1	// SPDX-License-Identifier: GPL-2.0
2	#include <linux/ceph/ceph_debug.h>
3
4	#include <linux/spinlock.h>
5	#include <linux/namei.h>
6	#include <linux/slab.h>
7	#include <linux/sched.h>
8	#include <linux/xattr.h>
9
10	#include "super.h"
11	#include "mds_client.h"
12	#include "crypto.h"
13
14	/*
15	* Directory operations: readdir, lookup, create, link, unlink,
16	* rename, etc.
17	*/
18
19	/*
20	* Ceph MDS operations are specified in terms of a base ino and
21	* relative path. Thus, the client can specify an operation on a
22	* specific inode (e.g., a getattr due to fstat(2)), or as a path
23	* relative to, say, the root directory.
24	*
25	* Normally, we limit ourselves to strict inode ops (no path component)
26	* or dentry operations (a single path component relative to an ino). The
27	* exception to this is open_root_dentry(), which will open the mount
28	* point by name.
29	*/
30
31	const struct dentry_operations ceph_dentry_ops;
32
33	static bool __dentry_lease_is_valid(struct ceph_dentry_info *di);
34	static int __dir_lease_try_check(const struct dentry *dentry);
35
36	/*
37	* Initialize ceph dentry state.
38	*/
39	static int ceph_d_init(struct dentry *dentry)
40	{
41	struct ceph_dentry_info *di;
42	struct ceph_mds_client *mdsc = ceph_sb_to_mdsc(sb: dentry->d_sb);
43
44	di = kmem_cache_zalloc(ceph_dentry_cachep, GFP_KERNEL);
45	if (!di)
46	return -ENOMEM; /* oh well */
47
48	di->dentry = dentry;
49	di->lease_session = NULL;
50	di->time = jiffies;
51	dentry->d_fsdata = di;
52	INIT_LIST_HEAD(list: &di->lease_list);
53
54	atomic64_inc(v: &mdsc->metric.total_dentries);
55
56	return 0;
57	}
58
59	/*
60	* for f_pos for readdir:
61	* - hash order:
62	* (0xff << 52) | ((24 bits hash) << 28) |
63	* (the nth entry has hash collision);
64	* - frag+name order;
65	* ((frag value) << 28) | (the nth entry in frag);
66	*/
67	#define OFFSET_BITS 28
68	#define OFFSET_MASK ((1 << OFFSET_BITS) - 1)
69	#define HASH_ORDER (0xffull << (OFFSET_BITS + 24))
70	loff_t ceph_make_fpos(unsigned high, unsigned off, bool hash_order)
71	{
72	loff_t fpos = ((loff_t)high << 28) | (loff_t)off;
73	if (hash_order)
74	fpos |= HASH_ORDER;
75	return fpos;
76	}
77
78	static bool is_hash_order(loff_t p)
79	{
80	return (p & HASH_ORDER) == HASH_ORDER;
81	}
82
83	static unsigned fpos_frag(loff_t p)
84	{
85	return p >> OFFSET_BITS;
86	}
87
88	static unsigned fpos_hash(loff_t p)
89	{
90	return ceph_frag_value(f: fpos_frag(p));
91	}
92
93	static unsigned fpos_off(loff_t p)
94	{
95	return p & OFFSET_MASK;
96	}
97
98	static int fpos_cmp(loff_t l, loff_t r)
99	{
100	int v = ceph_frag_compare(a: fpos_frag(p: l), b: fpos_frag(p: r));
101	if (v)
102	return v;
103	return (int)(fpos_off(p: l) - fpos_off(p: r));
104	}
105
106	/*
107	* make note of the last dentry we read, so we can
108	* continue at the same lexicographical point,
109	* regardless of what dir changes take place on the
110	* server.
111	*/
112	static int note_last_dentry(struct ceph_fs_client *fsc,
113	struct ceph_dir_file_info *dfi,
114	const char *name,
115	int len, unsigned next_offset)
116	{
117	char *buf = kmalloc(len+1, GFP_KERNEL);
118	if (!buf)
119	return -ENOMEM;
120	kfree(objp: dfi->last_name);
121	dfi->last_name = buf;
122	memcpy(dfi->last_name, name, len);
123	dfi->last_name[len] = 0;
124	dfi->next_offset = next_offset;
125	doutc(fsc->client, "'%s'\n", dfi->last_name);
126	return 0;
127	}
128
129
130	static struct dentry *
131	__dcache_find_get_entry(struct dentry *parent, u64 idx,
132	struct ceph_readdir_cache_control *cache_ctl)
133	{
134	struct inode *dir = d_inode(dentry: parent);
135	struct ceph_client *cl = ceph_inode_to_client(inode: dir);
136	struct dentry *dentry;
137	unsigned idx_mask = (PAGE_SIZE / sizeof(struct dentry *)) - 1;
138	loff_t ptr_pos = idx * sizeof(struct dentry *);
139	pgoff_t ptr_pgoff = ptr_pos >> PAGE_SHIFT;
140
141	if (ptr_pos >= i_size_read(inode: dir))
142	return NULL;
143
144	if (!cache_ctl->folio || ptr_pgoff != cache_ctl->folio->index) {
145	ceph_readdir_cache_release(ctl: cache_ctl);
146	cache_ctl->folio = filemap_lock_folio(mapping: &dir->i_data, index: ptr_pgoff);
147	if (IS_ERR(ptr: cache_ctl->folio)) {
148	cache_ctl->folio = NULL;
149	doutc(cl, " folio %lu not found\n", ptr_pgoff);
150	return ERR_PTR(error: -EAGAIN);
151	}
152	/* reading/filling the cache are serialized by
153	i_rwsem, no need to use folio lock */
154	folio_unlock(folio: cache_ctl->folio);
155	cache_ctl->dentries = kmap_local_folio(folio: cache_ctl->folio, offset: 0);
156	}
157
158	cache_ctl->index = idx & idx_mask;
159
160	rcu_read_lock();
161	spin_lock(lock: &parent->d_lock);
162	/* check i_size again here, because empty directory can be
163	* marked as complete while not holding the i_rwsem. */
164	if (ceph_dir_is_complete_ordered(inode: dir) && ptr_pos < i_size_read(inode: dir))
165	dentry = cache_ctl->dentries[cache_ctl->index];
166	else
167	dentry = NULL;
168	spin_unlock(lock: &parent->d_lock);
169	if (dentry && !lockref_get_not_dead(lockref: &dentry->d_lockref))
170	dentry = NULL;
171	rcu_read_unlock();
172	return dentry ? : ERR_PTR(error: -EAGAIN);
173	}
174
175	/*
176	* When possible, we try to satisfy a readdir by peeking at the
177	* dcache. We make this work by carefully ordering dentries on
178	* d_children when we initially get results back from the MDS, and
179	* falling back to a "normal" sync readdir if any dentries in the dir
180	* are dropped.
