]> git.ipfire.org Git - thirdparty/kernel/stable.git/blob - fs/ceph/mds_client.c
projects / thirdparty / kernel / stable.git / blob
? search:
summary | shortlog | log | commit | commitdiff | tree
blame | history | raw | HEAD
Merge tag 'kvm-x86-docs-6.7' of https://github.com/kvm-x86/linux into HEAD
[thirdparty/kernel/stable.git] / fs / ceph / mds_client.c
1 // SPDX-License-Identifier: GPL-2.0
2 #include <linux/ceph/ceph_debug.h>
3
4 #include <linux/fs.h>
5 #include <linux/wait.h>
6 #include <linux/slab.h>
7 #include <linux/gfp.h>
8 #include <linux/sched.h>
9 #include <linux/debugfs.h>
10 #include <linux/seq_file.h>
11 #include <linux/ratelimit.h>
12 #include <linux/bits.h>
13 #include <linux/ktime.h>
14 #include <linux/bitmap.h>
15
16 #include "super.h"
17 #include "mds_client.h"
18 #include "crypto.h"
19
20 #include <linux/ceph/ceph_features.h>
21 #include <linux/ceph/messenger.h>
22 #include <linux/ceph/decode.h>
23 #include <linux/ceph/pagelist.h>
24 #include <linux/ceph/auth.h>
25 #include <linux/ceph/debugfs.h>
26
27 #define RECONNECT_MAX_SIZE (INT_MAX - PAGE_SIZE)
28
29 /*
30 * A cluster of MDS (metadata server) daemons is responsible for
31 * managing the file system namespace (the directory hierarchy and
32 * inodes) and for coordinating shared access to storage. Metadata is
33 * partitioning hierarchically across a number of servers, and that
34 * partition varies over time as the cluster adjusts the distribution
35 * in order to balance load.
36 *
37 * The MDS client is primarily responsible to managing synchronous
38 * metadata requests for operations like open, unlink, and so forth.
39 * If there is a MDS failure, we find out about it when we (possibly
40 * request and) receive a new MDS map, and can resubmit affected
41 * requests.
42 *
43 * For the most part, though, we take advantage of a lossless
44 * communications channel to the MDS, and do not need to worry about
45 * timing out or resubmitting requests.
46 *
47 * We maintain a stateful "session" with each MDS we interact with.
48 * Within each session, we sent periodic heartbeat messages to ensure
49 * any capabilities or leases we have been issues remain valid. If
50 * the session times out and goes stale, our leases and capabilities
51 * are no longer valid.
52 */
53
54 struct ceph_reconnect_state {
55 struct ceph_mds_session *session;
56 int nr_caps, nr_realms;
57 struct ceph_pagelist *pagelist;
58 unsigned msg_version;
59 bool allow_multi;
60 };
61
62 static void __wake_requests(struct ceph_mds_client *mdsc,
63 struct list_head *head);
64 static void ceph_cap_release_work(struct work_struct *work);
65 static void ceph_cap_reclaim_work(struct work_struct *work);
66
67 static const struct ceph_connection_operations mds_con_ops;
68
69
70 /*
71 * mds reply parsing
72 */
73
74 static int parse_reply_info_quota(void **p, void *end,
75 struct ceph_mds_reply_info_in *info)
76 {
77 u8 struct_v, struct_compat;
78 u32 struct_len;
79
80 ceph_decode_8_safe(p, end, struct_v, bad);
81 ceph_decode_8_safe(p, end, struct_compat, bad);
82 /* struct_v is expected to be >= 1. we only
83 * understand encoding with struct_compat == 1. */
84 if (!struct_v || struct_compat != 1)
85 goto bad;
86 ceph_decode_32_safe(p, end, struct_len, bad);
87 ceph_decode_need(p, end, struct_len, bad);
88 end = *p + struct_len;
89 ceph_decode_64_safe(p, end, info->max_bytes, bad);
90 ceph_decode_64_safe(p, end, info->max_files, bad);
91 *p = end;
92 return 0;
93 bad:
94 return -EIO;
95 }
96
97 /*
98 * parse individual inode info
99 */
100 static int parse_reply_info_in(void **p, void *end,
101 struct ceph_mds_reply_info_in *info,
102 u64 features)
103 {
104 int err = 0;
105 u8 struct_v = 0;
106
107 if (features == (u64)-1) {
108 u32 struct_len;
109 u8 struct_compat;
110 ceph_decode_8_safe(p, end, struct_v, bad);
111 ceph_decode_8_safe(p, end, struct_compat, bad);
112 /* struct_v is expected to be >= 1. we only understand
113 * encoding with struct_compat == 1. */
114 if (!struct_v || struct_compat != 1)
115 goto bad;
116 ceph_decode_32_safe(p, end, struct_len, bad);
117 ceph_decode_need(p, end, struct_len, bad);
118 end = *p + struct_len;
119 }
120
121 ceph_decode_need(p, end, sizeof(struct ceph_mds_reply_inode), bad);
122 info->in = *p;
123 *p += sizeof(struct ceph_mds_reply_inode) +
124 sizeof(*info->in->fragtree.splits) *
125 le32_to_cpu(info->in->fragtree.nsplits);
126
127 ceph_decode_32_safe(p, end, info->symlink_len, bad);
128 ceph_decode_need(p, end, info->symlink_len, bad);
129 info->symlink = *p;
130 *p += info->symlink_len;
131
132 ceph_decode_copy_safe(p, end, &info->dir_layout,
133 sizeof(info->dir_layout), bad);
134 ceph_decode_32_safe(p, end, info->xattr_len, bad);
135 ceph_decode_need(p, end, info->xattr_len, bad);
136 info->xattr_data = *p;
137 *p += info->xattr_len;
138
139 if (features == (u64)-1) {
140 /* inline data */
141 ceph_decode_64_safe(p, end, info->inline_version, bad);
142 ceph_decode_32_safe(p, end, info->inline_len, bad);
143 ceph_decode_need(p, end, info->inline_len, bad);
144 info->inline_data = *p;
145 *p += info->inline_len;
146 /* quota */
147 err = parse_reply_info_quota(p, end, info);
148 if (err < 0)
149 goto out_bad;
150 /* pool namespace */
151 ceph_decode_32_safe(p, end, info->pool_ns_len, bad);
152 if (info->pool_ns_len > 0) {
153 ceph_decode_need(p, end, info->pool_ns_len, bad);
154 info->pool_ns_data = *p;
155 *p += info->pool_ns_len;
156 }
157
158 /* btime */
159 ceph_decode_need(p, end, sizeof(info->btime), bad);
160 ceph_decode_copy(p, &info->btime, sizeof(info->btime));
161
162 /* change attribute */
163 ceph_decode_64_safe(p, end, info->change_attr, bad);
164
165 /* dir pin */
166 if (struct_v >= 2) {
167 ceph_decode_32_safe(p, end, info->dir_pin, bad);
168 } else {
169 info->dir_pin = -ENODATA;
170 }
171
172 /* snapshot birth time, remains zero for v<=2 */
173 if (struct_v >= 3) {
174 ceph_decode_need(p, end, sizeof(info->snap_btime), bad);
175 ceph_decode_copy(p, &info->snap_btime,
176 sizeof(info->snap_btime));
177 } else {
178 memset(&info->snap_btime, 0, sizeof(info->snap_btime));
179 }
180
181 /* snapshot count, remains zero for v<=3 */
182 if (struct_v >= 4) {
183 ceph_decode_64_safe(p, end, info->rsnaps, bad);
184 } else {
185 info->rsnaps = 0;
186 }
187
188 if (struct_v >= 5) {
189 u32 alen;
190
191 ceph_decode_32_safe(p, end, alen, bad);
192
193 while (alen--) {
194 u32 len;
195
196 /* key */
197 ceph_decode_32_safe(p, end, len, bad);
198 ceph_decode_skip_n(p, end, len, bad);
199 /* value */
200 ceph_decode_32_safe(p, end, len, bad);
201 ceph_decode_skip_n(p, end, len, bad);
202 }
203 }
204
205 /* fscrypt flag -- ignore */
206 if (struct_v >= 6)
207 ceph_decode_skip_8(p, end, bad);
208
209 info->fscrypt_auth = NULL;
210 info->fscrypt_auth_len = 0;
211 info->fscrypt_file = NULL;
212 info->fscrypt_file_len = 0;
213 if (struct_v >= 7) {
214 ceph_decode_32_safe(p, end, info->fscrypt_auth_len, bad);
215 if (info->fscrypt_auth_len) {
216 info->fscrypt_auth = kmalloc(info->fscrypt_auth_len,
217 GFP_KERNEL);
218 if (!info->fscrypt_auth)
219 return -ENOMEM;
220 ceph_decode_copy_safe(p, end, info->fscrypt_auth,
221 info->fscrypt_auth_len, bad);
222 }
223 ceph_decode_32_safe(p, end, info->fscrypt_file_len, bad);
224 if (info->fscrypt_file_len) {
225 info->fscrypt_file = kmalloc(info->fscrypt_file_len,
226 GFP_KERNEL);
227 if (!info->fscrypt_file)
228 return -ENOMEM;
229 ceph_decode_copy_safe(p, end, info->fscrypt_file,
230 info->fscrypt_file_len, bad);
231 }
232 }
233 *p = end;
234 } else {
235 /* legacy (unversioned) struct */
236 if (features & CEPH_FEATURE_MDS_INLINE_DATA) {
237 ceph_decode_64_safe(p, end, info->inline_version, bad);
238 ceph_decode_32_safe(p, end, info->inline_len, bad);
239 ceph_decode_need(p, end, info->inline_len, bad);
240 info->inline_data = *p;
241 *p += info->inline_len;
242 } else
243 info->inline_version = CEPH_INLINE_NONE;
244
245 if (features & CEPH_FEATURE_MDS_QUOTA) {
246 err = parse_reply_info_quota(p, end, info);
247 if (err < 0)
248 goto out_bad;
249 } else {
250 info->max_bytes = 0;
251 info->max_files = 0;
252 }
253
254 info->pool_ns_len = 0;
255 info->pool_ns_data = NULL;
256 if (features & CEPH_FEATURE_FS_FILE_LAYOUT_V2) {
257 ceph_decode_32_safe(p, end, info->pool_ns_len, bad);
258 if (info->pool_ns_len > 0) {
259 ceph_decode_need(p, end, info->pool_ns_len, bad);
260 info->pool_ns_data = *p;
261 *p += info->pool_ns_len;
262 }
263 }
264
265 if (features & CEPH_FEATURE_FS_BTIME) {
266 ceph_decode_need(p, end, sizeof(info->btime), bad);
267 ceph_decode_copy(p, &info->btime, sizeof(info->btime));
268 ceph_decode_64_safe(p, end, info->change_attr, bad);
269 }
270
271 info->dir_pin = -ENODATA;
272 /* info->snap_btime and info->rsnaps remain zero */
273 }
274 return 0;
275 bad:
276 err = -EIO;
277 out_bad:
278 return err;
279 }
280
281 static int parse_reply_info_dir(void **p, void *end,
282 struct ceph_mds_reply_dirfrag **dirfrag,
283 u64 features)
284 {
285 if (features == (u64)-1) {
286 u8 struct_v, struct_compat;
287 u32 struct_len;
288 ceph_decode_8_safe(p, end, struct_v, bad);
289 ceph_decode_8_safe(p, end, struct_compat, bad);
290 /* struct_v is expected to be >= 1. we only understand
291 * encoding whose struct_compat == 1. */
292 if (!struct_v || struct_compat != 1)
293 goto bad;
294 ceph_decode_32_safe(p, end, struct_len, bad);
295 ceph_decode_need(p, end, struct_len, bad);
296 end = *p + struct_len;
297 }
298
299 ceph_decode_need(p, end, sizeof(**dirfrag), bad);
300 *dirfrag = *p;
301 *p += sizeof(**dirfrag) + sizeof(u32) * le32_to_cpu((*dirfrag)->ndist);
302 if (unlikely(*p > end))
303 goto bad;
304 if (features == (u64)-1)
305 *p = end;
306 return 0;
307 bad:
308 return -EIO;
309 }
310
311 static int parse_reply_info_lease(void **p, void *end,
312 struct ceph_mds_reply_lease **lease,
313 u64 features, u32 *altname_len, u8 **altname)
314 {
315 u8 struct_v;
316 u32 struct_len;
317 void *lend;
318
319 if (features == (u64)-1) {
320 u8 struct_compat;
321
322 ceph_decode_8_safe(p, end, struct_v, bad);
323 ceph_decode_8_safe(p, end, struct_compat, bad);
324
325 /* struct_v is expected to be >= 1. we only understand
326 * encoding whose struct_compat == 1. */
327 if (!struct_v || struct_compat != 1)
328 goto bad;
329
330 ceph_decode_32_safe(p, end, struct_len, bad);
331 } else {
332 struct_len = sizeof(**lease);
333 *altname_len = 0;
334 *altname = NULL;
335 }
336
337 lend = *p + struct_len;
338 ceph_decode_need(p, end, struct_len, bad);
339 *lease = *p;
340 *p += sizeof(**lease);
341
342 if (features == (u64)-1) {
343 if (struct_v >= 2) {
344 ceph_decode_32_safe(p, end, *altname_len, bad);
345 ceph_decode_need(p, end, *altname_len, bad);
346 *altname = *p;
347 *p += *altname_len;
348 } else {
349 *altname = NULL;
350 *altname_len = 0;
351 }
352 }
353 *p = lend;
354 return 0;
355 bad:
356 return -EIO;
357 }
358
359 /*
360 * parse a normal reply, which may contain a (dir+)dentry and/or a
361 * target inode.
362 */
363 static int parse_reply_info_trace(void **p, void *end,
364 struct ceph_mds_reply_info_parsed *info,
365 u64 features)
366 {
367 int err;
368
369 if (info->head->is_dentry) {
370 err = parse_reply_info_in(p, end, &info->diri, features);
371 if (err < 0)
372 goto out_bad;
373
374 err = parse_reply_info_dir(p, end, &info->dirfrag, features);
375 if (err < 0)
376 goto out_bad;
377
378 ceph_decode_32_safe(p, end, info->dname_len, bad);
379 ceph_decode_need(p, end, info->dname_len, bad);
380 info->dname = *p;
381 *p += info->dname_len;
382
383 err = parse_reply_info_lease(p, end, &info->dlease, features,
384 &info->altname_len, &info->altname);
385 if (err < 0)
386 goto out_bad;
387 }
388
389 if (info->head->is_target) {
390 err = parse_reply_info_in(p, end, &info->targeti, features);
391 if (err < 0)
392 goto out_bad;
393 }
394
395 if (unlikely(*p != end))
396 goto bad;
397 return 0;
398
399 bad:
400 err = -EIO;
401 out_bad:
402 pr_err("problem parsing mds trace %d\n", err);
403 return err;
404 }
405
406 /*
407 * parse readdir results
408 */
409 static int parse_reply_info_readdir(void **p, void *end,
410 struct ceph_mds_request *req,
411 u64 features)
412 {
413 struct ceph_mds_reply_info_parsed *info = &req->r_reply_info;
414 u32 num, i = 0;
415 int err;
416
417 err = parse_reply_info_dir(p, end, &info->dir_dir, features);
418 if (err < 0)
419 goto out_bad;
420
421 ceph_decode_need(p, end, sizeof(num) + 2, bad);
422 num = ceph_decode_32(p);
423 {
424 u16 flags = ceph_decode_16(p);
425 info->dir_end = !!(flags & CEPH_READDIR_FRAG_END);
426 info->dir_complete = !!(flags & CEPH_READDIR_FRAG_COMPLETE);
427 info->hash_order = !!(flags & CEPH_READDIR_HASH_ORDER);
428 info->offset_hash = !!(flags & CEPH_READDIR_OFFSET_HASH);
429 }
430 if (num == 0)
431 goto done;
432
433 BUG_ON(!info->dir_entries);
434 if ((unsigned long)(info->dir_entries + num) >
435 (unsigned long)info->dir_entries + info->dir_buf_size) {
436 pr_err("dir contents are larger than expected\n");
437 WARN_ON(1);
438 goto bad;
439 }
440
441 info->dir_nr = num;
442 while (num) {
443 struct inode *inode = d_inode(req->r_dentry);
444 struct ceph_inode_info *ci = ceph_inode(inode);
445 struct ceph_mds_reply_dir_entry *rde = info->dir_entries + i;
446 struct fscrypt_str tname = FSTR_INIT(NULL, 0);
447 struct fscrypt_str oname = FSTR_INIT(NULL, 0);
448 struct ceph_fname fname;
449 u32 altname_len, _name_len;
450 u8 *altname, *_name;
451
452 /* dentry */
453 ceph_decode_32_safe(p, end, _name_len, bad);
454 ceph_decode_need(p, end, _name_len, bad);
455 _name = *p;
456 *p += _name_len;
457 dout("parsed dir dname '%.*s'\n", _name_len, _name);
458
459 if (info->hash_order)
460 rde->raw_hash = ceph_str_hash(ci->i_dir_layout.dl_dir_hash,
461 _name, _name_len);
462
463 /* dentry lease */
464 err = parse_reply_info_lease(p, end, &rde->lease, features,
465 &altname_len, &altname);
466 if (err)
467 goto out_bad;
468
469 /*
470 * Try to dencrypt the dentry names and update them
471 * in the ceph_mds_reply_dir_entry struct.
472 */
473 fname.dir = inode;
474 fname.name = _name;
475 fname.name_len = _name_len;
476 fname.ctext = altname;
477 fname.ctext_len = altname_len;
478 /*
479 * The _name_len maybe larger than altname_len, such as
480 * when the human readable name length is in range of
481 * (CEPH_NOHASH_NAME_MAX, CEPH_NOHASH_NAME_MAX + SHA256_DIGEST_SIZE),
482 * then the copy in ceph_fname_to_usr will corrupt the
483 * data if there has no encryption key.
484 *
485 * Just set the no_copy flag and then if there has no
486 * encryption key the oname.name will be assigned to
487 * _name always.
488 */
489 fname.no_copy = true;
490 if (altname_len == 0) {
491 /*
492 * Set tname to _name, and this will be used
493 * to do the base64_decode in-place. It's
494 * safe because the decoded string should
495 * always be shorter, which is 3/4 of origin
496 * string.
497 */
498 tname.name = _name;
499
500 /*
501 * Set oname to _name too, and this will be
502 * used to do the dencryption in-place.
503 */
504 oname.name = _name;
505 oname.len = _name_len;
506 } else {
507 /*
508 * This will do the decryption only in-place
509 * from altname cryptext directly.
510 */
511 oname.name = altname;
512 oname.len = altname_len;
513 }
514 rde->is_nokey = false;
515 err = ceph_fname_to_usr(&fname, &tname, &oname, &rde->is_nokey);
516 if (err) {
517 pr_err("%s unable to decode %.*s, got %d\n", __func__,
518 _name_len, _name, err);
519 goto out_bad;
520 }
521 rde->name = oname.name;
522 rde->name_len = oname.len;
523
524 /* inode */
525 err = parse_reply_info_in(p, end, &rde->inode, features);
526 if (err < 0)
527 goto out_bad;
528 /* ceph_readdir_prepopulate() will update it */
529 rde->offset = 0;
530 i++;
531 num--;
532 }
533
534 done:
535 /* Skip over any unrecognized fields */
536 *p = end;
537 return 0;
538
539 bad:
540 err = -EIO;
541 out_bad:
542 pr_err("problem parsing dir contents %d\n", err);
543 return err;
544 }
545
546 /*
547 * parse fcntl F_GETLK results
548 */
549 static int parse_reply_info_filelock(void **p, void *end,
550 struct ceph_mds_reply_info_parsed *info,
551 u64 features)
552 {
553 if (*p + sizeof(*info->filelock_reply) > end)
554 goto bad;
555
556 info->filelock_reply = *p;
557
558 /* Skip over any unrecognized fields */
559 *p = end;
560 return 0;
561 bad:
562 return -EIO;
563 }
564
565
566 #if BITS_PER_LONG == 64
567
568 #define DELEGATED_INO_AVAILABLE xa_mk_value(1)
569
570 static int ceph_parse_deleg_inos(void **p, void *end,
571 struct ceph_mds_session *s)
572 {
573 u32 sets;
574
575 ceph_decode_32_safe(p, end, sets, bad);
576 dout("got %u sets of delegated inodes\n", sets);
577 while (sets--) {
578 u64 start, len;
579
580 ceph_decode_64_safe(p, end, start, bad);
581 ceph_decode_64_safe(p, end, len, bad);
582
583 /* Don't accept a delegation of system inodes */
584 if (start < CEPH_INO_SYSTEM_BASE) {
585 pr_warn_ratelimited("ceph: ignoring reserved inode range delegation (start=0x%llx len=0x%llx)\n",
586 start, len);
587 continue;
588 }
589 while (len--) {
590 int err = xa_insert(&s->s_delegated_inos, start++,
591 DELEGATED_INO_AVAILABLE,
592 GFP_KERNEL);
593 if (!err) {
594 dout("added delegated inode 0x%llx\n",
595 start - 1);
596 } else if (err == -EBUSY) {
597 pr_warn("MDS delegated inode 0x%llx more than once.\n",
598 start - 1);
599 } else {
600 return err;
601 }
602 }
603 }
604 return 0;
605 bad:
606 return -EIO;
607 }
608
609 u64 ceph_get_deleg_ino(struct ceph_mds_session *s)
610 {
611 unsigned long ino;
612 void *val;
613
614 xa_for_each(&s->s_delegated_inos, ino, val) {
615 val = xa_erase(&s->s_delegated_inos, ino);
616 if (val == DELEGATED_INO_AVAILABLE)
617 return ino;
618 }
619 return 0;
620 }
621
622 int ceph_restore_deleg_ino(struct ceph_mds_session *s, u64 ino)
623 {
624 return xa_insert(&s->s_delegated_inos, ino, DELEGATED_INO_AVAILABLE,
625 GFP_KERNEL);
626 }
627 #else /* BITS_PER_LONG == 64 */
628 /*
629 * FIXME: xarrays can't handle 64-bit indexes on a 32-bit arch. For now, just
630 * ignore delegated_inos on 32 bit arch. Maybe eventually add xarrays for top
631 * and bottom words?
632 */
633 static int ceph_parse_deleg_inos(void **p, void *end,
634 struct ceph_mds_session *s)
635 {
636 u32 sets;
637
638 ceph_decode_32_safe(p, end, sets, bad);
639 if (sets)
640 ceph_decode_skip_n(p, end, sets * 2 * sizeof(__le64), bad);
641 return 0;
642 bad:
643 return -EIO;
644 }
645
646 u64 ceph_get_deleg_ino(struct ceph_mds_session *s)
647 {
648 return 0;
649 }
650
651 int ceph_restore_deleg_ino(struct ceph_mds_session *s, u64 ino)
652 {
653 return 0;
654 }
655 #endif /* BITS_PER_LONG == 64 */
656
657 /*
658 * parse create results
659 */
660 static int parse_reply_info_create(void **p, void *end,
661 struct ceph_mds_reply_info_parsed *info,
662 u64 features, struct ceph_mds_session *s)
663 {
664 int ret;
665
666 if (features == (u64)-1 ||
667 (features & CEPH_FEATURE_REPLY_CREATE_INODE)) {
668 if (*p == end) {
669 /* Malformed reply? */
670 info->has_create_ino = false;
671 } else if (test_bit(CEPHFS_FEATURE_DELEG_INO, &s->s_features)) {
672 info->has_create_ino = true;
673 /* struct_v, struct_compat, and len */
674 ceph_decode_skip_n(p, end, 2 + sizeof(u32), bad);
675 ceph_decode_64_safe(p, end, info->ino, bad);
676 ret = ceph_parse_deleg_inos(p, end, s);
677 if (ret)
678 return ret;
679 } else {
680 /* legacy */
681 ceph_decode_64_safe(p, end, info->ino, bad);
682 info->has_create_ino = true;
683 }
684 } else {
685 if (*p != end)
686 goto bad;
687 }
688
689 /* Skip over any unrecognized fields */
690 *p = end;
691 return 0;
692 bad:
693 return -EIO;
694 }
695
696 static int parse_reply_info_getvxattr(void **p, void *end,
697 struct ceph_mds_reply_info_parsed *info,
698 u64 features)
699 {
700 u32 value_len;
701
702 ceph_decode_skip_8(p, end, bad); /* skip current version: 1 */
703 ceph_decode_skip_8(p, end, bad); /* skip first version: 1 */
704 ceph_decode_skip_32(p, end, bad); /* skip payload length */
705
706 ceph_decode_32_safe(p, end, value_len, bad);
707
708 if (value_len == end - *p) {
709 info->xattr_info.xattr_value = *p;
710 info->xattr_info.xattr_value_len = value_len;
711 *p = end;
712 return value_len;
713 }
714 bad:
715 return -EIO;
716 }
717
718 /*
719 * parse extra results
720 */
721 static int parse_reply_info_extra(void **p, void *end,
722 struct ceph_mds_request *req,
723 u64 features, struct ceph_mds_session *s)
724 {
725 struct ceph_mds_reply_info_parsed *info = &req->r_reply_info;
726 u32 op = le32_to_cpu(info->head->op);
727
728 if (op == CEPH_MDS_OP_GETFILELOCK)
729 return parse_reply_info_filelock(p, end, info, features);
730 else if (op == CEPH_MDS_OP_READDIR || op == CEPH_MDS_OP_LSSNAP)
731 return parse_reply_info_readdir(p, end, req, features);
732 else if (op == CEPH_MDS_OP_CREATE)
733 return parse_reply_info_create(p, end, info, features, s);
734 else if (op == CEPH_MDS_OP_GETVXATTR)
735 return parse_reply_info_getvxattr(p, end, info, features);
736 else
737 return -EIO;
738 }
739
740 /*
741 * parse entire mds reply
742 */
743 static int parse_reply_info(struct ceph_mds_session *s, struct ceph_msg *msg,
744 struct ceph_mds_request *req, u64 features)
745 {
746 struct ceph_mds_reply_info_parsed *info = &req->r_reply_info;
747 void *p, *end;
748 u32 len;
749 int err;
750
751 info->head = msg->front.iov_base;
752 p = msg->front.iov_base + sizeof(struct ceph_mds_reply_head);
753 end = p + msg->front.iov_len - sizeof(struct ceph_mds_reply_head);
754
755 /* trace */
756 ceph_decode_32_safe(&p, end, len, bad);
757 if (len > 0) {
758 ceph_decode_need(&p, end, len, bad);
759 err = parse_reply_info_trace(&p, p+len, info, features);
760 if (err < 0)
761 goto out_bad;
762 }
763
764 /* extra */
765 ceph_decode_32_safe(&p, end, len, bad);
766 if (len > 0) {
767 ceph_decode_need(&p, end, len, bad);
768 err = parse_reply_info_extra(&p, p+len, req, features, s);
769 if (err < 0)
770 goto out_bad;
771 }
772
773 /* snap blob */
774 ceph_decode_32_safe(&p, end, len, bad);
775 info->snapblob_len = len;
776 info->snapblob = p;
777 p += len;
778
779 if (p != end)
780 goto bad;
781 return 0;
782
783 bad:
784 err = -EIO;
785 out_bad:
786 pr_err("mds parse_reply err %d\n", err);
787 ceph_msg_dump(msg);
788 return err;
789 }
790
791 static void destroy_reply_info(struct ceph_mds_reply_info_parsed *info)
792 {
793 int i;
794
795 kfree(info->diri.fscrypt_auth);
796 kfree(info->diri.fscrypt_file);
797 kfree(info->targeti.fscrypt_auth);
798 kfree(info->targeti.fscrypt_file);
799 if (!info->dir_entries)
800 return;
801
802 for (i = 0; i < info->dir_nr; i++) {
803 struct ceph_mds_reply_dir_entry *rde = info->dir_entries + i;
804
805 kfree(rde->inode.fscrypt_auth);
806 kfree(rde->inode.fscrypt_file);
807 }
808 free_pages((unsigned long)info->dir_entries, get_order(info->dir_buf_size));
809 }
810
811 /*
812 * In async unlink case the kclient won't wait for the first reply
813 * from MDS and just drop all the links and unhash the dentry and then
814 * succeeds immediately.
815 *
816 * For any new create/link/rename,etc requests followed by using the
817 * same file names we must wait for the first reply of the inflight
818 * unlink request, or the MDS possibly will fail these following
819 * requests with -EEXIST if the inflight async unlink request was
820 * delayed for some reasons.
