1 // SPDX-License-Identifier: GPL-2.0
2 #include <linux/ceph/ceph_debug.h>
3
4 #include <linux/exportfs.h>
5 #include <linux/slab.h>
6 #include <asm/unaligned.h>
7
8 #include "super.h"
9 #include "mds_client.h"
10
11 /*
12 * Basic fh
13 */
14 struct ceph_nfs_fh {
15 u64 ino;
16 } __attribute__ ((packed));
17
18 /*
19 * Larger fh that includes parent ino.
20 */
21 struct ceph_nfs_confh {
22 u64 ino, parent_ino;
23 } __attribute__ ((packed));
24
25 /*
26 * fh for snapped inode
27 */
28 struct ceph_nfs_snapfh {
29 u64 ino;
30 u64 snapid;
31 u64 parent_ino;
32 u32 hash;
33 } __attribute__ ((packed));
34
ceph_encode_snapfh(struct inode * inode,u32 * rawfh,int * max_len,struct inode * parent_inode)35 static int ceph_encode_snapfh(struct inode *inode, u32 *rawfh, int *max_len,
36 struct inode *parent_inode)
37 {
38 static const int snap_handle_length =
39 sizeof(struct ceph_nfs_snapfh) >> 2;
40 struct ceph_nfs_snapfh *sfh = (void *)rawfh;
41 u64 snapid = ceph_snap(inode);
42 int ret;
43 bool no_parent = true;
44
45 if (*max_len < snap_handle_length) {
46 *max_len = snap_handle_length;
47 ret = FILEID_INVALID;
48 goto out;
49 }
50
51 ret = -EINVAL;
52 if (snapid != CEPH_SNAPDIR) {
53 struct inode *dir;
54 struct dentry *dentry = d_find_alias(inode);
55 if (!dentry)
56 goto out;
57
58 rcu_read_lock();
59 dir = d_inode_rcu(dentry->d_parent);
60 if (ceph_snap(dir) != CEPH_SNAPDIR) {
61 sfh->parent_ino = ceph_ino(dir);
62 sfh->hash = ceph_dentry_hash(dir, dentry);
63 no_parent = false;
64 }
65 rcu_read_unlock();
66 dput(dentry);
67 }
68
69 if (no_parent) {
70 if (!S_ISDIR(inode->i_mode))
71 goto out;
72 sfh->parent_ino = sfh->ino;
73 sfh->hash = 0;
74 }
75 sfh->ino = ceph_ino(inode);
76 sfh->snapid = snapid;
77
78 *max_len = snap_handle_length;
79 ret = FILEID_BTRFS_WITH_PARENT;
80 out:
81 dout("encode_snapfh %llx.%llx ret=%d\n", ceph_vinop(inode), ret);
82 return ret;
83 }
84
ceph_encode_fh(struct inode * inode,u32 * rawfh,int * max_len,struct inode * parent_inode)85 static int ceph_encode_fh(struct inode *inode, u32 *rawfh, int *max_len,
86 struct inode *parent_inode)
87 {
88 static const int handle_length =
89 sizeof(struct ceph_nfs_fh) >> 2;
90 static const int connected_handle_length =
91 sizeof(struct ceph_nfs_confh) >> 2;
92 int type;
93
94 if (ceph_snap(inode) != CEPH_NOSNAP)
95 return ceph_encode_snapfh(inode, rawfh, max_len, parent_inode);
96
97 if (parent_inode && (*max_len < connected_handle_length)) {
98 *max_len = connected_handle_length;
99 return FILEID_INVALID;
100 } else if (*max_len < handle_length) {
101 *max_len = handle_length;
102 return FILEID_INVALID;
103 }
104
105 if (parent_inode) {
106 struct ceph_nfs_confh *cfh = (void *)rawfh;
107 dout("encode_fh %llx with parent %llx\n",
108 ceph_ino(inode), ceph_ino(parent_inode));
109 cfh->ino = ceph_ino(inode);
110 cfh->parent_ino = ceph_ino(parent_inode);
111 *max_len = connected_handle_length;
112 type = FILEID_INO32_GEN_PARENT;
113 } else {
114 struct ceph_nfs_fh *fh = (void *)rawfh;
115 dout("encode_fh %llx\n", ceph_ino(inode));
116 fh->ino = ceph_ino(inode);
117 *max_len = handle_length;
118 type = FILEID_INO32_GEN;
119 }
120 return type;
121 }
122
__lookup_inode(struct super_block * sb,u64 ino)123 static struct inode *__lookup_inode(struct super_block *sb, u64 ino)
124 {
125 struct ceph_mds_client *mdsc = ceph_sb_to_client(sb)->mdsc;
126 struct inode *inode;
127 struct ceph_vino vino;
128 int err;
129
130 vino.ino = ino;
131 vino.snap = CEPH_NOSNAP;
