num_fwd, fwd_seq);
2578 } else {
2579 /* resend. forward race not possible; mds would drop */
2a8e5e36
SW		 2580 dout("forward tid %llu to mds%d (we resend)\n", tid, next_mds);
2581 BUG_ON(req->r_err);
2582 BUG_ON(req->r_got_result);
3de22be6 2583 req->r_attempts = 0;
2f2dc053
SW		 2584 req->r_num_fwd = fwd_seq;
2585 req->r_resend_mds = next_mds;
2586 put_request_session(req);
2587 __do_request(mdsc, req);
2588 }
2589 ceph_mdsc_put_request(req);
2590out:
2591 mutex_unlock(&mdsc->mutex);
2592 return;
2593
2594bad:
2595 pr_err("mdsc_handle_forward decode error err=%d\n", err);
2596}
2597
2598/*
2599 * handle a mds session control message
2600 */
2601static void handle_session(struct ceph_mds_session *session,
2602 struct ceph_msg *msg)
2603{
2604 struct ceph_mds_client *mdsc = session->s_mdsc;
2605 u32 op;
2606 u64 seq;
2600d2dd 2607 int mds = session->s_mds;
2f2dc053
SW		 2608 struct ceph_mds_session_head *h = msg->front.iov_base;
2609 int wake = 0;
2610
2f2dc053
SW		 2611 /* decode */
2612 if (msg->front.iov_len != sizeof(*h))
2613 goto bad;
2614 op = le32_to_cpu(h->op);
2615 seq = le64_to_cpu(h->seq);
2616
2617 mutex_lock(&mdsc->mutex);
2600d2dd
SW		 2618 if (op == CEPH_SESSION_CLOSE)
2619 __unregister_session(mdsc, session);
2f2dc053
SW		 2620 /* FIXME: this ttl calculation is generous */
2621 session->s_ttl = jiffies + HZ*mdsc->mdsmap->m_session_autoclose;
2622 mutex_unlock(&mdsc->mutex);
2623
2624 mutex_lock(&session->s_mutex);
2625
2626 dout("handle_session mds%d %s %p state %s seq %llu\n",
2627 mds, ceph_session_op_name(op), session,
a687ecaf 2628 ceph_session_state_name(session->s_state), seq);
2f2dc053
SW		 2629
2630 if (session->s_state == CEPH_MDS_SESSION_HUNG) {
2631 session->s_state = CEPH_MDS_SESSION_OPEN;
2632 pr_info("mds%d came back\n", session->s_mds);
2633 }
2634
2635 switch (op) {
2636 case CEPH_SESSION_OPEN:
29790f26
SW		 2637 if (session->s_state == CEPH_MDS_SESSION_RECONNECTING)
2638 pr_info("mds%d reconnect success\n", session->s_mds);
2f2dc053
SW		 2639 session->s_state = CEPH_MDS_SESSION_OPEN;
2640 renewed_caps(mdsc, session, 0);
2641 wake = 1;
2642 if (mdsc->stopping)
2643 __close_session(mdsc, session);
2644 break;
2645
2646 case CEPH_SESSION_RENEWCAPS:
2647 if (session->s_renew_seq == seq)
2648 renewed_caps(mdsc, session, 1);
2649 break;
2650
2651 case CEPH_SESSION_CLOSE:
29790f26
SW		 2652 if (session->s_state == CEPH_MDS_SESSION_RECONNECTING)
2653 pr_info("mds%d reconnect denied\n", session->s_mds);
1c841a96 2654 cleanup_session_requests(mdsc, session);
2f2dc053 2655 remove_session_caps(session);
656e4382 2656 wake = 2; /* for good measure */
f3c60c59 2657 wake_up_all(&mdsc->session_close_wq);
2f2dc053
SW		 2658 break;
2659
2660 case CEPH_SESSION_STALE:
2661 pr_info("mds%d caps went stale, renewing\n",
2662 session->s_mds);
d8fb02ab 2663 spin_lock(&session->s_gen_ttl_lock);
2f2dc053 2664 session->s_cap_gen++;
1ce208a6 2665 session->s_cap_ttl = jiffies - 1;
d8fb02ab 2666 spin_unlock(&session->s_gen_ttl_lock);
2f2dc053
SW		 2667 send_renew_caps(mdsc, session);
2668 break;
2669
2670 case CEPH_SESSION_RECALL_STATE:
2671 trim_caps(mdsc, session, le32_to_cpu(h->max_caps));
2672 break;
2673
186e4f7a
YZ		 2674 case CEPH_SESSION_FLUSHMSG:
2675 send_flushmsg_ack(mdsc, session, seq);
2676 break;
2677
03f4fcb0
YZ		 2678 case CEPH_SESSION_FORCE_RO:
2679 dout("force_session_readonly %p\n", session);
2680 spin_lock(&session->s_cap_lock);
2681 session->s_readonly = true;
2682 spin_unlock(&session->s_cap_lock);
2683 wake_up_session_caps(session, 0);
2684 break;
2685
2f2dc053
SW		 2686 default:
2687 pr_err("mdsc_handle_session bad op %d mds%d\n", op, mds);
2688 WARN_ON(1);
2689 }
2690
2691 mutex_unlock(&session->s_mutex);
2692 if (wake) {
2693 mutex_lock(&mdsc->mutex);
2694 __wake_requests(mdsc, &session->s_waiting);
656e4382
YZ		 2695 if (wake == 2)
2696 kick_requests(mdsc, mds);
2f2dc053
SW		 2697 mutex_unlock(&mdsc->mutex);
2698 }
2699 return;
2700
2701bad:
2702 pr_err("mdsc_handle_session corrupt message mds%d len %d\n", mds,
2703 (int)msg->front.iov_len);
9ec7cab1 2704 ceph_msg_dump(msg);
2f2dc053
SW		 2705 return;
2706}
2707
2708
2709/*
2710 * called under session->mutex.
2711 */
2712static void replay_unsafe_requests(struct ceph_mds_client *mdsc,
2713 struct ceph_mds_session *session)
2714{
2715 struct ceph_mds_request *req, *nreq;
3de22be6 2716 struct rb_node *p;
2f2dc053
SW		 2717 int err;
2718
2719 dout("replay_unsafe_requests mds%d\n", session->s_mds);
2720
2721 mutex_lock(&mdsc->mutex);
2722 list_for_each_entry_safe(req, nreq, &session->s_unsafe, r_unsafe_item) {
6e6f0923 2723 err = __prepare_send_request(mdsc, req, session->s_mds, true);
2f2dc053
SW		 2724 if (!err) {
2725 ceph_msg_get(req->r_request);
2726 ceph_con_send(&session->s_con, req->r_request);
2727 }
2728 }
3de22be6
YZ		 2729
2730 /*
2731 * also re-send old requests when MDS enters reconnect stage. So that MDS
2732 * can process completed request in clientreplay stage.
2733 */
2734 p = rb_first(&mdsc->request_tree);
2735 while (p) {
2736 req = rb_entry(p, struct ceph_mds_request, r_node);
2737 p = rb_next(p);
2738 if (req->r_got_unsafe)
2739 continue;
2740 if (req->r_attempts == 0)
2741 continue; /* only old requests */
2742 if (req->r_session &&
2743 req->r_session->s_mds == session->s_mds) {
6e6f0923
YZ		 2744 err = __prepare_send_request(mdsc, req,
2745 session->s_mds, true);
3de22be6
YZ		 2746 if (!err) {
2747 ceph_msg_get(req->r_request);
2748 ceph_con_send(&session->s_con, req->r_request);
2749 }
2750 }
2751 }
2f2dc053
SW		 2752 mutex_unlock(&mdsc->mutex);
2753}
2754
2755/*
2756 * Encode information about a cap for a reconnect with the MDS.