821 *
822 * And the worst case is that for the none async openc request it will
823 * successfully open the file if the CDentry hasn't been unlinked yet,
824 * but later the previous delayed async unlink request will remove the
825 * CDenty. That means the just created file is possiblly deleted later
826 * by accident.
827 *
828 * We need to wait for the inflight async unlink requests to finish
829 * when creating new files/directories by using the same file names.
830 */
831 int ceph_wait_on_conflict_unlink(struct dentry *dentry)
832 {
833 struct ceph_fs_client *fsc = ceph_sb_to_client(dentry->d_sb);
834 struct dentry *pdentry = dentry->d_parent;
835 struct dentry *udentry, *found = NULL;
836 struct ceph_dentry_info *di;
837 struct qstr dname;
838 u32 hash = dentry->d_name.hash;
839 int err;
840
841 dname.name = dentry->d_name.name;
842 dname.len = dentry->d_name.len;
843
844 rcu_read_lock();
845 hash_for_each_possible_rcu(fsc->async_unlink_conflict, di,
846 hnode, hash) {
847 udentry = di->dentry;
848
849 spin_lock(&udentry->d_lock);
850 if (udentry->d_name.hash != hash)
851 goto next;
852 if (unlikely(udentry->d_parent != pdentry))
853 goto next;
854 if (!hash_hashed(&di->hnode))
855 goto next;
856
857 if (!test_bit(CEPH_DENTRY_ASYNC_UNLINK_BIT, &di->flags))
858 pr_warn("%s dentry %p:%pd async unlink bit is not set\n",
859 __func__, dentry, dentry);
860
861 if (!d_same_name(udentry, pdentry, &dname))
862 goto next;
863
864 found = dget_dlock(udentry);
865 spin_unlock(&udentry->d_lock);
866 break;
867 next:
868 spin_unlock(&udentry->d_lock);
869 }
870 rcu_read_unlock();
871
872 if (likely(!found))
873 return 0;
874
875 dout("%s dentry %p:%pd conflict with old %p:%pd\n", __func__,
876 dentry, dentry, found, found);
877
878 err = wait_on_bit(&di->flags, CEPH_DENTRY_ASYNC_UNLINK_BIT,
879 TASK_KILLABLE);
880 dput(found);
881 return err;
882 }
883
884
885 /*
886 * sessions
887 */
888 const char *ceph_session_state_name(int s)
889 {
890 switch (s) {
891 case CEPH_MDS_SESSION_NEW: return "new";
892 case CEPH_MDS_SESSION_OPENING: return "opening";
893 case CEPH_MDS_SESSION_OPEN: return "open";
894 case CEPH_MDS_SESSION_HUNG: return "hung";
895 case CEPH_MDS_SESSION_CLOSING: return "closing";
896 case CEPH_MDS_SESSION_CLOSED: return "closed";
897 case CEPH_MDS_SESSION_RESTARTING: return "restarting";
898 case CEPH_MDS_SESSION_RECONNECTING: return "reconnecting";
899 case CEPH_MDS_SESSION_REJECTED: return "rejected";
900 default: return "???";
901 }
902 }
903
904 struct ceph_mds_session *ceph_get_mds_session(struct ceph_mds_session *s)
905 {
906 if (refcount_inc_not_zero(&s->s_ref))
907 return s;
908 return NULL;
909 }
910
911 void ceph_put_mds_session(struct ceph_mds_session *s)
912 {
913 if (IS_ERR_OR_NULL(s))
914 return;
915
916 if (refcount_dec_and_test(&s->s_ref)) {
917 if (s->s_auth.authorizer)
918 ceph_auth_destroy_authorizer(s->s_auth.authorizer);
919 WARN_ON(mutex_is_locked(&s->s_mutex));
920 xa_destroy(&s->s_delegated_inos);
921 kfree(s);
922 }
923 }
924
925 /*
926 * called under mdsc->mutex
927 */
928 struct ceph_mds_session *__ceph_lookup_mds_session(struct ceph_mds_client *mdsc,
929 int mds)
930 {
931 if (mds >= mdsc->max_sessions || !mdsc->sessions[mds])
932 return NULL;
933 return ceph_get_mds_session(mdsc->sessions[mds]);
934 }
935
936 static bool __have_session(struct ceph_mds_client *mdsc, int mds)
937 {
938 if (mds >= mdsc->max_sessions || !mdsc->sessions[mds])
939 return false;
940 else
941 return true;
942 }
943
944 static int __verify_registered_session(struct ceph_mds_client *mdsc,
945 struct ceph_mds_session *s)
946 {
947 if (s->s_mds >= mdsc->max_sessions ||
948 mdsc->sessions[s->s_mds] != s)
949 return -ENOENT;
950 return 0;
951 }
952
953 /*
954 * create+register a new session for given mds.
955 * called under mdsc->mutex.
956 */
957 static struct ceph_mds_session *register_session(struct ceph_mds_client *mdsc,
958 int mds)
959 {
960 struct ceph_mds_session *s;
961
962 if (READ_ONCE(mdsc->fsc->mount_state) == CEPH_MOUNT_FENCE_IO)
963 return ERR_PTR(-EIO);
964
965 if (mds >= mdsc->mdsmap->possible_max_rank)
966 return ERR_PTR(-EINVAL);
967
968 s = kzalloc(sizeof(*s), GFP_NOFS);
969 if (!s)
970 return ERR_PTR(-ENOMEM);
971
972 if (mds >= mdsc->max_sessions) {
973 int newmax = 1 << get_count_order(mds + 1);
974 struct ceph_mds_session **sa;
975
976 dout("%s: realloc to %d\n", __func__, newmax);
977 sa = kcalloc(newmax, sizeof(void *), GFP_NOFS);
978 if (!sa)
979 goto fail_realloc;
980 if (mdsc->sessions) {
981 memcpy(sa, mdsc->sessions,
982 mdsc->max_sessions * sizeof(void *));
983 kfree(mdsc->sessions);
984 }
985 mdsc->sessions = sa;
986 mdsc->max_sessions = newmax;
987 }
988
989 dout("%s: mds%d\n", __func__, mds);
990 s->s_mdsc = mdsc;
991 s->s_mds = mds;
992 s->s_state = CEPH_MDS_SESSION_NEW;
993 mutex_init(&s->s_mutex);
994
995 ceph_con_init(&s->s_con, s, &mds_con_ops, &mdsc->fsc->client->msgr);
996
997 atomic_set(&s->s_cap_gen, 1);
998 s->s_cap_ttl = jiffies - 1;
999
1000 spin_lock_init(&s->s_cap_lock);
1001 INIT_LIST_HEAD(&s->s_caps);
1002 refcount_set(&s->s_ref, 1);
1003 INIT_LIST_HEAD(&s->s_waiting);
1004 INIT_LIST_HEAD(&s->s_unsafe);
1005 xa_init(&s->s_delegated_inos);
1006 INIT_LIST_HEAD(&s->s_cap_releases);
1007 INIT_WORK(&s->s_cap_release_work, ceph_cap_release_work);
1008
1009 INIT_LIST_HEAD(&s->s_cap_dirty);
1010 INIT_LIST_HEAD(&s->s_cap_flushing);
1011
1012 mdsc->sessions[mds] = s;
1013 atomic_inc(&mdsc->num_sessions);
1014 refcount_inc(&s->s_ref); /* one ref to sessions[], one to caller */
1015
1016 ceph_con_open(&s->s_con, CEPH_ENTITY_TYPE_MDS, mds,
1017 ceph_mdsmap_get_addr(mdsc->mdsmap, mds));
1018
1019 return s;
1020
1021 fail_realloc:
1022 kfree(s);
1023 return ERR_PTR(-ENOMEM);
1024 }
1025
1026 /*
1027 * called under mdsc->mutex
1028 */
1029 static void __unregister_session(struct ceph_mds_client *mdsc,
1030 struct ceph_mds_session *s)
1031 {
1032 dout("__unregister_session mds%d %p\n", s->s_mds, s);
1033 BUG_ON(mdsc->sessions[s->s_mds] != s);
1034 mdsc->sessions[s->s_mds] = NULL;
1035 ceph_con_close(&s->s_con);
1036 ceph_put_mds_session(s);
1037 atomic_dec(&mdsc->num_sessions);
1038 }
1039
1040 /*
1041 * drop session refs in request.
1042 *
1043 * should be last request ref, or hold mdsc->mutex
1044 */
1045 static void put_request_session(struct ceph_mds_request *req)
1046 {
1047 if (req->r_session) {
1048 ceph_put_mds_session(req->r_session);
1049 req->r_session = NULL;
1050 }
1051 }
1052
1053 void ceph_mdsc_iterate_sessions(struct ceph_mds_client *mdsc,
1054 void (*cb)(struct ceph_mds_session *),
1055 bool check_state)
1056 {
1057 int mds;
1058
1059 mutex_lock(&mdsc->mutex);
1060 for (mds = 0; mds < mdsc->max_sessions; ++mds) {
1061 struct ceph_mds_session *s;
1062
1063 s = __ceph_lookup_mds_session(mdsc, mds);
1064 if (!s)
1065 continue;
1066
1067 if (check_state && !check_session_state(s)) {
1068 ceph_put_mds_session(s);
1069 continue;
1070 }
1071
1072 mutex_unlock(&mdsc->mutex);
1073 cb(s);
1074 ceph_put_mds_session(s);
1075 mutex_lock(&mdsc->mutex);
1076 }
1077 mutex_unlock(&mdsc->mutex);
1078 }
1079
1080 void ceph_mdsc_release_request(struct kref *kref)
1081 {
1082 struct ceph_mds_request *req = container_of(kref,
1083 struct ceph_mds_request,
1084 r_kref);
1085 ceph_mdsc_release_dir_caps_no_check(req);
1086 destroy_reply_info(&req->r_reply_info);
1087 if (req->r_request)
1088 ceph_msg_put(req->r_request);
1089 if (req->r_reply)
1090 ceph_msg_put(req->r_reply);
1091 if (req->r_inode) {
1092 ceph_put_cap_refs(ceph_inode(req->r_inode), CEPH_CAP_PIN);
1093 iput(req->r_inode);
1094 }
1095 if (req->r_parent) {
1096 ceph_put_cap_refs(ceph_inode(req->r_parent), CEPH_CAP_PIN);
1097 iput(req->r_parent);
1098 }
1099 iput(req->r_target_inode);
1100 iput(req->r_new_inode);
1101 if (req->r_dentry)
1102 dput(req->r_dentry);
1103 if (req->r_old_dentry)
1104 dput(req->r_old_dentry);
1105 if (req->r_old_dentry_dir) {
1106 /*
1107 * track (and drop pins for) r_old_dentry_dir
1108 * separately, since r_old_dentry's d_parent may have
1109 * changed between the dir mutex being dropped and
1110 * this request being freed.
1111 */
1112 ceph_put_cap_refs(ceph_inode(req->r_old_dentry_dir),
1113 CEPH_CAP_PIN);
1114 iput(req->r_old_dentry_dir);
1115 }
1116 kfree(req->r_path1);
1117 kfree(req->r_path2);
1118 put_cred(req->r_cred);
1119 if (req->r_pagelist)
1120 ceph_pagelist_release(req->r_pagelist);
1121 kfree(req->r_fscrypt_auth);
1122 kfree(req->r_altname);
1123 put_request_session(req);
1124 ceph_unreserve_caps(req->r_mdsc, &req->r_caps_reservation);
1125 WARN_ON_ONCE(!list_empty(&req->r_wait));
1126 kmem_cache_free(ceph_mds_request_cachep, req);
1127 }
1128
1129 DEFINE_RB_FUNCS(request, struct ceph_mds_request, r_tid, r_node)
1130
1131 /*
1132 * lookup session, bump ref if found.
1133 *
1134 * called under mdsc->mutex.
1135 */
1136 static struct ceph_mds_request *
1137 lookup_get_request(struct ceph_mds_client *mdsc, u64 tid)
1138 {
1139 struct ceph_mds_request *req;
1140
1141 req = lookup_request(&mdsc->request_tree, tid);
1142 if (req)
1143 ceph_mdsc_get_request(req);
1144
1145 return req;
1146 }
1147
1148 /*
1149 * Register an in-flight request, and assign a tid. Link to directory
1150 * are modifying (if any).
1151 *
1152 * Called under mdsc->mutex.
1153 */
1154 static void __register_request(struct ceph_mds_client *mdsc,
1155 struct ceph_mds_request *req,
1156 struct inode *dir)
1157 {
1158 int ret = 0;
1159
1160 req->r_tid = ++mdsc->last_tid;
1161 if (req->r_num_caps) {
1162 ret = ceph_reserve_caps(mdsc, &req->r_caps_reservation,
1163 req->r_num_caps);
1164 if (ret < 0) {
1165 pr_err("__register_request %p "
1166 "failed to reserve caps: %d\n", req, ret);
1167 /* set req->r_err to fail early from __do_request */
1168 req->r_err = ret;
1169 return;
1170 }
1171 }
1172 dout("__register_request %p tid %lld\n", req, req->r_tid);
1173 ceph_mdsc_get_request(req);
1174 insert_request(&mdsc->request_tree, req);
1175
1176 req->r_cred = get_current_cred();
1177
1178 if (mdsc->oldest_tid == 0 && req->r_op != CEPH_MDS_OP_SETFILELOCK)
1179 mdsc->oldest_tid = req->r_tid;
1180
1181 if (dir) {
1182 struct ceph_inode_info *ci = ceph_inode(dir);
1183
1184 ihold(dir);
1185 req->r_unsafe_dir = dir;
1186 spin_lock(&ci->i_unsafe_lock);
1187 list_add_tail(&req->r_unsafe_dir_item, &ci->i_unsafe_dirops);
1188 spin_unlock(&ci->i_unsafe_lock);
1189 }
1190 }
1191
1192 static void __unregister_request(struct ceph_mds_client *mdsc,
1193 struct ceph_mds_request *req)
1194 {
1195 dout("__unregister_request %p tid %lld\n", req, req->r_tid);
1196
1197 /* Never leave an unregistered request on an unsafe list! */
1198 list_del_init(&req->r_unsafe_item);
1199
1200 if (req->r_tid == mdsc->oldest_tid) {
1201 struct rb_node *p = rb_next(&req->r_node);
1202 mdsc->oldest_tid = 0;
1203 while (p) {
1204 struct ceph_mds_request *next_req =
1205 rb_entry(p, struct ceph_mds_request, r_node);
1206 if (next_req->r_op != CEPH_MDS_OP_SETFILELOCK) {
1207 mdsc->oldest_tid = next_req->r_tid;
1208 break;
1209 }
1210 p = rb_next(p);
1211 }
1212 }
1213
1214 erase_request(&mdsc->request_tree, req);
1215
1216 if (req->r_unsafe_dir) {
1217 struct ceph_inode_info *ci = ceph_inode(req->r_unsafe_dir);
1218 spin_lock(&ci->i_unsafe_lock);
1219 list_del_init(&req->r_unsafe_dir_item);
1220 spin_unlock(&ci->i_unsafe_lock);
1221 }
1222 if (req->r_target_inode &&
1223 test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags)) {
1224 struct ceph_inode_info *ci = ceph_inode(req->r_target_inode);
1225 spin_lock(&ci->i_unsafe_lock);
1226 list_del_init(&req->r_unsafe_target_item);
1227 spin_unlock(&ci->i_unsafe_lock);
1228 }
1229
1230 if (req->r_unsafe_dir) {
1231 iput(req->r_unsafe_dir);
1232 req->r_unsafe_dir = NULL;
1233 }
1234
1235 complete_all(&req->r_safe_completion);
1236
1237 ceph_mdsc_put_request(req);
1238 }
1239
1240 /*
1241 * Walk back up the dentry tree until we hit a dentry representing a
1242 * non-snapshot inode. We do this using the rcu_read_lock (which must be held
1243 * when calling this) to ensure that the objects won't disappear while we're
1244 * working with them. Once we hit a candidate dentry, we attempt to take a
1245 * reference to it, and return that as the result.
1246 */
1247 static struct inode *get_nonsnap_parent(struct dentry *dentry)
1248 {
1249 struct inode *inode = NULL;
1250
1251 while (dentry && !IS_ROOT(dentry)) {
1252 inode = d_inode_rcu(dentry);
1253 if (!inode || ceph_snap(inode) == CEPH_NOSNAP)
1254 break;
1255 dentry = dentry->d_parent;
1256 }
1257 if (inode)
1258 inode = igrab(inode);
1259 return inode;
1260 }
1261
1262 /*
1263 * Choose mds to send request to next. If there is a hint set in the
1264 * request (e.g., due to a prior forward hint from the mds), use that.
1265 * Otherwise, consult frag tree and/or caps to identify the
1266 * appropriate mds. If all else fails, choose randomly.
1267 *
1268 * Called under mdsc->mutex.
1269 */
1270 static int __choose_mds(struct ceph_mds_client *mdsc,
1271 struct ceph_mds_request *req,
1272 bool *random)
1273 {
1274 struct inode *inode;
1275 struct ceph_inode_info *ci;
1276 struct ceph_cap *cap;
1277 int mode = req->r_direct_mode;
1278 int mds = -1;
1279 u32 hash = req->r_direct_hash;
1280 bool is_hash = test_bit(CEPH_MDS_R_DIRECT_IS_HASH, &req->r_req_flags);
1281
1282 if (random)
1283 *random = false;
1284
1285 /*
1286 * is there a specific mds we should try? ignore hint if we have
1287 * no session and the mds is not up (active or recovering).
1288 */
1289 if (req->r_resend_mds >= 0 &&
1290 (__have_session(mdsc, req->r_resend_mds) ||
1291 ceph_mdsmap_get_state(mdsc->mdsmap, req->r_resend_mds) > 0)) {
1292 dout("%s using resend_mds mds%d\n", __func__,
1293 req->r_resend_mds);
1294 return req->r_resend_mds;
1295 }
1296
1297 if (mode == USE_RANDOM_MDS)
1298 goto random;
1299
1300 inode = NULL;
1301 if (req->r_inode) {
1302 if (ceph_snap(req->r_inode) != CEPH_SNAPDIR) {
1303 inode = req->r_inode;
1304 ihold(inode);
1305 } else {
1306 /* req->r_dentry is non-null for LSSNAP request */
1307 rcu_read_lock();
1308 inode = get_nonsnap_parent(req->r_dentry);
1309 rcu_read_unlock();
1310 dout("%s using snapdir's parent %p\n", __func__, inode);
1311 }
1312 } else if (req->r_dentry) {
1313 /* ignore race with rename; old or new d_parent is okay */
1314 struct dentry *parent;
1315 struct inode *dir;
1316
1317 rcu_read_lock();
1318 parent = READ_ONCE(req->r_dentry->d_parent);
1319 dir = req->r_parent ? : d_inode_rcu(parent);
1320
1321 if (!dir || dir->i_sb != mdsc->fsc->sb) {
1322 /* not this fs or parent went negative */
1323 inode = d_inode(req->r_dentry);
1324 if (inode)
1325 ihold(inode);
1326 } else if (ceph_snap(dir) != CEPH_NOSNAP) {
1327 /* direct snapped/virtual snapdir requests
1328 * based on parent dir inode */
1329 inode = get_nonsnap_parent(parent);
1330 dout("%s using nonsnap parent %p\n", __func__, inode);
1331 } else {
1332 /* dentry target */
1333 inode = d_inode(req->r_dentry);
1334 if (!inode || mode == USE_AUTH_MDS) {
1335 /* dir + name */
1336 inode = igrab(dir);
1337 hash = ceph_dentry_hash(dir, req->r_dentry);
1338 is_hash = true;
1339 } else {
1340 ihold(inode);
1341 }
1342 }
1343 rcu_read_unlock();
1344 }
1345
1346 dout("%s %p is_hash=%d (0x%x) mode %d\n", __func__, inode, (int)is_hash,
1347 hash, mode);
1348 if (!inode)
1349 goto random;
1350 ci = ceph_inode(inode);
1351
1352 if (is_hash && S_ISDIR(inode->i_mode)) {
1353 struct ceph_inode_frag frag;
1354 int found;
1355
1356 ceph_choose_frag(ci, hash, &frag, &found);
1357 if (found) {
1358 if (mode == USE_ANY_MDS && frag.ndist > 0) {
1359 u8 r;
1360
1361 /* choose a random replica */
1362 get_random_bytes(&r, 1);
1363 r %= frag.ndist;
1364 mds = frag.dist[r];
1365 dout("%s %p %llx.%llx frag %u mds%d (%d/%d)\n",
1366 __func__, inode, ceph_vinop(inode),
1367 frag.frag, mds, (int)r, frag.ndist);
1368 if (ceph_mdsmap_get_state(mdsc->mdsmap, mds) >=
1369 CEPH_MDS_STATE_ACTIVE &&
1370 !ceph_mdsmap_is_laggy(mdsc->mdsmap, mds))
1371 goto out;
1372 }
1373
1374 /* since this file/dir wasn't known to be
1375 * replicated, then we want to look for the
1376 * authoritative mds. */
1377 if (frag.mds >= 0) {
1378 /* choose auth mds */
1379 mds = frag.mds;
1380 dout("%s %p %llx.%llx frag %u mds%d (auth)\n",
1381 __func__, inode, ceph_vinop(inode),
1382 frag.frag, mds);
1383 if (ceph_mdsmap_get_state(mdsc->mdsmap, mds) >=
1384 CEPH_MDS_STATE_ACTIVE) {
1385 if (!ceph_mdsmap_is_laggy(mdsc->mdsmap,
1386 mds))
1387 goto out;
1388 }
1389 }
1390 mode = USE_AUTH_MDS;
1391 }
1392 }
1393
1394 spin_lock(&ci->i_ceph_lock);
1395 cap = NULL;
1396 if (mode == USE_AUTH_MDS)
1397 cap = ci->i_auth_cap;
1398 if (!cap && !RB_EMPTY_ROOT(&ci->i_caps))
1399 cap = rb_entry(rb_first(&ci->i_caps), struct ceph_cap, ci_node);
1400 if (!cap) {
1401 spin_unlock(&ci->i_ceph_lock);
1402 iput(inode);
1403 goto random;
1404 }
1405 mds = cap->session->s_mds;
1406 dout("%s %p %llx.%llx mds%d (%scap %p)\n", __func__,
1407 inode, ceph_vinop(inode), mds,
1408 cap == ci->i_auth_cap ? "auth " : "", cap);
1409 spin_unlock(&ci->i_ceph_lock);
1410 out:
1411 iput(inode);
1412 return mds;
1413
1414 random:
1415 if (random)
1416 *random = true;
1417
1418 mds = ceph_mdsmap_get_random_mds(mdsc->mdsmap);
1419 dout("%s chose random mds%d\n", __func__, mds);
1420 return mds;
1421 }
1422
1423
1424 /*
1425 * session messages
1426 */
1427 struct ceph_msg *ceph_create_session_msg(u32 op, u64 seq)
1428 {
1429 struct ceph_msg *msg;
1430 struct ceph_mds_session_head *h;
1431
1432 msg = ceph_msg_new(CEPH_MSG_CLIENT_SESSION, sizeof(*h), GFP_NOFS,
1433 false);
1434 if (!msg) {
1435 pr_err("ENOMEM creating session %s msg\n",
1436 ceph_session_op_name(op));
1437 return NULL;
1438 }
1439 h = msg->front.iov_base;
1440 h->op = cpu_to_le32(op);
1441 h->seq = cpu_to_le64(seq);
1442
1443 return msg;
1444 }
1445
1446 static const unsigned char feature_bits[] = CEPHFS_FEATURES_CLIENT_SUPPORTED;
1447 #define FEATURE_BYTES(c) (DIV_ROUND_UP((size_t)feature_bits[c - 1] + 1, 64) * 8)
1448 static int encode_supported_features(void **p, void *end)
1449 {
1450 static const size_t count = ARRAY_SIZE(feature_bits);
1451
1452 if (count > 0) {
1453 size_t i;
1454 size_t size = FEATURE_BYTES(count);
1455 unsigned long bit;
1456
1457 if (WARN_ON_ONCE(*p + 4 + size > end))
1458 return -ERANGE;
1459
1460 ceph_encode_32(p, size);
1461 memset(*p, 0, size);
1462 for (i = 0; i < count; i++) {
1463 bit = feature_bits[i];
1464 ((unsigned char *)(*p))[bit / 8] |= BIT(bit % 8);
1465 }
1466 *p += size;
1467 } else {
1468 if (WARN_ON_ONCE(*p + 4 > end))
1469 return -ERANGE;
1470
1471 ceph_encode_32(p, 0);
1472 }
1473
1474 return 0;
1475 }
1476
1477 static const unsigned char metric_bits[] = CEPHFS_METRIC_SPEC_CLIENT_SUPPORTED;
1478 #define METRIC_BYTES(cnt) (DIV_ROUND_UP((size_t)metric_bits[cnt - 1] + 1, 64) * 8)
1479 static int encode_metric_spec(void **p, void *end)
1480 {
1481 static const size_t count = ARRAY_SIZE(metric_bits);
1482
1483 /* header */
1484 if (WARN_ON_ONCE(*p + 2 > end))
1485 return -ERANGE;
1486
1487 ceph_encode_8(p, 1); /* version */
1488 ceph_encode_8(p, 1); /* compat */
1489
1490 if (count > 0) {
1491 size_t i;
1492 size_t size = METRIC_BYTES(count);
1493
1494 if (WARN_ON_ONCE(*p + 4 + 4 + size > end))
1495 return -ERANGE;
1496
1497 /* metric spec info length */
1498 ceph_encode_32(p, 4 + size);
1499
1500 /* metric spec */
1501 ceph_encode_32(p, size);
1502 memset(*p, 0, size);
1503 for (i = 0; i < count; i++)
1504 ((unsigned char *)(*p))[i / 8] |= BIT(metric_bits[i] % 8);
1505 *p += size;
1506 } else {
1507 if (WARN_ON_ONCE(*p + 4 + 4 > end))
1508 return -ERANGE;
1509
1510 /* metric spec info length */
1511 ceph_encode_32(p, 4);
1512 /* metric spec */
1513 ceph_encode_32(p, 0);
1514 }
1515
1516 return 0;
1517 }
1518
1519 /*
1520 * session message, specialization for CEPH_SESSION_REQUEST_OPEN
1521 * to include additional client metadata fields.
1522 */
1523 static struct ceph_msg *create_session_open_msg(struct ceph_mds_client *mdsc, u64 seq)
1524 {
1525 struct ceph_msg *msg;
1526 struct ceph_mds_session_head *h;
1527 int i;
1528 int extra_bytes = 0;
1529 int metadata_key_count = 0;
1530 struct ceph_options *opt = mdsc->fsc->client->options;
1531 struct ceph_mount_options *fsopt = mdsc->fsc->mount_options;
1532 size_t size, count;
1533 void *p, *end;
1534 int ret;
1535
1536 const char* metadata[][2] = {
1537 {"hostname", mdsc->nodename},
1538 {"kernel_version", init_utsname()->release},
1539 {"entity_id", opt->name ? : ""},
1540 {"root", fsopt->server_path ? : "/"},
1541 {NULL, NULL}
1542 };
1543
1544 /* Calculate serialized length of metadata */
1545 extra_bytes = 4; /* map length */
1546 for (i = 0; metadata[i][0]; ++i) {
1547 extra_bytes += 8 + strlen(metadata[i][0]) +
1548 strlen(metadata[i][1]);
1549 metadata_key_count++;
1550 }
1551
1552 /* supported feature */
1553 size = 0;
1554 count = ARRAY_SIZE(feature_bits);
1555 if (count > 0)
1556 size = FEATURE_BYTES(count);
1557 extra_bytes += 4 + size;
1558
1559 /* metric spec */
1560 size = 0;
1561 count = ARRAY_SIZE(metric_bits);
1562 if (count > 0)
1563 size = METRIC_BYTES(count);
1564 extra_bytes += 2 + 4 + 4 + size;
1565
1566 /* Allocate the message */
1567 msg = ceph_msg_new(CEPH_MSG_CLIENT_SESSION, sizeof(*h) + extra_bytes,
1568 GFP_NOFS, false);
1569 if (!msg) {
1570 pr_err("ENOMEM creating session open msg\n");
1571 return ERR_PTR(-ENOMEM);
1572 }
1573 p = msg->front.iov_base;
1574 end = p + msg->front.iov_len;
1575
1576 h = p;
1577 h->op = cpu_to_le32(CEPH_SESSION_REQUEST_OPEN);
1578 h->seq = cpu_to_le64(seq);
1579
1580 /*
1581 * Serialize client metadata into waiting buffer space, using
1582 * the format that userspace expects for map<string, string>
1583 *
1584 * ClientSession messages with metadata are v4
1585 */
1586 msg->hdr.version = cpu_to_le16(4);
1587 msg->hdr.compat_version = cpu_to_le16(1);
1588
1589 /* The write pointer, following the session_head structure */
1590 p += sizeof(*h);
1591
1592 /* Number of entries in the map */
1593 ceph_encode_32(&p, metadata_key_count);
1594
1595 /* Two length-prefixed strings for each entry in the map */
1596 for (i = 0; metadata[i][0]; ++i) {
1597 size_t const key_len = strlen(metadata[i][0]);
1598 size_t const val_len = strlen(metadata[i][1]);
1599
1600 ceph_encode_32(&p, key_len);
1601 memcpy(p, metadata[i][0], key_len);
1602 p += key_len;
1603 ceph_encode_32(&p, val_len);
1604 memcpy(p, metadata[i][1], val_len);
1605 p += val_len;
1606 }
1607
1608 ret = encode_supported_features(&p, end);
1609 if (ret) {
1610 pr_err("encode_supported_features failed!\n");
1611 ceph_msg_put(msg);
1612 return ERR_PTR(ret);
1613 }
1614
1615 ret = encode_metric_spec(&p, end);
1616 if (ret) {
1617 pr_err("encode_metric_spec failed!\n");
1618 ceph_msg_put(msg);
1619 return ERR_PTR(ret);
1620 }
1621
1622 msg->front.iov_len = p - msg->front.iov_base;
1623 msg->hdr.front_len = cpu_to_le32(msg->front.iov_len);
1624
1625 return msg;
1626 }
1627
1628 /*
1629 * send session open request.