181	*
182	* Complete dir indicates that we have all dentries in the dir. It is
183	* defined IFF we hold CEPH_CAP_FILE_SHARED (which will be revoked by
184	* the MDS if/when the directory is modified).
185	*/
186	static int __dcache_readdir(struct file *file, struct dir_context *ctx,
187	int shared_gen)
188	{
189	struct ceph_dir_file_info *dfi = file->private_data;
190	struct dentry *parent = file->f_path.dentry;
191	struct inode *dir = d_inode(dentry: parent);
192	struct ceph_fs_client *fsc = ceph_inode_to_fs_client(inode: dir);
193	struct ceph_client *cl = ceph_inode_to_client(inode: dir);
194	struct dentry *dentry, *last = NULL;
195	struct ceph_dentry_info *di;
196	struct ceph_readdir_cache_control cache_ctl = {};
197	u64 idx = 0;
198	int err = 0;
199
200	doutc(cl, "%p %llx.%llx v%u at %llx\n", dir, ceph_vinop(dir),
201	(unsigned)shared_gen, ctx->pos);
202
203	/* search start position */
204	if (ctx->pos > 2) {
205	u64 count = div_u64(dividend: i_size_read(inode: dir), divisor: sizeof(struct dentry *));
206	while (count > 0) {
207	u64 step = count >> 1;
208	dentry = __dcache_find_get_entry(parent, idx: idx + step,
209	cache_ctl: &cache_ctl);
210	if (!dentry) {
211	/* use linear search */
212	idx = 0;
213	break;
214	}
215	if (IS_ERR(ptr: dentry)) {
216	err = PTR_ERR(ptr: dentry);
217	goto out;
218	}
219	di = ceph_dentry(dentry);
220	spin_lock(lock: &dentry->d_lock);
221	if (fpos_cmp(l: di->offset, r: ctx->pos) < 0) {
222	idx += step + 1;
223	count -= step + 1;
224	} else {
225	count = step;
226	}
227	spin_unlock(lock: &dentry->d_lock);
228	dput(dentry);
229	}
230
231	doutc(cl, "%p %llx.%llx cache idx %llu\n", dir,
232	ceph_vinop(dir), idx);
233	}
234
235
236	for (;;) {
237	bool emit_dentry = false;
238	dentry = __dcache_find_get_entry(parent, idx: idx++, cache_ctl: &cache_ctl);
239	if (!dentry) {
240	dfi->file_info.flags |= CEPH_F_ATEND;
241	err = 0;
242	break;
243	}
244	if (IS_ERR(ptr: dentry)) {
245	err = PTR_ERR(ptr: dentry);
246	goto out;
247	}
248
249	spin_lock(lock: &dentry->d_lock);
250	di = ceph_dentry(dentry);
251	if (d_unhashed(dentry) ||
252	d_really_is_negative(dentry) ||
253	di->lease_shared_gen != shared_gen ||
254	((dentry->d_flags & DCACHE_NOKEY_NAME) &&
255	fscrypt_has_encryption_key(inode: dir))) {
256	spin_unlock(lock: &dentry->d_lock);
257	dput(dentry);
258	err = -EAGAIN;
259	goto out;
260	}
261	if (fpos_cmp(l: ctx->pos, r: di->offset) <= 0) {
262	__ceph_dentry_dir_lease_touch(di);
263	emit_dentry = true;
264	}
265	spin_unlock(lock: &dentry->d_lock);
266
267	if (emit_dentry) {
268	doutc(cl, " %llx dentry %p %pd %p\n", di->offset,
269	dentry, dentry, d_inode(dentry));
270	ctx->pos = di->offset;
271	if (!dir_emit(ctx, name: dentry->d_name.name,
272	namelen: dentry->d_name.len, ino: ceph_present_inode(inode: d_inode(dentry)),
273	type: d_inode(dentry)->i_mode >> 12)) {
274	dput(dentry);
275	err = 0;
276	break;
277	}
278	ctx->pos++;
279
280	if (last)
281	dput(last);
282	last = dentry;
283	} else {
284	dput(dentry);
285	}
286	}
287	out:
288	ceph_readdir_cache_release(ctl: &cache_ctl);
289	if (last) {
290	int ret;
291	di = ceph_dentry(dentry: last);
292	ret = note_last_dentry(fsc, dfi, name: last->d_name.name,
293	len: last->d_name.len,
294	next_offset: fpos_off(p: di->offset) + 1);
295	if (ret < 0)
296	err = ret;
297	dput(last);
298	/* last_name no longer match cache index */
299	if (dfi->readdir_cache_idx >= 0) {
300	dfi->readdir_cache_idx = -1;
301	dfi->dir_release_count = 0;
302	}
303	}
304	return err;
305	}
306
307	static bool need_send_readdir(struct ceph_dir_file_info *dfi, loff_t pos)
308	{
309	if (!dfi->last_readdir)
310	return true;
311	if (is_hash_order(p: pos))
312	return !ceph_frag_contains_value(f: dfi->frag, v: fpos_hash(p: pos));
313	else
314	return dfi->frag != fpos_frag(p: pos);
315	}
316
317	static int ceph_readdir(struct file *file, struct dir_context *ctx)
318	{
319	struct ceph_dir_file_info *dfi = file->private_data;
320	struct inode *inode = file_inode(f: file);