2757 */
2f2dc053
SW		 2758static int encode_caps_cb(struct inode *inode, struct ceph_cap *cap,
2759 void *arg)
2760{
20cb34ae
SW		 2761 union {
2762 struct ceph_mds_cap_reconnect v2;
2763 struct ceph_mds_cap_reconnect_v1 v1;
2764 } rec;
2765 size_t reclen;
2f2dc053 2766 struct ceph_inode_info *ci;
20cb34ae
SW		 2767 struct ceph_reconnect_state *recon_state = arg;
2768 struct ceph_pagelist *pagelist = recon_state->pagelist;
2f2dc053
SW		 2769 char *path;
2770 int pathlen, err;
2771 u64 pathbase;
2772 struct dentry *dentry;
2773
2774 ci = cap->ci;
2775
2776 dout(" adding %p ino %llx.%llx cap %p %lld %s\n",
2777 inode, ceph_vinop(inode), cap, cap->cap_id,
2778 ceph_cap_string(cap->issued));
93cea5be
SW		 2779 err = ceph_pagelist_encode_64(pagelist, ceph_ino(inode));
2780 if (err)
2781 return err;
2f2dc053
SW		 2782
2783 dentry = d_find_alias(inode);
2784 if (dentry) {
2785 path = ceph_mdsc_build_path(dentry, &pathlen, &pathbase, 0);
2786 if (IS_ERR(path)) {
2787 err = PTR_ERR(path);
e072f8aa 2788 goto out_dput;
2f2dc053
SW		 2789 }
2790 } else {
2791 path = NULL;
2792 pathlen = 0;
2793 }
93cea5be
SW		 2794 err = ceph_pagelist_encode_string(pagelist, path, pathlen);
2795 if (err)
e072f8aa 2796 goto out_free;
2f2dc053 2797
be655596 2798 spin_lock(&ci->i_ceph_lock);
2f2dc053
SW		 2799 cap->seq = 0; /* reset cap seq */
2800 cap->issue_seq = 0; /* and issue_seq */
667ca05c 2801 cap->mseq = 0; /* and migrate_seq */
99a9c273 2802 cap->cap_gen = cap->session->s_cap_gen;
20cb34ae
SW		 2803
2804 if (recon_state->flock) {
2805 rec.v2.cap_id = cpu_to_le64(cap->cap_id);
2806 rec.v2.wanted = cpu_to_le32(__ceph_caps_wanted(ci));
2807 rec.v2.issued = cpu_to_le32(cap->issued);
2808 rec.v2.snaprealm = cpu_to_le64(ci->i_snap_realm->ino);
2809 rec.v2.pathbase = cpu_to_le64(pathbase);
2810 rec.v2.flock_len = 0;
2811 reclen = sizeof(rec.v2);
2812 } else {
2813 rec.v1.cap_id = cpu_to_le64(cap->cap_id);
2814 rec.v1.wanted = cpu_to_le32(__ceph_caps_wanted(ci));
2815 rec.v1.issued = cpu_to_le32(cap->issued);
2816 rec.v1.size = cpu_to_le64(inode->i_size);
2817 ceph_encode_timespec(&rec.v1.mtime, &inode->i_mtime);
2818 ceph_encode_timespec(&rec.v1.atime, &inode->i_atime);
2819 rec.v1.snaprealm = cpu_to_le64(ci->i_snap_realm->ino);
2820 rec.v1.pathbase = cpu_to_le64(pathbase);
2821 reclen = sizeof(rec.v1);
2822 }
be655596 2823 spin_unlock(&ci->i_ceph_lock);
2f2dc053 2824
40819f6f
GF		 2825 if (recon_state->flock) {
2826 int num_fcntl_locks, num_flock_locks;
39be95e9
JS		 2827 struct ceph_filelock *flocks;
2828
2829encode_again:
39be95e9 2830 ceph_count_locks(inode, &num_fcntl_locks, &num_flock_locks);
39be95e9
JS		 2831 flocks = kmalloc((num_fcntl_locks+num_flock_locks) *
2832 sizeof(struct ceph_filelock), GFP_NOFS);
2833 if (!flocks) {
2834 err = -ENOMEM;
2835 goto out_free;
2836 }
39be95e9
JS		 2837 err = ceph_encode_locks_to_buffer(inode, flocks,
2838 num_fcntl_locks,
2839 num_flock_locks);
39be95e9
JS		 2840 if (err) {
2841 kfree(flocks);
2842 if (err == -ENOSPC)
2843 goto encode_again;
2844 goto out_free;
2845 }
2846 /*
2847 * number of encoded locks is stable, so copy to pagelist
2848 */
2849 rec.v2.flock_len = cpu_to_le32(2*sizeof(u32) +
2850 (num_fcntl_locks+num_flock_locks) *
2851 sizeof(struct ceph_filelock));
2852 err = ceph_pagelist_append(pagelist, &rec, reclen);
2853 if (!err)
2854 err = ceph_locks_to_pagelist(flocks, pagelist,
2855 num_fcntl_locks,
2856 num_flock_locks);
2857 kfree(flocks);
3612abbd
SW		 2858 } else {
2859 err = ceph_pagelist_append(pagelist, &rec, reclen);
40819f6f 2860 }
44c99757
YZ		 2861
2862 recon_state->nr_caps++;
e072f8aa 2863out_free:
2f2dc053 2864 kfree(path);
e072f8aa 2865out_dput:
2f2dc053 2866 dput(dentry);
93cea5be 2867 return err;
2f2dc053
SW		 2868}
2869
2870
2871/*
2872 * If an MDS fails and recovers, clients need to reconnect in order to
2873 * reestablish shared state. This includes all caps issued through
2874 * this session _and_ the snap_realm hierarchy. Because it's not
2875 * clear which snap realms the mds cares about, we send everything we
2876 * know about.. that ensures we'll then get any new info the
2877 * recovering MDS might have.
2878 *
2879 * This is a relatively heavyweight operation, but it's rare.
2880 *
2881 * called with mdsc->mutex held.
2882 */
34b6c855
SW		 2883static void send_mds_reconnect(struct ceph_mds_client *mdsc,
2884 struct ceph_mds_session *session)
2f2dc053 2885{
2f2dc053 2886 struct ceph_msg *reply;
a105f00c 2887 struct rb_node *p;
34b6c855 2888 int mds = session->s_mds;
9abf82b8 2889 int err = -ENOMEM;
44c99757 2890 int s_nr_caps;
93cea5be 2891 struct ceph_pagelist *pagelist;
20cb34ae 2892 struct ceph_reconnect_state recon_state;
2f2dc053 2893
34b6c855 2894 pr_info("mds%d reconnect start\n", mds);
2f2dc053 2895
93cea5be
SW		 2896 pagelist = kmalloc(sizeof(*pagelist), GFP_NOFS);
2897 if (!pagelist)
2898 goto fail_nopagelist;
2899 ceph_pagelist_init(pagelist);
2900
b61c2763 2901 reply = ceph_msg_new(CEPH_MSG_CLIENT_RECONNECT, 0, GFP_NOFS, false);
a79832f2 2902 if (!reply)
93cea5be 2903 goto fail_nomsg;
93cea5be 2904
34b6c855
SW		 2905 mutex_lock(&session->s_mutex);
2906 session->s_state = CEPH_MDS_SESSION_RECONNECTING;
2907 session->s_seq = 0;
2f2dc053 2908
2f2dc053 2909 dout("session %p state %s\n", session,
a687ecaf 2910 ceph_session_state_name(session->s_state));
2f2dc053 2911
99a9c273
YZ		 2912 spin_lock(&session->s_gen_ttl_lock);
2913 session->s_cap_gen++;
2914 spin_unlock(&session->s_gen_ttl_lock);
2915
2916 spin_lock(&session->s_cap_lock);
03f4fcb0
YZ		 2917 /* don't know if session is readonly */
2918 session->s_readonly = 0;
99a9c273
YZ		 2919 /*
2920 * notify __ceph_remove_cap() that we are composing cap reconnect.
2921 * If a cap get released before being added to the cap reconnect,
2922 * __ceph_remove_cap() should skip queuing cap release.