1630 *
1631 * called under mdsc->mutex
1632 */
1633 static int __open_session(struct ceph_mds_client *mdsc,
1634 struct ceph_mds_session *session)
1635 {
1636 struct ceph_msg *msg;
1637 int mstate;
1638 int mds = session->s_mds;
1639
1640 if (READ_ONCE(mdsc->fsc->mount_state) == CEPH_MOUNT_FENCE_IO)
1641 return -EIO;
1642
1643 /* wait for mds to go active? */
1644 mstate = ceph_mdsmap_get_state(mdsc->mdsmap, mds);
1645 dout("open_session to mds%d (%s)\n", mds,
1646 ceph_mds_state_name(mstate));
1647 session->s_state = CEPH_MDS_SESSION_OPENING;
1648 session->s_renew_requested = jiffies;
1649
1650 /* send connect message */
1651 msg = create_session_open_msg(mdsc, session->s_seq);
1652 if (IS_ERR(msg))
1653 return PTR_ERR(msg);
1654 ceph_con_send(&session->s_con, msg);
1655 return 0;
1656 }
1657
1658 /*
1659 * open sessions for any export targets for the given mds
1660 *
1661 * called under mdsc->mutex
1662 */
1663 static struct ceph_mds_session *
1664 __open_export_target_session(struct ceph_mds_client *mdsc, int target)
1665 {
1666 struct ceph_mds_session *session;
1667 int ret;
1668
1669 session = __ceph_lookup_mds_session(mdsc, target);
1670 if (!session) {
1671 session = register_session(mdsc, target);
1672 if (IS_ERR(session))
1673 return session;
1674 }
1675 if (session->s_state == CEPH_MDS_SESSION_NEW ||
1676 session->s_state == CEPH_MDS_SESSION_CLOSING) {
1677 ret = __open_session(mdsc, session);
1678 if (ret)
1679 return ERR_PTR(ret);
1680 }
1681
1682 return session;
1683 }
1684
1685 struct ceph_mds_session *
1686 ceph_mdsc_open_export_target_session(struct ceph_mds_client *mdsc, int target)
1687 {
1688 struct ceph_mds_session *session;
1689
1690 dout("open_export_target_session to mds%d\n", target);
1691
1692 mutex_lock(&mdsc->mutex);
1693 session = __open_export_target_session(mdsc, target);
1694 mutex_unlock(&mdsc->mutex);
1695
1696 return session;
1697 }
1698
1699 static void __open_export_target_sessions(struct ceph_mds_client *mdsc,
1700 struct ceph_mds_session *session)
1701 {
1702 struct ceph_mds_info *mi;
1703 struct ceph_mds_session *ts;
1704 int i, mds = session->s_mds;
1705
1706 if (mds >= mdsc->mdsmap->possible_max_rank)
1707 return;
1708
1709 mi = &mdsc->mdsmap->m_info[mds];
1710 dout("open_export_target_sessions for mds%d (%d targets)\n",
1711 session->s_mds, mi->num_export_targets);
1712
1713 for (i = 0; i < mi->num_export_targets; i++) {
1714 ts = __open_export_target_session(mdsc, mi->export_targets[i]);
1715 ceph_put_mds_session(ts);
1716 }
1717 }
1718
1719 void ceph_mdsc_open_export_target_sessions(struct ceph_mds_client *mdsc,
1720 struct ceph_mds_session *session)
1721 {
1722 mutex_lock(&mdsc->mutex);
1723 __open_export_target_sessions(mdsc, session);
1724 mutex_unlock(&mdsc->mutex);
1725 }
1726
1727 /*
1728 * session caps
1729 */
1730
1731 static void detach_cap_releases(struct ceph_mds_session *session,
1732 struct list_head *target)
1733 {
1734 lockdep_assert_held(&session->s_cap_lock);
1735
1736 list_splice_init(&session->s_cap_releases, target);
1737 session->s_num_cap_releases = 0;
1738 dout("dispose_cap_releases mds%d\n", session->s_mds);
1739 }
1740
1741 static void dispose_cap_releases(struct ceph_mds_client *mdsc,
1742 struct list_head *dispose)
1743 {
1744 while (!list_empty(dispose)) {
1745 struct ceph_cap *cap;
1746 /* zero out the in-progress message */
1747 cap = list_first_entry(dispose, struct ceph_cap, session_caps);
1748 list_del(&cap->session_caps);
1749 ceph_put_cap(mdsc, cap);
1750 }
1751 }
1752
1753 static void cleanup_session_requests(struct ceph_mds_client *mdsc,
1754 struct ceph_mds_session *session)
1755 {
1756 struct ceph_mds_request *req;
1757 struct rb_node *p;
1758
1759 dout("cleanup_session_requests mds%d\n", session->s_mds);
1760 mutex_lock(&mdsc->mutex);
1761 while (!list_empty(&session->s_unsafe)) {
1762 req = list_first_entry(&session->s_unsafe,
1763 struct ceph_mds_request, r_unsafe_item);
1764 pr_warn_ratelimited(" dropping unsafe request %llu\n",
1765 req->r_tid);
1766 if (req->r_target_inode)
1767 mapping_set_error(req->r_target_inode->i_mapping, -EIO);
1768 if (req->r_unsafe_dir)
1769 mapping_set_error(req->r_unsafe_dir->i_mapping, -EIO);
1770 __unregister_request(mdsc, req);
1771 }
1772 /* zero r_attempts, so kick_requests() will re-send requests */
1773 p = rb_first(&mdsc->request_tree);
1774 while (p) {
1775 req = rb_entry(p, struct ceph_mds_request, r_node);
1776 p = rb_next(p);
1777 if (req->r_session &&
1778 req->r_session->s_mds == session->s_mds)
1779 req->r_attempts = 0;
1780 }
1781 mutex_unlock(&mdsc->mutex);
1782 }
1783
1784 /*
1785 * Helper to safely iterate over all caps associated with a session, with
1786 * special care taken to handle a racing __ceph_remove_cap().
1787 *
1788 * Caller must hold session s_mutex.
1789 */
1790 int ceph_iterate_session_caps(struct ceph_mds_session *session,
1791 int (*cb)(struct inode *, int mds, void *),
1792 void *arg)
1793 {
1794 struct list_head *p;
1795 struct ceph_cap *cap;
1796 struct inode *inode, *last_inode = NULL;
1797 struct ceph_cap *old_cap = NULL;
1798 int ret;
1799
1800 dout("iterate_session_caps %p mds%d\n", session, session->s_mds);
1801 spin_lock(&session->s_cap_lock);
1802 p = session->s_caps.next;
1803 while (p != &session->s_caps) {
1804 int mds;
1805
1806 cap = list_entry(p, struct ceph_cap, session_caps);
1807 inode = igrab(&cap->ci->netfs.inode);
1808 if (!inode) {
1809 p = p->next;
1810 continue;
1811 }
1812 session->s_cap_iterator = cap;
1813 mds = cap->mds;
1814 spin_unlock(&session->s_cap_lock);
1815
1816 if (last_inode) {
1817 iput(last_inode);
1818 last_inode = NULL;
1819 }
1820 if (old_cap) {
1821 ceph_put_cap(session->s_mdsc, old_cap);
1822 old_cap = NULL;
1823 }
1824
1825 ret = cb(inode, mds, arg);
1826 last_inode = inode;
1827
1828 spin_lock(&session->s_cap_lock);
1829 p = p->next;
1830 if (!cap->ci) {
1831 dout("iterate_session_caps finishing cap %p removal\n",
1832 cap);
1833 BUG_ON(cap->session != session);
1834 cap->session = NULL;
1835 list_del_init(&cap->session_caps);
1836 session->s_nr_caps--;
1837 atomic64_dec(&session->s_mdsc->metric.total_caps);
1838 if (cap->queue_release)
1839 __ceph_queue_cap_release(session, cap);
1840 else
1841 old_cap = cap; /* put_cap it w/o locks held */
1842 }
1843 if (ret < 0)
1844 goto out;
1845 }
1846 ret = 0;
1847 out:
1848 session->s_cap_iterator = NULL;
1849 spin_unlock(&session->s_cap_lock);
1850
1851 iput(last_inode);
1852 if (old_cap)
1853 ceph_put_cap(session->s_mdsc, old_cap);
1854
1855 return ret;
1856 }
1857
1858 static int remove_session_caps_cb(struct inode *inode, int mds, void *arg)
1859 {
1860 struct ceph_inode_info *ci = ceph_inode(inode);
1861 bool invalidate = false;
1862 struct ceph_cap *cap;
1863 int iputs = 0;
1864
1865 spin_lock(&ci->i_ceph_lock);
1866 cap = __get_cap_for_mds(ci, mds);
1867 if (cap) {
1868 dout(" removing cap %p, ci is %p, inode is %p\n",
1869 cap, ci, &ci->netfs.inode);
1870
1871 iputs = ceph_purge_inode_cap(inode, cap, &invalidate);
1872 }
1873 spin_unlock(&ci->i_ceph_lock);
1874
1875 if (cap)
1876 wake_up_all(&ci->i_cap_wq);
1877 if (invalidate)
1878 ceph_queue_invalidate(inode);
1879 while (iputs--)
1880 iput(inode);
1881 return 0;
1882 }
1883
1884 /*
1885 * caller must hold session s_mutex
1886 */
1887 static void remove_session_caps(struct ceph_mds_session *session)
1888 {
1889 struct ceph_fs_client *fsc = session->s_mdsc->fsc;
1890 struct super_block *sb = fsc->sb;
1891 LIST_HEAD(dispose);
1892
1893 dout("remove_session_caps on %p\n", session);
1894 ceph_iterate_session_caps(session, remove_session_caps_cb, fsc);
1895
1896 wake_up_all(&fsc->mdsc->cap_flushing_wq);
1897
1898 spin_lock(&session->s_cap_lock);
1899 if (session->s_nr_caps > 0) {
1900 struct inode *inode;
1901 struct ceph_cap *cap, *prev = NULL;
1902 struct ceph_vino vino;
1903 /*
1904 * iterate_session_caps() skips inodes that are being
1905 * deleted, we need to wait until deletions are complete.
1906 * __wait_on_freeing_inode() is designed for the job,
1907 * but it is not exported, so use lookup inode function
1908 * to access it.
1909 */
1910 while (!list_empty(&session->s_caps)) {
1911 cap = list_entry(session->s_caps.next,
1912 struct ceph_cap, session_caps);
1913 if (cap == prev)
1914 break;
1915 prev = cap;
1916 vino = cap->ci->i_vino;
1917 spin_unlock(&session->s_cap_lock);
1918
1919 inode = ceph_find_inode(sb, vino);
1920 iput(inode);
1921
1922 spin_lock(&session->s_cap_lock);
1923 }
1924 }
1925
1926 // drop cap expires and unlock s_cap_lock
1927 detach_cap_releases(session, &dispose);
1928
1929 BUG_ON(session->s_nr_caps > 0);
1930 BUG_ON(!list_empty(&session->s_cap_flushing));
1931 spin_unlock(&session->s_cap_lock);
1932 dispose_cap_releases(session->s_mdsc, &dispose);
1933 }
1934
1935 enum {
1936 RECONNECT,
1937 RENEWCAPS,
1938 FORCE_RO,
1939 };
1940
1941 /*
1942 * wake up any threads waiting on this session's caps. if the cap is
1943 * old (didn't get renewed on the client reconnect), remove it now.
1944 *
1945 * caller must hold s_mutex.
1946 */
1947 static int wake_up_session_cb(struct inode *inode, int mds, void *arg)
1948 {
1949 struct ceph_inode_info *ci = ceph_inode(inode);
1950 unsigned long ev = (unsigned long)arg;
1951
1952 if (ev == RECONNECT) {
1953 spin_lock(&ci->i_ceph_lock);
1954 ci->i_wanted_max_size = 0;
1955 ci->i_requested_max_size = 0;
1956 spin_unlock(&ci->i_ceph_lock);
1957 } else if (ev == RENEWCAPS) {
1958 struct ceph_cap *cap;
1959
1960 spin_lock(&ci->i_ceph_lock);
1961 cap = __get_cap_for_mds(ci, mds);
1962 /* mds did not re-issue stale cap */
1963 if (cap && cap->cap_gen < atomic_read(&cap->session->s_cap_gen))
1964 cap->issued = cap->implemented = CEPH_CAP_PIN;
1965 spin_unlock(&ci->i_ceph_lock);
1966 } else if (ev == FORCE_RO) {
1967 }
1968 wake_up_all(&ci->i_cap_wq);
1969 return 0;
1970 }
1971
1972 static void wake_up_session_caps(struct ceph_mds_session *session, int ev)
1973 {
1974 dout("wake_up_session_caps %p mds%d\n", session, session->s_mds);
1975 ceph_iterate_session_caps(session, wake_up_session_cb,
1976 (void *)(unsigned long)ev);
1977 }
1978
1979 /*
1980 * Send periodic message to MDS renewing all currently held caps. The
1981 * ack will reset the expiration for all caps from this session.
1982 *
1983 * caller holds s_mutex
1984 */
1985 static int send_renew_caps(struct ceph_mds_client *mdsc,
1986 struct ceph_mds_session *session)
1987 {
1988 struct ceph_msg *msg;
1989 int state;
1990
1991 if (time_after_eq(jiffies, session->s_cap_ttl) &&
1992 time_after_eq(session->s_cap_ttl, session->s_renew_requested))
1993 pr_info("mds%d caps stale\n", session->s_mds);
1994 session->s_renew_requested = jiffies;
1995
1996 /* do not try to renew caps until a recovering mds has reconnected
1997 * with its clients. */
1998 state = ceph_mdsmap_get_state(mdsc->mdsmap, session->s_mds);
1999 if (state < CEPH_MDS_STATE_RECONNECT) {
2000 dout("send_renew_caps ignoring mds%d (%s)\n",
2001 session->s_mds, ceph_mds_state_name(state));
2002 return 0;
2003 }
2004
2005 dout("send_renew_caps to mds%d (%s)\n", session->s_mds,
2006 ceph_mds_state_name(state));
2007 msg = ceph_create_session_msg(CEPH_SESSION_REQUEST_RENEWCAPS,
2008 ++session->s_renew_seq);
2009 if (!msg)
2010 return -ENOMEM;
2011 ceph_con_send(&session->s_con, msg);
2012 return 0;
2013 }
2014
2015 static int send_flushmsg_ack(struct ceph_mds_client *mdsc,
2016 struct ceph_mds_session *session, u64 seq)
2017 {
2018 struct ceph_msg *msg;
2019
2020 dout("send_flushmsg_ack to mds%d (%s)s seq %lld\n",
2021 session->s_mds, ceph_session_state_name(session->s_state), seq);
2022 msg = ceph_create_session_msg(CEPH_SESSION_FLUSHMSG_ACK, seq);
2023 if (!msg)
2024 return -ENOMEM;
2025 ceph_con_send(&session->s_con, msg);
2026 return 0;
2027 }
2028
2029
2030 /*
2031 * Note new cap ttl, and any transition from stale -> not stale (fresh?).
2032 *
2033 * Called under session->s_mutex
2034 */
2035 static void renewed_caps(struct ceph_mds_client *mdsc,
2036 struct ceph_mds_session *session, int is_renew)
2037 {
2038 int was_stale;
2039 int wake = 0;
2040
2041 spin_lock(&session->s_cap_lock);
2042 was_stale = is_renew && time_after_eq(jiffies, session->s_cap_ttl);
2043
2044 session->s_cap_ttl = session->s_renew_requested +
2045 mdsc->mdsmap->m_session_timeout*HZ;
2046
2047 if (was_stale) {
2048 if (time_before(jiffies, session->s_cap_ttl)) {
2049 pr_info("mds%d caps renewed\n", session->s_mds);
2050 wake = 1;
2051 } else {
2052 pr_info("mds%d caps still stale\n", session->s_mds);
2053 }
2054 }
2055 dout("renewed_caps mds%d ttl now %lu, was %s, now %s\n",
2056 session->s_mds, session->s_cap_ttl, was_stale ? "stale" : "fresh",
2057 time_before(jiffies, session->s_cap_ttl) ? "stale" : "fresh");
2058 spin_unlock(&session->s_cap_lock);
2059
2060 if (wake)
2061 wake_up_session_caps(session, RENEWCAPS);
2062 }
2063
2064 /*
2065 * send a session close request
2066 */
2067 static int request_close_session(struct ceph_mds_session *session)
2068 {
2069 struct ceph_msg *msg;
2070
2071 dout("request_close_session mds%d state %s seq %lld\n",
2072 session->s_mds, ceph_session_state_name(session->s_state),
2073 session->s_seq);
2074 msg = ceph_create_session_msg(CEPH_SESSION_REQUEST_CLOSE,
2075 session->s_seq);
2076 if (!msg)
2077 return -ENOMEM;
2078 ceph_con_send(&session->s_con, msg);
2079 return 1;
2080 }
2081
2082 /*
2083 * Called with s_mutex held.
2084 */
2085 static int __close_session(struct ceph_mds_client *mdsc,
2086 struct ceph_mds_session *session)
2087 {
2088 if (session->s_state >= CEPH_MDS_SESSION_CLOSING)
2089 return 0;
2090 session->s_state = CEPH_MDS_SESSION_CLOSING;
2091 return request_close_session(session);
2092 }
2093
2094 static bool drop_negative_children(struct dentry *dentry)
2095 {
2096 struct dentry *child;
2097 bool all_negative = true;
2098
2099 if (!d_is_dir(dentry))
2100 goto out;
2101
2102 spin_lock(&dentry->d_lock);
2103 list_for_each_entry(child, &dentry->d_subdirs, d_child) {
2104 if (d_really_is_positive(child)) {
2105 all_negative = false;
2106 break;
2107 }
2108 }
2109 spin_unlock(&dentry->d_lock);
2110
2111 if (all_negative)
2112 shrink_dcache_parent(dentry);
2113 out:
2114 return all_negative;
2115 }
2116
2117 /*
2118 * Trim old(er) caps.
2119 *
2120 * Because we can't cache an inode without one or more caps, we do
2121 * this indirectly: if a cap is unused, we prune its aliases, at which
2122 * point the inode will hopefully get dropped to.
2123 *
2124 * Yes, this is a bit sloppy. Our only real goal here is to respond to
2125 * memory pressure from the MDS, though, so it needn't be perfect.
2126 */
2127 static int trim_caps_cb(struct inode *inode, int mds, void *arg)
2128 {
2129 int *remaining = arg;
2130 struct ceph_inode_info *ci = ceph_inode(inode);
2131 int used, wanted, oissued, mine;
2132 struct ceph_cap *cap;
2133
2134 if (*remaining <= 0)
2135 return -1;
2136
2137 spin_lock(&ci->i_ceph_lock);
2138 cap = __get_cap_for_mds(ci, mds);
2139 if (!cap) {
2140 spin_unlock(&ci->i_ceph_lock);
2141 return 0;
2142 }
2143 mine = cap->issued | cap->implemented;
2144 used = __ceph_caps_used(ci);
2145 wanted = __ceph_caps_file_wanted(ci);
2146 oissued = __ceph_caps_issued_other(ci, cap);
2147
2148 dout("trim_caps_cb %p cap %p mine %s oissued %s used %s wanted %s\n",
2149 inode, cap, ceph_cap_string(mine), ceph_cap_string(oissued),
2150 ceph_cap_string(used), ceph_cap_string(wanted));
2151 if (cap == ci->i_auth_cap) {
2152 if (ci->i_dirty_caps || ci->i_flushing_caps ||
2153 !list_empty(&ci->i_cap_snaps))
2154 goto out;
2155 if ((used | wanted) & CEPH_CAP_ANY_WR)
2156 goto out;
2157 /* Note: it's possible that i_filelock_ref becomes non-zero
2158 * after dropping auth caps. It doesn't hurt because reply
2159 * of lock mds request will re-add auth caps. */
2160 if (atomic_read(&ci->i_filelock_ref) > 0)
2161 goto out;
2162 }
2163 /* The inode has cached pages, but it's no longer used.
2164 * we can safely drop it */
2165 if (S_ISREG(inode->i_mode) &&
2166 wanted == 0 && used == CEPH_CAP_FILE_CACHE &&
2167 !(oissued & CEPH_CAP_FILE_CACHE)) {
2168 used = 0;
2169 oissued = 0;
2170 }
2171 if ((used | wanted) & ~oissued & mine)
2172 goto out; /* we need these caps */
2173
2174 if (oissued) {
2175 /* we aren't the only cap.. just remove us */
2176 ceph_remove_cap(cap, true);
2177 (*remaining)--;
2178 } else {
2179 struct dentry *dentry;
2180 /* try dropping referring dentries */
2181 spin_unlock(&ci->i_ceph_lock);
2182 dentry = d_find_any_alias(inode);
2183 if (dentry && drop_negative_children(dentry)) {
2184 int count;
2185 dput(dentry);
2186 d_prune_aliases(inode);
2187 count = atomic_read(&inode->i_count);
2188 if (count == 1)
2189 (*remaining)--;
2190 dout("trim_caps_cb %p cap %p pruned, count now %d\n",
2191 inode, cap, count);
2192 } else {
2193 dput(dentry);
2194 }
2195 return 0;
2196 }
2197
2198 out:
2199 spin_unlock(&ci->i_ceph_lock);
2200 return 0;
2201 }
2202
2203 /*
2204 * Trim session cap count down to some max number.
2205 */
2206 int ceph_trim_caps(struct ceph_mds_client *mdsc,
2207 struct ceph_mds_session *session,
2208 int max_caps)
2209 {
2210 int trim_caps = session->s_nr_caps - max_caps;
2211
2212 dout("trim_caps mds%d start: %d / %d, trim %d\n",
2213 session->s_mds, session->s_nr_caps, max_caps, trim_caps);
2214 if (trim_caps > 0) {
2215 int remaining = trim_caps;
2216
2217 ceph_iterate_session_caps(session, trim_caps_cb, &remaining);
2218 dout("trim_caps mds%d done: %d / %d, trimmed %d\n",
2219 session->s_mds, session->s_nr_caps, max_caps,
2220 trim_caps - remaining);
2221 }
2222
2223 ceph_flush_cap_releases(mdsc, session);
2224 return 0;
2225 }
2226
2227 static int check_caps_flush(struct ceph_mds_client *mdsc,
2228 u64 want_flush_tid)
2229 {
2230 int ret = 1;
2231
2232 spin_lock(&mdsc->cap_dirty_lock);
2233 if (!list_empty(&mdsc->cap_flush_list)) {
2234 struct ceph_cap_flush *cf =
2235 list_first_entry(&mdsc->cap_flush_list,
2236 struct ceph_cap_flush, g_list);
2237 if (cf->tid <= want_flush_tid) {
2238 dout("check_caps_flush still flushing tid "
2239 "%llu <= %llu\n", cf->tid, want_flush_tid);
2240 ret = 0;
2241 }
2242 }
2243 spin_unlock(&mdsc->cap_dirty_lock);
2244 return ret;
2245 }
2246
2247 /*
2248 * flush all dirty inode data to disk.