132
133 if (ceph_vino_is_reserved(vino))
134 return ERR_PTR(-ESTALE);
135
136 inode = ceph_find_inode(sb, vino);
137 if (!inode) {
138 struct ceph_mds_request *req;
139 int mask;
140
141 req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_LOOKUPINO,
142 USE_ANY_MDS);
143 if (IS_ERR(req))
144 return ERR_CAST(req);
145
146 mask = CEPH_STAT_CAP_INODE;
147 if (ceph_security_xattr_wanted(d_inode(sb->s_root)))
148 mask |= CEPH_CAP_XATTR_SHARED;
149 req->r_args.lookupino.mask = cpu_to_le32(mask);
150
151 req->r_ino1 = vino;
152 req->r_num_caps = 1;
153 err = ceph_mdsc_do_request(mdsc, NULL, req);
154 inode = req->r_target_inode;
155 if (inode)
156 ihold(inode);
157 ceph_mdsc_put_request(req);
158 if (!inode)
159 return err < 0 ? ERR_PTR(err) : ERR_PTR(-ESTALE);
160 } else {
161 if (ceph_inode_is_shutdown(inode)) {
162 iput(inode);
163 return ERR_PTR(-ESTALE);
164 }
165 }
166 return inode;
167 }
168
ceph_lookup_inode(struct super_block * sb,u64 ino)169 struct inode *ceph_lookup_inode(struct super_block *sb, u64 ino)
170 {
171 struct inode *inode = __lookup_inode(sb, ino);
172 if (IS_ERR(inode))
173 return inode;
174 if (inode->i_nlink == 0) {
175 iput(inode);
176 return ERR_PTR(-ESTALE);
177 }
178 return inode;
179 }
180
__fh_to_dentry(struct super_block * sb,u64 ino)181 static struct dentry *__fh_to_dentry(struct super_block *sb, u64 ino)
182 {
183 struct inode *inode = __lookup_inode(sb, ino);
184 struct ceph_inode_info *ci = ceph_inode(inode);
185 int err;
186
187 if (IS_ERR(inode))
188 return ERR_CAST(inode);
189 /* We need LINK caps to reliably check i_nlink */
190 err = ceph_do_getattr(inode, CEPH_CAP_LINK_SHARED, false);
191 if (err) {
192 iput(inode);
193 return ERR_PTR(err);
194 }
195 /* -ESTALE if inode as been unlinked and no file is open */
196 if ((inode->i_nlink == 0) && !__ceph_is_file_opened(ci)) {
197 iput(inode);
198 return ERR_PTR(-ESTALE);
199 }
200 return d_obtain_alias(inode);
201 }
202
__snapfh_to_dentry(struct super_block * sb,struct ceph_nfs_snapfh * sfh,bool want_parent)203 static struct dentry *__snapfh_to_dentry(struct super_block *sb,
204 struct ceph_nfs_snapfh *sfh,
205 bool want_parent)
206 {
207 struct ceph_mds_client *mdsc = ceph_sb_to_client(sb)->mdsc;
208 struct ceph_mds_request *req;
209 struct inode *inode;
210 struct ceph_vino vino;
211 int mask;
212 int err;
213 bool unlinked = false;
214
215 if (want_parent) {
216 vino.ino = sfh->parent_ino;
217 if (sfh->snapid == CEPH_SNAPDIR)
218 vino.snap = CEPH_NOSNAP;
219 else if (sfh->ino == sfh->parent_ino)
220 vino.snap = CEPH_SNAPDIR;
221 else
222 vino.snap = sfh->snapid;
223 } else {
224 vino.ino = sfh->ino;
225 vino.snap = sfh->snapid;
226 }
227
228 if (ceph_vino_is_reserved(vino))
229 return ERR_PTR(-ESTALE);
230
231 inode = ceph_find_inode(sb, vino);
232 if (inode) {
233 if (ceph_inode_is_shutdown(inode)) {
234 iput(inode);
235 return ERR_PTR(-ESTALE);
236 }
237 return d_obtain_alias(inode);
238 }
239
240 req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_LOOKUPINO,
241 USE_ANY_MDS);
242 if (IS_ERR(req))
243 return ERR_CAST(req);
244
245 mask = CEPH_STAT_CAP_INODE;
246 if (ceph_security_xattr_wanted(d_inode(sb->s_root)))
247 mask |= CEPH_CAP_XATTR_SHARED;
248 req->r_args.lookupino.mask = cpu_to_le32(mask);
249 if (vino.snap < CEPH_NOSNAP) {
250 req->r_args.lookupino.snapid = cpu_to_le64(vino.snap);
251 if (!want_parent && sfh->ino != sfh->parent_ino) {
252 req->r_args.lookupino.parent =
253 cpu_to_le64(sfh->parent_ino);