2923 */
2924 session->s_cap_reconnect = 1;
e01a5946 2925 /* drop old cap expires; we're about to reestablish that state */
745a8e3b 2926 cleanup_cap_releases(mdsc, session);
e01a5946 2927
5d23371f 2928 /* trim unused caps to reduce MDS's cache rejoin time */
c0bd50e2
YZ		 2929 if (mdsc->fsc->sb->s_root)
2930 shrink_dcache_parent(mdsc->fsc->sb->s_root);
5d23371f
YZ		 2931
2932 ceph_con_close(&session->s_con);
2933 ceph_con_open(&session->s_con,
2934 CEPH_ENTITY_TYPE_MDS, mds,
2935 ceph_mdsmap_get_addr(mdsc->mdsmap, mds));
2936
2937 /* replay unsafe requests */
2938 replay_unsafe_requests(mdsc, session);
2939
2940 down_read(&mdsc->snap_rwsem);
2941
2f2dc053 2942 /* traverse this session's caps */
44c99757
YZ		 2943 s_nr_caps = session->s_nr_caps;
2944 err = ceph_pagelist_encode_32(pagelist, s_nr_caps);
93cea5be
SW		 2945 if (err)
2946 goto fail;
20cb34ae 2947
44c99757 2948 recon_state.nr_caps = 0;
20cb34ae
SW		 2949 recon_state.pagelist = pagelist;
2950 recon_state.flock = session->s_con.peer_features & CEPH_FEATURE_FLOCK;
2951 err = iterate_session_caps(session, encode_caps_cb, &recon_state);
2f2dc053 2952 if (err < 0)
9abf82b8 2953 goto fail;
2f2dc053 2954
99a9c273
YZ		 2955 spin_lock(&session->s_cap_lock);
2956 session->s_cap_reconnect = 0;
2957 spin_unlock(&session->s_cap_lock);
2958
2f2dc053
SW		 2959 /*
2960 * snaprealms. we provide mds with the ino, seq (version), and
2961 * parent for all of our realms. If the mds has any newer info,
2962 * it will tell us.
2963 */
a105f00c
SW		 2964 for (p = rb_first(&mdsc->snap_realms); p; p = rb_next(p)) {
2965 struct ceph_snap_realm *realm =
2966 rb_entry(p, struct ceph_snap_realm, node);
93cea5be 2967 struct ceph_mds_snaprealm_reconnect sr_rec;
2f2dc053
SW		 2968
2969 dout(" adding snap realm %llx seq %lld parent %llx\n",
2970 realm->ino, realm->seq, realm->parent_ino);
93cea5be
SW		 2971 sr_rec.ino = cpu_to_le64(realm->ino);
2972 sr_rec.seq = cpu_to_le64(realm->seq);
2973 sr_rec.parent = cpu_to_le64(realm->parent_ino);
2974 err = ceph_pagelist_append(pagelist, &sr_rec, sizeof(sr_rec));
2975 if (err)
2976 goto fail;
2f2dc053 2977 }
2f2dc053 2978
20cb34ae
SW		 2979 if (recon_state.flock)
2980 reply->hdr.version = cpu_to_le16(2);
44c99757
YZ		 2981
2982 /* raced with cap release? */
2983 if (s_nr_caps != recon_state.nr_caps) {
2984 struct page *page = list_first_entry(&pagelist->head,
2985 struct page, lru);
2986 __le32 *addr = kmap_atomic(page);
2987 *addr = cpu_to_le32(recon_state.nr_caps);
2988 kunmap_atomic(addr);
ebf18f47 2989 }
44c99757
YZ		 2990
2991 reply->hdr.data_len = cpu_to_le32(pagelist->length);
2992 ceph_msg_data_add_pagelist(reply, pagelist);
e548e9b9
YZ		 2993
2994 ceph_early_kick_flushing_caps(mdsc, session);
2995
2f2dc053
SW		 2996 ceph_con_send(&session->s_con, reply);
2997
9abf82b8
SW		 2998 mutex_unlock(&session->s_mutex);
2999
3000 mutex_lock(&mdsc->mutex);
3001 __wake_requests(mdsc, &session->s_waiting);
3002 mutex_unlock(&mdsc->mutex);
3003
2f2dc053 3004 up_read(&mdsc->snap_rwsem);
2f2dc053
SW		 3005 return;
3006
93cea5be 3007fail:
2f2dc053 3008 ceph_msg_put(reply);
9abf82b8
SW		 3009 up_read(&mdsc->snap_rwsem);
3010 mutex_unlock(&session->s_mutex);
93cea5be
SW		 3011fail_nomsg:
3012 ceph_pagelist_release(pagelist);
93cea5be 3013fail_nopagelist:
9abf82b8 3014 pr_err("error %d preparing reconnect for mds%d\n", err, mds);
9abf82b8 3015 return;
2f2dc053
SW		 3016}
3017
3018
3019/*
3020 * compare old and new mdsmaps, kicking requests
3021 * and closing out old connections as necessary
3022 *
3023 * called under mdsc->mutex.
3024 */
3025static void check_new_map(struct ceph_mds_client *mdsc,
3026 struct ceph_mdsmap *newmap,
3027 struct ceph_mdsmap *oldmap)
3028{
3029 int i;
3030 int oldstate, newstate;
3031 struct ceph_mds_session *s;
3032
3033 dout("check_new_map new %u old %u\n",
3034 newmap->m_epoch, oldmap->m_epoch);
3035
3036 for (i = 0; i < oldmap->m_max_mds && i < mdsc->max_sessions; i++) {
3037 if (mdsc->sessions[i] == NULL)
3038 continue;
3039 s = mdsc->sessions[i];
3040 oldstate = ceph_mdsmap_get_state(oldmap, i);
3041 newstate = ceph_mdsmap_get_state(newmap, i);
3042
0deb01c9 3043 dout("check_new_map mds%d state %s%s -> %s%s (session %s)\n",
2f2dc053 3044 i, ceph_mds_state_name(oldstate),
0deb01c9 3045 ceph_mdsmap_is_laggy(oldmap, i) ? " (laggy)" : "",
2f2dc053 3046 ceph_mds_state_name(newstate),
0deb01c9 3047 ceph_mdsmap_is_laggy(newmap, i) ? " (laggy)" : "",
a687ecaf 3048 ceph_session_state_name(s->s_state));
2f2dc053 3049
3e8f43a0
YZ		 3050 if (i >= newmap->m_max_mds ||
3051 memcmp(ceph_mdsmap_get_addr(oldmap, i),
2f2dc053
SW		 3052 ceph_mdsmap_get_addr(newmap, i),
3053 sizeof(struct ceph_entity_addr))) {
3054 if (s->s_state == CEPH_MDS_SESSION_OPENING) {
3055 /* the session never opened, just close it
3056 * out now */
3057 __wake_requests(mdsc, &s->s_waiting);
2600d2dd 3058 __unregister_session(mdsc, s);
2f2dc053
SW		 3059 } else {
3060 /* just close it */
3061 mutex_unlock(&mdsc->mutex);
3062 mutex_lock(&s->s_mutex);
3063 mutex_lock(&mdsc->mutex);
3064 ceph_con_close(&s->s_con);
3065 mutex_unlock(&s->s_mutex);
3066 s->s_state = CEPH_MDS_SESSION_RESTARTING;
3067 }
2f2dc053
SW		 3068 } else if (oldstate == newstate) {
3069 continue; /* nothing new with this mds */
3070 }
3071
3072 /*
3073 * send reconnect?
3074 */
3075 if (s->s_state == CEPH_MDS_SESSION_RESTARTING &&
34b6c855
SW		 3076 newstate >= CEPH_MDS_STATE_RECONNECT) {
3077 mutex_unlock(&mdsc->mutex);
3078 send_mds_reconnect(mdsc, s);
3079 mutex_lock(&mdsc->mutex);
3080 }
2f2dc053
SW		 3081
3082 /*
29790f26 3083 * kick request on any mds that has gone active.