2249 *
2250 * returns true if we've flushed through want_flush_tid
2251 */
2252 static void wait_caps_flush(struct ceph_mds_client *mdsc,
2253 u64 want_flush_tid)
2254 {
2255 dout("check_caps_flush want %llu\n", want_flush_tid);
2256
2257 wait_event(mdsc->cap_flushing_wq,
2258 check_caps_flush(mdsc, want_flush_tid));
2259
2260 dout("check_caps_flush ok, flushed thru %llu\n", want_flush_tid);
2261 }
2262
2263 /*
2264 * called under s_mutex
2265 */
2266 static void ceph_send_cap_releases(struct ceph_mds_client *mdsc,
2267 struct ceph_mds_session *session)
2268 {
2269 struct ceph_msg *msg = NULL;
2270 struct ceph_mds_cap_release *head;
2271 struct ceph_mds_cap_item *item;
2272 struct ceph_osd_client *osdc = &mdsc->fsc->client->osdc;
2273 struct ceph_cap *cap;
2274 LIST_HEAD(tmp_list);
2275 int num_cap_releases;
2276 __le32 barrier, *cap_barrier;
2277
2278 down_read(&osdc->lock);
2279 barrier = cpu_to_le32(osdc->epoch_barrier);
2280 up_read(&osdc->lock);
2281
2282 spin_lock(&session->s_cap_lock);
2283 again:
2284 list_splice_init(&session->s_cap_releases, &tmp_list);
2285 num_cap_releases = session->s_num_cap_releases;
2286 session->s_num_cap_releases = 0;
2287 spin_unlock(&session->s_cap_lock);
2288
2289 while (!list_empty(&tmp_list)) {
2290 if (!msg) {
2291 msg = ceph_msg_new(CEPH_MSG_CLIENT_CAPRELEASE,
2292 PAGE_SIZE, GFP_NOFS, false);
2293 if (!msg)
2294 goto out_err;
2295 head = msg->front.iov_base;
2296 head->num = cpu_to_le32(0);
2297 msg->front.iov_len = sizeof(*head);
2298
2299 msg->hdr.version = cpu_to_le16(2);
2300 msg->hdr.compat_version = cpu_to_le16(1);
2301 }
2302
2303 cap = list_first_entry(&tmp_list, struct ceph_cap,
2304 session_caps);
2305 list_del(&cap->session_caps);
2306 num_cap_releases--;
2307
2308 head = msg->front.iov_base;
2309 put_unaligned_le32(get_unaligned_le32(&head->num) + 1,
2310 &head->num);
2311 item = msg->front.iov_base + msg->front.iov_len;
2312 item->ino = cpu_to_le64(cap->cap_ino);
2313 item->cap_id = cpu_to_le64(cap->cap_id);
2314 item->migrate_seq = cpu_to_le32(cap->mseq);
2315 item->seq = cpu_to_le32(cap->issue_seq);
2316 msg->front.iov_len += sizeof(*item);
2317
2318 ceph_put_cap(mdsc, cap);
2319
2320 if (le32_to_cpu(head->num) == CEPH_CAPS_PER_RELEASE) {
2321 // Append cap_barrier field
2322 cap_barrier = msg->front.iov_base + msg->front.iov_len;
2323 *cap_barrier = barrier;
2324 msg->front.iov_len += sizeof(*cap_barrier);
2325
2326 msg->hdr.front_len = cpu_to_le32(msg->front.iov_len);
2327 dout("send_cap_releases mds%d %p\n", session->s_mds, msg);
2328 ceph_con_send(&session->s_con, msg);
2329 msg = NULL;
2330 }
2331 }
2332
2333 BUG_ON(num_cap_releases != 0);
2334
2335 spin_lock(&session->s_cap_lock);
2336 if (!list_empty(&session->s_cap_releases))
2337 goto again;
2338 spin_unlock(&session->s_cap_lock);
2339
2340 if (msg) {
2341 // Append cap_barrier field
2342 cap_barrier = msg->front.iov_base + msg->front.iov_len;
2343 *cap_barrier = barrier;
2344 msg->front.iov_len += sizeof(*cap_barrier);
2345
2346 msg->hdr.front_len = cpu_to_le32(msg->front.iov_len);
2347 dout("send_cap_releases mds%d %p\n", session->s_mds, msg);
2348 ceph_con_send(&session->s_con, msg);
2349 }
2350 return;
2351 out_err:
2352 pr_err("send_cap_releases mds%d, failed to allocate message\n",
2353 session->s_mds);
2354 spin_lock(&session->s_cap_lock);
2355 list_splice(&tmp_list, &session->s_cap_releases);
2356 session->s_num_cap_releases += num_cap_releases;
2357 spin_unlock(&session->s_cap_lock);
2358 }
2359
2360 static void ceph_cap_release_work(struct work_struct *work)
2361 {
2362 struct ceph_mds_session *session =
2363 container_of(work, struct ceph_mds_session, s_cap_release_work);
2364
2365 mutex_lock(&session->s_mutex);
2366 if (session->s_state == CEPH_MDS_SESSION_OPEN ||
2367 session->s_state == CEPH_MDS_SESSION_HUNG)
2368 ceph_send_cap_releases(session->s_mdsc, session);
2369 mutex_unlock(&session->s_mutex);
2370 ceph_put_mds_session(session);
2371 }
2372
2373 void ceph_flush_cap_releases(struct ceph_mds_client *mdsc,
2374 struct ceph_mds_session *session)
2375 {
2376 if (mdsc->stopping)
2377 return;
2378
2379 ceph_get_mds_session(session);
2380 if (queue_work(mdsc->fsc->cap_wq,
2381 &session->s_cap_release_work)) {
2382 dout("cap release work queued\n");
2383 } else {
2384 ceph_put_mds_session(session);
2385 dout("failed to queue cap release work\n");
2386 }
2387 }
2388
2389 /*
2390 * caller holds session->s_cap_lock
2391 */
2392 void __ceph_queue_cap_release(struct ceph_mds_session *session,
2393 struct ceph_cap *cap)
2394 {
2395 list_add_tail(&cap->session_caps, &session->s_cap_releases);
2396 session->s_num_cap_releases++;
2397
2398 if (!(session->s_num_cap_releases % CEPH_CAPS_PER_RELEASE))
2399 ceph_flush_cap_releases(session->s_mdsc, session);
2400 }
2401
2402 static void ceph_cap_reclaim_work(struct work_struct *work)
2403 {
2404 struct ceph_mds_client *mdsc =
2405 container_of(work, struct ceph_mds_client, cap_reclaim_work);
2406 int ret = ceph_trim_dentries(mdsc);
2407 if (ret == -EAGAIN)
2408 ceph_queue_cap_reclaim_work(mdsc);
2409 }
2410
2411 void ceph_queue_cap_reclaim_work(struct ceph_mds_client *mdsc)
2412 {
2413 if (mdsc->stopping)
2414 return;
2415
2416 if (queue_work(mdsc->fsc->cap_wq, &mdsc->cap_reclaim_work)) {
2417 dout("caps reclaim work queued\n");
2418 } else {
2419 dout("failed to queue caps release work\n");
2420 }
2421 }
2422
2423 void ceph_reclaim_caps_nr(struct ceph_mds_client *mdsc, int nr)
2424 {
2425 int val;
2426 if (!nr)
2427 return;
2428 val = atomic_add_return(nr, &mdsc->cap_reclaim_pending);
2429 if ((val % CEPH_CAPS_PER_RELEASE) < nr) {
2430 atomic_set(&mdsc->cap_reclaim_pending, 0);
2431 ceph_queue_cap_reclaim_work(mdsc);
2432 }
2433 }
2434
2435 /*
2436 * requests
2437 */
2438
2439 int ceph_alloc_readdir_reply_buffer(struct ceph_mds_request *req,
2440 struct inode *dir)
2441 {
2442 struct ceph_inode_info *ci = ceph_inode(dir);
2443 struct ceph_mds_reply_info_parsed *rinfo = &req->r_reply_info;
2444 struct ceph_mount_options *opt = req->r_mdsc->fsc->mount_options;
2445 size_t size = sizeof(struct ceph_mds_reply_dir_entry);
2446 unsigned int num_entries;
2447 int order;
2448
2449 spin_lock(&ci->i_ceph_lock);
2450 num_entries = ci->i_files + ci->i_subdirs;
2451 spin_unlock(&ci->i_ceph_lock);
2452 num_entries = max(num_entries, 1U);
2453 num_entries = min(num_entries, opt->max_readdir);
2454
2455 order = get_order(size * num_entries);
2456 while (order >= 0) {
2457 rinfo->dir_entries = (void*)__get_free_pages(GFP_KERNEL |
2458 __GFP_NOWARN |
2459 __GFP_ZERO,
2460 order);
2461 if (rinfo->dir_entries)
2462 break;
2463 order--;
2464 }
2465 if (!rinfo->dir_entries)
2466 return -ENOMEM;
2467
2468 num_entries = (PAGE_SIZE << order) / size;
2469 num_entries = min(num_entries, opt->max_readdir);
2470
2471 rinfo->dir_buf_size = PAGE_SIZE << order;
2472 req->r_num_caps = num_entries + 1;
2473 req->r_args.readdir.max_entries = cpu_to_le32(num_entries);
2474 req->r_args.readdir.max_bytes = cpu_to_le32(opt->max_readdir_bytes);
2475 return 0;
2476 }
2477
2478 /*
2479 * Create an mds request.
2480 */
2481 struct ceph_mds_request *
2482 ceph_mdsc_create_request(struct ceph_mds_client *mdsc, int op, int mode)
2483 {
2484 struct ceph_mds_request *req;
2485
2486 req = kmem_cache_zalloc(ceph_mds_request_cachep, GFP_NOFS);
2487 if (!req)
2488 return ERR_PTR(-ENOMEM);
2489
2490 mutex_init(&req->r_fill_mutex);
2491 req->r_mdsc = mdsc;
2492 req->r_started = jiffies;
2493 req->r_start_latency = ktime_get();
2494 req->r_resend_mds = -1;
2495 INIT_LIST_HEAD(&req->r_unsafe_dir_item);
2496 INIT_LIST_HEAD(&req->r_unsafe_target_item);
2497 req->r_fmode = -1;
2498 req->r_feature_needed = -1;
2499 kref_init(&req->r_kref);
2500 RB_CLEAR_NODE(&req->r_node);
2501 INIT_LIST_HEAD(&req->r_wait);
2502 init_completion(&req->r_completion);
2503 init_completion(&req->r_safe_completion);
2504 INIT_LIST_HEAD(&req->r_unsafe_item);
2505
2506 ktime_get_coarse_real_ts64(&req->r_stamp);
2507
2508 req->r_op = op;
2509 req->r_direct_mode = mode;
2510 return req;
2511 }
2512
2513 /*
2514 * return oldest (lowest) request, tid in request tree, 0 if none.
2515 *
2516 * called under mdsc->mutex.
2517 */
2518 static struct ceph_mds_request *__get_oldest_req(struct ceph_mds_client *mdsc)
2519 {
2520 if (RB_EMPTY_ROOT(&mdsc->request_tree))
2521 return NULL;
2522 return rb_entry(rb_first(&mdsc->request_tree),
2523 struct ceph_mds_request, r_node);
2524 }
2525
2526 static inline u64 __get_oldest_tid(struct ceph_mds_client *mdsc)
2527 {
2528 return mdsc->oldest_tid;
2529 }
2530
2531 #if IS_ENABLED(CONFIG_FS_ENCRYPTION)
2532 static u8 *get_fscrypt_altname(const struct ceph_mds_request *req, u32 *plen)
2533 {
2534 struct inode *dir = req->r_parent;
2535 struct dentry *dentry = req->r_dentry;
2536 u8 *cryptbuf = NULL;
2537 u32 len = 0;
2538 int ret = 0;
2539
2540 /* only encode if we have parent and dentry */
2541 if (!dir || !dentry)
2542 goto success;
2543
2544 /* No-op unless this is encrypted */
2545 if (!IS_ENCRYPTED(dir))
2546 goto success;
2547
2548 ret = ceph_fscrypt_prepare_readdir(dir);
2549 if (ret < 0)
2550 return ERR_PTR(ret);
2551
2552 /* No key? Just ignore it. */
2553 if (!fscrypt_has_encryption_key(dir))
2554 goto success;
2555
2556 if (!fscrypt_fname_encrypted_size(dir, dentry->d_name.len, NAME_MAX,
2557 &len)) {
2558 WARN_ON_ONCE(1);
2559 return ERR_PTR(-ENAMETOOLONG);
2560 }
2561
2562 /* No need to append altname if name is short enough */
2563 if (len <= CEPH_NOHASH_NAME_MAX) {
2564 len = 0;
2565 goto success;
2566 }
2567
2568 cryptbuf = kmalloc(len, GFP_KERNEL);
2569 if (!cryptbuf)
2570 return ERR_PTR(-ENOMEM);
2571
2572 ret = fscrypt_fname_encrypt(dir, &dentry->d_name, cryptbuf, len);
2573 if (ret) {
2574 kfree(cryptbuf);
2575 return ERR_PTR(ret);
2576 }
2577 success:
2578 *plen = len;
2579 return cryptbuf;
2580 }
2581 #else
2582 static u8 *get_fscrypt_altname(const struct ceph_mds_request *req, u32 *plen)
2583 {
2584 *plen = 0;
2585 return NULL;
2586 }
2587 #endif
2588
2589 /**
2590 * ceph_mdsc_build_path - build a path string to a given dentry
2591 * @dentry: dentry to which path should be built
2592 * @plen: returned length of string
2593 * @pbase: returned base inode number
2594 * @for_wire: is this path going to be sent to the MDS?
2595 *
2596 * Build a string that represents the path to the dentry. This is mostly called
2597 * for two different purposes:
2598 *
2599 * 1) we need to build a path string to send to the MDS (for_wire == true)
2600 * 2) we need a path string for local presentation (e.g. debugfs)
2601 * (for_wire == false)
2602 *
2603 * The path is built in reverse, starting with the dentry. Walk back up toward
2604 * the root, building the path until the first non-snapped inode is reached
2605 * (for_wire) or the root inode is reached (!for_wire).
2606 *
2607 * Encode hidden .snap dirs as a double /, i.e.
2608 * foo/.snap/bar -> foo//bar
2609 */
2610 char *ceph_mdsc_build_path(struct dentry *dentry, int *plen, u64 *pbase,
2611 int for_wire)
2612 {
2613 struct dentry *cur;
2614 struct inode *inode;
2615 char *path;
2616 int pos;
2617 unsigned seq;
2618 u64 base;
2619
2620 if (!dentry)
2621 return ERR_PTR(-EINVAL);
2622
2623 path = __getname();
2624 if (!path)
2625 return ERR_PTR(-ENOMEM);
2626 retry:
2627 pos = PATH_MAX - 1;
2628 path[pos] = '\0';
2629
2630 seq = read_seqbegin(&rename_lock);
2631 cur = dget(dentry);
2632 for (;;) {
2633 struct dentry *parent;
2634
2635 spin_lock(&cur->d_lock);
2636 inode = d_inode(cur);
2637 if (inode && ceph_snap(inode) == CEPH_SNAPDIR) {
2638 dout("build_path path+%d: %p SNAPDIR\n",
2639 pos, cur);
2640 spin_unlock(&cur->d_lock);
2641 parent = dget_parent(cur);
2642 } else if (for_wire && inode && dentry != cur &&
2643 ceph_snap(inode) == CEPH_NOSNAP) {
2644 spin_unlock(&cur->d_lock);
2645 pos++; /* get rid of any prepended '/' */
2646 break;
2647 } else if (!for_wire || !IS_ENCRYPTED(d_inode(cur->d_parent))) {
2648 pos -= cur->d_name.len;
2649 if (pos < 0) {
2650 spin_unlock(&cur->d_lock);
2651 break;
2652 }
2653 memcpy(path + pos, cur->d_name.name, cur->d_name.len);
2654 spin_unlock(&cur->d_lock);
2655 parent = dget_parent(cur);
2656 } else {
2657 int len, ret;
2658 char buf[NAME_MAX];
2659
2660 /*
2661 * Proactively copy name into buf, in case we need to
2662 * present it as-is.
2663 */
2664 memcpy(buf, cur->d_name.name, cur->d_name.len);
2665 len = cur->d_name.len;
2666 spin_unlock(&cur->d_lock);
2667 parent = dget_parent(cur);
2668
2669 ret = ceph_fscrypt_prepare_readdir(d_inode(parent));
2670 if (ret < 0) {
2671 dput(parent);
2672 dput(cur);
2673 return ERR_PTR(ret);
2674 }
2675
2676 if (fscrypt_has_encryption_key(d_inode(parent))) {
2677 len = ceph_encode_encrypted_fname(d_inode(parent),
2678 cur, buf);
2679 if (len < 0) {
2680 dput(parent);
2681 dput(cur);
2682 return ERR_PTR(len);
2683 }
2684 }
2685 pos -= len;
2686 if (pos < 0) {
2687 dput(parent);
2688 break;
2689 }
2690 memcpy(path + pos, buf, len);
2691 }
2692 dput(cur);
2693 cur = parent;
2694
2695 /* Are we at the root? */
2696 if (IS_ROOT(cur))
2697 break;
2698
2699 /* Are we out of buffer? */
2700 if (--pos < 0)
2701 break;
2702
2703 path[pos] = '/';
2704 }
2705 inode = d_inode(cur);
2706 base = inode ? ceph_ino(inode) : 0;
2707 dput(cur);
2708
2709 if (read_seqretry(&rename_lock, seq))
2710 goto retry;
2711
2712 if (pos < 0) {
2713 /*
2714 * A rename didn't occur, but somehow we didn't end up where
2715 * we thought we would. Throw a warning and try again.
2716 */
2717 pr_warn("build_path did not end path lookup where expected (pos = %d)\n",
2718 pos);
2719 goto retry;
2720 }
2721
2722 *pbase = base;
2723 *plen = PATH_MAX - 1 - pos;
2724 dout("build_path on %p %d built %llx '%.*s'\n",
2725 dentry, d_count(dentry), base, *plen, path + pos);
2726 return path + pos;
2727 }
2728
2729 static int build_dentry_path(struct dentry *dentry, struct inode *dir,
2730 const char **ppath, int *ppathlen, u64 *pino,
2731 bool *pfreepath, bool parent_locked)
2732 {
2733 char *path;
2734
2735 rcu_read_lock();
2736 if (!dir)
2737 dir = d_inode_rcu(dentry->d_parent);
2738 if (dir && parent_locked && ceph_snap(dir) == CEPH_NOSNAP &&
2739 !IS_ENCRYPTED(dir)) {
2740 *pino = ceph_ino(dir);
2741 rcu_read_unlock();
2742 *ppath = dentry->d_name.name;
2743 *ppathlen = dentry->d_name.len;
2744 return 0;
2745 }
2746 rcu_read_unlock();
2747 path = ceph_mdsc_build_path(dentry, ppathlen, pino, 1);
2748 if (IS_ERR(path))
2749 return PTR_ERR(path);
2750 *ppath = path;
2751 *pfreepath = true;
2752 return 0;
2753 }
2754
2755 static int build_inode_path(struct inode *inode,
2756 const char **ppath, int *ppathlen, u64 *pino,
2757 bool *pfreepath)
2758 {
2759 struct dentry *dentry;
2760 char *path;
2761
2762 if (ceph_snap(inode) == CEPH_NOSNAP) {
2763 *pino = ceph_ino(inode);
2764 *ppathlen = 0;
2765 return 0;
2766 }
2767 dentry = d_find_alias(inode);
2768 path = ceph_mdsc_build_path(dentry, ppathlen, pino, 1);
2769 dput(dentry);
2770 if (IS_ERR(path))
2771 return PTR_ERR(path);
2772 *ppath = path;
2773 *pfreepath = true;
2774 return 0;
2775 }
2776
2777 /*
2778 * request arguments may be specified via an inode *, a dentry *, or
2779 * an explicit ino+path.
2780 */
2781 static int set_request_path_attr(struct inode *rinode, struct dentry *rdentry,
2782 struct inode *rdiri, const char *rpath,
2783 u64 rino, const char **ppath, int *pathlen,
2784 u64 *ino, bool *freepath, bool parent_locked)
2785 {
2786 int r = 0;
2787
2788 if (rinode) {
2789 r = build_inode_path(rinode, ppath, pathlen, ino, freepath);
2790 dout(" inode %p %llx.%llx\n", rinode, ceph_ino(rinode),
2791 ceph_snap(rinode));
2792 } else if (rdentry) {
2793 r = build_dentry_path(rdentry, rdiri, ppath, pathlen, ino,
2794 freepath, parent_locked);
2795 dout(" dentry %p %llx/%.*s\n", rdentry, *ino, *pathlen,
2796 *ppath);
2797 } else if (rpath || rino) {
2798 *ino = rino;
2799 *ppath = rpath;
2800 *pathlen = rpath ? strlen(rpath) : 0;
2801 dout(" path %.*s\n", *pathlen, rpath);
2802 }
2803
2804 return r;
2805 }
2806
2807 static void encode_mclientrequest_tail(void **p,
2808 const struct ceph_mds_request *req)
2809 {
2810 struct ceph_timespec ts;
2811 int i;
2812
2813 ceph_encode_timespec64(&ts, &req->r_stamp);
2814 ceph_encode_copy(p, &ts, sizeof(ts));
2815
2816 /* v4: gid_list */
2817 ceph_encode_32(p, req->r_cred->group_info->ngroups);
2818 for (i = 0; i < req->r_cred->group_info->ngroups; i++)
2819 ceph_encode_64(p, from_kgid(&init_user_ns,
2820 req->r_cred->group_info->gid[i]));
2821
2822 /* v5: altname */
2823 ceph_encode_32(p, req->r_altname_len);
2824 ceph_encode_copy(p, req->r_altname, req->r_altname_len);
2825
2826 /* v6: fscrypt_auth and fscrypt_file */
2827 if (req->r_fscrypt_auth) {
2828 u32 authlen = ceph_fscrypt_auth_len(req->r_fscrypt_auth);
2829
2830 ceph_encode_32(p, authlen);
2831 ceph_encode_copy(p, req->r_fscrypt_auth, authlen);
2832 } else {
2833 ceph_encode_32(p, 0);
2834 }
2835 if (test_bit(CEPH_MDS_R_FSCRYPT_FILE, &req->r_req_flags)) {
2836 ceph_encode_32(p, sizeof(__le64));
2837 ceph_encode_64(p, req->r_fscrypt_file);
2838 } else {
2839 ceph_encode_32(p, 0);
2840 }
2841 }
2842
2843 static struct ceph_mds_request_head_legacy *
2844 find_legacy_request_head(void *p, u64 features)
2845 {
2846 bool legacy = !(features & CEPH_FEATURE_FS_BTIME);
2847 struct ceph_mds_request_head_old *ohead;
2848
2849 if (legacy)
2850 return (struct ceph_mds_request_head_legacy *)p;
2851 ohead = (struct ceph_mds_request_head_old *)p;
2852 return (struct ceph_mds_request_head_legacy *)&ohead->oldest_client_tid;
2853 }
2854
2855 /*
2856 * called under mdsc->mutex
2857 */
2858 static struct ceph_msg *create_request_message(struct ceph_mds_session *session,
2859 struct ceph_mds_request *req,
2860 bool drop_cap_releases)
2861 {
2862 int mds = session->s_mds;
2863 struct ceph_mds_client *mdsc = session->s_mdsc;
2864 struct ceph_msg *msg;
2865 struct ceph_mds_request_head_legacy *lhead;
2866 const char *path1 = NULL;
2867 const char *path2 = NULL;
2868 u64 ino1 = 0, ino2 = 0;
2869 int pathlen1 = 0, pathlen2 = 0;
2870 bool freepath1 = false, freepath2 = false;
2871 struct dentry *old_dentry = NULL;
2872 int len;
2873 u16 releases;
2874 void *p, *end;
2875 int ret;
2876 bool legacy = !(session->s_con.peer_features & CEPH_FEATURE_FS_BTIME);
2877 bool old_version = !test_bit(CEPHFS_FEATURE_32BITS_RETRY_FWD,
2878 &session->s_features);
2879
2880 ret = set_request_path_attr(req->r_inode, req->r_dentry,
2881 req->r_parent, req->r_path1, req->r_ino1.ino,
2882 &path1, &pathlen1, &ino1, &freepath1,
2883 test_bit(CEPH_MDS_R_PARENT_LOCKED,
2884 &req->r_req_flags));
2885 if (ret < 0) {
2886 msg = ERR_PTR(ret);
2887 goto out;
2888 }
2889
2890 /* If r_old_dentry is set, then assume that its parent is locked */
2891 if (req->r_old_dentry &&
2892 !(req->r_old_dentry->d_flags & DCACHE_DISCONNECTED))
2893 old_dentry = req->r_old_dentry;
2894 ret = set_request_path_attr(NULL, old_dentry,
2895 req->r_old_dentry_dir,
2896 req->r_path2, req->r_ino2.ino,
2897 &path2, &pathlen2, &ino2, &freepath2, true);
2898 if (ret < 0) {
2899 msg = ERR_PTR(ret);
2900 goto out_free1;
2901 }
2902
2903 req->r_altname = get_fscrypt_altname(req, &req->r_altname_len);
2904 if (IS_ERR(req->r_altname)) {
2905 msg = ERR_CAST(req->r_altname);
2906 req->r_altname = NULL;
2907 goto out_free2;
2908 }
2909
2910 /*
2911 * For old cephs without supporting the 32bit retry/fwd feature
2912 * it will copy the raw memories directly when decoding the
2913 * requests. While new cephs will decode the head depending the
2914 * version member, so we need to make sure it will be compatible
2915 * with them both.
2916 */
2917 if (legacy)
2918 len = sizeof(struct ceph_mds_request_head_legacy);
2919 else if (old_version)
2920 len = sizeof(struct ceph_mds_request_head_old);
2921 else
2922 len = sizeof(struct ceph_mds_request_head);
2923
2924 /* filepaths */
2925 len += 2 * (1 + sizeof(u32) + sizeof(u64));
2926 len += pathlen1 + pathlen2;
2927
2928 /* cap releases */
2929 len += sizeof(struct ceph_mds_request_release) *
2930 (!!req->r_inode_drop + !!req->r_dentry_drop +
2931 !!req->r_old_inode_drop + !!req->r_old_dentry_drop);
2932
2933 if (req->r_dentry_drop)
2934 len += pathlen1;
2935 if (req->r_old_dentry_drop)
2936 len += pathlen2;
2937
2938 /* MClientRequest tail */
2939
2940 /* req->r_stamp */
2941 len += sizeof(struct ceph_timespec);
2942
2943 /* gid list */
2944 len += sizeof(u32) + (sizeof(u64) * req->r_cred->group_info->ngroups);
2945
2946 /* alternate name */
2947 len += sizeof(u32) + req->r_altname_len;
2948
2949 /* fscrypt_auth */
2950 len += sizeof(u32); // fscrypt_auth
2951 if (req->r_fscrypt_auth)
2952 len += ceph_fscrypt_auth_len(req->r_fscrypt_auth);
2953
2954 /* fscrypt_file */
2955 len += sizeof(u32);
2956 if (test_bit(CEPH_MDS_R_FSCRYPT_FILE, &req->r_req_flags))
2957 len += sizeof(__le64);
2958
2959 msg = ceph_msg_new2(CEPH_MSG_CLIENT_REQUEST, len, 1, GFP_NOFS, false);
2960 if (!msg) {
2961 msg = ERR_PTR(-ENOMEM);
2962 goto out_free2;
2963 }
2964
2965 msg->hdr.tid = cpu_to_le64(req->r_tid);
2966
2967 lhead = find_legacy_request_head(msg->front.iov_base,
2968 session->s_con.peer_features);
2969
2970 /*
2971 * The ceph_mds_request_head_legacy didn't contain a version field, and
2972 * one was added when we moved the message version from 3->4.
2973 */
2974 if (legacy) {
2975 msg->hdr.version = cpu_to_le16(3);
2976 p = msg->front.iov_base + sizeof(*lhead);
2977 } else if (old_version) {
2978 struct ceph_mds_request_head_old *ohead = msg->front.iov_base;
2979
2980 msg->hdr.version = cpu_to_le16(4);
2981 ohead->version = cpu_to_le16(1);
2982 p = msg->front.iov_base + sizeof(*ohead);
2983 } else {
2984 struct ceph_mds_request_head *nhead = msg->front.iov_base;
2985
2986 msg->hdr.version = cpu_to_le16(6);
2987 nhead->version = cpu_to_le16(CEPH_MDS_REQUEST_HEAD_VERSION);
2988 p = msg->front.iov_base + sizeof(*nhead);
2989 }
2990
2991 end = msg->front.iov_base + msg->front.iov_len;
2992
2993 lhead->mdsmap_epoch = cpu_to_le32(mdsc->mdsmap->m_epoch);
2994 lhead->op = cpu_to_le32(req->r_op);
2995 lhead->caller_uid = cpu_to_le32(from_kuid(&init_user_ns,
2996 req->r_cred->fsuid));
2997 lhead->caller_gid = cpu_to_le32(from_kgid(&init_user_ns,
2998 req->r_cred->fsgid));
2999 lhead->ino = cpu_to_le64(req->r_deleg_ino);
3000 lhead->args = req->r_args;
3001
3002 ceph_encode_filepath(&p, end, ino1, path1);
3003 ceph_encode_filepath(&p, end, ino2, path2);
3004
3005 /* make note of release offset, in case we need to replay */
3006 req->r_request_release_offset = p - msg->front.iov_base;
3007
3008 /* cap releases */
3009 releases = 0;
3010 if (req->r_inode_drop)
3011 releases += ceph_encode_inode_release(&p,
3012 req->r_inode ? req->r_inode : d_inode(req->r_dentry),
3013 mds, req->r_inode_drop, req->r_inode_unless,
3014 req->r_op == CEPH_MDS_OP_READDIR);
3015 if (req->r_dentry_drop) {
3016 ret = ceph_encode_dentry_release(&p, req->r_dentry,
3017 req->r_parent, mds, req->r_dentry_drop,
3018 req->r_dentry_unless);
3019 if (ret < 0)
3020 goto out_err;
3021 releases += ret;
3022 }
3023 if (req->r_old_dentry_drop) {
3024 ret = ceph_encode_dentry_release(&p, req->r_old_dentry,
3025 req->r_old_dentry_dir, mds,
3026 req->r_old_dentry_drop,
3027 req->r_old_dentry_unless);
3028 if (ret < 0)
3029 goto out_err;
3030 releases += ret;
3031 }
3032 if (req->r_old_inode_drop)
3033 releases += ceph_encode_inode_release(&p,
3034 d_inode(req->r_old_dentry),
3035 mds, req->r_old_inode_drop, req->r_old_inode_unless, 0);
3036
3037 if (drop_cap_releases) {
3038 releases = 0;
3039 p = msg->front.iov_base + req->r_request_release_offset;
3040 }
3041
3042 lhead->num_releases = cpu_to_le16(releases);
3043
3044 encode_mclientrequest_tail(&p, req);
3045
3046 if (WARN_ON_ONCE(p > end)) {
3047 ceph_msg_put(msg);
3048 msg = ERR_PTR(-ERANGE);
3049 goto out_free2;
3050 }
3051
3052 msg->front.iov_len = p - msg->front.iov_base;
3053 msg->hdr.front_len = cpu_to_le32(msg->front.iov_len);
3054
3055 if (req->r_pagelist) {
3056 struct ceph_pagelist *pagelist = req->r_pagelist;
3057 ceph_msg_data_add_pagelist(msg, pagelist);
3058 msg->hdr.data_len = cpu_to_le32(pagelist->length);
3059 } else {
3060 msg->hdr.data_len = 0;
3061 }
3062
3063 msg->hdr.data_off = cpu_to_le16(0);
3064
3065 out_free2:
3066 if (freepath2)
3067 ceph_mdsc_free_path((char *)path2, pathlen2);
3068 out_free1:
3069 if (freepath1)
3070 ceph_mdsc_free_path((char *)path1, pathlen1);
3071 out:
3072 return msg;
3073 out_err:
3074 ceph_msg_put(msg);
3075 msg = ERR_PTR(ret);
3076 goto out_free2;
3077 }
3078
3079 /*
3080 * called under mdsc->mutex if error, under no mutex if
3081 * success.