321	struct ceph_inode_info *ci = ceph_inode(inode);
322	struct ceph_fs_client *fsc = ceph_inode_to_fs_client(inode);
323	struct ceph_mds_client *mdsc = fsc->mdsc;
324	struct ceph_client *cl = fsc->client;
325	int i;
326	int err;
327	unsigned frag = -1;
328	struct ceph_mds_reply_info_parsed *rinfo;
329
330	doutc(cl, "%p %llx.%llx file %p pos %llx\n", inode,
331	ceph_vinop(inode), file, ctx->pos);
332	if (dfi->file_info.flags & CEPH_F_ATEND)
333	return 0;
334
335	/* always start with . and .. */
336	if (ctx->pos == 0) {
337	doutc(cl, "%p %llx.%llx off 0 -> '.'\n", inode,
338	ceph_vinop(inode));
339	if (!dir_emit(ctx, name: ".", namelen: 1, ino: ceph_present_inode(inode),
340	type: inode->i_mode >> 12))
341	return 0;
342	ctx->pos = 1;
343	}
344	if (ctx->pos == 1) {
345	u64 ino;
346	struct dentry *dentry = file->f_path.dentry;
347
348	spin_lock(lock: &dentry->d_lock);
349	ino = ceph_present_inode(inode: dentry->d_parent->d_inode);
350	spin_unlock(lock: &dentry->d_lock);
351
352	doutc(cl, "%p %llx.%llx off 1 -> '..'\n", inode,
353	ceph_vinop(inode));
354	if (!dir_emit(ctx, name: "..", namelen: 2, ino, type: inode->i_mode >> 12))
355	return 0;
356	ctx->pos = 2;
357	}
358
359	err = ceph_fscrypt_prepare_readdir(dir: inode);
360	if (err < 0)
361	return err;
362
363	spin_lock(lock: &ci->i_ceph_lock);
364	/* request Fx cap. if have Fx, we don't need to release Fs cap
365	* for later create/unlink. */
366	__ceph_touch_fmode(ci, mdsc, CEPH_FILE_MODE_WR);
367	/* can we use the dcache? */
368	if (ceph_test_mount_opt(fsc, DCACHE) &&
369	!ceph_test_mount_opt(fsc, NOASYNCREADDIR) &&
370	ceph_snap(inode) != CEPH_SNAPDIR &&
371	__ceph_dir_is_complete_ordered(ci) &&
372	__ceph_caps_issued_mask_metric(ci, CEPH_CAP_FILE_SHARED, t: 1)) {
373	int shared_gen = atomic_read(v: &ci->i_shared_gen);
374
375	spin_unlock(lock: &ci->i_ceph_lock);
376	err = __dcache_readdir(file, ctx, shared_gen);
377	if (err != -EAGAIN)
378	return err;
379	} else {
380	spin_unlock(lock: &ci->i_ceph_lock);
381	}
382
383	/* proceed with a normal readdir */
384	more:
385	/* do we have the correct frag content buffered? */
386	if (need_send_readdir(dfi, pos: ctx->pos)) {
387	struct ceph_mds_request *req;
388	int op = ceph_snap(inode) == CEPH_SNAPDIR ?
389	CEPH_MDS_OP_LSSNAP : CEPH_MDS_OP_READDIR;
390
391	/* discard old result, if any */
392	if (dfi->last_readdir) {
393	ceph_mdsc_put_request(req: dfi->last_readdir);
394	dfi->last_readdir = NULL;
395	}
396
397	if (is_hash_order(p: ctx->pos)) {
398	/* fragtree isn't always accurate. choose frag
399	* based on previous reply when possible. */
400	if (frag == (unsigned)-1)
401	frag = ceph_choose_frag(ci, v: fpos_hash(p: ctx->pos),
402	NULL, NULL);
403	} else {
404	frag = fpos_frag(p: ctx->pos);
405	}
406
407	doutc(cl, "fetching %p %llx.%llx frag %x offset '%s'\n",
408	inode, ceph_vinop(inode), frag, dfi->last_name);
409	req = ceph_mdsc_create_request(mdsc, op, mode: USE_AUTH_MDS);
410	if (IS_ERR(ptr: req))
411	return PTR_ERR(ptr: req);
412
413	err = ceph_alloc_readdir_reply_buffer(req, dir: inode);
414	if (err) {
415	ceph_mdsc_put_request(req);
416	return err;
417	}
418	/* hints to request -> mds selection code */
419	req->r_direct_mode = USE_AUTH_MDS;
420	if (op == CEPH_MDS_OP_READDIR) {
421	req->r_direct_hash = ceph_frag_value(f: frag);
422	__set_bit(CEPH_MDS_R_DIRECT_IS_HASH, &req->r_req_flags);
423	req->r_inode_drop = CEPH_CAP_FILE_EXCL;
424	}
425	if (dfi->last_name) {
426	int len = strlen(dfi->last_name);
427
428	req->r_path2 = kzalloc(NAME_MAX + 1, GFP_KERNEL);
429	if (!req->r_path2) {
430	ceph_mdsc_put_request(req);
431	return -ENOMEM;
432	}
433	memcpy(req->r_path2, dfi->last_name, len);
434
435	err = ceph_encode_encrypted_dname(parent: inode, buf: req->r_path2, len);
436	if (err < 0) {
437	ceph_mdsc_put_request(req);
438	return err;
439	}
440	} else if (is_hash_order(p: ctx->pos)) {
441	req->r_args.readdir.offset_hash =
442	cpu_to_le32(fpos_hash(ctx->pos));
443	}
444
445	req->r_dir_release_cnt = dfi->dir_release_count;
446	req->r_dir_ordered_cnt = dfi->dir_ordered_count;