2f2dc053
SW		 3084 */
3085 if (oldstate < CEPH_MDS_STATE_ACTIVE &&
3086 newstate >= CEPH_MDS_STATE_ACTIVE) {
29790f26
SW		 3087 if (oldstate != CEPH_MDS_STATE_CREATING &&
3088 oldstate != CEPH_MDS_STATE_STARTING)
3089 pr_info("mds%d recovery completed\n", s->s_mds);
3090 kick_requests(mdsc, i);
2f2dc053 3091 ceph_kick_flushing_caps(mdsc, s);
0dc2570f 3092 wake_up_session_caps(s, 1);
2f2dc053
SW		 3093 }
3094 }
cb170a22
SW		 3095
3096 for (i = 0; i < newmap->m_max_mds && i < mdsc->max_sessions; i++) {
3097 s = mdsc->sessions[i];
3098 if (!s)
3099 continue;
3100 if (!ceph_mdsmap_is_laggy(newmap, i))
3101 continue;
3102 if (s->s_state == CEPH_MDS_SESSION_OPEN ||
3103 s->s_state == CEPH_MDS_SESSION_HUNG ||
3104 s->s_state == CEPH_MDS_SESSION_CLOSING) {
3105 dout(" connecting to export targets of laggy mds%d\n",
3106 i);
3107 __open_export_target_sessions(mdsc, s);
3108 }
3109 }
2f2dc053
SW		 3110}
3111
3112
3113
3114/*
3115 * leases
3116 */
3117
3118/*
3119 * caller must hold session s_mutex, dentry->d_lock
3120 */
3121void __ceph_mdsc_drop_dentry_lease(struct dentry *dentry)
3122{
3123 struct ceph_dentry_info *di = ceph_dentry(dentry);
3124
3125 ceph_put_mds_session(di->lease_session);
3126 di->lease_session = NULL;
3127}
3128
2600d2dd
SW		 3129static void handle_lease(struct ceph_mds_client *mdsc,
3130 struct ceph_mds_session *session,
3131 struct ceph_msg *msg)
2f2dc053 3132{
3d14c5d2 3133 struct super_block *sb = mdsc->fsc->sb;
2f2dc053 3134 struct inode *inode;
2f2dc053
SW		 3135 struct dentry *parent, *dentry;
3136 struct ceph_dentry_info *di;
2600d2dd 3137 int mds = session->s_mds;
2f2dc053 3138 struct ceph_mds_lease *h = msg->front.iov_base;
1e5ea23d 3139 u32 seq;
2f2dc053 3140 struct ceph_vino vino;
2f2dc053
SW		 3141 struct qstr dname;
3142 int release = 0;
3143
2f2dc053
SW		 3144 dout("handle_lease from mds%d\n", mds);
3145
3146 /* decode */
3147 if (msg->front.iov_len < sizeof(*h) + sizeof(u32))
3148 goto bad;
3149 vino.ino = le64_to_cpu(h->ino);
3150 vino.snap = CEPH_NOSNAP;
1e5ea23d 3151 seq = le32_to_cpu(h->seq);
2f2dc053
SW		 3152 dname.name = (void *)h + sizeof(*h) + sizeof(u32);
3153 dname.len = msg->front.iov_len - sizeof(*h) - sizeof(u32);
3154 if (dname.len != get_unaligned_le32(h+1))
3155 goto bad;
3156
2f2dc053
SW		 3157 /* lookup inode */
3158 inode = ceph_find_inode(sb, vino);
2f90b852
SW		 3159 dout("handle_lease %s, ino %llx %p %.*s\n",
3160 ceph_lease_op_name(h->action), vino.ino, inode,
1e5ea23d 3161 dname.len, dname.name);
6cd3bcad
YZ		 3162
3163 mutex_lock(&session->s_mutex);
3164 session->s_seq++;
3165
2f2dc053
SW		 3166 if (inode == NULL) {
3167 dout("handle_lease no inode %llx\n", vino.ino);
3168 goto release;
3169 }
2f2dc053
SW		 3170
3171 /* dentry */
3172 parent = d_find_alias(inode);
3173 if (!parent) {
3174 dout("no parent dentry on inode %p\n", inode);
3175 WARN_ON(1);
3176 goto release; /* hrm... */
3177 }
3178 dname.hash = full_name_hash(dname.name, dname.len);
3179 dentry = d_lookup(parent, &dname);
3180 dput(parent);
3181 if (!dentry)
3182 goto release;
3183
3184 spin_lock(&dentry->d_lock);
3185 di = ceph_dentry(dentry);
3186 switch (h->action) {
3187 case CEPH_MDS_LEASE_REVOKE:
3d8eb7a9 3188 if (di->lease_session == session) {
1e5ea23d
SW		 3189 if (ceph_seq_cmp(di->lease_seq, seq) > 0)
3190 h->seq = cpu_to_le32(di->lease_seq);
2f2dc053
SW		 3191 __ceph_mdsc_drop_dentry_lease(dentry);
3192 }
3193 release = 1;
3194 break;
3195
3196 case CEPH_MDS_LEASE_RENEW:
3d8eb7a9 3197 if (di->lease_session == session &&
2f2dc053
SW		 3198 di->lease_gen == session->s_cap_gen &&
3199 di->lease_renew_from &&
3200 di->lease_renew_after == 0) {
3201 unsigned long duration =
3563dbdd 3202 msecs_to_jiffies(le32_to_cpu(h->duration_ms));
2f2dc053 3203
1e5ea23d 3204 di->lease_seq = seq;
2f2dc053
SW		 3205 dentry->d_time = di->lease_renew_from + duration;
3206 di->lease_renew_after = di->lease_renew_from +
3207 (duration >> 1);
3208 di->lease_renew_from = 0;
3209 }
3210 break;
3211 }
3212 spin_unlock(&dentry->d_lock);
3213 dput(dentry);
3214
3215 if (!release)
3216 goto out;
3217
3218release:
3219 /* let's just reuse the same message */
3220 h->action = CEPH_MDS_LEASE_REVOKE_ACK;
3221 ceph_msg_get(msg);
3222 ceph_con_send(&session->s_con, msg);
3223
3224out:
3225 iput(inode);
3226 mutex_unlock(&session->s_mutex);
2f2dc053
SW		 3227 return;
3228
3229bad:
3230 pr_err("corrupt lease message\n");
9ec7cab1 3231 ceph_msg_dump(msg);
2f2dc053
SW		 3232}
3233
3234void ceph_mdsc_lease_send_msg(struct ceph_mds_session *session,
3235 struct inode *inode,
3236 struct dentry *dentry, char action,
3237 u32 seq)
3238{
3239 struct ceph_msg *msg;
3240 struct ceph_mds_lease *lease;
3241 int len = sizeof(*lease) + sizeof(u32);
3242 int dnamelen = 0;
3243
3244 dout("lease_send_msg inode %p dentry %p %s to mds%d\n",
3245 inode, dentry, ceph_lease_op_name(action), session->s_mds);
3246 dnamelen = dentry->d_name.len;
3247 len += dnamelen;
3248
b61c2763 3249 msg = ceph_msg_new(CEPH_MSG_CLIENT_LEASE, len, GFP_NOFS, false);
a79832f2 3250 if (!msg)
2f2dc053
SW		 3251 return;
3252 lease = msg->front.iov_base;
3253 lease->action = action;
2f2dc053
SW		 3254 lease->ino = cpu_to_le64(ceph_vino(inode).ino);
3255 lease->first = lease->last = cpu_to_le64(ceph_vino(inode).snap);
3256 lease->seq = cpu_to_le32(seq);
3257 put_unaligned_le32(dnamelen, lease + 1);
3258 memcpy((void *)(lease + 1) + 4, dentry->d_name.name, dnamelen);
3259
3260 /*
3261 * if this is a preemptive lease RELEASE, no need to
3262 * flush request stream, since the actual request will
3263 * soon follow.