254 req->r_args.lookupino.hash =
255 cpu_to_le32(sfh->hash);
256 }
257 }
258
259 req->r_ino1 = vino;
260 req->r_num_caps = 1;
261 err = ceph_mdsc_do_request(mdsc, NULL, req);
262 inode = req->r_target_inode;
263 if (inode) {
264 if (vino.snap == CEPH_SNAPDIR) {
265 if (inode->i_nlink == 0)
266 unlinked = true;
267 inode = ceph_get_snapdir(inode);
268 } else if (ceph_snap(inode) == vino.snap) {
269 ihold(inode);
270 } else {
271 /* mds does not support lookup snapped inode */
272 inode = ERR_PTR(-EOPNOTSUPP);
273 }
274 } else {
275 inode = ERR_PTR(-ESTALE);
276 }
277 ceph_mdsc_put_request(req);
278
279 if (want_parent) {
280 dout("snapfh_to_parent %llx.%llx\n err=%d\n",
281 vino.ino, vino.snap, err);
282 } else {
283 dout("snapfh_to_dentry %llx.%llx parent %llx hash %x err=%d",
284 vino.ino, vino.snap, sfh->parent_ino, sfh->hash, err);
285 }
286 if (IS_ERR(inode))
287 return ERR_CAST(inode);
288 /* see comments in ceph_get_parent() */
289 return unlinked ? d_obtain_root(inode) : d_obtain_alias(inode);
290 }
291
292 /*
293 * convert regular fh to dentry
294 */
ceph_fh_to_dentry(struct super_block * sb,struct fid * fid,int fh_len,int fh_type)295 static struct dentry *ceph_fh_to_dentry(struct super_block *sb,
296 struct fid *fid,
297 int fh_len, int fh_type)
298 {
299 struct ceph_nfs_fh *fh = (void *)fid->raw;
300
301 if (fh_type == FILEID_BTRFS_WITH_PARENT) {
302 struct ceph_nfs_snapfh *sfh = (void *)fid->raw;
303 return __snapfh_to_dentry(sb, sfh, false);
304 }
305
306 if (fh_type != FILEID_INO32_GEN &&
307 fh_type != FILEID_INO32_GEN_PARENT)
308 return NULL;
309 if (fh_len < sizeof(*fh) / 4)
310 return NULL;
311
312 dout("fh_to_dentry %llx\n", fh->ino);
313 return __fh_to_dentry(sb, fh->ino);
314 }
315
__get_parent(struct super_block * sb,struct dentry * child,u64 ino)316 static struct dentry *__get_parent(struct super_block *sb,
317 struct dentry *child, u64 ino)
318 {
319 struct ceph_mds_client *mdsc = ceph_sb_to_client(sb)->mdsc;
320 struct ceph_mds_request *req;
321 struct inode *inode;
322 int mask;
323 int err;
324
325 req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_LOOKUPPARENT,
326 USE_ANY_MDS);
327 if (IS_ERR(req))
328 return ERR_CAST(req);
329
330 if (child) {
331 req->r_inode = d_inode(child);
332 ihold(d_inode(child));
333 } else {
334 req->r_ino1 = (struct ceph_vino) {
335 .ino = ino,
336 .snap = CEPH_NOSNAP,
337 };
338 }
339
340 mask = CEPH_STAT_CAP_INODE;
341 if (ceph_security_xattr_wanted(d_inode(sb->s_root)))
342 mask |= CEPH_CAP_XATTR_SHARED;
343 req->r_args.getattr.mask = cpu_to_le32(mask);
344
345 req->r_num_caps = 1;
346 err = ceph_mdsc_do_request(mdsc, NULL, req);
347 if (err) {
348 ceph_mdsc_put_request(req);
349 return ERR_PTR(err);
350 }
351
352 inode = req->r_target_inode;
353 if (inode)
354 ihold(inode);
355 ceph_mdsc_put_request(req);
356 if (!inode)
357 return ERR_PTR(-ENOENT);
358
359 return d_obtain_alias(inode);
360 }
361
ceph_get_parent(struct dentry * child)362 static struct dentry *ceph_get_parent(struct dentry *child)
363 {
364 struct inode *inode = d_inode(child);
365 struct dentry *dn;
366
367 if (ceph_snap(inode) != CEPH_NOSNAP) {
368 struct inode* dir;
369 bool unlinked = false;
370 /* do not support non-directory */
371 if (!d_is_dir(child)) {
372 dn = ERR_PTR(-EINVAL);
373 goto out;
374 }
375 dir = __lookup_inode(inode->i_sb, ceph_ino(inode));
376 if (IS_ERR(dir)) {
377 dn = ERR_CAST(dir);
378 goto out;
379 }
380 /* There can be multiple paths to access snapped inode.