447	req->r_readdir_cache_idx = dfi->readdir_cache_idx;
448	req->r_readdir_offset = dfi->next_offset;
449	req->r_args.readdir.frag = cpu_to_le32(frag);
450	req->r_args.readdir.flags =
451	cpu_to_le16(CEPH_READDIR_REPLY_BITFLAGS);
452
453	req->r_inode = inode;
454	ihold(inode);
455	req->r_dentry = dget(dentry: file->f_path.dentry);
456	err = ceph_mdsc_do_request(mdsc, NULL, req);
457	if (err < 0) {
458	ceph_mdsc_put_request(req);
459	return err;
460	}
461	doutc(cl, "%p %llx.%llx got and parsed readdir result=%d"
462	"on frag %x, end=%d, complete=%d, hash_order=%d\n",
463	inode, ceph_vinop(inode), err, frag,
464	(int)req->r_reply_info.dir_end,
465	(int)req->r_reply_info.dir_complete,
466	(int)req->r_reply_info.hash_order);
467
468	rinfo = &req->r_reply_info;
469	if (le32_to_cpu(rinfo->dir_dir->frag) != frag) {
470	frag = le32_to_cpu(rinfo->dir_dir->frag);
471	if (!rinfo->hash_order) {
472	dfi->next_offset = req->r_readdir_offset;
473	/* adjust ctx->pos to beginning of frag */
474	ctx->pos = ceph_make_fpos(high: frag,
475	off: dfi->next_offset,
476	hash_order: false);
477	}
478	}
479
480	dfi->frag = frag;
481	dfi->last_readdir = req;
482
483	if (test_bit(CEPH_MDS_R_DID_PREPOPULATE, &req->r_req_flags)) {
484	dfi->readdir_cache_idx = req->r_readdir_cache_idx;
485	if (dfi->readdir_cache_idx < 0) {
486	/* preclude from marking dir ordered */
487	dfi->dir_ordered_count = 0;
488	} else if (ceph_frag_is_leftmost(f: frag) &&
489	dfi->next_offset == 2) {
490	/* note dir version at start of readdir so
491	* we can tell if any dentries get dropped */
492	dfi->dir_release_count = req->r_dir_release_cnt;
493	dfi->dir_ordered_count = req->r_dir_ordered_cnt;
494	}
495	} else {
496	doutc(cl, "%p %llx.%llx !did_prepopulate\n", inode,
497	ceph_vinop(inode));
498	/* disable readdir cache */
499	dfi->readdir_cache_idx = -1;
500	/* preclude from marking dir complete */
501	dfi->dir_release_count = 0;
502	}
503
504	/* note next offset and last dentry name */
505	if (rinfo->dir_nr > 0) {
506	struct ceph_mds_reply_dir_entry *rde =
507	rinfo->dir_entries + (rinfo->dir_nr-1);
508	unsigned next_offset = req->r_reply_info.dir_end ?
509	2 : (fpos_off(p: rde->offset) + 1);
510	err = note_last_dentry(fsc, dfi, name: rde->name,
511	len: rde->name_len, next_offset);
512	if (err) {
513	ceph_mdsc_put_request(req: dfi->last_readdir);
514	dfi->last_readdir = NULL;
515	return err;
516	}
517	} else if (req->r_reply_info.dir_end) {
518	dfi->next_offset = 2;
519	/* keep last name */
520	}
521	}
522
523	rinfo = &dfi->last_readdir->r_reply_info;
524	doutc(cl, "%p %llx.%llx frag %x num %d pos %llx chunk first %llx\n",
525	inode, ceph_vinop(inode), dfi->frag, rinfo->dir_nr, ctx->pos,
526	rinfo->dir_nr ? rinfo->dir_entries[0].offset : 0LL);
527
528	i = 0;
529	/* search start position */
530	if (rinfo->dir_nr > 0) {
531	int step, nr = rinfo->dir_nr;
532	while (nr > 0) {
533	step = nr >> 1;
534	if (rinfo->dir_entries[i + step].offset < ctx->pos) {
535	i += step + 1;
536	nr -= step + 1;
537	} else {
538	nr = step;
539	}
540	}
541	}
542	for (; i < rinfo->dir_nr; i++) {
543	struct ceph_mds_reply_dir_entry *rde = rinfo->dir_entries + i;
544
545	if (rde->offset < ctx->pos) {
546	pr_warn_client(cl,
547	"%p %llx.%llx rde->offset 0x%llx ctx->pos 0x%llx\n",
548	inode, ceph_vinop(inode), rde->offset, ctx->pos);
549	return -EIO;
550	}
551
552	if (WARN_ON_ONCE(!rde->inode.in))
553	return -EIO;
554
555	ctx->pos = rde->offset;
556	doutc(cl, "%p %llx.%llx (%d/%d) -> %llx '%.*s' %p\n", inode,
557	ceph_vinop(inode), i, rinfo->dir_nr, ctx->pos,
558	rde->name_len, rde->name, &rde->inode.in);
559
560	if (!dir_emit(ctx, name: rde->name, namelen: rde->name_len,
561	ino: ceph_present_ino(sb: inode->i_sb, le64_to_cpu(rde->inode.in->ino)),
562	le32_to_cpu(rde->inode.in->mode) >> 12)) {
563	/*
564	* NOTE: Here no need to put the 'dfi->last_readdir',
565	* because when dir_emit stops us it's most likely
566	* doesn't have enough memory, etc. So for next readdir
567	* it will continue.