3264 */
3265 msg->more_to_follow = (action == CEPH_MDS_LEASE_RELEASE);
3266
3267 ceph_con_send(&session->s_con, msg);
3268}
3269
3270/*
3271 * Preemptively release a lease we expect to invalidate anyway.
3272 * Pass @inode always, @dentry is optional.
3273 */
3274void ceph_mdsc_lease_release(struct ceph_mds_client *mdsc, struct inode *inode,
2f90b852 3275 struct dentry *dentry)
2f2dc053
SW		 3276{
3277 struct ceph_dentry_info *di;
3278 struct ceph_mds_session *session;
3279 u32 seq;
3280
3281 BUG_ON(inode == NULL);
3282 BUG_ON(dentry == NULL);
2f2dc053
SW		 3283
3284 /* is dentry lease valid? */
3285 spin_lock(&dentry->d_lock);
3286 di = ceph_dentry(dentry);
3287 if (!di || !di->lease_session ||
3288 di->lease_session->s_mds < 0 ||
3289 di->lease_gen != di->lease_session->s_cap_gen ||
3290 !time_before(jiffies, dentry->d_time)) {
3291 dout("lease_release inode %p dentry %p -- "
2f90b852
SW		 3292 "no lease\n",
3293 inode, dentry);
2f2dc053
SW		 3294 spin_unlock(&dentry->d_lock);
3295 return;
3296 }
3297
3298 /* we do have a lease on this dentry; note mds and seq */
3299 session = ceph_get_mds_session(di->lease_session);
3300 seq = di->lease_seq;
3301 __ceph_mdsc_drop_dentry_lease(dentry);
3302 spin_unlock(&dentry->d_lock);
3303
2f90b852
SW		 3304 dout("lease_release inode %p dentry %p to mds%d\n",
3305 inode, dentry, session->s_mds);
2f2dc053
SW		 3306 ceph_mdsc_lease_send_msg(session, inode, dentry,
3307 CEPH_MDS_LEASE_RELEASE, seq);
3308 ceph_put_mds_session(session);
3309}
3310
3311/*
3312 * drop all leases (and dentry refs) in preparation for umount
3313 */
3314static void drop_leases(struct ceph_mds_client *mdsc)
3315{
3316 int i;
3317
3318 dout("drop_leases\n");
3319 mutex_lock(&mdsc->mutex);
3320 for (i = 0; i < mdsc->max_sessions; i++) {
3321 struct ceph_mds_session *s = __ceph_lookup_mds_session(mdsc, i);
3322 if (!s)
3323 continue;
3324 mutex_unlock(&mdsc->mutex);
3325 mutex_lock(&s->s_mutex);
3326 mutex_unlock(&s->s_mutex);
3327 ceph_put_mds_session(s);
3328 mutex_lock(&mdsc->mutex);
3329 }
3330 mutex_unlock(&mdsc->mutex);
3331}
3332
3333
3334
3335/*
3336 * delayed work -- periodically trim expired leases, renew caps with mds
3337 */
3338static void schedule_delayed(struct ceph_mds_client *mdsc)
3339{
3340 int delay = 5;
3341 unsigned hz = round_jiffies_relative(HZ * delay);
3342 schedule_delayed_work(&mdsc->delayed_work, hz);
3343}
3344
3345static void delayed_work(struct work_struct *work)
3346{
3347 int i;
3348 struct ceph_mds_client *mdsc =
3349 container_of(work, struct ceph_mds_client, delayed_work.work);
3350 int renew_interval;
3351 int renew_caps;
3352
3353 dout("mdsc delayed_work\n");
afcdaea3 3354 ceph_check_delayed_caps(mdsc);
2f2dc053
SW		 3355
3356 mutex_lock(&mdsc->mutex);
3357 renew_interval = mdsc->mdsmap->m_session_timeout >> 2;
3358 renew_caps = time_after_eq(jiffies, HZ*renew_interval +
3359 mdsc->last_renew_caps);
3360 if (renew_caps)
3361 mdsc->last_renew_caps = jiffies;
3362
3363 for (i = 0; i < mdsc->max_sessions; i++) {
3364 struct ceph_mds_session *s = __ceph_lookup_mds_session(mdsc, i);
3365 if (s == NULL)
3366 continue;
3367 if (s->s_state == CEPH_MDS_SESSION_CLOSING) {
3368 dout("resending session close request for mds%d\n",
3369 s->s_mds);
3370 request_close_session(mdsc, s);
3371 ceph_put_mds_session(s);
3372 continue;
3373 }
3374 if (s->s_ttl && time_after(jiffies, s->s_ttl)) {
3375 if (s->s_state == CEPH_MDS_SESSION_OPEN) {
3376 s->s_state = CEPH_MDS_SESSION_HUNG;
3377 pr_info("mds%d hung\n", s->s_mds);
3378 }
3379 }
3380 if (s->s_state < CEPH_MDS_SESSION_OPEN) {
3381 /* this mds is failed or recovering, just wait */
3382 ceph_put_mds_session(s);
3383 continue;
3384 }
3385 mutex_unlock(&mdsc->mutex);
3386
3387 mutex_lock(&s->s_mutex);
3388 if (renew_caps)
3389 send_renew_caps(mdsc, s);
3390 else
3391 ceph_con_keepalive(&s->s_con);
aab53dd9
SW		 3392 if (s->s_state == CEPH_MDS_SESSION_OPEN ||
3393 s->s_state == CEPH_MDS_SESSION_HUNG)
3d7ded4d 3394 ceph_send_cap_releases(mdsc, s);
2f2dc053
SW		 3395 mutex_unlock(&s->s_mutex);
3396 ceph_put_mds_session(s);
3397
3398 mutex_lock(&mdsc->mutex);
3399 }
3400 mutex_unlock(&mdsc->mutex);
3401
3402 schedule_delayed(mdsc);
3403}
3404
3d14c5d2 3405int ceph_mdsc_init(struct ceph_fs_client *fsc)
2f2dc053 3406
2f2dc053 3407{
3d14c5d2
YS		 3408 struct ceph_mds_client *mdsc;
3409
3410 mdsc = kzalloc(sizeof(struct ceph_mds_client), GFP_NOFS);
3411 if (!mdsc)
3412 return -ENOMEM;
3413 mdsc->fsc = fsc;
3414 fsc->mdsc = mdsc;
2f2dc053
SW		 3415 mutex_init(&mdsc->mutex);
3416 mdsc->mdsmap = kzalloc(sizeof(*mdsc->mdsmap), GFP_NOFS);
fb3101b6 3417 if (mdsc->mdsmap == NULL) {
3418 kfree(mdsc);
2d06eeb8 3419 return -ENOMEM;
fb3101b6 3420 }
2d06eeb8 3421
2f2dc053 3422 init_completion(&mdsc->safe_umount_waiters);
f3c60c59 3423 init_waitqueue_head(&mdsc->session_close_wq);
2f2dc053
SW		 3424 INIT_LIST_HEAD(&mdsc->waiting_for_map);
3425 mdsc->sessions = NULL;
86d8f67b 3426 atomic_set(&mdsc->num_sessions, 0);
2f2dc053
SW		 3427 mdsc->max_sessions = 0;
3428 mdsc->stopping = 0;
affbc19a 3429 mdsc->last_snap_seq = 0;
2f2dc053 3430 init_rwsem(&mdsc->snap_rwsem);
a105f00c 3431 mdsc->snap_realms = RB_ROOT;
2f2dc053
SW		 3432 INIT_LIST_HEAD(&mdsc->snap_empty);
3433 spin_lock_init(&mdsc->snap_empty_lock);
3434 mdsc->last_tid = 0;
e8a7b8b1 3435 mdsc->oldest_tid = 0;
44ca18f2 3436 mdsc->request_tree = RB_ROOT;
2f2dc053
SW		 3437 INIT_DELAYED_WORK(&mdsc->delayed_work, delayed_work);
3438 mdsc->last_renew_caps = jiffies;
3439 INIT_LIST_HEAD(&mdsc->cap_delay_list);
3440 spin_lock_init(&mdsc->cap_delay_lock);
3441 INIT_LIST_HEAD(&mdsc->snap_flush_list);
3442 spin_lock_init(&mdsc->snap_flush_lock);
553adfd9 3443 mdsc->last_cap_flush_tid = 1;
8310b089 3444 mdsc->cap_flush_tree = RB_ROOT;
2f2dc053 3445 INIT_LIST_HEAD(&mdsc->cap_dirty);
db354052 3446 INIT_LIST_HEAD(&mdsc->cap_dirty_migrating);
2f2dc053
SW		 3447 mdsc->num_cap_flushing = 0;
3448 spin_lock_init(&mdsc->cap_dirty_lock);
3449 init_waitqueue_head(&mdsc->cap_flushing_wq);
3450 spin_lock_init(&mdsc->dentry_lru_lock);
3451 INIT_LIST_HEAD(&mdsc->dentry_lru);
2d06eeb8 3452
37151668 3453 ceph_caps_init(mdsc);
3d14c5d2 3454 ceph_adjust_min_caps(mdsc, fsc->min_caps);
37151668 3455
10183a69
YZ		 3456 init_rwsem(&mdsc->pool_perm_rwsem);
3457 mdsc->pool_perm_tree = RB_ROOT;
3458
5f44f142 3459 return 0;
2f2dc053
SW		 3460}
3461
3462/*
3463 * Wait for safe replies on open mds requests. If we time out, drop
3464 * all requests from the tree to avoid dangling dentry refs.