3082 */
3083 static void complete_request(struct ceph_mds_client *mdsc,
3084 struct ceph_mds_request *req)
3085 {
3086 req->r_end_latency = ktime_get();
3087
3088 if (req->r_callback)
3089 req->r_callback(mdsc, req);
3090 complete_all(&req->r_completion);
3091 }
3092
3093 /*
3094 * called under mdsc->mutex
3095 */
3096 static int __prepare_send_request(struct ceph_mds_session *session,
3097 struct ceph_mds_request *req,
3098 bool drop_cap_releases)
3099 {
3100 int mds = session->s_mds;
3101 struct ceph_mds_client *mdsc = session->s_mdsc;
3102 struct ceph_mds_request_head_legacy *lhead;
3103 struct ceph_mds_request_head *nhead;
3104 struct ceph_msg *msg;
3105 int flags = 0, old_max_retry;
3106 bool old_version = !test_bit(CEPHFS_FEATURE_32BITS_RETRY_FWD,
3107 &session->s_features);
3108
3109 /*
3110 * Avoid inifinite retrying after overflow. The client will
3111 * increase the retry count and if the MDS is old version,
3112 * so we limit to retry at most 256 times.
3113 */
3114 if (req->r_attempts) {
3115 old_max_retry = sizeof_field(struct ceph_mds_request_head_old,
3116 num_retry);
3117 old_max_retry = 1 << (old_max_retry * BITS_PER_BYTE);
3118 if ((old_version && req->r_attempts >= old_max_retry) ||
3119 ((uint32_t)req->r_attempts >= U32_MAX)) {
3120 pr_warn_ratelimited("%s request tid %llu seq overflow\n",
3121 __func__, req->r_tid);
3122 return -EMULTIHOP;
3123 }
3124 }
3125
3126 req->r_attempts++;
3127 if (req->r_inode) {
3128 struct ceph_cap *cap =
3129 ceph_get_cap_for_mds(ceph_inode(req->r_inode), mds);
3130
3131 if (cap)
3132 req->r_sent_on_mseq = cap->mseq;
3133 else
3134 req->r_sent_on_mseq = -1;
3135 }
3136 dout("%s %p tid %lld %s (attempt %d)\n", __func__, req,
3137 req->r_tid, ceph_mds_op_name(req->r_op), req->r_attempts);
3138
3139 if (test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags)) {
3140 void *p;
3141
3142 /*
3143 * Replay. Do not regenerate message (and rebuild
3144 * paths, etc.); just use the original message.
3145 * Rebuilding paths will break for renames because
3146 * d_move mangles the src name.
3147 */
3148 msg = req->r_request;
3149 lhead = find_legacy_request_head(msg->front.iov_base,
3150 session->s_con.peer_features);
3151
3152 flags = le32_to_cpu(lhead->flags);
3153 flags |= CEPH_MDS_FLAG_REPLAY;
3154 lhead->flags = cpu_to_le32(flags);
3155
3156 if (req->r_target_inode)
3157 lhead->ino = cpu_to_le64(ceph_ino(req->r_target_inode));
3158
3159 lhead->num_retry = req->r_attempts - 1;
3160 if (!old_version) {
3161 nhead = (struct ceph_mds_request_head*)msg->front.iov_base;
3162 nhead->ext_num_retry = cpu_to_le32(req->r_attempts - 1);
3163 }
3164
3165 /* remove cap/dentry releases from message */
3166 lhead->num_releases = 0;
3167
3168 p = msg->front.iov_base + req->r_request_release_offset;
3169 encode_mclientrequest_tail(&p, req);
3170
3171 msg->front.iov_len = p - msg->front.iov_base;
3172 msg->hdr.front_len = cpu_to_le32(msg->front.iov_len);
3173 return 0;
3174 }
3175
3176 if (req->r_request) {
3177 ceph_msg_put(req->r_request);
3178 req->r_request = NULL;
3179 }
3180 msg = create_request_message(session, req, drop_cap_releases);
3181 if (IS_ERR(msg)) {
3182 req->r_err = PTR_ERR(msg);
3183 return PTR_ERR(msg);
3184 }
3185 req->r_request = msg;
3186
3187 lhead = find_legacy_request_head(msg->front.iov_base,
3188 session->s_con.peer_features);
3189 lhead->oldest_client_tid = cpu_to_le64(__get_oldest_tid(mdsc));
3190 if (test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags))
3191 flags |= CEPH_MDS_FLAG_REPLAY;
3192 if (test_bit(CEPH_MDS_R_ASYNC, &req->r_req_flags))
3193 flags |= CEPH_MDS_FLAG_ASYNC;
3194 if (req->r_parent)
3195 flags |= CEPH_MDS_FLAG_WANT_DENTRY;
3196 lhead->flags = cpu_to_le32(flags);
3197 lhead->num_fwd = req->r_num_fwd;
3198 lhead->num_retry = req->r_attempts - 1;
3199 if (!old_version) {
3200 nhead = (struct ceph_mds_request_head*)msg->front.iov_base;
3201 nhead->ext_num_fwd = cpu_to_le32(req->r_num_fwd);
3202 nhead->ext_num_retry = cpu_to_le32(req->r_attempts - 1);
3203 }
3204
3205 dout(" r_parent = %p\n", req->r_parent);
3206 return 0;
3207 }
3208
3209 /*
3210 * called under mdsc->mutex
3211 */
3212 static int __send_request(struct ceph_mds_session *session,
3213 struct ceph_mds_request *req,
3214 bool drop_cap_releases)
3215 {
3216 int err;
3217
3218 err = __prepare_send_request(session, req, drop_cap_releases);
3219 if (!err) {
3220 ceph_msg_get(req->r_request);
3221 ceph_con_send(&session->s_con, req->r_request);
3222 }
3223
3224 return err;
3225 }
3226
3227 /*
3228 * send request, or put it on the appropriate wait list.
3229 */
3230 static void __do_request(struct ceph_mds_client *mdsc,
3231 struct ceph_mds_request *req)
3232 {
3233 struct ceph_mds_session *session = NULL;
3234 int mds = -1;
3235 int err = 0;
3236 bool random;
3237
3238 if (req->r_err || test_bit(CEPH_MDS_R_GOT_RESULT, &req->r_req_flags)) {
3239 if (test_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags))
3240 __unregister_request(mdsc, req);
3241 return;
3242 }
3243
3244 if (READ_ONCE(mdsc->fsc->mount_state) == CEPH_MOUNT_FENCE_IO) {
3245 dout("do_request metadata corrupted\n");
3246 err = -EIO;
3247 goto finish;
3248 }
3249 if (req->r_timeout &&
3250 time_after_eq(jiffies, req->r_started + req->r_timeout)) {
3251 dout("do_request timed out\n");
3252 err = -ETIMEDOUT;
3253 goto finish;
3254 }
3255 if (READ_ONCE(mdsc->fsc->mount_state) == CEPH_MOUNT_SHUTDOWN) {
3256 dout("do_request forced umount\n");
3257 err = -EIO;
3258 goto finish;
3259 }
3260 if (READ_ONCE(mdsc->fsc->mount_state) == CEPH_MOUNT_MOUNTING) {
3261 if (mdsc->mdsmap_err) {
3262 err = mdsc->mdsmap_err;
3263 dout("do_request mdsmap err %d\n", err);
3264 goto finish;
3265 }
3266 if (mdsc->mdsmap->m_epoch == 0) {
3267 dout("do_request no mdsmap, waiting for map\n");
3268 list_add(&req->r_wait, &mdsc->waiting_for_map);
3269 return;
3270 }
3271 if (!(mdsc->fsc->mount_options->flags &
3272 CEPH_MOUNT_OPT_MOUNTWAIT) &&
3273 !ceph_mdsmap_is_cluster_available(mdsc->mdsmap)) {
3274 err = -EHOSTUNREACH;
3275 goto finish;
3276 }
3277 }
3278
3279 put_request_session(req);
3280
3281 mds = __choose_mds(mdsc, req, &random);
3282 if (mds < 0 ||
3283 ceph_mdsmap_get_state(mdsc->mdsmap, mds) < CEPH_MDS_STATE_ACTIVE) {
3284 if (test_bit(CEPH_MDS_R_ASYNC, &req->r_req_flags)) {
3285 err = -EJUKEBOX;
3286 goto finish;
3287 }
3288 dout("do_request no mds or not active, waiting for map\n");
3289 list_add(&req->r_wait, &mdsc->waiting_for_map);
3290 return;
3291 }
3292
3293 /* get, open session */
3294 session = __ceph_lookup_mds_session(mdsc, mds);
3295 if (!session) {
3296 session = register_session(mdsc, mds);
3297 if (IS_ERR(session)) {
3298 err = PTR_ERR(session);
3299 goto finish;
3300 }
3301 }
3302 req->r_session = ceph_get_mds_session(session);
3303
3304 dout("do_request mds%d session %p state %s\n", mds, session,
3305 ceph_session_state_name(session->s_state));
3306
3307 /*
3308 * The old ceph will crash the MDSs when see unknown OPs
3309 */
3310 if (req->r_feature_needed > 0 &&
3311 !test_bit(req->r_feature_needed, &session->s_features)) {
3312 err = -EOPNOTSUPP;
3313 goto out_session;
3314 }
3315
3316 if (session->s_state != CEPH_MDS_SESSION_OPEN &&
3317 session->s_state != CEPH_MDS_SESSION_HUNG) {
3318 /*
3319 * We cannot queue async requests since the caps and delegated
3320 * inodes are bound to the session. Just return -EJUKEBOX and
3321 * let the caller retry a sync request in that case.
3322 */
3323 if (test_bit(CEPH_MDS_R_ASYNC, &req->r_req_flags)) {
3324 err = -EJUKEBOX;
3325 goto out_session;
3326 }
3327
3328 /*
3329 * If the session has been REJECTED, then return a hard error,
3330 * unless it's a CLEANRECOVER mount, in which case we'll queue
3331 * it to the mdsc queue.
3332 */
3333 if (session->s_state == CEPH_MDS_SESSION_REJECTED) {
3334 if (ceph_test_mount_opt(mdsc->fsc, CLEANRECOVER))
3335 list_add(&req->r_wait, &mdsc->waiting_for_map);
3336 else
3337 err = -EACCES;
3338 goto out_session;
3339 }
3340
3341 if (session->s_state == CEPH_MDS_SESSION_NEW ||
3342 session->s_state == CEPH_MDS_SESSION_CLOSING) {
3343 err = __open_session(mdsc, session);
3344 if (err)
3345 goto out_session;
3346 /* retry the same mds later */
3347 if (random)
3348 req->r_resend_mds = mds;
3349 }
3350 list_add(&req->r_wait, &session->s_waiting);
3351 goto out_session;
3352 }
3353
3354 /* send request */
3355 req->r_resend_mds = -1; /* forget any previous mds hint */
3356
3357 if (req->r_request_started == 0) /* note request start time */
3358 req->r_request_started = jiffies;
3359
3360 /*
3361 * For async create we will choose the auth MDS of frag in parent
3362 * directory to send the request and ususally this works fine, but
3363 * if the migrated the dirtory to another MDS before it could handle
3364 * it the request will be forwarded.
3365 *
3366 * And then the auth cap will be changed.
3367 */
3368 if (test_bit(CEPH_MDS_R_ASYNC, &req->r_req_flags) && req->r_num_fwd) {
3369 struct ceph_dentry_info *di = ceph_dentry(req->r_dentry);
3370 struct ceph_inode_info *ci;
3371 struct ceph_cap *cap;
3372
3373 /*
3374 * The request maybe handled very fast and the new inode
3375 * hasn't been linked to the dentry yet. We need to wait
3376 * for the ceph_finish_async_create(), which shouldn't be
3377 * stuck too long or fail in thoery, to finish when forwarding
3378 * the request.
3379 */
3380 if (!d_inode(req->r_dentry)) {
3381 err = wait_on_bit(&di->flags, CEPH_DENTRY_ASYNC_CREATE_BIT,
3382 TASK_KILLABLE);
3383 if (err) {
3384 mutex_lock(&req->r_fill_mutex);
3385 set_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags);
3386 mutex_unlock(&req->r_fill_mutex);
3387 goto out_session;
3388 }
3389 }
3390
3391 ci = ceph_inode(d_inode(req->r_dentry));
3392
3393 spin_lock(&ci->i_ceph_lock);
3394 cap = ci->i_auth_cap;
3395 if (ci->i_ceph_flags & CEPH_I_ASYNC_CREATE && mds != cap->mds) {
3396 dout("do_request session changed for auth cap %d -> %d\n",
3397 cap->session->s_mds, session->s_mds);
3398
3399 /* Remove the auth cap from old session */
3400 spin_lock(&cap->session->s_cap_lock);
3401 cap->session->s_nr_caps--;
3402 list_del_init(&cap->session_caps);
3403 spin_unlock(&cap->session->s_cap_lock);
3404
3405 /* Add the auth cap to the new session */
3406 cap->mds = mds;
3407 cap->session = session;
3408 spin_lock(&session->s_cap_lock);
3409 session->s_nr_caps++;
3410 list_add_tail(&cap->session_caps, &session->s_caps);
3411 spin_unlock(&session->s_cap_lock);
3412
3413 change_auth_cap_ses(ci, session);
3414 }
3415 spin_unlock(&ci->i_ceph_lock);
3416 }
3417
3418 err = __send_request(session, req, false);
3419
3420 out_session:
3421 ceph_put_mds_session(session);
3422 finish:
3423 if (err) {
3424 dout("__do_request early error %d\n", err);
3425 req->r_err = err;
3426 complete_request(mdsc, req);
3427 __unregister_request(mdsc, req);
3428 }
3429 return;
3430 }
3431
3432 /*
3433 * called under mdsc->mutex
3434 */
3435 static void __wake_requests(struct ceph_mds_client *mdsc,
3436 struct list_head *head)
3437 {
3438 struct ceph_mds_request *req;
3439 LIST_HEAD(tmp_list);
3440
3441 list_splice_init(head, &tmp_list);
3442
3443 while (!list_empty(&tmp_list)) {
3444 req = list_entry(tmp_list.next,
3445 struct ceph_mds_request, r_wait);
3446 list_del_init(&req->r_wait);
3447 dout(" wake request %p tid %llu\n", req, req->r_tid);
3448 __do_request(mdsc, req);
3449 }
3450 }
3451
3452 /*
3453 * Wake up threads with requests pending for @mds, so that they can
3454 * resubmit their requests to a possibly different mds.
3455 */
3456 static void kick_requests(struct ceph_mds_client *mdsc, int mds)
3457 {
3458 struct ceph_mds_request *req;
3459 struct rb_node *p = rb_first(&mdsc->request_tree);
3460
3461 dout("kick_requests mds%d\n", mds);
3462 while (p) {
3463 req = rb_entry(p, struct ceph_mds_request, r_node);
3464 p = rb_next(p);
3465 if (test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags))
3466 continue;
3467 if (req->r_attempts > 0)
3468 continue; /* only new requests */
3469 if (req->r_session &&
3470 req->r_session->s_mds == mds) {
3471 dout(" kicking tid %llu\n", req->r_tid);
3472 list_del_init(&req->r_wait);
3473 __do_request(mdsc, req);
3474 }
3475 }
3476 }
3477
3478 int ceph_mdsc_submit_request(struct ceph_mds_client *mdsc, struct inode *dir,
3479 struct ceph_mds_request *req)
3480 {
3481 int err = 0;
3482
3483 /* take CAP_PIN refs for r_inode, r_parent, r_old_dentry */
3484 if (req->r_inode)
3485 ceph_get_cap_refs(ceph_inode(req->r_inode), CEPH_CAP_PIN);
3486 if (req->r_parent) {
3487 struct ceph_inode_info *ci = ceph_inode(req->r_parent);
3488 int fmode = (req->r_op & CEPH_MDS_OP_WRITE) ?
3489 CEPH_FILE_MODE_WR : CEPH_FILE_MODE_RD;
3490 spin_lock(&ci->i_ceph_lock);
3491 ceph_take_cap_refs(ci, CEPH_CAP_PIN, false);
3492 __ceph_touch_fmode(ci, mdsc, fmode);
3493 spin_unlock(&ci->i_ceph_lock);
3494 }
3495 if (req->r_old_dentry_dir)
3496 ceph_get_cap_refs(ceph_inode(req->r_old_dentry_dir),
3497 CEPH_CAP_PIN);
3498
3499 if (req->r_inode) {
3500 err = ceph_wait_on_async_create(req->r_inode);
3501 if (err) {
3502 dout("%s: wait for async create returned: %d\n",
3503 __func__, err);
3504 return err;
3505 }
3506 }
3507
3508 if (!err && req->r_old_inode) {
3509 err = ceph_wait_on_async_create(req->r_old_inode);
3510 if (err) {
3511 dout("%s: wait for async create returned: %d\n",
3512 __func__, err);
3513 return err;
3514 }
3515 }
3516
3517 dout("submit_request on %p for inode %p\n", req, dir);
3518 mutex_lock(&mdsc->mutex);
3519 __register_request(mdsc, req, dir);
3520 __do_request(mdsc, req);
3521 err = req->r_err;
3522 mutex_unlock(&mdsc->mutex);
3523 return err;
3524 }
3525
3526 int ceph_mdsc_wait_request(struct ceph_mds_client *mdsc,
3527 struct ceph_mds_request *req,
3528 ceph_mds_request_wait_callback_t wait_func)
3529 {
3530 int err;
3531
3532 /* wait */
3533 dout("do_request waiting\n");
3534 if (wait_func) {
3535 err = wait_func(mdsc, req);
3536 } else {
3537 long timeleft = wait_for_completion_killable_timeout(
3538 &req->r_completion,
3539 ceph_timeout_jiffies(req->r_timeout));
3540 if (timeleft > 0)
3541 err = 0;
3542 else if (!timeleft)
3543 err = -ETIMEDOUT; /* timed out */
3544 else
3545 err = timeleft; /* killed */
3546 }
3547 dout("do_request waited, got %d\n", err);
3548 mutex_lock(&mdsc->mutex);
3549
3550 /* only abort if we didn't race with a real reply */
3551 if (test_bit(CEPH_MDS_R_GOT_RESULT, &req->r_req_flags)) {
3552 err = le32_to_cpu(req->r_reply_info.head->result);
3553 } else if (err < 0) {
3554 dout("aborted request %lld with %d\n", req->r_tid, err);
3555
3556 /*
3557 * ensure we aren't running concurrently with
3558 * ceph_fill_trace or ceph_readdir_prepopulate, which
3559 * rely on locks (dir mutex) held by our caller.
3560 */
3561 mutex_lock(&req->r_fill_mutex);
3562 req->r_err = err;
3563 set_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags);
3564 mutex_unlock(&req->r_fill_mutex);
3565
3566 if (req->r_parent &&
3567 (req->r_op & CEPH_MDS_OP_WRITE))
3568 ceph_invalidate_dir_request(req);
3569 } else {
3570 err = req->r_err;
3571 }
3572
3573 mutex_unlock(&mdsc->mutex);
3574 return err;
3575 }
3576
3577 /*
3578 * Synchrously perform an mds request. Take care of all of the
3579 * session setup, forwarding, retry details.
3580 */
3581 int ceph_mdsc_do_request(struct ceph_mds_client *mdsc,
3582 struct inode *dir,
3583 struct ceph_mds_request *req)
3584 {
3585 int err;
3586
3587 dout("do_request on %p\n", req);
3588
3589 /* issue */
3590 err = ceph_mdsc_submit_request(mdsc, dir, req);
3591 if (!err)
3592 err = ceph_mdsc_wait_request(mdsc, req, NULL);
3593 dout("do_request %p done, result %d\n", req, err);
3594 return err;
3595 }
3596
3597 /*
3598 * Invalidate dir's completeness, dentry lease state on an aborted MDS
3599 * namespace request.
3600 */
3601 void ceph_invalidate_dir_request(struct ceph_mds_request *req)
3602 {
3603 struct inode *dir = req->r_parent;
3604 struct inode *old_dir = req->r_old_dentry_dir;
3605
3606 dout("invalidate_dir_request %p %p (complete, lease(s))\n", dir, old_dir);
3607
3608 ceph_dir_clear_complete(dir);
3609 if (old_dir)
3610 ceph_dir_clear_complete(old_dir);
3611 if (req->r_dentry)
3612 ceph_invalidate_dentry_lease(req->r_dentry);
3613 if (req->r_old_dentry)
3614 ceph_invalidate_dentry_lease(req->r_old_dentry);
3615 }
3616
3617 /*
3618 * Handle mds reply.
3619 *
3620 * We take the session mutex and parse and process the reply immediately.
3621 * This preserves the logical ordering of replies, capabilities, etc., sent
3622 * by the MDS as they are applied to our local cache.
3623 */
3624 static void handle_reply(struct ceph_mds_session *session, struct ceph_msg *msg)
3625 {
3626 struct ceph_mds_client *mdsc = session->s_mdsc;
3627 struct ceph_mds_request *req;
3628 struct ceph_mds_reply_head *head = msg->front.iov_base;
3629 struct ceph_mds_reply_info_parsed *rinfo; /* parsed reply info */
3630 struct ceph_snap_realm *realm;
3631 u64 tid;
3632 int err, result;
3633 int mds = session->s_mds;
3634 bool close_sessions = false;
3635
3636 if (msg->front.iov_len < sizeof(*head)) {
3637 pr_err("mdsc_handle_reply got corrupt (short) reply\n");
3638 ceph_msg_dump(msg);
3639 return;
3640 }
3641
3642 /* get request, session */
3643 tid = le64_to_cpu(msg->hdr.tid);
3644 mutex_lock(&mdsc->mutex);
3645 req = lookup_get_request(mdsc, tid);
3646 if (!req) {
3647 dout("handle_reply on unknown tid %llu\n", tid);
3648 mutex_unlock(&mdsc->mutex);
3649 return;
3650 }
3651 dout("handle_reply %p\n", req);
3652
3653 /* correct session? */
3654 if (req->r_session != session) {
3655 pr_err("mdsc_handle_reply got %llu on session mds%d"
3656 " not mds%d\n", tid, session->s_mds,
3657 req->r_session ? req->r_session->s_mds : -1);
3658 mutex_unlock(&mdsc->mutex);
3659 goto out;
3660 }
3661
3662 /* dup? */
3663 if ((test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags) && !head->safe) ||
3664 (test_bit(CEPH_MDS_R_GOT_SAFE, &req->r_req_flags) && head->safe)) {
3665 pr_warn("got a dup %s reply on %llu from mds%d\n",
3666 head->safe ? "safe" : "unsafe", tid, mds);
3667 mutex_unlock(&mdsc->mutex);
3668 goto out;
3669 }
3670 if (test_bit(CEPH_MDS_R_GOT_SAFE, &req->r_req_flags)) {
3671 pr_warn("got unsafe after safe on %llu from mds%d\n",
3672 tid, mds);
3673 mutex_unlock(&mdsc->mutex);
3674 goto out;
3675 }
3676
3677 result = le32_to_cpu(head->result);
3678
3679 if (head->safe) {
3680 set_bit(CEPH_MDS_R_GOT_SAFE, &req->r_req_flags);
3681 __unregister_request(mdsc, req);
3682
3683 /* last request during umount? */
3684 if (mdsc->stopping && !__get_oldest_req(mdsc))
3685 complete_all(&mdsc->safe_umount_waiters);
3686
3687 if (test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags)) {
3688 /*
3689 * We already handled the unsafe response, now do the
3690 * cleanup. No need to examine the response; the MDS
3691 * doesn't include any result info in the safe
3692 * response. And even if it did, there is nothing
3693 * useful we could do with a revised return value.
3694 */
3695 dout("got safe reply %llu, mds%d\n", tid, mds);
3696
3697 mutex_unlock(&mdsc->mutex);
3698 goto out;
3699 }
3700 } else {
3701 set_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags);
3702 list_add_tail(&req->r_unsafe_item, &req->r_session->s_unsafe);
3703 }
3704
3705 dout("handle_reply tid %lld result %d\n", tid, result);
3706 if (test_bit(CEPHFS_FEATURE_REPLY_ENCODING, &session->s_features))
3707 err = parse_reply_info(session, msg, req, (u64)-1);
3708 else
3709 err = parse_reply_info(session, msg, req,
3710 session->s_con.peer_features);
3711 mutex_unlock(&mdsc->mutex);
3712
3713 /* Must find target inode outside of mutexes to avoid deadlocks */
3714 rinfo = &req->r_reply_info;
3715 if ((err >= 0) && rinfo->head->is_target) {
3716 struct inode *in = xchg(&req->r_new_inode, NULL);
3717 struct ceph_vino tvino = {
3718 .ino = le64_to_cpu(rinfo->targeti.in->ino),
3719 .snap = le64_to_cpu(rinfo->targeti.in->snapid)
3720 };
3721
3722 /*
3723 * If we ended up opening an existing inode, discard
3724 * r_new_inode
3725 */
3726 if (req->r_op == CEPH_MDS_OP_CREATE &&
3727 !req->r_reply_info.has_create_ino) {
3728 /* This should never happen on an async create */
3729 WARN_ON_ONCE(req->r_deleg_ino);
3730 iput(in);
3731 in = NULL;
3732 }
3733
3734 in = ceph_get_inode(mdsc->fsc->sb, tvino, in);
3735 if (IS_ERR(in)) {
3736 err = PTR_ERR(in);
3737 mutex_lock(&session->s_mutex);
3738 goto out_err;
3739 }
3740 req->r_target_inode = in;
3741 }
3742
3743 mutex_lock(&session->s_mutex);
3744 if (err < 0) {
3745 pr_err("mdsc_handle_reply got corrupt reply mds%d(tid:%lld)\n", mds, tid);
3746 ceph_msg_dump(msg);
3747 goto out_err;
3748 }
3749
3750 /* snap trace */
3751 realm = NULL;
3752 if (rinfo->snapblob_len) {
3753 down_write(&mdsc->snap_rwsem);
3754 err = ceph_update_snap_trace(mdsc, rinfo->snapblob,
3755 rinfo->snapblob + rinfo->snapblob_len,
3756 le32_to_cpu(head->op) == CEPH_MDS_OP_RMSNAP,
3757 &realm);
3758 if (err) {
3759 up_write(&mdsc->snap_rwsem);
3760 close_sessions = true;
3761 if (err == -EIO)
3762 ceph_msg_dump(msg);
3763 goto out_err;
3764 }
3765 downgrade_write(&mdsc->snap_rwsem);
3766 } else {
3767 down_read(&mdsc->snap_rwsem);
3768 }
3769
3770 /* insert trace into our cache */
3771 mutex_lock(&req->r_fill_mutex);
3772 current->journal_info = req;
3773 err = ceph_fill_trace(mdsc->fsc->sb, req);
3774 if (err == 0) {
3775 if (result == 0 && (req->r_op == CEPH_MDS_OP_READDIR ||
3776 req->r_op == CEPH_MDS_OP_LSSNAP))
3777 err = ceph_readdir_prepopulate(req, req->r_session);
3778 }
3779 current->journal_info = NULL;
3780 mutex_unlock(&req->r_fill_mutex);
3781
3782 up_read(&mdsc->snap_rwsem);
3783 if (realm)
3784 ceph_put_snap_realm(mdsc, realm);
3785
3786 if (err == 0) {
3787 if (req->r_target_inode &&
3788 test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags)) {
3789 struct ceph_inode_info *ci =
3790 ceph_inode(req->r_target_inode);
3791 spin_lock(&ci->i_unsafe_lock);
3792 list_add_tail(&req->r_unsafe_target_item,
3793 &ci->i_unsafe_iops);
3794 spin_unlock(&ci->i_unsafe_lock);
3795 }
3796
3797 ceph_unreserve_caps(mdsc, &req->r_caps_reservation);
3798 }
3799 out_err:
3800 mutex_lock(&mdsc->mutex);
3801 if (!test_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags)) {
3802 if (err) {
3803 req->r_err = err;
3804 } else {
3805 req->r_reply = ceph_msg_get(msg);
3806 set_bit(CEPH_MDS_R_GOT_RESULT, &req->r_req_flags);
3807 }
3808 } else {
3809 dout("reply arrived after request %lld was aborted\n", tid);
3810 }
3811 mutex_unlock(&mdsc->mutex);
3812
3813 mutex_unlock(&session->s_mutex);
3814
3815 /* kick calling process */
3816 complete_request(mdsc, req);
3817
3818 ceph_update_metadata_metrics(&mdsc->metric, req->r_start_latency,
3819 req->r_end_latency, err);
3820 out:
3821 ceph_mdsc_put_request(req);
3822
3823 /* Defer closing the sessions after s_mutex lock being released */
3824 if (close_sessions)
3825 ceph_mdsc_close_sessions(mdsc);
3826 return;
3827 }
3828
3829
3830
3831 /*
3832 * handle mds notification that our request has been forwarded.