568	*/
569	doutc(cl, "filldir stopping us...\n");
570	return 0;
571	}
572
573	/* Reset the lengths to their original allocated vals */
574	ctx->pos++;
575	}
576
577	ceph_mdsc_put_request(req: dfi->last_readdir);
578	dfi->last_readdir = NULL;
579
580	if (dfi->next_offset > 2) {
581	frag = dfi->frag;
582	goto more;
583	}
584
585	/* more frags? */
586	if (!ceph_frag_is_rightmost(f: dfi->frag)) {
587	frag = ceph_frag_next(f: dfi->frag);
588	if (is_hash_order(p: ctx->pos)) {
589	loff_t new_pos = ceph_make_fpos(high: ceph_frag_value(f: frag),
590	off: dfi->next_offset, hash_order: true);
591	if (new_pos > ctx->pos)
592	ctx->pos = new_pos;
593	/* keep last_name */
594	} else {
595	ctx->pos = ceph_make_fpos(high: frag, off: dfi->next_offset,
596	hash_order: false);
597	kfree(objp: dfi->last_name);
598	dfi->last_name = NULL;
599	}
600	doutc(cl, "%p %llx.%llx next frag is %x\n", inode,
601	ceph_vinop(inode), frag);
602	goto more;
603	}
604	dfi->file_info.flags |= CEPH_F_ATEND;
605
606	/*
607	* if dir_release_count still matches the dir, no dentries
608	* were released during the whole readdir, and we should have
609	* the complete dir contents in our cache.
610	*/
611	if (atomic64_read(v: &ci->i_release_count) ==
612	dfi->dir_release_count) {
613	spin_lock(lock: &ci->i_ceph_lock);
614	if (dfi->dir_ordered_count ==
615	atomic64_read(v: &ci->i_ordered_count)) {
616	doutc(cl, " marking %p %llx.%llx complete and ordered\n",
617	inode, ceph_vinop(inode));
618	/* use i_size to track number of entries in
619	* readdir cache */
620	BUG_ON(dfi->readdir_cache_idx < 0);
621	i_size_write(inode, i_size: dfi->readdir_cache_idx *
622	sizeof(struct dentry*));
623	} else {
624	doutc(cl, " marking %llx.%llx complete\n",
625	ceph_vinop(inode));
626	}
627	__ceph_dir_set_complete(ci, release_count: dfi->dir_release_count,
628	ordered_count: dfi->dir_ordered_count);
629	spin_unlock(lock: &ci->i_ceph_lock);
630	}
631	doutc(cl, "%p %llx.%llx file %p done.\n", inode, ceph_vinop(inode),
632	file);
633	return 0;
634	}
635
636	static void reset_readdir(struct ceph_dir_file_info *dfi)
637	{
638	if (dfi->last_readdir) {
639	ceph_mdsc_put_request(req: dfi->last_readdir);
640	dfi->last_readdir = NULL;
641	}
642	kfree(objp: dfi->last_name);
643	dfi->last_name = NULL;
644	dfi->dir_release_count = 0;
645	dfi->readdir_cache_idx = -1;
646	dfi->next_offset = 2; /* compensate for . and .. */
647	dfi->file_info.flags &= ~CEPH_F_ATEND;
648	}
649
650	/*
651	* discard buffered readdir content on seekdir(0), or seek to new frag,
652	* or seek prior to current chunk
653	*/
654	static bool need_reset_readdir(struct ceph_dir_file_info *dfi, loff_t new_pos)
655	{
656	struct ceph_mds_reply_info_parsed *rinfo;
657	loff_t chunk_offset;
658	if (new_pos == 0)
659	return true;
660	if (is_hash_order(p: new_pos)) {
661	/* no need to reset last_name for a forward seek when
662	* dentries are sorted in hash order */
663	} else if (dfi->frag != fpos_frag(p: new_pos)) {
664	return true;
665	}
666	rinfo = dfi->last_readdir ? &dfi->last_readdir->r_reply_info : NULL;
667	if (!rinfo || !rinfo->dir_nr)
668	return true;
669	chunk_offset = rinfo->dir_entries[0].offset;
670	return new_pos < chunk_offset ||
671	is_hash_order(p: new_pos) != is_hash_order(p: chunk_offset);
672	}
673
674	static loff_t ceph_dir_llseek(struct file *file, loff_t offset, int whence)
675	{
676	struct ceph_dir_file_info *dfi = file->private_data;
677	struct inode *inode = file->f_mapping->host;
678	struct ceph_client *cl = ceph_inode_to_client(inode);
679	loff_t retval;
680
681	inode_lock(inode);
682	retval = -EINVAL;
683	switch (whence) {
684	case SEEK_CUR:
685	offset += file->f_pos;
686	break;
687	case SEEK_SET:
688	break;
689	case SEEK_END:
690	retval = -EOPNOTSUPP;
691	goto out;
692	default:
693	goto out;
694	}
695
696	if (offset >= 0) {
697	if (need_reset_readdir(dfi, new_pos: offset)) {
698	doutc(cl, "%p %llx.%llx dropping %p content\n",
699	inode, ceph_vinop(inode), file);
700	reset_readdir(dfi);
701	} else if (is_hash_order(p: offset) && offset > file->f_pos) {
702	/* for hash offset, we don't know if a forward seek
703	* is within same frag */
704	dfi->dir_release_count = 0;
705	dfi->readdir_cache_idx = -1;
706	}
707
708	if (offset != file->f_pos) {
709	file->f_pos = offset;
710	dfi->file_info.flags &= ~CEPH_F_ATEND;
711	}
712	retval = offset;
713	}
714	out:
715	inode_unlock(inode);
716	return retval;
717	}
718
719	/*
720	* Handle lookups for the hidden .snap directory.
721	*/
722	struct dentry *ceph_handle_snapdir(struct ceph_mds_request *req,
723	struct dentry *dentry)
724	{
725	struct ceph_fs_client *fsc = ceph_sb_to_fs_client(sb: dentry->d_sb);
726	struct inode *parent = d_inode(dentry: dentry->d_parent); /* we hold i_rwsem */
727	struct ceph_client *cl = ceph_inode_to_client(inode: parent);
728
729	/* .snap dir? */
730	if (ceph_snap(inode: parent) == CEPH_NOSNAP &&
731	strcmp(dentry->d_name.name, fsc->mount_options->snapdir_name) == 0) {
732	struct dentry *res;
733	struct inode *inode = ceph_get_snapdir(parent);
734
735	res = d_splice_alias(inode, dentry);
736	doutc(cl, "ENOENT on snapdir %p '%pd', linking to "
737	"snapdir %p %llx.%llx. Spliced dentry %p\n",
738	dentry, dentry, inode, ceph_vinop(inode), res);
739	if (res)
740	dentry = res;
741	}
742	return dentry;
743	}
744
745	/*
746	* Figure out final result of a lookup/open request.