3465 */
3466static void wait_requests(struct ceph_mds_client *mdsc)
3467{
a319bf56 3468 struct ceph_options *opts = mdsc->fsc->client->options;
2f2dc053 3469 struct ceph_mds_request *req;
2f2dc053
SW		 3470
3471 mutex_lock(&mdsc->mutex);
44ca18f2 3472 if (__get_oldest_req(mdsc)) {
2f2dc053 3473 mutex_unlock(&mdsc->mutex);
44ca18f2 3474
2f2dc053
SW		 3475 dout("wait_requests waiting for requests\n");
3476 wait_for_completion_timeout(&mdsc->safe_umount_waiters,
a319bf56 3477 ceph_timeout_jiffies(opts->mount_timeout));
2f2dc053
SW		 3478
3479 /* tear down remaining requests */
44ca18f2
SW		 3480 mutex_lock(&mdsc->mutex);
3481 while ((req = __get_oldest_req(mdsc))) {
2f2dc053
SW		 3482 dout("wait_requests timed out on tid %llu\n",
3483 req->r_tid);
44ca18f2 3484 __unregister_request(mdsc, req);
2f2dc053
SW		 3485 }
3486 }
3487 mutex_unlock(&mdsc->mutex);
3488 dout("wait_requests done\n");
3489}
3490
3491/*
3492 * called before mount is ro, and before dentries are torn down.
3493 * (hmm, does this still race with new lookups?)
3494 */
3495void ceph_mdsc_pre_umount(struct ceph_mds_client *mdsc)
3496{
3497 dout("pre_umount\n");
3498 mdsc->stopping = 1;
3499
3500 drop_leases(mdsc);
afcdaea3 3501 ceph_flush_dirty_caps(mdsc);
2f2dc053 3502 wait_requests(mdsc);
17c688c3
SW		 3503
3504 /*
3505 * wait for reply handlers to drop their request refs and
3506 * their inode/dcache refs
3507 */
3508 ceph_msgr_flush();
2f2dc053
SW		 3509}
3510
3511/*
3512 * wait for all write mds requests to flush.
3513 */
3514static void wait_unsafe_requests(struct ceph_mds_client *mdsc, u64 want_tid)
3515{
80fc7314 3516 struct ceph_mds_request *req = NULL, *nextreq;
44ca18f2 3517 struct rb_node *n;
2f2dc053
SW		 3518
3519 mutex_lock(&mdsc->mutex);
3520 dout("wait_unsafe_requests want %lld\n", want_tid);
80fc7314 3521restart:
44ca18f2
SW		 3522 req = __get_oldest_req(mdsc);
3523 while (req && req->r_tid <= want_tid) {
80fc7314
SW		 3524 /* find next request */
3525 n = rb_next(&req->r_node);
3526 if (n)
3527 nextreq = rb_entry(n, struct ceph_mds_request, r_node);
3528 else
3529 nextreq = NULL;
e8a7b8b1
YZ		 3530 if (req->r_op != CEPH_MDS_OP_SETFILELOCK &&
3531 (req->r_op & CEPH_MDS_OP_WRITE)) {
44ca18f2
SW		 3532 /* write op */
3533 ceph_mdsc_get_request(req);
80fc7314
SW		 3534 if (nextreq)
3535 ceph_mdsc_get_request(nextreq);
44ca18f2
SW		 3536 mutex_unlock(&mdsc->mutex);
3537 dout("wait_unsafe_requests wait on %llu (want %llu)\n",
3538 req->r_tid, want_tid);
3539 wait_for_completion(&req->r_safe_completion);
3540 mutex_lock(&mdsc->mutex);
44ca18f2 3541 ceph_mdsc_put_request(req);
80fc7314
SW		 3542 if (!nextreq)
3543 break; /* next dne before, so we're done! */
3544 if (RB_EMPTY_NODE(&nextreq->r_node)) {
3545 /* next request was removed from tree */
3546 ceph_mdsc_put_request(nextreq);
3547 goto restart;
3548 }
3549 ceph_mdsc_put_request(nextreq); /* won't go away */
44ca18f2 3550 }
80fc7314 3551 req = nextreq;
2f2dc053
SW		 3552 }
3553 mutex_unlock(&mdsc->mutex);
3554 dout("wait_unsafe_requests done\n");
3555}
3556
3557void ceph_mdsc_sync(struct ceph_mds_client *mdsc)
3558{
affbc19a 3559 u64 want_tid, want_flush, want_snap;
2f2dc053 3560
48fec5d0 3561 if (ACCESS_ONCE(mdsc->fsc->mount_state) == CEPH_MOUNT_SHUTDOWN)
56b7cf95
SW		 3562 return;
3563
2f2dc053
SW		 3564 dout("sync\n");
3565 mutex_lock(&mdsc->mutex);
3566 want_tid = mdsc->last_tid;
2f2dc053 3567 mutex_unlock(&mdsc->mutex);
2f2dc053 3568
afcdaea3 3569 ceph_flush_dirty_caps(mdsc);
d3383a8e 3570 spin_lock(&mdsc->cap_dirty_lock);
8310b089 3571 want_flush = mdsc->last_cap_flush_tid;
d3383a8e
YZ		 3572 spin_unlock(&mdsc->cap_dirty_lock);
3573
affbc19a
YZ		 3574 down_read(&mdsc->snap_rwsem);
3575 want_snap = mdsc->last_snap_seq;
3576 up_read(&mdsc->snap_rwsem);
3577
3578 dout("sync want tid %lld flush_seq %lld snap_seq %lld\n",
3579 want_tid, want_flush, want_snap);
2f2dc053
SW		 3580
3581 wait_unsafe_requests(mdsc, want_tid);
affbc19a 3582 wait_caps_flush(mdsc, want_flush, want_snap);
2f2dc053
SW		 3583}
3584
f3c60c59
SW		 3585/*
3586 * true if all sessions are closed, or we force unmount
3587 */
7fd7d101 3588static bool done_closing_sessions(struct ceph_mds_client *mdsc)
f3c60c59 3589{
48fec5d0 3590 if (ACCESS_ONCE(mdsc->fsc->mount_state) == CEPH_MOUNT_SHUTDOWN)
f3c60c59 3591 return true;
86d8f67b 3592 return atomic_read(&mdsc->num_sessions) == 0;
f3c60c59 3593}
2f2dc053
SW		 3594
3595/*
3596 * called after sb is ro.