381 * For simplicity, treat snapdir of head inode as parent */
382 if (ceph_snap(inode) != CEPH_SNAPDIR) {
383 struct inode *snapdir = ceph_get_snapdir(dir);
384 if (dir->i_nlink == 0)
385 unlinked = true;
386 iput(dir);
387 if (IS_ERR(snapdir)) {
388 dn = ERR_CAST(snapdir);
389 goto out;
390 }
391 dir = snapdir;
392 }
393 /* If directory has already been deleted, futher get_parent
394 * will fail. Do not mark snapdir dentry as disconnected,
395 * this prevent exportfs from doing futher get_parent. */
396 if (unlinked)
397 dn = d_obtain_root(dir);
398 else
399 dn = d_obtain_alias(dir);
400 } else {
401 dn = __get_parent(child->d_sb, child, 0);
402 }
403 out:
404 dout("get_parent %p ino %llx.%llx err=%ld\n",
405 child, ceph_vinop(inode), (long)PTR_ERR_OR_ZERO(dn));
406 return dn;
407 }
408
409 /*
410 * convert regular fh to parent
411 */
ceph_fh_to_parent(struct super_block * sb,struct fid * fid,int fh_len,int fh_type)412 static struct dentry *ceph_fh_to_parent(struct super_block *sb,
413 struct fid *fid,
414 int fh_len, int fh_type)
415 {
416 struct ceph_nfs_confh *cfh = (void *)fid->raw;
417 struct dentry *dentry;
418
419 if (fh_type == FILEID_BTRFS_WITH_PARENT) {
420 struct ceph_nfs_snapfh *sfh = (void *)fid->raw;
421 return __snapfh_to_dentry(sb, sfh, true);
422 }
423
424 if (fh_type != FILEID_INO32_GEN_PARENT)
425 return NULL;
426 if (fh_len < sizeof(*cfh) / 4)
427 return NULL;
428
429 dout("fh_to_parent %llx\n", cfh->parent_ino);
430 dentry = __get_parent(sb, NULL, cfh->ino);
431 if (unlikely(dentry == ERR_PTR(-ENOENT)))
432 dentry = __fh_to_dentry(sb, cfh->parent_ino);
433 return dentry;
434 }
435
__get_snap_name(struct dentry * parent,char * name,struct dentry * child)436 static int __get_snap_name(struct dentry *parent, char *name,
437 struct dentry *child)
438 {
439 struct inode *inode = d_inode(child);
440 struct inode *dir = d_inode(parent);
441 struct ceph_fs_client *fsc = ceph_inode_to_client(inode);
442 struct ceph_mds_request *req = NULL;
443 char *last_name = NULL;
444 unsigned next_offset = 2;
445 int err = -EINVAL;
446
447 if (ceph_ino(inode) != ceph_ino(dir))
448 goto out;
449 if (ceph_snap(inode) == CEPH_SNAPDIR) {
450 if (ceph_snap(dir) == CEPH_NOSNAP) {
451 strcpy(name, fsc->mount_options->snapdir_name);
452 err = 0;
453 }
454 goto out;
455 }
456 if (ceph_snap(dir) != CEPH_SNAPDIR)
457 goto out;
458
459 while (1) {
460 struct ceph_mds_reply_info_parsed *rinfo;
461 struct ceph_mds_reply_dir_entry *rde;
462 int i;
463
464 req = ceph_mdsc_create_request(fsc->mdsc, CEPH_MDS_OP_LSSNAP,
465 USE_AUTH_MDS);
466 if (IS_ERR(req)) {
467 err = PTR_ERR(req);
468 req = NULL;
469 goto out;
470 }
471 err = ceph_alloc_readdir_reply_buffer(req, inode);
472 if (err)
473 goto out;
474
475 req->r_direct_mode = USE_AUTH_MDS;
476 req->r_readdir_offset = next_offset;