3833 */
3834 static void handle_forward(struct ceph_mds_client *mdsc,
3835 struct ceph_mds_session *session,
3836 struct ceph_msg *msg)
3837 {
3838 struct ceph_mds_request *req;
3839 u64 tid = le64_to_cpu(msg->hdr.tid);
3840 u32 next_mds;
3841 u32 fwd_seq;
3842 int err = -EINVAL;
3843 void *p = msg->front.iov_base;
3844 void *end = p + msg->front.iov_len;
3845 bool aborted = false;
3846
3847 ceph_decode_need(&p, end, 2*sizeof(u32), bad);
3848 next_mds = ceph_decode_32(&p);
3849 fwd_seq = ceph_decode_32(&p);
3850
3851 mutex_lock(&mdsc->mutex);
3852 req = lookup_get_request(mdsc, tid);
3853 if (!req) {
3854 mutex_unlock(&mdsc->mutex);
3855 dout("forward tid %llu to mds%d - req dne\n", tid, next_mds);
3856 return; /* dup reply? */
3857 }
3858
3859 if (test_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags)) {
3860 dout("forward tid %llu aborted, unregistering\n", tid);
3861 __unregister_request(mdsc, req);
3862 } else if (fwd_seq <= req->r_num_fwd || (uint32_t)fwd_seq >= U32_MAX) {
3863 /*
3864 * Avoid inifinite retrying after overflow.
3865 *
3866 * The MDS will increase the fwd count and in client side
3867 * if the num_fwd is less than the one saved in request
3868 * that means the MDS is an old version and overflowed of
3869 * 8 bits.
3870 */
3871 mutex_lock(&req->r_fill_mutex);
3872 req->r_err = -EMULTIHOP;
3873 set_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags);
3874 mutex_unlock(&req->r_fill_mutex);
3875 aborted = true;
3876 pr_warn_ratelimited("forward tid %llu seq overflow\n", tid);
3877 } else {
3878 /* resend. forward race not possible; mds would drop */
3879 dout("forward tid %llu to mds%d (we resend)\n", tid, next_mds);
3880 BUG_ON(req->r_err);
3881 BUG_ON(test_bit(CEPH_MDS_R_GOT_RESULT, &req->r_req_flags));
3882 req->r_attempts = 0;
3883 req->r_num_fwd = fwd_seq;
3884 req->r_resend_mds = next_mds;
3885 put_request_session(req);
3886 __do_request(mdsc, req);
3887 }
3888 mutex_unlock(&mdsc->mutex);
3889
3890 /* kick calling process */
3891 if (aborted)
3892 complete_request(mdsc, req);
3893 ceph_mdsc_put_request(req);
3894 return;
3895
3896 bad:
3897 pr_err("mdsc_handle_forward decode error err=%d\n", err);
3898 ceph_msg_dump(msg);
3899 }
3900
3901 static int __decode_session_metadata(void **p, void *end,
3902 bool *blocklisted)
3903 {
3904 /* map<string,string> */
3905 u32 n;
3906 bool err_str;
3907 ceph_decode_32_safe(p, end, n, bad);
3908 while (n-- > 0) {
3909 u32 len;
3910 ceph_decode_32_safe(p, end, len, bad);
3911 ceph_decode_need(p, end, len, bad);
3912 err_str = !strncmp(*p, "error_string", len);
3913 *p += len;
3914 ceph_decode_32_safe(p, end, len, bad);
3915 ceph_decode_need(p, end, len, bad);
3916 /*
3917 * Match "blocklisted (blacklisted)" from newer MDSes,
3918 * or "blacklisted" from older MDSes.
3919 */
3920 if (err_str && strnstr(*p, "blacklisted", len))
3921 *blocklisted = true;
3922 *p += len;
3923 }
3924 return 0;
3925 bad:
3926 return -1;
3927 }
3928
3929 /*
3930 * handle a mds session control message
3931 */
3932 static void handle_session(struct ceph_mds_session *session,
3933 struct ceph_msg *msg)
3934 {
3935 struct ceph_mds_client *mdsc = session->s_mdsc;
3936 int mds = session->s_mds;
3937 int msg_version = le16_to_cpu(msg->hdr.version);
3938 void *p = msg->front.iov_base;
3939 void *end = p + msg->front.iov_len;
3940 struct ceph_mds_session_head *h;
3941 u32 op;
3942 u64 seq, features = 0;
3943 int wake = 0;
3944 bool blocklisted = false;
3945
3946 /* decode */
3947 ceph_decode_need(&p, end, sizeof(*h), bad);
3948 h = p;
3949 p += sizeof(*h);
3950
3951 op = le32_to_cpu(h->op);
3952 seq = le64_to_cpu(h->seq);
3953
3954 if (msg_version >= 3) {
3955 u32 len;
3956 /* version >= 2 and < 5, decode metadata, skip otherwise
3957 * as it's handled via flags.
3958 */
3959 if (msg_version >= 5)
3960 ceph_decode_skip_map(&p, end, string, string, bad);
3961 else if (__decode_session_metadata(&p, end, &blocklisted) < 0)
3962 goto bad;
3963
3964 /* version >= 3, feature bits */
3965 ceph_decode_32_safe(&p, end, len, bad);
3966 if (len) {
3967 ceph_decode_64_safe(&p, end, features, bad);
3968 p += len - sizeof(features);
3969 }
3970 }
3971
3972 if (msg_version >= 5) {
3973 u32 flags, len;
3974
3975 /* version >= 4 */
3976 ceph_decode_skip_16(&p, end, bad); /* struct_v, struct_cv */
3977 ceph_decode_32_safe(&p, end, len, bad); /* len */
3978 ceph_decode_skip_n(&p, end, len, bad); /* metric_spec */
3979
3980 /* version >= 5, flags */
3981 ceph_decode_32_safe(&p, end, flags, bad);
3982 if (flags & CEPH_SESSION_BLOCKLISTED) {
3983 pr_warn("mds%d session blocklisted\n", session->s_mds);
3984 blocklisted = true;
3985 }
3986 }
3987
3988 mutex_lock(&mdsc->mutex);
3989 if (op == CEPH_SESSION_CLOSE) {
3990 ceph_get_mds_session(session);
3991 __unregister_session(mdsc, session);
3992 }
3993 /* FIXME: this ttl calculation is generous */
3994 session->s_ttl = jiffies + HZ*mdsc->mdsmap->m_session_autoclose;
3995 mutex_unlock(&mdsc->mutex);
3996
3997 mutex_lock(&session->s_mutex);
3998
3999 dout("handle_session mds%d %s %p state %s seq %llu\n",
4000 mds, ceph_session_op_name(op), session,
4001 ceph_session_state_name(session->s_state), seq);
4002
4003 if (session->s_state == CEPH_MDS_SESSION_HUNG) {
4004 session->s_state = CEPH_MDS_SESSION_OPEN;
4005 pr_info("mds%d came back\n", session->s_mds);
4006 }
4007
4008 switch (op) {
4009 case CEPH_SESSION_OPEN:
4010 if (session->s_state == CEPH_MDS_SESSION_RECONNECTING)
4011 pr_info("mds%d reconnect success\n", session->s_mds);
4012
4013 if (session->s_state == CEPH_MDS_SESSION_OPEN) {
4014 pr_notice("mds%d is already opened\n", session->s_mds);
4015 } else {
4016 session->s_state = CEPH_MDS_SESSION_OPEN;
4017 session->s_features = features;
4018 renewed_caps(mdsc, session, 0);
4019 if (test_bit(CEPHFS_FEATURE_METRIC_COLLECT,
4020 &session->s_features))
4021 metric_schedule_delayed(&mdsc->metric);
4022 }
4023
4024 /*
4025 * The connection maybe broken and the session in client
4026 * side has been reinitialized, need to update the seq
4027 * anyway.
4028 */
4029 if (!session->s_seq && seq)
4030 session->s_seq = seq;
4031
4032 wake = 1;
4033 if (mdsc->stopping)
4034 __close_session(mdsc, session);
4035 break;
4036
4037 case CEPH_SESSION_RENEWCAPS:
4038 if (session->s_renew_seq == seq)
4039 renewed_caps(mdsc, session, 1);
4040 break;
4041
4042 case CEPH_SESSION_CLOSE:
4043 if (session->s_state == CEPH_MDS_SESSION_RECONNECTING)
4044 pr_info("mds%d reconnect denied\n", session->s_mds);
4045 session->s_state = CEPH_MDS_SESSION_CLOSED;
4046 cleanup_session_requests(mdsc, session);
4047 remove_session_caps(session);
4048 wake = 2; /* for good measure */
4049 wake_up_all(&mdsc->session_close_wq);
4050 break;
4051
4052 case CEPH_SESSION_STALE:
4053 pr_info("mds%d caps went stale, renewing\n",
4054 session->s_mds);
4055 atomic_inc(&session->s_cap_gen);
4056 session->s_cap_ttl = jiffies - 1;
4057 send_renew_caps(mdsc, session);
4058 break;
4059
4060 case CEPH_SESSION_RECALL_STATE:
4061 ceph_trim_caps(mdsc, session, le32_to_cpu(h->max_caps));
4062 break;
4063
4064 case CEPH_SESSION_FLUSHMSG:
4065 /* flush cap releases */
4066 spin_lock(&session->s_cap_lock);
4067 if (session->s_num_cap_releases)
4068 ceph_flush_cap_releases(mdsc, session);
4069 spin_unlock(&session->s_cap_lock);
4070
4071 send_flushmsg_ack(mdsc, session, seq);
4072 break;
4073
4074 case CEPH_SESSION_FORCE_RO:
4075 dout("force_session_readonly %p\n", session);
4076 spin_lock(&session->s_cap_lock);
4077 session->s_readonly = true;
4078 spin_unlock(&session->s_cap_lock);
4079 wake_up_session_caps(session, FORCE_RO);
4080 break;
4081
4082 case CEPH_SESSION_REJECT:
4083 WARN_ON(session->s_state != CEPH_MDS_SESSION_OPENING);
4084 pr_info("mds%d rejected session\n", session->s_mds);
4085 session->s_state = CEPH_MDS_SESSION_REJECTED;
4086 cleanup_session_requests(mdsc, session);
4087 remove_session_caps(session);
4088 if (blocklisted)
4089 mdsc->fsc->blocklisted = true;
4090 wake = 2; /* for good measure */
4091 break;
4092
4093 default:
4094 pr_err("mdsc_handle_session bad op %d mds%d\n", op, mds);
4095 WARN_ON(1);
4096 }
4097
4098 mutex_unlock(&session->s_mutex);
4099 if (wake) {
4100 mutex_lock(&mdsc->mutex);
4101 __wake_requests(mdsc, &session->s_waiting);
4102 if (wake == 2)
4103 kick_requests(mdsc, mds);
4104 mutex_unlock(&mdsc->mutex);
4105 }
4106 if (op == CEPH_SESSION_CLOSE)
4107 ceph_put_mds_session(session);
4108 return;
4109
4110 bad:
4111 pr_err("mdsc_handle_session corrupt message mds%d len %d\n", mds,
4112 (int)msg->front.iov_len);
4113 ceph_msg_dump(msg);
4114 return;
4115 }
4116
4117 void ceph_mdsc_release_dir_caps(struct ceph_mds_request *req)
4118 {
4119 int dcaps;
4120
4121 dcaps = xchg(&req->r_dir_caps, 0);
4122 if (dcaps) {
4123 dout("releasing r_dir_caps=%s\n", ceph_cap_string(dcaps));
4124 ceph_put_cap_refs(ceph_inode(req->r_parent), dcaps);
4125 }
4126 }
4127
4128 void ceph_mdsc_release_dir_caps_no_check(struct ceph_mds_request *req)
4129 {
4130 int dcaps;
4131
4132 dcaps = xchg(&req->r_dir_caps, 0);
4133 if (dcaps) {
4134 dout("releasing r_dir_caps=%s\n", ceph_cap_string(dcaps));
4135 ceph_put_cap_refs_no_check_caps(ceph_inode(req->r_parent),
4136 dcaps);
4137 }
4138 }
4139
4140 /*
4141 * called under session->mutex.
4142 */
4143 static void replay_unsafe_requests(struct ceph_mds_client *mdsc,
4144 struct ceph_mds_session *session)
4145 {
4146 struct ceph_mds_request *req, *nreq;
4147 struct rb_node *p;
4148
4149 dout("replay_unsafe_requests mds%d\n", session->s_mds);
4150
4151 mutex_lock(&mdsc->mutex);
4152 list_for_each_entry_safe(req, nreq, &session->s_unsafe, r_unsafe_item)
4153 __send_request(session, req, true);
4154
4155 /*
4156 * also re-send old requests when MDS enters reconnect stage. So that MDS
4157 * can process completed request in clientreplay stage.
4158 */
4159 p = rb_first(&mdsc->request_tree);
4160 while (p) {
4161 req = rb_entry(p, struct ceph_mds_request, r_node);
4162 p = rb_next(p);
4163 if (test_bit(CEPH_MDS_R_GOT_UNSAFE, &req->r_req_flags))
4164 continue;
4165 if (req->r_attempts == 0)
4166 continue; /* only old requests */
4167 if (!req->r_session)
4168 continue;
4169 if (req->r_session->s_mds != session->s_mds)
4170 continue;
4171
4172 ceph_mdsc_release_dir_caps_no_check(req);
4173
4174 __send_request(session, req, true);
4175 }
4176 mutex_unlock(&mdsc->mutex);
4177 }
4178
4179 static int send_reconnect_partial(struct ceph_reconnect_state *recon_state)
4180 {
4181 struct ceph_msg *reply;
4182 struct ceph_pagelist *_pagelist;
4183 struct page *page;
4184 __le32 *addr;
4185 int err = -ENOMEM;
4186
4187 if (!recon_state->allow_multi)
4188 return -ENOSPC;
4189
4190 /* can't handle message that contains both caps and realm */
4191 BUG_ON(!recon_state->nr_caps == !recon_state->nr_realms);
4192
4193 /* pre-allocate new pagelist */
4194 _pagelist = ceph_pagelist_alloc(GFP_NOFS);
4195 if (!_pagelist)
4196 return -ENOMEM;
4197
4198 reply = ceph_msg_new2(CEPH_MSG_CLIENT_RECONNECT, 0, 1, GFP_NOFS, false);
4199 if (!reply)
4200 goto fail_msg;
4201
4202 /* placeholder for nr_caps */
4203 err = ceph_pagelist_encode_32(_pagelist, 0);
4204 if (err < 0)
4205 goto fail;
4206
4207 if (recon_state->nr_caps) {
4208 /* currently encoding caps */
4209 err = ceph_pagelist_encode_32(recon_state->pagelist, 0);
4210 if (err)
4211 goto fail;
4212 } else {
4213 /* placeholder for nr_realms (currently encoding relams) */
4214 err = ceph_pagelist_encode_32(_pagelist, 0);
4215 if (err < 0)
4216 goto fail;
4217 }
4218
4219 err = ceph_pagelist_encode_8(recon_state->pagelist, 1);
4220 if (err)
4221 goto fail;
4222
4223 page = list_first_entry(&recon_state->pagelist->head, struct page, lru);
4224 addr = kmap_atomic(page);
4225 if (recon_state->nr_caps) {
4226 /* currently encoding caps */
4227 *addr = cpu_to_le32(recon_state->nr_caps);
4228 } else {
4229 /* currently encoding relams */
4230 *(addr + 1) = cpu_to_le32(recon_state->nr_realms);
4231 }
4232 kunmap_atomic(addr);
4233
4234 reply->hdr.version = cpu_to_le16(5);
4235 reply->hdr.compat_version = cpu_to_le16(4);
4236
4237 reply->hdr.data_len = cpu_to_le32(recon_state->pagelist->length);
4238 ceph_msg_data_add_pagelist(reply, recon_state->pagelist);
4239
4240 ceph_con_send(&recon_state->session->s_con, reply);
4241 ceph_pagelist_release(recon_state->pagelist);
4242
4243 recon_state->pagelist = _pagelist;
4244 recon_state->nr_caps = 0;
4245 recon_state->nr_realms = 0;
4246 recon_state->msg_version = 5;
4247 return 0;
4248 fail:
4249 ceph_msg_put(reply);
4250 fail_msg:
4251 ceph_pagelist_release(_pagelist);
4252 return err;
4253 }
4254
4255 static struct dentry* d_find_primary(struct inode *inode)
4256 {
4257 struct dentry *alias, *dn = NULL;
4258
4259 if (hlist_empty(&inode->i_dentry))
4260 return NULL;
4261
4262 spin_lock(&inode->i_lock);
4263 if (hlist_empty(&inode->i_dentry))
4264 goto out_unlock;
4265
4266 if (S_ISDIR(inode->i_mode)) {
4267 alias = hlist_entry(inode->i_dentry.first, struct dentry, d_u.d_alias);
4268 if (!IS_ROOT(alias))
4269 dn = dget(alias);
4270 goto out_unlock;
4271 }
4272
4273 hlist_for_each_entry(alias, &inode->i_dentry, d_u.d_alias) {
4274 spin_lock(&alias->d_lock);
4275 if (!d_unhashed(alias) &&
4276 (ceph_dentry(alias)->flags & CEPH_DENTRY_PRIMARY_LINK)) {
4277 dn = dget_dlock(alias);
4278 }
4279 spin_unlock(&alias->d_lock);
4280 if (dn)
4281 break;
4282 }
4283 out_unlock:
4284 spin_unlock(&inode->i_lock);
4285 return dn;
4286 }
4287
4288 /*
4289 * Encode information about a cap for a reconnect with the MDS.
4290 */
4291 static int reconnect_caps_cb(struct inode *inode, int mds, void *arg)
4292 {
4293 union {
4294 struct ceph_mds_cap_reconnect v2;
4295 struct ceph_mds_cap_reconnect_v1 v1;
4296 } rec;
4297 struct ceph_inode_info *ci = ceph_inode(inode);
4298 struct ceph_reconnect_state *recon_state = arg;
4299 struct ceph_pagelist *pagelist = recon_state->pagelist;
4300 struct dentry *dentry;
4301 struct ceph_cap *cap;
4302 char *path;
4303 int pathlen = 0, err;
4304 u64 pathbase;
4305 u64 snap_follows;
4306
4307 dentry = d_find_primary(inode);
4308 if (dentry) {
4309 /* set pathbase to parent dir when msg_version >= 2 */
4310 path = ceph_mdsc_build_path(dentry, &pathlen, &pathbase,
4311 recon_state->msg_version >= 2);
4312 dput(dentry);
4313 if (IS_ERR(path)) {
4314 err = PTR_ERR(path);
4315 goto out_err;
4316 }
4317 } else {
4318 path = NULL;
4319 pathbase = 0;
4320 }
4321
4322 spin_lock(&ci->i_ceph_lock);
4323 cap = __get_cap_for_mds(ci, mds);
4324 if (!cap) {
4325 spin_unlock(&ci->i_ceph_lock);
4326 err = 0;
4327 goto out_err;
4328 }
4329 dout(" adding %p ino %llx.%llx cap %p %lld %s\n",
4330 inode, ceph_vinop(inode), cap, cap->cap_id,
4331 ceph_cap_string(cap->issued));
4332
4333 cap->seq = 0; /* reset cap seq */
4334 cap->issue_seq = 0; /* and issue_seq */
4335 cap->mseq = 0; /* and migrate_seq */
4336 cap->cap_gen = atomic_read(&cap->session->s_cap_gen);
4337
4338 /* These are lost when the session goes away */
4339 if (S_ISDIR(inode->i_mode)) {
4340 if (cap->issued & CEPH_CAP_DIR_CREATE) {
4341 ceph_put_string(rcu_dereference_raw(ci->i_cached_layout.pool_ns));
4342 memset(&ci->i_cached_layout, 0, sizeof(ci->i_cached_layout));
4343 }
4344 cap->issued &= ~CEPH_CAP_ANY_DIR_OPS;
4345 }
4346
4347 if (recon_state->msg_version >= 2) {
4348 rec.v2.cap_id = cpu_to_le64(cap->cap_id);
4349 rec.v2.wanted = cpu_to_le32(__ceph_caps_wanted(ci));
4350 rec.v2.issued = cpu_to_le32(cap->issued);
4351 rec.v2.snaprealm = cpu_to_le64(ci->i_snap_realm->ino);
4352 rec.v2.pathbase = cpu_to_le64(pathbase);
4353 rec.v2.flock_len = (__force __le32)
4354 ((ci->i_ceph_flags & CEPH_I_ERROR_FILELOCK) ? 0 : 1);
4355 } else {
4356 rec.v1.cap_id = cpu_to_le64(cap->cap_id);
4357 rec.v1.wanted = cpu_to_le32(__ceph_caps_wanted(ci));
4358 rec.v1.issued = cpu_to_le32(cap->issued);
4359 rec.v1.size = cpu_to_le64(i_size_read(inode));
4360 ceph_encode_timespec64(&rec.v1.mtime, &inode->i_mtime);
4361 ceph_encode_timespec64(&rec.v1.atime, &inode->i_atime);
4362 rec.v1.snaprealm = cpu_to_le64(ci->i_snap_realm->ino);
4363 rec.v1.pathbase = cpu_to_le64(pathbase);
4364 }
4365
4366 if (list_empty(&ci->i_cap_snaps)) {
4367 snap_follows = ci->i_head_snapc ? ci->i_head_snapc->seq : 0;
4368 } else {
4369 struct ceph_cap_snap *capsnap =
4370 list_first_entry(&ci->i_cap_snaps,
4371 struct ceph_cap_snap, ci_item);
4372 snap_follows = capsnap->follows;
4373 }
4374 spin_unlock(&ci->i_ceph_lock);
4375
4376 if (recon_state->msg_version >= 2) {
4377 int num_fcntl_locks, num_flock_locks;
4378 struct ceph_filelock *flocks = NULL;
4379 size_t struct_len, total_len = sizeof(u64);
4380 u8 struct_v = 0;
4381
4382 encode_again:
4383 if (rec.v2.flock_len) {
4384 ceph_count_locks(inode, &num_fcntl_locks, &num_flock_locks);
4385 } else {
4386 num_fcntl_locks = 0;
4387 num_flock_locks = 0;
4388 }
4389 if (num_fcntl_locks + num_flock_locks > 0) {
4390 flocks = kmalloc_array(num_fcntl_locks + num_flock_locks,
4391 sizeof(struct ceph_filelock),
4392 GFP_NOFS);
4393 if (!flocks) {
4394 err = -ENOMEM;
4395 goto out_err;
4396 }
4397 err = ceph_encode_locks_to_buffer(inode, flocks,
4398 num_fcntl_locks,
4399 num_flock_locks);
4400 if (err) {
4401 kfree(flocks);
4402 flocks = NULL;
4403 if (err == -ENOSPC)
4404 goto encode_again;
4405 goto out_err;
4406 }
4407 } else {
4408 kfree(flocks);
4409 flocks = NULL;
4410 }
4411
4412 if (recon_state->msg_version >= 3) {
4413 /* version, compat_version and struct_len */
4414 total_len += 2 * sizeof(u8) + sizeof(u32);
4415 struct_v = 2;
4416 }
4417 /*
4418 * number of encoded locks is stable, so copy to pagelist
4419 */
4420 struct_len = 2 * sizeof(u32) +
4421 (num_fcntl_locks + num_flock_locks) *
4422 sizeof(struct ceph_filelock);
4423 rec.v2.flock_len = cpu_to_le32(struct_len);
4424
4425 struct_len += sizeof(u32) + pathlen + sizeof(rec.v2);
4426
4427 if (struct_v >= 2)
4428 struct_len += sizeof(u64); /* snap_follows */
4429
4430 total_len += struct_len;
4431
4432 if (pagelist->length + total_len > RECONNECT_MAX_SIZE) {
4433 err = send_reconnect_partial(recon_state);
4434 if (err)
4435 goto out_freeflocks;
4436 pagelist = recon_state->pagelist;
4437 }
4438
4439 err = ceph_pagelist_reserve(pagelist, total_len);
4440 if (err)
4441 goto out_freeflocks;
4442
4443 ceph_pagelist_encode_64(pagelist, ceph_ino(inode));
4444 if (recon_state->msg_version >= 3) {
4445 ceph_pagelist_encode_8(pagelist, struct_v);
4446 ceph_pagelist_encode_8(pagelist, 1);
4447 ceph_pagelist_encode_32(pagelist, struct_len);
4448 }
4449 ceph_pagelist_encode_string(pagelist, path, pathlen);
4450 ceph_pagelist_append(pagelist, &rec, sizeof(rec.v2));
4451 ceph_locks_to_pagelist(flocks, pagelist,
4452 num_fcntl_locks, num_flock_locks);
4453 if (struct_v >= 2)
4454 ceph_pagelist_encode_64(pagelist, snap_follows);
4455 out_freeflocks:
4456 kfree(flocks);
4457 } else {
4458 err = ceph_pagelist_reserve(pagelist,
4459 sizeof(u64) + sizeof(u32) +
4460 pathlen + sizeof(rec.v1));
4461 if (err)
4462 goto out_err;
4463
4464 ceph_pagelist_encode_64(pagelist, ceph_ino(inode));
4465 ceph_pagelist_encode_string(pagelist, path, pathlen);
4466 ceph_pagelist_append(pagelist, &rec, sizeof(rec.v1));
4467 }
4468
4469 out_err:
4470 ceph_mdsc_free_path(path, pathlen);
4471 if (!err)
4472 recon_state->nr_caps++;
4473 return err;
4474 }
4475
4476 static int encode_snap_realms(struct ceph_mds_client *mdsc,
4477 struct ceph_reconnect_state *recon_state)
4478 {
4479 struct rb_node *p;
4480 struct ceph_pagelist *pagelist = recon_state->pagelist;
4481 int err = 0;
4482
4483 if (recon_state->msg_version >= 4) {
4484 err = ceph_pagelist_encode_32(pagelist, mdsc->num_snap_realms);
4485 if (err < 0)
4486 goto fail;
4487 }
4488
4489 /*
4490 * snaprealms. we provide mds with the ino, seq (version), and
4491 * parent for all of our realms. If the mds has any newer info,
4492 * it will tell us.
4493 */
4494 for (p = rb_first(&mdsc->snap_realms); p; p = rb_next(p)) {
4495 struct ceph_snap_realm *realm =
4496 rb_entry(p, struct ceph_snap_realm, node);
4497 struct ceph_mds_snaprealm_reconnect sr_rec;
4498
4499 if (recon_state->msg_version >= 4) {
4500 size_t need = sizeof(u8) * 2 + sizeof(u32) +
4501 sizeof(sr_rec);
4502
4503 if (pagelist->length + need > RECONNECT_MAX_SIZE) {
4504 err = send_reconnect_partial(recon_state);
4505 if (err)
4506 goto fail;
4507 pagelist = recon_state->pagelist;
4508 }
4509
4510 err = ceph_pagelist_reserve(pagelist, need);
4511 if (err)
4512 goto fail;
4513
4514 ceph_pagelist_encode_8(pagelist, 1);
4515 ceph_pagelist_encode_8(pagelist, 1);
4516 ceph_pagelist_encode_32(pagelist, sizeof(sr_rec));
4517 }
4518
4519 dout(" adding snap realm %llx seq %lld parent %llx\n",
4520 realm->ino, realm->seq, realm->parent_ino);
4521 sr_rec.ino = cpu_to_le64(realm->ino);
4522 sr_rec.seq = cpu_to_le64(realm->seq);
4523 sr_rec.parent = cpu_to_le64(realm->parent_ino);
4524
4525 err = ceph_pagelist_append(pagelist, &sr_rec, sizeof(sr_rec));
4526 if (err)
4527 goto fail;
4528
4529 recon_state->nr_realms++;
4530 }
4531 fail:
4532 return err;
4533 }
4534
4535
4536 /*
4537 * If an MDS fails and recovers, clients need to reconnect in order to
4538 * reestablish shared state. This includes all caps issued through
4539 * this session _and_ the snap_realm hierarchy. Because it's not
4540 * clear which snap realms the mds cares about, we send everything we
4541 * know about.. that ensures we'll then get any new info the
4542 * recovering MDS might have.