3597 */
3598void ceph_mdsc_close_sessions(struct ceph_mds_client *mdsc)
3599{
a319bf56 3600 struct ceph_options *opts = mdsc->fsc->client->options;
2f2dc053
SW		 3601 struct ceph_mds_session *session;
3602 int i;
2f2dc053
SW		 3603
3604 dout("close_sessions\n");
3605
2f2dc053 3606 /* close sessions */
f3c60c59
SW		 3607 mutex_lock(&mdsc->mutex);
3608 for (i = 0; i < mdsc->max_sessions; i++) {
3609 session = __ceph_lookup_mds_session(mdsc, i);
3610 if (!session)
3611 continue;
2f2dc053 3612 mutex_unlock(&mdsc->mutex);
f3c60c59
SW		 3613 mutex_lock(&session->s_mutex);
3614 __close_session(mdsc, session);
3615 mutex_unlock(&session->s_mutex);
3616 ceph_put_mds_session(session);
2f2dc053
SW		 3617 mutex_lock(&mdsc->mutex);
3618 }
f3c60c59
SW		 3619 mutex_unlock(&mdsc->mutex);
3620
3621 dout("waiting for sessions to close\n");
3622 wait_event_timeout(mdsc->session_close_wq, done_closing_sessions(mdsc),
a319bf56 3623 ceph_timeout_jiffies(opts->mount_timeout));
2f2dc053
SW		 3624
3625 /* tear down remaining sessions */
f3c60c59 3626 mutex_lock(&mdsc->mutex);
2f2dc053
SW		 3627 for (i = 0; i < mdsc->max_sessions; i++) {
3628 if (mdsc->sessions[i]) {
3629 session = get_session(mdsc->sessions[i]);
2600d2dd 3630 __unregister_session(mdsc, session);
2f2dc053
SW		 3631 mutex_unlock(&mdsc->mutex);
3632 mutex_lock(&session->s_mutex);
3633 remove_session_caps(session);
3634 mutex_unlock(&session->s_mutex);
3635 ceph_put_mds_session(session);
3636 mutex_lock(&mdsc->mutex);
3637 }
3638 }
2f2dc053 3639 WARN_ON(!list_empty(&mdsc->cap_delay_list));
2f2dc053
SW		 3640 mutex_unlock(&mdsc->mutex);
3641
3642 ceph_cleanup_empty_realms(mdsc);
3643
3644 cancel_delayed_work_sync(&mdsc->delayed_work); /* cancel timer */
3645
3646 dout("stopped\n");
3647}
3648
48fec5d0
YZ		 3649void ceph_mdsc_force_umount(struct ceph_mds_client *mdsc)
3650{
3651 struct ceph_mds_session *session;
3652 int mds;
3653
3654 dout("force umount\n");
3655
3656 mutex_lock(&mdsc->mutex);
3657 for (mds = 0; mds < mdsc->max_sessions; mds++) {
3658 session = __ceph_lookup_mds_session(mdsc, mds);
3659 if (!session)
3660 continue;
3661 mutex_unlock(&mdsc->mutex);
3662 mutex_lock(&session->s_mutex);
3663 __close_session(mdsc, session);
3664 if (session->s_state == CEPH_MDS_SESSION_CLOSING) {
3665 cleanup_session_requests(mdsc, session);
3666 remove_session_caps(session);
3667 }
3668 mutex_unlock(&session->s_mutex);
3669 ceph_put_mds_session(session);
3670 mutex_lock(&mdsc->mutex);
3671 kick_requests(mdsc, mds);
3672 }
3673 __wake_requests(mdsc, &mdsc->waiting_for_map);
3674 mutex_unlock(&mdsc->mutex);
3675}
3676
3d14c5d2 3677static void ceph_mdsc_stop(struct ceph_mds_client *mdsc)
2f2dc053
SW		 3678{
3679 dout("stop\n");
3680 cancel_delayed_work_sync(&mdsc->delayed_work); /* cancel timer */
3681 if (mdsc->mdsmap)
3682 ceph_mdsmap_destroy(mdsc->mdsmap);
3683 kfree(mdsc->sessions);
37151668 3684 ceph_caps_finalize(mdsc);
10183a69 3685 ceph_pool_perm_destroy(mdsc);
2f2dc053
SW		 3686}
3687
3d14c5d2
YS		 3688void ceph_mdsc_destroy(struct ceph_fs_client *fsc)
3689{
3690 struct ceph_mds_client *mdsc = fsc->mdsc;
3691
ef550f6f 3692 dout("mdsc_destroy %p\n", mdsc);
3d14c5d2 3693 ceph_mdsc_stop(mdsc);
ef550f6f
SW		 3694
3695 /* flush out any connection work with references to us */
3696 ceph_msgr_flush();
3697
3d14c5d2
YS		 3698 fsc->mdsc = NULL;
3699 kfree(mdsc);
ef550f6f 3700 dout("mdsc_destroy %p done\n", mdsc);
3d14c5d2
YS		 3701}
3702
2f2dc053
SW		 3703
3704/*
3705 * handle mds map update.
3706 */
3707void ceph_mdsc_handle_map(struct ceph_mds_client *mdsc, struct ceph_msg *msg)
3708{
3709 u32 epoch;
3710 u32 maplen;
3711 void *p = msg->front.iov_base;
3712 void *end = p + msg->front.iov_len;
3713 struct ceph_mdsmap *newmap, *oldmap;
3714 struct ceph_fsid fsid;
3715 int err = -EINVAL;
3716
3717 ceph_decode_need(&p, end, sizeof(fsid)+2*sizeof(u32), bad);
3718 ceph_decode_copy(&p, &fsid, sizeof(fsid));
3d14c5d2 3719 if (ceph_check_fsid(mdsc->fsc->client, &fsid) < 0)
0743304d 3720 return;
c89136ea
SW		 3721 epoch = ceph_decode_32(&p);
3722 maplen = ceph_decode_32(&p);
2f2dc053
SW		 3723 dout("handle_map epoch %u len %d\n", epoch, (int)maplen);
3724
3725 /* do we need it? */
3d14c5d2 3726 ceph_monc_got_mdsmap(&mdsc->fsc->client->monc, epoch);
2f2dc053
SW		 3727 mutex_lock(&mdsc->mutex);
3728 if (mdsc->mdsmap && epoch <= mdsc->mdsmap->m_epoch) {
3729 dout("handle_map epoch %u <= our %u\n",
3730 epoch, mdsc->mdsmap->m_epoch);
3731 mutex_unlock(&mdsc->mutex);
3732 return;
3733 }
3734
3735 newmap = ceph_mdsmap_decode(&p, end);
3736 if (IS_ERR(newmap)) {
3737 err = PTR_ERR(newmap);
3738 goto bad_unlock;
3739 }
3740
3741 /* swap into place */
3742 if (mdsc->mdsmap) {
3743 oldmap = mdsc->mdsmap;
3744 mdsc->mdsmap = newmap;
3745 check_new_map(mdsc, newmap, oldmap);
3746 ceph_mdsmap_destroy(oldmap);
3747 } else {
3748 mdsc->mdsmap = newmap; /* first mds map */
3749 }
3d14c5d2 3750 mdsc->fsc->sb->s_maxbytes = mdsc->mdsmap->m_max_file_size;
2f2dc053
SW		 3751
3752 __wake_requests(mdsc, &mdsc->waiting_for_map);
3753
3754 mutex_unlock(&mdsc->mutex);
3755 schedule_delayed(mdsc);
3756 return;
3757
3758bad_unlock:
3759 mutex_unlock(&mdsc->mutex);
3760bad:
3761 pr_err("error decoding mdsmap %d\n", err);
3762 return;
3763}
3764
3765static struct ceph_connection *con_get(struct ceph_connection *con)
3766{
3767 struct ceph_mds_session *s = con->private;
3768
3769 if (get_session(s)) {
2600d2dd 3770 dout("mdsc con_get %p ok (%d)\n", s, atomic_read(&s->s_ref));
2f2dc053
SW		 3771 return con;
3772 }
3773 dout("mdsc con_get %p FAIL\n", s);
3774 return NULL;
3775}
3776
3777static void con_put(struct ceph_connection *con)
3778{
3779 struct ceph_mds_session *s = con->private;
3780
7d8e18a6 3781 dout("mdsc con_put %p (%d)\n", s, atomic_read(&s->s_ref) - 1);
2f2dc053
SW		 3782 ceph_put_mds_session(s);
3783}
3784
3785/*
3786 * if the client is unresponsive for long enough, the mds will kill
3787 * the session entirely.