477 req->r_args.readdir.flags =
478 cpu_to_le16(CEPH_READDIR_REPLY_BITFLAGS);
479 if (last_name) {
480 req->r_path2 = last_name;
481 last_name = NULL;
482 }
483
484 req->r_inode = dir;
485 ihold(dir);
486 req->r_dentry = dget(parent);
487
488 inode_lock(dir);
489 err = ceph_mdsc_do_request(fsc->mdsc, NULL, req);
490 inode_unlock(dir);
491
492 if (err < 0)
493 goto out;
494
495 rinfo = &req->r_reply_info;
496 for (i = 0; i < rinfo->dir_nr; i++) {
497 rde = rinfo->dir_entries + i;
498 BUG_ON(!rde->inode.in);
499 if (ceph_snap(inode) ==
500 le64_to_cpu(rde->inode.in->snapid)) {
501 memcpy(name, rde->name, rde->name_len);
502 name[rde->name_len] = '\0';
503 err = 0;
504 goto out;
505 }
506 }
507
508 if (rinfo->dir_end)
509 break;
510
511 BUG_ON(rinfo->dir_nr <= 0);
512 rde = rinfo->dir_entries + (rinfo->dir_nr - 1);
513 next_offset += rinfo->dir_nr;
514 last_name = kstrndup(rde->name, rde->name_len, GFP_KERNEL);
515 if (!last_name) {
516 err = -ENOMEM;
517 goto out;
518 }
519
520 ceph_mdsc_put_request(req);
521 req = NULL;
522 }
523 err = -ENOENT;
524 out:
525 if (req)
526 ceph_mdsc_put_request(req);
527 kfree(last_name);
528 dout("get_snap_name %p ino %llx.%llx err=%d\n",
529 child, ceph_vinop(inode), err);
530 return err;
531 }
532
ceph_get_name(struct dentry * parent,char * name,struct dentry * child)533 static int ceph_get_name(struct dentry *parent, char *name,
534 struct dentry *child)
535 {
536 struct ceph_mds_client *mdsc;
537 struct ceph_mds_request *req;
538 struct inode *inode = d_inode(child);
539 int err;
540
541 if (ceph_snap(inode) != CEPH_NOSNAP)
542 return __get_snap_name(parent, name, child);
543
544 mdsc = ceph_inode_to_client(inode)->mdsc;
545 req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_LOOKUPNAME,
546 USE_ANY_MDS);
547 if (IS_ERR(req))
548 return PTR_ERR(req);
549
550 inode_lock(d_inode(parent));
551
552 req->r_inode = inode;
553 ihold(inode);
554 req->r_ino2 = ceph_vino(d_inode(parent));
555 req->r_parent = d_inode(parent);
556 ihold(req->r_parent);
557 set_bit(CEPH_MDS_R_PARENT_LOCKED, &req->r_req_flags);
558 req->r_num_caps = 2;
559 err = ceph_mdsc_do_request(mdsc, NULL, req);
560
561 inode_unlock(d_inode(parent));
562
563 if (!err) {
564 struct ceph_mds_reply_info_parsed *rinfo = &req->r_reply_info;
565 memcpy(name, rinfo->dname, rinfo->dname_len);
566 name[rinfo->dname_len] = 0;
567 dout("get_name %p ino %llx.%llx name %s\n",
568 child, ceph_vinop(inode), name);
569 } else {
570 dout("get_name %p ino %llx.%llx err %d\n",
571 child, ceph_vinop(inode), err);
572 }
573
574 ceph_mdsc_put_request(req);
575 return err;
576 }
577
578 const struct export_operations ceph_export_ops = {
579 .encode_fh = ceph_encode_fh,
580 .fh_to_dentry = ceph_fh_to_dentry,
581 .fh_to_parent = ceph_fh_to_parent,
582 .get_parent = ceph_get_parent,
583 .get_name = ceph_get_name,
584 };
585