4543 *
4544 * This is a relatively heavyweight operation, but it's rare.
4545 */
4546 static void send_mds_reconnect(struct ceph_mds_client *mdsc,
4547 struct ceph_mds_session *session)
4548 {
4549 struct ceph_msg *reply;
4550 int mds = session->s_mds;
4551 int err = -ENOMEM;
4552 struct ceph_reconnect_state recon_state = {
4553 .session = session,
4554 };
4555 LIST_HEAD(dispose);
4556
4557 pr_info("mds%d reconnect start\n", mds);
4558
4559 recon_state.pagelist = ceph_pagelist_alloc(GFP_NOFS);
4560 if (!recon_state.pagelist)
4561 goto fail_nopagelist;
4562
4563 reply = ceph_msg_new2(CEPH_MSG_CLIENT_RECONNECT, 0, 1, GFP_NOFS, false);
4564 if (!reply)
4565 goto fail_nomsg;
4566
4567 xa_destroy(&session->s_delegated_inos);
4568
4569 mutex_lock(&session->s_mutex);
4570 session->s_state = CEPH_MDS_SESSION_RECONNECTING;
4571 session->s_seq = 0;
4572
4573 dout("session %p state %s\n", session,
4574 ceph_session_state_name(session->s_state));
4575
4576 atomic_inc(&session->s_cap_gen);
4577
4578 spin_lock(&session->s_cap_lock);
4579 /* don't know if session is readonly */
4580 session->s_readonly = 0;
4581 /*
4582 * notify __ceph_remove_cap() that we are composing cap reconnect.
4583 * If a cap get released before being added to the cap reconnect,
4584 * __ceph_remove_cap() should skip queuing cap release.
4585 */
4586 session->s_cap_reconnect = 1;
4587 /* drop old cap expires; we're about to reestablish that state */
4588 detach_cap_releases(session, &dispose);
4589 spin_unlock(&session->s_cap_lock);
4590 dispose_cap_releases(mdsc, &dispose);
4591
4592 /* trim unused caps to reduce MDS's cache rejoin time */
4593 if (mdsc->fsc->sb->s_root)
4594 shrink_dcache_parent(mdsc->fsc->sb->s_root);
4595
4596 ceph_con_close(&session->s_con);
4597 ceph_con_open(&session->s_con,
4598 CEPH_ENTITY_TYPE_MDS, mds,
4599 ceph_mdsmap_get_addr(mdsc->mdsmap, mds));
4600
4601 /* replay unsafe requests */
4602 replay_unsafe_requests(mdsc, session);
4603
4604 ceph_early_kick_flushing_caps(mdsc, session);
4605
4606 down_read(&mdsc->snap_rwsem);
4607
4608 /* placeholder for nr_caps */
4609 err = ceph_pagelist_encode_32(recon_state.pagelist, 0);
4610 if (err)
4611 goto fail;
4612
4613 if (test_bit(CEPHFS_FEATURE_MULTI_RECONNECT, &session->s_features)) {
4614 recon_state.msg_version = 3;
4615 recon_state.allow_multi = true;
4616 } else if (session->s_con.peer_features & CEPH_FEATURE_MDSENC) {
4617 recon_state.msg_version = 3;
4618 } else {
4619 recon_state.msg_version = 2;
4620 }
4621 /* trsaverse this session's caps */
4622 err = ceph_iterate_session_caps(session, reconnect_caps_cb, &recon_state);
4623
4624 spin_lock(&session->s_cap_lock);
4625 session->s_cap_reconnect = 0;
4626 spin_unlock(&session->s_cap_lock);
4627
4628 if (err < 0)
4629 goto fail;
4630
4631 /* check if all realms can be encoded into current message */
4632 if (mdsc->num_snap_realms) {
4633 size_t total_len =
4634 recon_state.pagelist->length +
4635 mdsc->num_snap_realms *
4636 sizeof(struct ceph_mds_snaprealm_reconnect);
4637 if (recon_state.msg_version >= 4) {
4638 /* number of realms */
4639 total_len += sizeof(u32);
4640 /* version, compat_version and struct_len */
4641 total_len += mdsc->num_snap_realms *
4642 (2 * sizeof(u8) + sizeof(u32));
4643 }
4644 if (total_len > RECONNECT_MAX_SIZE) {
4645 if (!recon_state.allow_multi) {
4646 err = -ENOSPC;
4647 goto fail;
4648 }
4649 if (recon_state.nr_caps) {
4650 err = send_reconnect_partial(&recon_state);
4651 if (err)
4652 goto fail;
4653 }
4654 recon_state.msg_version = 5;
4655 }
4656 }
4657
4658 err = encode_snap_realms(mdsc, &recon_state);
4659 if (err < 0)
4660 goto fail;
4661
4662 if (recon_state.msg_version >= 5) {
4663 err = ceph_pagelist_encode_8(recon_state.pagelist, 0);
4664 if (err < 0)
4665 goto fail;
4666 }
4667
4668 if (recon_state.nr_caps || recon_state.nr_realms) {
4669 struct page *page =
4670 list_first_entry(&recon_state.pagelist->head,
4671 struct page, lru);
4672 __le32 *addr = kmap_atomic(page);
4673 if (recon_state.nr_caps) {
4674 WARN_ON(recon_state.nr_realms != mdsc->num_snap_realms);
4675 *addr = cpu_to_le32(recon_state.nr_caps);
4676 } else if (recon_state.msg_version >= 4) {
4677 *(addr + 1) = cpu_to_le32(recon_state.nr_realms);
4678 }
4679 kunmap_atomic(addr);
4680 }
4681
4682 reply->hdr.version = cpu_to_le16(recon_state.msg_version);
4683 if (recon_state.msg_version >= 4)
4684 reply->hdr.compat_version = cpu_to_le16(4);
4685
4686 reply->hdr.data_len = cpu_to_le32(recon_state.pagelist->length);
4687 ceph_msg_data_add_pagelist(reply, recon_state.pagelist);
4688
4689 ceph_con_send(&session->s_con, reply);
4690
4691 mutex_unlock(&session->s_mutex);
4692
4693 mutex_lock(&mdsc->mutex);
4694 __wake_requests(mdsc, &session->s_waiting);
4695 mutex_unlock(&mdsc->mutex);
4696
4697 up_read(&mdsc->snap_rwsem);
4698 ceph_pagelist_release(recon_state.pagelist);
4699 return;
4700
4701 fail:
4702 ceph_msg_put(reply);
4703 up_read(&mdsc->snap_rwsem);
4704 mutex_unlock(&session->s_mutex);
4705 fail_nomsg:
4706 ceph_pagelist_release(recon_state.pagelist);
4707 fail_nopagelist:
4708 pr_err("error %d preparing reconnect for mds%d\n", err, mds);
4709 return;
4710 }
4711
4712
4713 /*
4714 * compare old and new mdsmaps, kicking requests
4715 * and closing out old connections as necessary
4716 *
4717 * called under mdsc->mutex.
4718 */
4719 static void check_new_map(struct ceph_mds_client *mdsc,
4720 struct ceph_mdsmap *newmap,
4721 struct ceph_mdsmap *oldmap)
4722 {
4723 int i, j, err;
4724 int oldstate, newstate;
4725 struct ceph_mds_session *s;
4726 unsigned long targets[DIV_ROUND_UP(CEPH_MAX_MDS, sizeof(unsigned long))] = {0};
4727
4728 dout("check_new_map new %u old %u\n",
4729 newmap->m_epoch, oldmap->m_epoch);
4730
4731 if (newmap->m_info) {
4732 for (i = 0; i < newmap->possible_max_rank; i++) {
4733 for (j = 0; j < newmap->m_info[i].num_export_targets; j++)
4734 set_bit(newmap->m_info[i].export_targets[j], targets);
4735 }
4736 }
4737
4738 for (i = 0; i < oldmap->possible_max_rank && i < mdsc->max_sessions; i++) {
4739 if (!mdsc->sessions[i])
4740 continue;
4741 s = mdsc->sessions[i];
4742 oldstate = ceph_mdsmap_get_state(oldmap, i);
4743 newstate = ceph_mdsmap_get_state(newmap, i);
4744
4745 dout("check_new_map mds%d state %s%s -> %s%s (session %s)\n",
4746 i, ceph_mds_state_name(oldstate),
4747 ceph_mdsmap_is_laggy(oldmap, i) ? " (laggy)" : "",
4748 ceph_mds_state_name(newstate),
4749 ceph_mdsmap_is_laggy(newmap, i) ? " (laggy)" : "",
4750 ceph_session_state_name(s->s_state));
4751
4752 if (i >= newmap->possible_max_rank) {
4753 /* force close session for stopped mds */
4754 ceph_get_mds_session(s);
4755 __unregister_session(mdsc, s);
4756 __wake_requests(mdsc, &s->s_waiting);
4757 mutex_unlock(&mdsc->mutex);
4758
4759 mutex_lock(&s->s_mutex);
4760 cleanup_session_requests(mdsc, s);
4761 remove_session_caps(s);
4762 mutex_unlock(&s->s_mutex);
4763
4764 ceph_put_mds_session(s);
4765
4766 mutex_lock(&mdsc->mutex);
4767 kick_requests(mdsc, i);
4768 continue;
4769 }
4770
4771 if (memcmp(ceph_mdsmap_get_addr(oldmap, i),
4772 ceph_mdsmap_get_addr(newmap, i),
4773 sizeof(struct ceph_entity_addr))) {
4774 /* just close it */
4775 mutex_unlock(&mdsc->mutex);
4776 mutex_lock(&s->s_mutex);
4777 mutex_lock(&mdsc->mutex);
4778 ceph_con_close(&s->s_con);
4779 mutex_unlock(&s->s_mutex);
4780 s->s_state = CEPH_MDS_SESSION_RESTARTING;
4781 } else if (oldstate == newstate) {
4782 continue; /* nothing new with this mds */
4783 }
4784
4785 /*
4786 * send reconnect?
4787 */
4788 if (s->s_state == CEPH_MDS_SESSION_RESTARTING &&
4789 newstate >= CEPH_MDS_STATE_RECONNECT) {
4790 mutex_unlock(&mdsc->mutex);
4791 clear_bit(i, targets);
4792 send_mds_reconnect(mdsc, s);
4793 mutex_lock(&mdsc->mutex);
4794 }
4795
4796 /*
4797 * kick request on any mds that has gone active.
4798 */
4799 if (oldstate < CEPH_MDS_STATE_ACTIVE &&
4800 newstate >= CEPH_MDS_STATE_ACTIVE) {
4801 if (oldstate != CEPH_MDS_STATE_CREATING &&
4802 oldstate != CEPH_MDS_STATE_STARTING)
4803 pr_info("mds%d recovery completed\n", s->s_mds);
4804 kick_requests(mdsc, i);
4805 mutex_unlock(&mdsc->mutex);
4806 mutex_lock(&s->s_mutex);
4807 mutex_lock(&mdsc->mutex);
4808 ceph_kick_flushing_caps(mdsc, s);
4809 mutex_unlock(&s->s_mutex);
4810 wake_up_session_caps(s, RECONNECT);
4811 }
4812 }
4813
4814 /*
4815 * Only open and reconnect sessions that don't exist yet.
4816 */
4817 for (i = 0; i < newmap->possible_max_rank; i++) {
4818 /*
4819 * In case the import MDS is crashed just after
4820 * the EImportStart journal is flushed, so when
4821 * a standby MDS takes over it and is replaying
4822 * the EImportStart journal the new MDS daemon
4823 * will wait the client to reconnect it, but the
4824 * client may never register/open the session yet.
4825 *
4826 * Will try to reconnect that MDS daemon if the
4827 * rank number is in the export targets array and
4828 * is the up:reconnect state.
4829 */
4830 newstate = ceph_mdsmap_get_state(newmap, i);
4831 if (!test_bit(i, targets) || newstate != CEPH_MDS_STATE_RECONNECT)
4832 continue;
4833
4834 /*
4835 * The session maybe registered and opened by some
4836 * requests which were choosing random MDSes during
4837 * the mdsc->mutex's unlock/lock gap below in rare
4838 * case. But the related MDS daemon will just queue
4839 * that requests and be still waiting for the client's
4840 * reconnection request in up:reconnect state.
4841 */
4842 s = __ceph_lookup_mds_session(mdsc, i);
4843 if (likely(!s)) {
4844 s = __open_export_target_session(mdsc, i);
4845 if (IS_ERR(s)) {
4846 err = PTR_ERR(s);
4847 pr_err("failed to open export target session, err %d\n",
4848 err);
4849 continue;
4850 }
4851 }
4852 dout("send reconnect to export target mds.%d\n", i);
4853 mutex_unlock(&mdsc->mutex);
4854 send_mds_reconnect(mdsc, s);
4855 ceph_put_mds_session(s);
4856 mutex_lock(&mdsc->mutex);
4857 }
4858
4859 for (i = 0; i < newmap->possible_max_rank && i < mdsc->max_sessions; i++) {
4860 s = mdsc->sessions[i];
4861 if (!s)
4862 continue;
4863 if (!ceph_mdsmap_is_laggy(newmap, i))
4864 continue;
4865 if (s->s_state == CEPH_MDS_SESSION_OPEN ||
4866 s->s_state == CEPH_MDS_SESSION_HUNG ||
4867 s->s_state == CEPH_MDS_SESSION_CLOSING) {
4868 dout(" connecting to export targets of laggy mds%d\n",
4869 i);
4870 __open_export_target_sessions(mdsc, s);
4871 }
4872 }
4873 }
4874
4875
4876
4877 /*
4878 * leases
4879 */
4880
4881 /*
4882 * caller must hold session s_mutex, dentry->d_lock
4883 */
4884 void __ceph_mdsc_drop_dentry_lease(struct dentry *dentry)
4885 {
4886 struct ceph_dentry_info *di = ceph_dentry(dentry);
4887
4888 ceph_put_mds_session(di->lease_session);
4889 di->lease_session = NULL;
4890 }
4891
4892 static void handle_lease(struct ceph_mds_client *mdsc,
4893 struct ceph_mds_session *session,
4894 struct ceph_msg *msg)
4895 {
4896 struct super_block *sb = mdsc->fsc->sb;
4897 struct inode *inode;
4898 struct dentry *parent, *dentry;
4899 struct ceph_dentry_info *di;
4900 int mds = session->s_mds;
4901 struct ceph_mds_lease *h = msg->front.iov_base;
4902 u32 seq;
4903 struct ceph_vino vino;
4904 struct qstr dname;
4905 int release = 0;
4906
4907 dout("handle_lease from mds%d\n", mds);
4908
4909 if (!ceph_inc_mds_stopping_blocker(mdsc, session))
4910 return;
4911
4912 /* decode */
4913 if (msg->front.iov_len < sizeof(*h) + sizeof(u32))
4914 goto bad;
4915 vino.ino = le64_to_cpu(h->ino);
4916 vino.snap = CEPH_NOSNAP;
4917 seq = le32_to_cpu(h->seq);
4918 dname.len = get_unaligned_le32(h + 1);
4919 if (msg->front.iov_len < sizeof(*h) + sizeof(u32) + dname.len)
4920 goto bad;
4921 dname.name = (void *)(h + 1) + sizeof(u32);
4922
4923 /* lookup inode */
4924 inode = ceph_find_inode(sb, vino);
4925 dout("handle_lease %s, ino %llx %p %.*s\n",
4926 ceph_lease_op_name(h->action), vino.ino, inode,
4927 dname.len, dname.name);
4928
4929 mutex_lock(&session->s_mutex);
4930 if (!inode) {
4931 dout("handle_lease no inode %llx\n", vino.ino);
4932 goto release;
4933 }
4934
4935 /* dentry */
4936 parent = d_find_alias(inode);
4937 if (!parent) {
4938 dout("no parent dentry on inode %p\n", inode);
4939 WARN_ON(1);
4940 goto release; /* hrm... */
4941 }
4942 dname.hash = full_name_hash(parent, dname.name, dname.len);
4943 dentry = d_lookup(parent, &dname);
4944 dput(parent);
4945 if (!dentry)
4946 goto release;
4947
4948 spin_lock(&dentry->d_lock);
4949 di = ceph_dentry(dentry);
4950 switch (h->action) {
4951 case CEPH_MDS_LEASE_REVOKE:
4952 if (di->lease_session == session) {
4953 if (ceph_seq_cmp(di->lease_seq, seq) > 0)
4954 h->seq = cpu_to_le32(di->lease_seq);
4955 __ceph_mdsc_drop_dentry_lease(dentry);
4956 }
4957 release = 1;
4958 break;
4959
4960 case CEPH_MDS_LEASE_RENEW:
4961 if (di->lease_session == session &&
4962 di->lease_gen == atomic_read(&session->s_cap_gen) &&
4963 di->lease_renew_from &&
4964 di->lease_renew_after == 0) {
4965 unsigned long duration =
4966 msecs_to_jiffies(le32_to_cpu(h->duration_ms));
4967
4968 di->lease_seq = seq;
4969 di->time = di->lease_renew_from + duration;
4970 di->lease_renew_after = di->lease_renew_from +
4971 (duration >> 1);
4972 di->lease_renew_from = 0;
4973 }
4974 break;
4975 }
4976 spin_unlock(&dentry->d_lock);
4977 dput(dentry);
4978
4979 if (!release)
4980 goto out;
4981
4982 release:
4983 /* let's just reuse the same message */
4984 h->action = CEPH_MDS_LEASE_REVOKE_ACK;
4985 ceph_msg_get(msg);
4986 ceph_con_send(&session->s_con, msg);
4987
4988 out:
4989 mutex_unlock(&session->s_mutex);
4990 iput(inode);
4991
4992 ceph_dec_mds_stopping_blocker(mdsc);
4993 return;
4994
4995 bad:
4996 ceph_dec_mds_stopping_blocker(mdsc);
4997
4998 pr_err("corrupt lease message\n");
4999 ceph_msg_dump(msg);
5000 }
5001
5002 void ceph_mdsc_lease_send_msg(struct ceph_mds_session *session,
5003 struct dentry *dentry, char action,
5004 u32 seq)
5005 {
5006 struct ceph_msg *msg;
5007 struct ceph_mds_lease *lease;
5008 struct inode *dir;
5009 int len = sizeof(*lease) + sizeof(u32) + NAME_MAX;
5010
5011 dout("lease_send_msg identry %p %s to mds%d\n",
5012 dentry, ceph_lease_op_name(action), session->s_mds);
5013
5014 msg = ceph_msg_new(CEPH_MSG_CLIENT_LEASE, len, GFP_NOFS, false);
5015 if (!msg)
5016 return;
5017 lease = msg->front.iov_base;
5018 lease->action = action;
5019 lease->seq = cpu_to_le32(seq);
5020
5021 spin_lock(&dentry->d_lock);
5022 dir = d_inode(dentry->d_parent);
5023 lease->ino = cpu_to_le64(ceph_ino(dir));
5024 lease->first = lease->last = cpu_to_le64(ceph_snap(dir));
5025
5026 put_unaligned_le32(dentry->d_name.len, lease + 1);
5027 memcpy((void *)(lease + 1) + 4,
5028 dentry->d_name.name, dentry->d_name.len);
5029 spin_unlock(&dentry->d_lock);
5030
5031 ceph_con_send(&session->s_con, msg);
5032 }
5033
5034 /*
5035 * lock unlock the session, to wait ongoing session activities
5036 */
5037 static void lock_unlock_session(struct ceph_mds_session *s)
5038 {
5039 mutex_lock(&s->s_mutex);
5040 mutex_unlock(&s->s_mutex);
5041 }
5042
5043 static void maybe_recover_session(struct ceph_mds_client *mdsc)
5044 {
5045 struct ceph_fs_client *fsc = mdsc->fsc;
5046
5047 if (!ceph_test_mount_opt(fsc, CLEANRECOVER))
5048 return;
5049
5050 if (READ_ONCE(fsc->mount_state) != CEPH_MOUNT_MOUNTED)
5051 return;
5052
5053 if (!READ_ONCE(fsc->blocklisted))
5054 return;
5055
5056 pr_info("auto reconnect after blocklisted\n");
5057 ceph_force_reconnect(fsc->sb);
5058 }
5059
5060 bool check_session_state(struct ceph_mds_session *s)
5061 {
5062 switch (s->s_state) {
5063 case CEPH_MDS_SESSION_OPEN:
5064 if (s->s_ttl && time_after(jiffies, s->s_ttl)) {
5065 s->s_state = CEPH_MDS_SESSION_HUNG;
5066 pr_info("mds%d hung\n", s->s_mds);
5067 }
5068 break;
5069 case CEPH_MDS_SESSION_CLOSING:
5070 case CEPH_MDS_SESSION_NEW:
5071 case CEPH_MDS_SESSION_RESTARTING:
5072 case CEPH_MDS_SESSION_CLOSED:
5073 case CEPH_MDS_SESSION_REJECTED:
5074 return false;
5075 }
5076
5077 return true;
5078 }
5079
5080 /*
5081 * If the sequence is incremented while we're waiting on a REQUEST_CLOSE reply,
5082 * then we need to retransmit that request.
5083 */
5084 void inc_session_sequence(struct ceph_mds_session *s)
5085 {
5086 lockdep_assert_held(&s->s_mutex);
5087
5088 s->s_seq++;
5089
5090 if (s->s_state == CEPH_MDS_SESSION_CLOSING) {
5091 int ret;
5092
5093 dout("resending session close request for mds%d\n", s->s_mds);
5094 ret = request_close_session(s);
5095 if (ret < 0)
5096 pr_err("unable to close session to mds%d: %d\n",
5097 s->s_mds, ret);
5098 }
5099 }
5100
5101 /*
5102 * delayed work -- periodically trim expired leases, renew caps with mds. If
5103 * the @delay parameter is set to 0 or if it's more than 5 secs, the default
5104 * workqueue delay value of 5 secs will be used.
5105 */
5106 static void schedule_delayed(struct ceph_mds_client *mdsc, unsigned long delay)
5107 {
5108 unsigned long max_delay = HZ * 5;
5109
5110 /* 5 secs default delay */
5111 if (!delay || (delay > max_delay))
5112 delay = max_delay;
5113 schedule_delayed_work(&mdsc->delayed_work,
5114 round_jiffies_relative(delay));
5115 }
5116
5117 static void delayed_work(struct work_struct *work)
5118 {
5119 struct ceph_mds_client *mdsc =
5120 container_of(work, struct ceph_mds_client, delayed_work.work);
5121 unsigned long delay;
5122 int renew_interval;
5123 int renew_caps;
5124 int i;
5125
5126 dout("mdsc delayed_work\n");
5127
5128 if (mdsc->stopping >= CEPH_MDSC_STOPPING_FLUSHED)
5129 return;
5130
5131 mutex_lock(&mdsc->mutex);
5132 renew_interval = mdsc->mdsmap->m_session_timeout >> 2;
5133 renew_caps = time_after_eq(jiffies, HZ*renew_interval +
5134 mdsc->last_renew_caps);
5135 if (renew_caps)
5136 mdsc->last_renew_caps = jiffies;
5137
5138 for (i = 0; i < mdsc->max_sessions; i++) {
5139 struct ceph_mds_session *s = __ceph_lookup_mds_session(mdsc, i);
5140 if (!s)
5141 continue;
5142
5143 if (!check_session_state(s)) {
5144 ceph_put_mds_session(s);
5145 continue;
5146 }
5147 mutex_unlock(&mdsc->mutex);
5148
5149 mutex_lock(&s->s_mutex);
5150 if (renew_caps)
5151 send_renew_caps(mdsc, s);
5152 else
5153 ceph_con_keepalive(&s->s_con);
5154 if (s->s_state == CEPH_MDS_SESSION_OPEN ||
5155 s->s_state == CEPH_MDS_SESSION_HUNG)
5156 ceph_send_cap_releases(mdsc, s);
5157 mutex_unlock(&s->s_mutex);
5158 ceph_put_mds_session(s);
5159
5160 mutex_lock(&mdsc->mutex);
5161 }
5162 mutex_unlock(&mdsc->mutex);
5163
5164 delay = ceph_check_delayed_caps(mdsc);
5165
5166 ceph_queue_cap_reclaim_work(mdsc);
5167
5168 ceph_trim_snapid_map(mdsc);
5169
5170 maybe_recover_session(mdsc);
5171
5172 schedule_delayed(mdsc, delay);
5173 }
5174
5175 int ceph_mdsc_init(struct ceph_fs_client *fsc)
5176
5177 {
5178 struct ceph_mds_client *mdsc;
5179 int err;
5180
5181 mdsc = kzalloc(sizeof(struct ceph_mds_client), GFP_NOFS);
5182 if (!mdsc)
5183 return -ENOMEM;
5184 mdsc->fsc = fsc;
5185 mutex_init(&mdsc->mutex);
5186 mdsc->mdsmap = kzalloc(sizeof(*mdsc->mdsmap), GFP_NOFS);
5187 if (!mdsc->mdsmap) {
5188 err = -ENOMEM;
5189 goto err_mdsc;
5190 }
5191
5192 init_completion(&mdsc->safe_umount_waiters);
5193 spin_lock_init(&mdsc->stopping_lock);
5194 atomic_set(&mdsc->stopping_blockers, 0);
5195 init_completion(&mdsc->stopping_waiter);
5196 init_waitqueue_head(&mdsc->session_close_wq);
5197 INIT_LIST_HEAD(&mdsc->waiting_for_map);
5198 mdsc->quotarealms_inodes = RB_ROOT;
5199 mutex_init(&mdsc->quotarealms_inodes_mutex);
5200 init_rwsem(&mdsc->snap_rwsem);
5201 mdsc->snap_realms = RB_ROOT;
5202 INIT_LIST_HEAD(&mdsc->snap_empty);
5203 spin_lock_init(&mdsc->snap_empty_lock);
5204 mdsc->request_tree = RB_ROOT;
5205 INIT_DELAYED_WORK(&mdsc->delayed_work, delayed_work);
5206 mdsc->last_renew_caps = jiffies;
5207 INIT_LIST_HEAD(&mdsc->cap_delay_list);
5208 INIT_LIST_HEAD(&mdsc->cap_wait_list);
5209 spin_lock_init(&mdsc->cap_delay_lock);
5210 INIT_LIST_HEAD(&mdsc->snap_flush_list);
5211 spin_lock_init(&mdsc->snap_flush_lock);
5212 mdsc->last_cap_flush_tid = 1;
5213 INIT_LIST_HEAD(&mdsc->cap_flush_list);
5214 INIT_LIST_HEAD(&mdsc->cap_dirty_migrating);
5215 spin_lock_init(&mdsc->cap_dirty_lock);
5216 init_waitqueue_head(&mdsc->cap_flushing_wq);
5217 INIT_WORK(&mdsc->cap_reclaim_work, ceph_cap_reclaim_work);
5218 err = ceph_metric_init(&mdsc->metric);
5219 if (err)
5220 goto err_mdsmap;
5221
5222 spin_lock_init(&mdsc->dentry_list_lock);
5223 INIT_LIST_HEAD(&mdsc->dentry_leases);
5224 INIT_LIST_HEAD(&mdsc->dentry_dir_leases);
5225
5226 ceph_caps_init(mdsc);
5227 ceph_adjust_caps_max_min(mdsc, fsc->mount_options);
5228
5229 spin_lock_init(&mdsc->snapid_map_lock);
5230 mdsc->snapid_map_tree = RB_ROOT;
5231 INIT_LIST_HEAD(&mdsc->snapid_map_lru);
5232
5233 init_rwsem(&mdsc->pool_perm_rwsem);
5234 mdsc->pool_perm_tree = RB_ROOT;
5235
5236 strscpy(mdsc->nodename, utsname()->nodename,
5237 sizeof(mdsc->nodename));
5238
5239 fsc->mdsc = mdsc;
5240 return 0;
5241
5242 err_mdsmap:
5243 kfree(mdsc->mdsmap);
5244 err_mdsc:
5245 kfree(mdsc);
5246 return err;
5247 }
5248
5249 /*
5250 * Wait for safe replies on open mds requests. If we time out, drop
5251 * all requests from the tree to avoid dangling dentry refs.