3788 */
3789static void peer_reset(struct ceph_connection *con)
3790{
3791 struct ceph_mds_session *s = con->private;
7e70f0ed 3792 struct ceph_mds_client *mdsc = s->s_mdsc;
2f2dc053 3793
f3ae1b97 3794 pr_warn("mds%d closed our session\n", s->s_mds);
7e70f0ed 3795 send_mds_reconnect(mdsc, s);
2f2dc053
SW		 3796}
3797
3798static void dispatch(struct ceph_connection *con, struct ceph_msg *msg)
3799{
3800 struct ceph_mds_session *s = con->private;
3801 struct ceph_mds_client *mdsc = s->s_mdsc;
3802 int type = le16_to_cpu(msg->hdr.type);
3803
2600d2dd
SW		 3804 mutex_lock(&mdsc->mutex);
3805 if (__verify_registered_session(mdsc, s) < 0) {
3806 mutex_unlock(&mdsc->mutex);
3807 goto out;
3808 }
3809 mutex_unlock(&mdsc->mutex);
3810
2f2dc053
SW		 3811 switch (type) {
3812 case CEPH_MSG_MDS_MAP:
3813 ceph_mdsc_handle_map(mdsc, msg);
3814 break;
3815 case CEPH_MSG_CLIENT_SESSION:
3816 handle_session(s, msg);
3817 break;
3818 case CEPH_MSG_CLIENT_REPLY:
3819 handle_reply(s, msg);
3820 break;
3821 case CEPH_MSG_CLIENT_REQUEST_FORWARD:
2600d2dd 3822 handle_forward(mdsc, s, msg);
2f2dc053
SW		 3823 break;
3824 case CEPH_MSG_CLIENT_CAPS:
3825 ceph_handle_caps(s, msg);
3826 break;
3827 case CEPH_MSG_CLIENT_SNAP:
2600d2dd 3828 ceph_handle_snap(mdsc, s, msg);
2f2dc053
SW		 3829 break;
3830 case CEPH_MSG_CLIENT_LEASE:
2600d2dd 3831 handle_lease(mdsc, s, msg);
2f2dc053
SW		 3832 break;
3833
3834 default:
3835 pr_err("received unknown message type %d %s\n", type,
3836 ceph_msg_type_name(type));
3837 }
2600d2dd 3838out:
2f2dc053
SW		 3839 ceph_msg_put(msg);
3840}
3841
4e7a5dcd
SW		 3842/*
3843 * authentication
3844 */
a3530df3
AE		 3845
3846/*
3847 * Note: returned pointer is the address of a structure that's
3848 * managed separately. Caller must *not* attempt to free it.
3849 */
3850static struct ceph_auth_handshake *get_authorizer(struct ceph_connection *con,
8f43fb53 3851 int *proto, int force_new)
4e7a5dcd
SW		 3852{
3853 struct ceph_mds_session *s = con->private;
3854 struct ceph_mds_client *mdsc = s->s_mdsc;
3d14c5d2 3855 struct ceph_auth_client *ac = mdsc->fsc->client->monc.auth;
74f1869f 3856 struct ceph_auth_handshake *auth = &s->s_auth;
4e7a5dcd 3857
74f1869f 3858 if (force_new && auth->authorizer) {
27859f97 3859 ceph_auth_destroy_authorizer(ac, auth->authorizer);
74f1869f 3860 auth->authorizer = NULL;
4e7a5dcd 3861 }
27859f97
SW		 3862 if (!auth->authorizer) {
3863 int ret = ceph_auth_create_authorizer(ac, CEPH_ENTITY_TYPE_MDS,
3864 auth);
0bed9b5c
SW		 3865 if (ret)
3866 return ERR_PTR(ret);
27859f97
SW		 3867 } else {
3868 int ret = ceph_auth_update_authorizer(ac, CEPH_ENTITY_TYPE_MDS,
3869 auth);
a255651d 3870 if (ret)
a3530df3 3871 return ERR_PTR(ret);
4e7a5dcd 3872 }
4e7a5dcd 3873 *proto = ac->protocol;
74f1869f 3874
a3530df3 3875 return auth;
4e7a5dcd
SW		 3876}
3877
3878
3879static int verify_authorizer_reply(struct ceph_connection *con, int len)
3880{
3881 struct ceph_mds_session *s = con->private;
3882 struct ceph_mds_client *mdsc = s->s_mdsc;
3d14c5d2 3883 struct ceph_auth_client *ac = mdsc->fsc->client->monc.auth;
4e7a5dcd 3884
27859f97 3885 return ceph_auth_verify_authorizer_reply(ac, s->s_auth.authorizer, len);
4e7a5dcd
SW		 3886}
3887
9bd2e6f8
SW		 3888static int invalidate_authorizer(struct ceph_connection *con)
3889{
3890 struct ceph_mds_session *s = con->private;
3891 struct ceph_mds_client *mdsc = s->s_mdsc;
3d14c5d2 3892 struct ceph_auth_client *ac = mdsc->fsc->client->monc.auth;
9bd2e6f8 3893
27859f97 3894 ceph_auth_invalidate_authorizer(ac, CEPH_ENTITY_TYPE_MDS);
9bd2e6f8 3895
3d14c5d2 3896 return ceph_monc_validate_auth(&mdsc->fsc->client->monc);
9bd2e6f8
SW		 3897}
3898
53ded495
AE		 3899static struct ceph_msg *mds_alloc_msg(struct ceph_connection *con,
3900 struct ceph_msg_header *hdr, int *skip)
3901{
3902 struct ceph_msg *msg;
3903 int type = (int) le16_to_cpu(hdr->type);
3904 int front_len = (int) le32_to_cpu(hdr->front_len);
3905
3906 if (con->in_msg)
3907 return con->in_msg;
3908
3909 *skip = 0;
3910 msg = ceph_msg_new(type, front_len, GFP_NOFS, false);
3911 if (!msg) {
3912 pr_err("unable to allocate msg type %d len %d\n",
3913 type, front_len);
3914 return NULL;
3915 }
53ded495
AE		 3916
3917 return msg;
3918}
3919
33d07337
YZ		 3920static int sign_message(struct ceph_connection *con, struct ceph_msg *msg)
3921{
3922 struct ceph_mds_session *s = con->private;
3923 struct ceph_auth_handshake *auth = &s->s_auth;
3924 return ceph_auth_sign_message(auth, msg);
3925}
3926
3927static int check_message_signature(struct ceph_connection *con, struct ceph_msg *msg)
3928{
3929 struct ceph_mds_session *s = con->private;
3930 struct ceph_auth_handshake *auth = &s->s_auth;
3931 return ceph_auth_check_message_signature(auth, msg);
3932}
3933
9e32789f 3934static const struct ceph_connection_operations mds_con_ops = {
2f2dc053
SW		 3935 .get = con_get,
3936 .put = con_put,
3937 .dispatch = dispatch,
4e7a5dcd
SW		 3938 .get_authorizer = get_authorizer,
3939 .verify_authorizer_reply = verify_authorizer_reply,
9bd2e6f8 3940 .invalidate_authorizer = invalidate_authorizer,
2f2dc053 3941 .peer_reset = peer_reset,
53ded495 3942 .alloc_msg = mds_alloc_msg,
33d07337
YZ		 3943 .sign_message = sign_message,
3944 .check_message_signature = check_message_signature,
2f2dc053
SW		 3945};
3946
2f2dc053 3947/* eof */
Ubuntu Hirsute Linux kernel mirror