747	*
748	* Mainly, make sure we return the final req->r_dentry (if it already
749	* existed) in place of the original VFS-provided dentry when they
750	* differ.
751	*
752	* Gracefully handle the case where the MDS replies with -ENOENT and
753	* no trace (which it may do, at its discretion, e.g., if it doesn't
754	* care to issue a lease on the negative dentry).
755	*/
756	struct dentry *ceph_finish_lookup(struct ceph_mds_request *req,
757	struct dentry *dentry, int err)
758	{
759	struct ceph_client *cl = req->r_mdsc->fsc->client;
760
761	if (err == -ENOENT) {
762	/* no trace? */
763	err = 0;
764	if (!req->r_reply_info.head->is_dentry) {
765	doutc(cl,
766	"ENOENT and no trace, dentry %p inode %llx.%llx\n",
767	dentry, ceph_vinop(d_inode(dentry)));
768	if (d_really_is_positive(dentry)) {
769	d_drop(dentry);
770	err = -ENOENT;
771	} else {
772	d_add(dentry, NULL);
773	}
774	}
775	}
776	if (err)
777	dentry = ERR_PTR(error: err);
778	else if (dentry != req->r_dentry)
779	dentry = dget(dentry: req->r_dentry); /* we got spliced */
780	else
781	dentry = NULL;
782	return dentry;
783	}
784
785	static bool is_root_ceph_dentry(struct inode *inode, struct dentry *dentry)
786	{
787	return ceph_ino(inode) == CEPH_INO_ROOT &&
788	strncmp(dentry->d_name.name, ".ceph", 5) == 0;
789	}
790
791	/*
792	* Look up a single dir entry. If there is a lookup intent, inform
793	* the MDS so that it gets our 'caps wanted' value in a single op.
794	*/
795	static struct dentry *ceph_lookup(struct inode *dir, struct dentry *dentry,
796	unsigned int flags)
797	{
798	struct ceph_fs_client *fsc = ceph_sb_to_fs_client(sb: dir->i_sb);
799	struct ceph_mds_client *mdsc = ceph_sb_to_mdsc(sb: dir->i_sb);
800	struct ceph_client *cl = fsc->client;
801	struct ceph_mds_request *req;
802	int op;
803	int mask;
804	int err;
805
806	doutc(cl, "%p %llx.%llx/'%pd' dentry %p\n", dir, ceph_vinop(dir),
807	dentry, dentry);
808
809	if (dentry->d_name.len > NAME_MAX)
810	return ERR_PTR(error: -ENAMETOOLONG);
811
812	if (IS_ENCRYPTED(dir)) {
813	bool had_key = fscrypt_has_encryption_key(inode: dir);
814
815	err = fscrypt_prepare_lookup_partial(dir, dentry);
816	if (err < 0)
817	return ERR_PTR(error: err);
818
819	/* mark directory as incomplete if it has been unlocked */
820	if (!had_key && fscrypt_has_encryption_key(inode: dir))
821	ceph_dir_clear_complete(inode: dir);
822	}
823
824	/* can we conclude ENOENT locally? */
825	if (d_really_is_negative(dentry)) {
826	struct ceph_inode_info *ci = ceph_inode(inode: dir);
827	struct ceph_dentry_info *di = ceph_dentry(dentry);
828
829	spin_lock(lock: &ci->i_ceph_lock);
830	doutc(cl, " dir %llx.%llx flags are 0x%lx\n",
831	ceph_vinop(dir), ci->i_ceph_flags);
832	if (strncmp(dentry->d_name.name,
833	fsc->mount_options->snapdir_name,
834	dentry->d_name.len) &&
835	!is_root_ceph_dentry(inode: dir, dentry) &&
836	ceph_test_mount_opt(fsc, DCACHE) &&
837	__ceph_dir_is_complete(ci) &&
838	__ceph_caps_issued_mask_metric(ci, CEPH_CAP_FILE_SHARED, t: 1)) {
839	__ceph_touch_fmode(ci, mdsc, CEPH_FILE_MODE_RD);
840	spin_unlock(lock: &ci->i_ceph_lock);
841	doutc(cl, " dir %llx.%llx complete, -ENOENT\n",
842	ceph_vinop(dir));
843	d_add(dentry, NULL);
844	di->lease_shared_gen = atomic_read(v: &ci->i_shared_gen);
845	return NULL;
846	}
847	spin_unlock(lock: &ci->i_ceph_lock);
848	}
849
850	op = ceph_snap(inode: dir) == CEPH_SNAPDIR ?
851	CEPH_MDS_OP_LOOKUPSNAP : CEPH_MDS_OP_LOOKUP;
852	req = ceph_mdsc_create_request(mdsc, op, mode: USE_ANY_MDS);
853	if (IS_ERR(ptr: req))
854	return ERR_CAST(ptr: req);
855	req->r_dentry = dget(dentry);
856	req->r_num_caps = 2;
857
858	mask = CEPH_STAT_CAP_INODE | CEPH_CAP_AUTH_SHARED;
859	if (ceph_security_xattr_wanted(in: dir))
860	mask |= CEPH_CAP_XATTR_SHARED;
861	req->r_args.getattr.mask = cpu_to_le32(mask);
862
863	ihold(inode: dir);
864	req->r_parent = dir;
865	set_bit(CEPH_MDS_R_PARENT_LOCKED, addr: &req->r_req_flags);
866	err = ceph_mdsc_do_request(mdsc, NULL, req);
867	if (err == -ENOENT) {
868	struct dentry *res;
869
870	res = ceph_handle_snapdir(req, dentry);
871	if (IS_ERR(ptr: res)) {
872	err = PTR_ERR(ptr: res);
873	} else {
874	dentry = res;
875	err = 0;
876	}
877	}
878	dentry = ceph_finish_lookup(req, dentry, err);
879	ceph_mdsc_put_request(req); /* will dput(dentry) */
880	doutc(cl, "result=%p\n", dentry);
881	return dentry;
882	}
883
884	/*
885	* If we do a create but get no trace back from the MDS, follow up with
886	* a lookup (the VFS expects us to link up the provided dentry).