5252 */
5253 static void wait_requests(struct ceph_mds_client *mdsc)
5254 {
5255 struct ceph_options *opts = mdsc->fsc->client->options;
5256 struct ceph_mds_request *req;
5257
5258 mutex_lock(&mdsc->mutex);
5259 if (__get_oldest_req(mdsc)) {
5260 mutex_unlock(&mdsc->mutex);
5261
5262 dout("wait_requests waiting for requests\n");
5263 wait_for_completion_timeout(&mdsc->safe_umount_waiters,
5264 ceph_timeout_jiffies(opts->mount_timeout));
5265
5266 /* tear down remaining requests */
5267 mutex_lock(&mdsc->mutex);
5268 while ((req = __get_oldest_req(mdsc))) {
5269 dout("wait_requests timed out on tid %llu\n",
5270 req->r_tid);
5271 list_del_init(&req->r_wait);
5272 __unregister_request(mdsc, req);
5273 }
5274 }
5275 mutex_unlock(&mdsc->mutex);
5276 dout("wait_requests done\n");
5277 }
5278
5279 void send_flush_mdlog(struct ceph_mds_session *s)
5280 {
5281 struct ceph_msg *msg;
5282
5283 /*
5284 * Pre-luminous MDS crashes when it sees an unknown session request
5285 */
5286 if (!CEPH_HAVE_FEATURE(s->s_con.peer_features, SERVER_LUMINOUS))
5287 return;
5288
5289 mutex_lock(&s->s_mutex);
5290 dout("request mdlog flush to mds%d (%s)s seq %lld\n", s->s_mds,
5291 ceph_session_state_name(s->s_state), s->s_seq);
5292 msg = ceph_create_session_msg(CEPH_SESSION_REQUEST_FLUSH_MDLOG,
5293 s->s_seq);
5294 if (!msg) {
5295 pr_err("failed to request mdlog flush to mds%d (%s) seq %lld\n",
5296 s->s_mds, ceph_session_state_name(s->s_state), s->s_seq);
5297 } else {
5298 ceph_con_send(&s->s_con, msg);
5299 }
5300 mutex_unlock(&s->s_mutex);
5301 }
5302
5303 /*
5304 * called before mount is ro, and before dentries are torn down.
5305 * (hmm, does this still race with new lookups?)
5306 */
5307 void ceph_mdsc_pre_umount(struct ceph_mds_client *mdsc)
5308 {
5309 dout("pre_umount\n");
5310 mdsc->stopping = CEPH_MDSC_STOPPING_BEGIN;
5311
5312 ceph_mdsc_iterate_sessions(mdsc, send_flush_mdlog, true);
5313 ceph_mdsc_iterate_sessions(mdsc, lock_unlock_session, false);
5314 ceph_flush_dirty_caps(mdsc);
5315 wait_requests(mdsc);
5316
5317 /*
5318 * wait for reply handlers to drop their request refs and
5319 * their inode/dcache refs
5320 */
5321 ceph_msgr_flush();
5322
5323 ceph_cleanup_quotarealms_inodes(mdsc);
5324 }
5325
5326 /*
5327 * flush the mdlog and wait for all write mds requests to flush.
5328 */
5329 static void flush_mdlog_and_wait_mdsc_unsafe_requests(struct ceph_mds_client *mdsc,
5330 u64 want_tid)
5331 {
5332 struct ceph_mds_request *req = NULL, *nextreq;
5333 struct ceph_mds_session *last_session = NULL;
5334 struct rb_node *n;
5335
5336 mutex_lock(&mdsc->mutex);
5337 dout("%s want %lld\n", __func__, want_tid);
5338 restart:
5339 req = __get_oldest_req(mdsc);
5340 while (req && req->r_tid <= want_tid) {
5341 /* find next request */
5342 n = rb_next(&req->r_node);
5343 if (n)
5344 nextreq = rb_entry(n, struct ceph_mds_request, r_node);
5345 else
5346 nextreq = NULL;
5347 if (req->r_op != CEPH_MDS_OP_SETFILELOCK &&
5348 (req->r_op & CEPH_MDS_OP_WRITE)) {
5349 struct ceph_mds_session *s = req->r_session;
5350
5351 if (!s) {
5352 req = nextreq;
5353 continue;
5354 }
5355
5356 /* write op */
5357 ceph_mdsc_get_request(req);
5358 if (nextreq)
5359 ceph_mdsc_get_request(nextreq);
5360 s = ceph_get_mds_session(s);
5361 mutex_unlock(&mdsc->mutex);
5362
5363 /* send flush mdlog request to MDS */
5364 if (last_session != s) {
5365 send_flush_mdlog(s);
5366 ceph_put_mds_session(last_session);
5367 last_session = s;
5368 } else {
5369 ceph_put_mds_session(s);
5370 }
5371 dout("%s wait on %llu (want %llu)\n", __func__,
5372 req->r_tid, want_tid);
5373 wait_for_completion(&req->r_safe_completion);
5374
5375 mutex_lock(&mdsc->mutex);
5376 ceph_mdsc_put_request(req);
5377 if (!nextreq)
5378 break; /* next dne before, so we're done! */
5379 if (RB_EMPTY_NODE(&nextreq->r_node)) {
5380 /* next request was removed from tree */
5381 ceph_mdsc_put_request(nextreq);
5382 goto restart;
5383 }
5384 ceph_mdsc_put_request(nextreq); /* won't go away */
5385 }
5386 req = nextreq;
5387 }
5388 mutex_unlock(&mdsc->mutex);
5389 ceph_put_mds_session(last_session);
5390 dout("%s done\n", __func__);
5391 }
5392
5393 void ceph_mdsc_sync(struct ceph_mds_client *mdsc)
5394 {
5395 u64 want_tid, want_flush;
5396
5397 if (READ_ONCE(mdsc->fsc->mount_state) >= CEPH_MOUNT_SHUTDOWN)
5398 return;
5399
5400 dout("sync\n");
5401 mutex_lock(&mdsc->mutex);
5402 want_tid = mdsc->last_tid;
5403 mutex_unlock(&mdsc->mutex);
5404
5405 ceph_flush_dirty_caps(mdsc);
5406 spin_lock(&mdsc->cap_dirty_lock);
5407 want_flush = mdsc->last_cap_flush_tid;
5408 if (!list_empty(&mdsc->cap_flush_list)) {
5409 struct ceph_cap_flush *cf =
5410 list_last_entry(&mdsc->cap_flush_list,
5411 struct ceph_cap_flush, g_list);
5412 cf->wake = true;
5413 }
5414 spin_unlock(&mdsc->cap_dirty_lock);
5415
5416 dout("sync want tid %lld flush_seq %lld\n",
5417 want_tid, want_flush);
5418
5419 flush_mdlog_and_wait_mdsc_unsafe_requests(mdsc, want_tid);
5420 wait_caps_flush(mdsc, want_flush);
5421 }
5422
5423 /*
5424 * true if all sessions are closed, or we force unmount
5425 */
5426 static bool done_closing_sessions(struct ceph_mds_client *mdsc, int skipped)
5427 {
5428 if (READ_ONCE(mdsc->fsc->mount_state) == CEPH_MOUNT_SHUTDOWN)
5429 return true;
5430 return atomic_read(&mdsc->num_sessions) <= skipped;
5431 }
5432
5433 /*
5434 * called after sb is ro or when metadata corrupted.
5435 */
5436 void ceph_mdsc_close_sessions(struct ceph_mds_client *mdsc)
5437 {
5438 struct ceph_options *opts = mdsc->fsc->client->options;
5439 struct ceph_mds_session *session;
5440 int i;
5441 int skipped = 0;
5442
5443 dout("close_sessions\n");
5444
5445 /* close sessions */
5446 mutex_lock(&mdsc->mutex);
5447 for (i = 0; i < mdsc->max_sessions; i++) {
5448 session = __ceph_lookup_mds_session(mdsc, i);
5449 if (!session)
5450 continue;
5451 mutex_unlock(&mdsc->mutex);
5452 mutex_lock(&session->s_mutex);
5453 if (__close_session(mdsc, session) <= 0)
5454 skipped++;
5455 mutex_unlock(&session->s_mutex);
5456 ceph_put_mds_session(session);
5457 mutex_lock(&mdsc->mutex);
5458 }
5459 mutex_unlock(&mdsc->mutex);
5460
5461 dout("waiting for sessions to close\n");
5462 wait_event_timeout(mdsc->session_close_wq,
5463 done_closing_sessions(mdsc, skipped),
5464 ceph_timeout_jiffies(opts->mount_timeout));
5465
5466 /* tear down remaining sessions */
5467 mutex_lock(&mdsc->mutex);
5468 for (i = 0; i < mdsc->max_sessions; i++) {
5469 if (mdsc->sessions[i]) {
5470 session = ceph_get_mds_session(mdsc->sessions[i]);
5471 __unregister_session(mdsc, session);
5472 mutex_unlock(&mdsc->mutex);
5473 mutex_lock(&session->s_mutex);
5474 remove_session_caps(session);
5475 mutex_unlock(&session->s_mutex);
5476 ceph_put_mds_session(session);
5477 mutex_lock(&mdsc->mutex);
5478 }
5479 }
5480 WARN_ON(!list_empty(&mdsc->cap_delay_list));
5481 mutex_unlock(&mdsc->mutex);
5482
5483 ceph_cleanup_snapid_map(mdsc);
5484 ceph_cleanup_global_and_empty_realms(mdsc);
5485
5486 cancel_work_sync(&mdsc->cap_reclaim_work);
5487 cancel_delayed_work_sync(&mdsc->delayed_work); /* cancel timer */
5488
5489 dout("stopped\n");
5490 }
5491
5492 void ceph_mdsc_force_umount(struct ceph_mds_client *mdsc)
5493 {
5494 struct ceph_mds_session *session;
5495 int mds;
5496
5497 dout("force umount\n");
5498
5499 mutex_lock(&mdsc->mutex);
5500 for (mds = 0; mds < mdsc->max_sessions; mds++) {
5501 session = __ceph_lookup_mds_session(mdsc, mds);
5502 if (!session)
5503 continue;
5504
5505 if (session->s_state == CEPH_MDS_SESSION_REJECTED)
5506 __unregister_session(mdsc, session);
5507 __wake_requests(mdsc, &session->s_waiting);
5508 mutex_unlock(&mdsc->mutex);
5509
5510 mutex_lock(&session->s_mutex);
5511 __close_session(mdsc, session);
5512 if (session->s_state == CEPH_MDS_SESSION_CLOSING) {
5513 cleanup_session_requests(mdsc, session);
5514 remove_session_caps(session);
5515 }
5516 mutex_unlock(&session->s_mutex);
5517 ceph_put_mds_session(session);
5518
5519 mutex_lock(&mdsc->mutex);
5520 kick_requests(mdsc, mds);
5521 }
5522 __wake_requests(mdsc, &mdsc->waiting_for_map);
5523 mutex_unlock(&mdsc->mutex);
5524 }
5525
5526 static void ceph_mdsc_stop(struct ceph_mds_client *mdsc)
5527 {
5528 dout("stop\n");
5529 /*
5530 * Make sure the delayed work stopped before releasing
5531 * the resources.
5532 *
5533 * Because the cancel_delayed_work_sync() will only
5534 * guarantee that the work finishes executing. But the
5535 * delayed work will re-arm itself again after that.
5536 */
5537 flush_delayed_work(&mdsc->delayed_work);
5538
5539 if (mdsc->mdsmap)
5540 ceph_mdsmap_destroy(mdsc->mdsmap);
5541 kfree(mdsc->sessions);
5542 ceph_caps_finalize(mdsc);
5543 ceph_pool_perm_destroy(mdsc);
5544 }
5545
5546 void ceph_mdsc_destroy(struct ceph_fs_client *fsc)
5547 {
5548 struct ceph_mds_client *mdsc = fsc->mdsc;
5549 dout("mdsc_destroy %p\n", mdsc);
5550
5551 if (!mdsc)
5552 return;
5553
5554 /* flush out any connection work with references to us */
5555 ceph_msgr_flush();
5556
5557 ceph_mdsc_stop(mdsc);
5558
5559 ceph_metric_destroy(&mdsc->metric);
5560
5561 fsc->mdsc = NULL;
5562 kfree(mdsc);
5563 dout("mdsc_destroy %p done\n", mdsc);
5564 }
5565
5566 void ceph_mdsc_handle_fsmap(struct ceph_mds_client *mdsc, struct ceph_msg *msg)
5567 {
5568 struct ceph_fs_client *fsc = mdsc->fsc;
5569 const char *mds_namespace = fsc->mount_options->mds_namespace;
5570 void *p = msg->front.iov_base;
5571 void *end = p + msg->front.iov_len;
5572 u32 epoch;
5573 u32 num_fs;
5574 u32 mount_fscid = (u32)-1;
5575 int err = -EINVAL;
5576
5577 ceph_decode_need(&p, end, sizeof(u32), bad);
5578 epoch = ceph_decode_32(&p);
5579
5580 dout("handle_fsmap epoch %u\n", epoch);
5581
5582 /* struct_v, struct_cv, map_len, epoch, legacy_client_fscid */
5583 ceph_decode_skip_n(&p, end, 2 + sizeof(u32) * 3, bad);
5584
5585 ceph_decode_32_safe(&p, end, num_fs, bad);
5586 while (num_fs-- > 0) {
5587 void *info_p, *info_end;
5588 u32 info_len;
5589 u32 fscid, namelen;
5590
5591 ceph_decode_need(&p, end, 2 + sizeof(u32), bad);
5592 p += 2; // info_v, info_cv
5593 info_len = ceph_decode_32(&p);
5594 ceph_decode_need(&p, end, info_len, bad);
5595 info_p = p;
5596 info_end = p + info_len;
5597 p = info_end;
5598
5599 ceph_decode_need(&info_p, info_end, sizeof(u32) * 2, bad);
5600 fscid = ceph_decode_32(&info_p);
5601 namelen = ceph_decode_32(&info_p);
5602 ceph_decode_need(&info_p, info_end, namelen, bad);
5603
5604 if (mds_namespace &&
5605 strlen(mds_namespace) == namelen &&
5606 !strncmp(mds_namespace, (char *)info_p, namelen)) {
5607 mount_fscid = fscid;
5608 break;
5609 }
5610 }
5611
5612 ceph_monc_got_map(&fsc->client->monc, CEPH_SUB_FSMAP, epoch);
5613 if (mount_fscid != (u32)-1) {
5614 fsc->client->monc.fs_cluster_id = mount_fscid;
5615 ceph_monc_want_map(&fsc->client->monc, CEPH_SUB_MDSMAP,
5616 0, true);
5617 ceph_monc_renew_subs(&fsc->client->monc);
5618 } else {
5619 err = -ENOENT;
5620 goto err_out;
5621 }
5622 return;
5623
5624 bad:
5625 pr_err("error decoding fsmap %d. Shutting down mount.\n", err);
5626 ceph_umount_begin(mdsc->fsc->sb);
5627 ceph_msg_dump(msg);
5628 err_out:
5629 mutex_lock(&mdsc->mutex);
5630 mdsc->mdsmap_err = err;
5631 __wake_requests(mdsc, &mdsc->waiting_for_map);
5632 mutex_unlock(&mdsc->mutex);
5633 }
5634
5635 /*
5636 * handle mds map update.
5637 */
5638 void ceph_mdsc_handle_mdsmap(struct ceph_mds_client *mdsc, struct ceph_msg *msg)
5639 {
5640 u32 epoch;
5641 u32 maplen;
5642 void *p = msg->front.iov_base;
5643 void *end = p + msg->front.iov_len;
5644 struct ceph_mdsmap *newmap, *oldmap;
5645 struct ceph_fsid fsid;
5646 int err = -EINVAL;
5647
5648 ceph_decode_need(&p, end, sizeof(fsid)+2*sizeof(u32), bad);
5649 ceph_decode_copy(&p, &fsid, sizeof(fsid));
5650 if (ceph_check_fsid(mdsc->fsc->client, &fsid) < 0)
5651 return;
5652 epoch = ceph_decode_32(&p);
5653 maplen = ceph_decode_32(&p);
5654 dout("handle_map epoch %u len %d\n", epoch, (int)maplen);
5655
5656 /* do we need it? */
5657 mutex_lock(&mdsc->mutex);
5658 if (mdsc->mdsmap && epoch <= mdsc->mdsmap->m_epoch) {
5659 dout("handle_map epoch %u <= our %u\n",
5660 epoch, mdsc->mdsmap->m_epoch);
5661 mutex_unlock(&mdsc->mutex);
5662 return;
5663 }
5664
5665 newmap = ceph_mdsmap_decode(&p, end, ceph_msgr2(mdsc->fsc->client));
5666 if (IS_ERR(newmap)) {
5667 err = PTR_ERR(newmap);
5668 goto bad_unlock;
5669 }
5670
5671 /* swap into place */
5672 if (mdsc->mdsmap) {
5673 oldmap = mdsc->mdsmap;
5674 mdsc->mdsmap = newmap;
5675 check_new_map(mdsc, newmap, oldmap);
5676 ceph_mdsmap_destroy(oldmap);
5677 } else {
5678 mdsc->mdsmap = newmap; /* first mds map */
5679 }
5680 mdsc->fsc->max_file_size = min((loff_t)mdsc->mdsmap->m_max_file_size,
5681 MAX_LFS_FILESIZE);
5682
5683 __wake_requests(mdsc, &mdsc->waiting_for_map);
5684 ceph_monc_got_map(&mdsc->fsc->client->monc, CEPH_SUB_MDSMAP,
5685 mdsc->mdsmap->m_epoch);
5686
5687 mutex_unlock(&mdsc->mutex);
5688 schedule_delayed(mdsc, 0);
5689 return;
5690
5691 bad_unlock:
5692 mutex_unlock(&mdsc->mutex);
5693 bad:
5694 pr_err("error decoding mdsmap %d. Shutting down mount.\n", err);
5695 ceph_umount_begin(mdsc->fsc->sb);
5696 ceph_msg_dump(msg);
5697 return;
5698 }
5699
5700 static struct ceph_connection *mds_get_con(struct ceph_connection *con)
5701 {
5702 struct ceph_mds_session *s = con->private;
5703
5704 if (ceph_get_mds_session(s))
5705 return con;
5706 return NULL;
5707 }
5708
5709 static void mds_put_con(struct ceph_connection *con)
5710 {
5711 struct ceph_mds_session *s = con->private;
5712
5713 ceph_put_mds_session(s);
5714 }
5715
5716 /*
5717 * if the client is unresponsive for long enough, the mds will kill
5718 * the session entirely.
5719 */
5720 static void mds_peer_reset(struct ceph_connection *con)
5721 {
5722 struct ceph_mds_session *s = con->private;
5723 struct ceph_mds_client *mdsc = s->s_mdsc;
5724
5725 pr_warn("mds%d closed our session\n", s->s_mds);
5726 if (READ_ONCE(mdsc->fsc->mount_state) != CEPH_MOUNT_FENCE_IO)
5727 send_mds_reconnect(mdsc, s);
5728 }
5729
5730 static void mds_dispatch(struct ceph_connection *con, struct ceph_msg *msg)
5731 {
5732 struct ceph_mds_session *s = con->private;
5733 struct ceph_mds_client *mdsc = s->s_mdsc;
5734 int type = le16_to_cpu(msg->hdr.type);
5735
5736 mutex_lock(&mdsc->mutex);
5737 if (__verify_registered_session(mdsc, s) < 0) {
5738 mutex_unlock(&mdsc->mutex);
5739 goto out;
5740 }
5741 mutex_unlock(&mdsc->mutex);
5742
5743 switch (type) {
5744 case CEPH_MSG_MDS_MAP:
5745 ceph_mdsc_handle_mdsmap(mdsc, msg);
5746 break;
5747 case CEPH_MSG_FS_MAP_USER:
5748 ceph_mdsc_handle_fsmap(mdsc, msg);
5749 break;
5750 case CEPH_MSG_CLIENT_SESSION:
5751 handle_session(s, msg);
5752 break;
5753 case CEPH_MSG_CLIENT_REPLY:
5754 handle_reply(s, msg);
5755 break;
5756 case CEPH_MSG_CLIENT_REQUEST_FORWARD:
5757 handle_forward(mdsc, s, msg);
5758 break;
5759 case CEPH_MSG_CLIENT_CAPS:
5760 ceph_handle_caps(s, msg);
5761 break;
5762 case CEPH_MSG_CLIENT_SNAP:
5763 ceph_handle_snap(mdsc, s, msg);
5764 break;
5765 case CEPH_MSG_CLIENT_LEASE:
5766 handle_lease(mdsc, s, msg);
5767 break;
5768 case CEPH_MSG_CLIENT_QUOTA:
5769 ceph_handle_quota(mdsc, s, msg);
5770 break;
5771
5772 default:
5773 pr_err("received unknown message type %d %s\n", type,
5774 ceph_msg_type_name(type));
5775 }
5776 out:
5777 ceph_msg_put(msg);
5778 }
5779
5780 /*
5781 * authentication
5782 */
5783
5784 /*
5785 * Note: returned pointer is the address of a structure that's
5786 * managed separately. Caller must *not* attempt to free it.
5787 */
5788 static struct ceph_auth_handshake *
5789 mds_get_authorizer(struct ceph_connection *con, int *proto, int force_new)
5790 {
5791 struct ceph_mds_session *s = con->private;
5792 struct ceph_mds_client *mdsc = s->s_mdsc;
5793 struct ceph_auth_client *ac = mdsc->fsc->client->monc.auth;
5794 struct ceph_auth_handshake *auth = &s->s_auth;
5795 int ret;
5796
5797 ret = __ceph_auth_get_authorizer(ac, auth, CEPH_ENTITY_TYPE_MDS,
5798 force_new, proto, NULL, NULL);
5799 if (ret)
5800 return ERR_PTR(ret);
5801
5802 return auth;
5803 }
5804
5805 static int mds_add_authorizer_challenge(struct ceph_connection *con,
5806 void *challenge_buf, int challenge_buf_len)
5807 {
5808 struct ceph_mds_session *s = con->private;
5809 struct ceph_mds_client *mdsc = s->s_mdsc;
5810 struct ceph_auth_client *ac = mdsc->fsc->client->monc.auth;
5811
5812 return ceph_auth_add_authorizer_challenge(ac, s->s_auth.authorizer,
5813 challenge_buf, challenge_buf_len);
5814 }
5815
5816 static int mds_verify_authorizer_reply(struct ceph_connection *con)
5817 {
5818 struct ceph_mds_session *s = con->private;
5819 struct ceph_mds_client *mdsc = s->s_mdsc;
5820 struct ceph_auth_client *ac = mdsc->fsc->client->monc.auth;
5821 struct ceph_auth_handshake *auth = &s->s_auth;
5822
5823 return ceph_auth_verify_authorizer_reply(ac, auth->authorizer,
5824 auth->authorizer_reply_buf, auth->authorizer_reply_buf_len,
5825 NULL, NULL, NULL, NULL);
5826 }
5827
5828 static int mds_invalidate_authorizer(struct ceph_connection *con)
5829 {
5830 struct ceph_mds_session *s = con->private;
5831 struct ceph_mds_client *mdsc = s->s_mdsc;
5832 struct ceph_auth_client *ac = mdsc->fsc->client->monc.auth;
5833
5834 ceph_auth_invalidate_authorizer(ac, CEPH_ENTITY_TYPE_MDS);
5835
5836 return ceph_monc_validate_auth(&mdsc->fsc->client->monc);
5837 }
5838
5839 static int mds_get_auth_request(struct ceph_connection *con,
5840 void *buf, int *buf_len,
5841 void **authorizer, int *authorizer_len)
5842 {
5843 struct ceph_mds_session *s = con->private;
5844 struct ceph_auth_client *ac = s->s_mdsc->fsc->client->monc.auth;
5845 struct ceph_auth_handshake *auth = &s->s_auth;
5846 int ret;
5847
5848 ret = ceph_auth_get_authorizer(ac, auth, CEPH_ENTITY_TYPE_MDS,
5849 buf, buf_len);
5850 if (ret)
5851 return ret;
5852
5853 *authorizer = auth->authorizer_buf;
5854 *authorizer_len = auth->authorizer_buf_len;
5855 return 0;
5856 }
5857
5858 static int mds_handle_auth_reply_more(struct ceph_connection *con,
5859 void *reply, int reply_len,
5860 void *buf, int *buf_len,
5861 void **authorizer, int *authorizer_len)
5862 {
5863 struct ceph_mds_session *s = con->private;
5864 struct ceph_auth_client *ac = s->s_mdsc->fsc->client->monc.auth;
5865 struct ceph_auth_handshake *auth = &s->s_auth;
5866 int ret;
5867
5868 ret = ceph_auth_handle_svc_reply_more(ac, auth, reply, reply_len,
5869 buf, buf_len);
5870 if (ret)
5871 return ret;
5872
5873 *authorizer = auth->authorizer_buf;
5874 *authorizer_len = auth->authorizer_buf_len;
5875 return 0;
5876 }
5877
5878 static int mds_handle_auth_done(struct ceph_connection *con,
5879 u64 global_id, void *reply, int reply_len,
5880 u8 *session_key, int *session_key_len,
5881 u8 *con_secret, int *con_secret_len)
5882 {
5883 struct ceph_mds_session *s = con->private;
5884 struct ceph_auth_client *ac = s->s_mdsc->fsc->client->monc.auth;
5885 struct ceph_auth_handshake *auth = &s->s_auth;
5886
5887 return ceph_auth_handle_svc_reply_done(ac, auth, reply, reply_len,
5888 session_key, session_key_len,
5889 con_secret, con_secret_len);
5890 }
5891
5892 static int mds_handle_auth_bad_method(struct ceph_connection *con,
5893 int used_proto, int result,
5894 const int *allowed_protos, int proto_cnt,
5895 const int *allowed_modes, int mode_cnt)
5896 {
5897 struct ceph_mds_session *s = con->private;
5898 struct ceph_mon_client *monc = &s->s_mdsc->fsc->client->monc;
5899 int ret;
5900
5901 if (ceph_auth_handle_bad_authorizer(monc->auth, CEPH_ENTITY_TYPE_MDS,
5902 used_proto, result,
5903 allowed_protos, proto_cnt,
5904 allowed_modes, mode_cnt)) {
5905 ret = ceph_monc_validate_auth(monc);
5906 if (ret)
5907 return ret;
5908 }
5909
5910 return -EACCES;
5911 }
5912
5913 static struct ceph_msg *mds_alloc_msg(struct ceph_connection *con,
5914 struct ceph_msg_header *hdr, int *skip)
5915 {
5916 struct ceph_msg *msg;
5917 int type = (int) le16_to_cpu(hdr->type);
5918 int front_len = (int) le32_to_cpu(hdr->front_len);
5919
5920 if (con->in_msg)
5921 return con->in_msg;
5922
5923 *skip = 0;
5924 msg = ceph_msg_new(type, front_len, GFP_NOFS, false);
5925 if (!msg) {
5926 pr_err("unable to allocate msg type %d len %d\n",
5927 type, front_len);
5928 return NULL;
5929 }
5930
5931 return msg;
5932 }
5933
5934 static int mds_sign_message(struct ceph_msg *msg)
5935 {
5936 struct ceph_mds_session *s = msg->con->private;
5937 struct ceph_auth_handshake *auth = &s->s_auth;
5938
5939 return ceph_auth_sign_message(auth, msg);
5940 }
5941
5942 static int mds_check_message_signature(struct ceph_msg *msg)
5943 {
5944 struct ceph_mds_session *s = msg->con->private;
5945 struct ceph_auth_handshake *auth = &s->s_auth;
5946
5947 return ceph_auth_check_message_signature(auth, msg);
5948 }
5949
5950 static const struct ceph_connection_operations mds_con_ops = {
5951 .get = mds_get_con,
5952 .put = mds_put_con,
5953 .alloc_msg = mds_alloc_msg,
5954 .dispatch = mds_dispatch,
5955 .peer_reset = mds_peer_reset,
5956 .get_authorizer = mds_get_authorizer,
5957 .add_authorizer_challenge = mds_add_authorizer_challenge,
5958 .verify_authorizer_reply = mds_verify_authorizer_reply,
5959 .invalidate_authorizer = mds_invalidate_authorizer,
5960 .sign_message = mds_sign_message,
5961 .check_message_signature = mds_check_message_signature,
5962 .get_auth_request = mds_get_auth_request,
5963 .handle_auth_reply_more = mds_handle_auth_reply_more,
5964 .handle_auth_done = mds_handle_auth_done,
5965 .handle_auth_bad_method = mds_handle_auth_bad_method,
5966 };
5967
5968 /* eof */
Mirror https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git