887	*/
888	int ceph_handle_notrace_create(struct inode *dir, struct dentry *dentry)
889	{
890	struct dentry *result = ceph_lookup(dir, dentry, flags: 0);
891
892	if (result && !IS_ERR(ptr: result)) {
893	/*
894	* We created the item, then did a lookup, and found
895	* it was already linked to another inode we already
896	* had in our cache (and thus got spliced). To not
897	* confuse VFS (especially when inode is a directory),
898	* we don't link our dentry to that inode, return an
899	* error instead.
900	*
901	* This event should be rare and it happens only when
902	* we talk to old MDS. Recent MDS does not send traceless
903	* reply for request that creates new inode.
904	*/
905	d_drop(dentry: result);
906	return -ESTALE;
907	}
908	return PTR_ERR(ptr: result);
909	}
910
911	static int ceph_mknod(struct mnt_idmap *idmap, struct inode *dir,
912	struct dentry *dentry, umode_t mode, dev_t rdev)
913	{
914	struct ceph_mds_client *mdsc = ceph_sb_to_mdsc(sb: dir->i_sb);
915	struct ceph_client *cl = mdsc->fsc->client;
916	struct ceph_mds_request *req;
917	struct ceph_acl_sec_ctx as_ctx = {};
918	int err;
919
920	if (ceph_snap(inode: dir) != CEPH_NOSNAP)
921	return -EROFS;
922
923	err = ceph_wait_on_conflict_unlink(dentry);
924	if (err)
925	return err;
926
927	if (ceph_quota_is_max_files_exceeded(inode: dir)) {
928	err = -EDQUOT;
929	goto out;
930	}
931
932	doutc(cl, "%p %llx.%llx/'%pd' dentry %p mode 0%ho rdev %d\n",
933	dir, ceph_vinop(dir), dentry, dentry, mode, rdev);
934	req = ceph_mdsc_create_request(mdsc, op: CEPH_MDS_OP_MKNOD, mode: USE_AUTH_MDS);
935	if (IS_ERR(ptr: req)) {
936	err = PTR_ERR(ptr: req);
937	goto out;
938	}
939
940	req->r_new_inode = ceph_new_inode(dir, dentry, mode: &mode, as_ctx: &as_ctx);
941	if (IS_ERR(ptr: req->r_new_inode)) {
942	err = PTR_ERR(ptr: req->r_new_inode);
943	req->r_new_inode = NULL;
944	goto out_req;
945	}
946
947	if (S_ISREG(mode) && IS_ENCRYPTED(dir))
948	set_bit(CEPH_MDS_R_FSCRYPT_FILE, addr: &req->r_req_flags);
949
950	req->r_dentry = dget(dentry);
951	req->r_num_caps = 2;
952	req->r_parent = dir;
953	ihold(inode: dir);
954	set_bit(CEPH_MDS_R_PARENT_LOCKED, addr: &req->r_req_flags);
955	req->r_mnt_idmap = mnt_idmap_get(idmap);
956	req->r_args.mknod.mode = cpu_to_le32(mode);
957	req->r_args.mknod.rdev = cpu_to_le32(rdev);
958	req->r_dentry_drop = CEPH_CAP_FILE_SHARED | CEPH_CAP_AUTH_EXCL |
959	CEPH_CAP_XATTR_EXCL;
960	req->r_dentry_unless = CEPH_CAP_FILE_EXCL;
961
962	ceph_as_ctx_to_req(req, as_ctx: &as_ctx);
963
964	err = ceph_mdsc_do_request(mdsc, dir, req);
965	if (!err && !req->r_reply_info.head->is_dentry)
966	err = ceph_handle_notrace_create(dir, dentry);
967	out_req:
968	ceph_mdsc_put_request(req);
969	out:
970	if (!err)
971	ceph_init_inode_acls(inode: d_inode(dentry), as_ctx: &as_ctx);
972	else
973	d_drop(dentry);
974	ceph_release_acl_sec_ctx(as_ctx: &as_ctx);
975	return err;
976	}
977
978	static int ceph_create(struct mnt_idmap *idmap, struct inode *dir,
979	struct dentry *dentry, umode_t mode, bool excl)
980	{
981	return ceph_mknod(idmap, dir, dentry, mode, rdev: 0);
982	}
983
984	#if IS_ENABLED(CONFIG_FS_ENCRYPTION)
985	static int prep_encrypted_symlink_target(struct ceph_mds_request *req,
986	const char *dest)
987	{
988	int err;
989	int len = strlen(dest);
990	struct fscrypt_str osd_link = FSTR_INIT(NULL, 0);
991
992	err = fscrypt_prepare_symlink(dir: req->r_parent, target: dest, len, PATH_MAX,
993	disk_link: &osd_link);
994	if (err)
995	goto out;
996
997	err = fscrypt_encrypt_symlink(inode: req->r_new_inode, target: dest, len, disk_link: &osd_link);
998	if (err)
999	goto out;
1000
1001	req->r_path2 = kmalloc(BASE64_CHARS(osd_link.len) + 1, GFP_KERNEL);
1002	if (!req->r_path2) {
1003	err = -ENOMEM;
1004	goto out;
1005	}
1006
1007	len = base64_encode(src: osd_link.name, len: osd_link.len,
1008	dst: req->r_path2, padding: false, variant: BASE64_IMAP);
1009	req->r_path2[len] = '\0';
1010	out:
1011	fscrypt_fname_free_buffer(crypto_str: &osd_link);
1012	return err;
1013	}
1014	#else
1015	static int prep_encrypted_symlink_target(struct ceph_mds_request *req,
1016	const char *dest)
1017	{
1018	return -EOPNOTSUPP;
1019	}
1020	#endif
1021
1022	static int ceph_symlink(struct mnt_idmap *idmap, struct inode *dir,
1023	struct dentry *dentry, const char *dest)
1024	{
1025	struct ceph_mds_client *mdsc = ceph_sb_to_mdsc(sb: dir->i_sb);
1026	struct ceph_client *cl = mdsc->fsc->client;
1027	struct ceph_mds_request *req;
1028	struct ceph_acl_sec_ctx as_ctx = {};
1029	umode_t mode = S_IFLNK | 0777;
1030	int err;
1031
1032	if (ceph_snap(inode: dir) != CEPH_NOSNAP)
1033	return -EROFS;
1034
1035	err = ceph_wait_on_conflict_unlink(dentry);
1036	if (err)
1037	return err;
1038
1039	if (ceph_quota_is_max_files_exceeded(inode: dir)) {
1040	err = -EDQUOT;
1041	goto out;
1042	}
1043
1044	doutc(cl, "%p %llx.%llx/'%pd' to '%s'\n", dir, ceph_vinop(dir), dentry,
1045	dest);
1046	req = ceph_mdsc_create_request(mdsc, op: CEPH_MDS_OP_SYMLINK, mode: USE_AUTH_MDS);
1047	if (IS_ERR(ptr: req)) {
1048	err = PTR_ERR(ptr: req);
1049	goto out;
1050	}
1051
1052	req->r_new_inode = ceph_new_inode(dir, dentry, mode: &mode, as_ctx: &as_ctx);
1053	if (IS_ERR(ptr: req->r_new_inode)) {
1054	err = PTR_ERR(ptr: req->r_new_inode);
1055	req->r_new_inode = NULL;
1056	goto out_req;
1057	}
1058
1059	req->r_parent = dir;
1060	ihold(inode: dir);
1061
1062	if (IS_ENCRYPTED(req->r_new_inode)) {
1063	err = prep_encrypted_symlink_target(req, dest);
1064	if (err)
1065	goto out_req;
1066	} else {
1067	req->r_path2 = kstrdup(s: dest, GFP_KERNEL);
1068	if (!req->r_path2) {
1069	err = -ENOMEM;
1070	goto out_req;
1071	}
1072	}
1073
1074	set_bit(CEPH_MDS_R_PARENT_LOCKED, addr: &req->r_req_flags);
1075	req->r_mnt_idmap = mnt_idmap_get(idmap);
1076	req->r_dentry = dget(dentry);
1077	req->r_num_caps = 2;
1078	req->r_dentry_drop = CEPH_CAP_FILE_SHARED | CEPH_CAP_AUTH_EXCL |
1079	CEPH_CAP_XATTR_EXCL;
1080	req->r_dentry_unless = CEPH_CAP_FILE_EXCL;
1081
1082	ceph_as_ctx_to_req(req, as_ctx: &as_ctx);
1083
1084	err = ceph_mdsc_do_request(mdsc, dir, req);
1085	if (!err && !req->r_reply_info.head->is_dentry)
1086	err = ceph_handle_notrace_create(dir, dentry);
1087	out_req:
1088	ceph_mdsc_put_request(req);
1089	out:
1090	if (err)
1091	d_drop(dentry);
1092	ceph_release_acl_sec_ctx(as_ctx: &as_ctx);
1093	return err;
1094	}
1095
1096	static struct dentry *ceph_mkdir(struct mnt_idmap *idmap, struct inode *dir,
1097	struct dentry *dentry, umode_t mode)
1098	{
1099	struct ceph_mds_client *mdsc = ceph_sb_to_mdsc(sb: dir->i_sb);
1100	struct ceph_client *cl = mdsc->fsc->client;
1101	struct ceph_mds_request *req;
1102	struct ceph_acl_sec_ctx as_ctx = {};
1103	struct dentry *ret;
1104	int err;
1105	int op;
1106
1107	err = ceph_wait_on_conflict_unlink(dentry);
1108	if (err)
1109	return ERR_PTR(error: err);
1110
1111	if (ceph_snap(inode: dir) == CEPH_SNAPDIR) {
1112	/* mkdir .snap/foo is a MKSNAP */
1113	op = CEPH_MDS_OP_MKSNAP;
1114	doutc(cl, "mksnap %llx.%llx/'%pd' dentry %p\n",
1115	ceph_vinop(dir), dentry, dentry);
1116	} else if (ceph_snap(inode: dir) == CEPH_NOSNAP) {
1117	doutc(cl, "mkdir %llx.%llx/'%pd' dentry %p mode 0%ho\n",
1118	ceph_vinop(dir), dentry, dentry, mode);
1119	op = CEPH_MDS_OP_MKDIR;
1120	} else {
1121	ret = ERR_PTR(error: -EROFS);
1122	goto out;
1123	}
1124
1125	if (op == CEPH_MDS_OP_MKDIR &&
1126	ceph_quota_is_max_files_exceeded(inode: dir)) {
1127	ret = ERR_PTR(error: -EDQUOT);
1128	goto out;
1129	}
1130	if ((op == CEPH_MDS_OP_MKSNAP) && IS_ENCRYPTED(dir) &&
1131	!fscrypt_has_encryption_key(inode: dir)) {
1132	ret = ERR_PTR(error: -ENOKEY);
1133	goto out;
1134	}
1135
1136
1137	req = ceph_mdsc_create_request(mdsc, op, mode: USE_AUTH_MDS);
1138	if (IS_ERR(ptr: req)) {
1139	ret = ERR_CAST(ptr: req);
1140	goto out;
1141	}
1142
1143	mode |= S_IFDIR;
1144	req->r_new_inode = ceph_new_inode(dir, dentry, mode: &mode, as_ctx: &as_ctx);
1145	if (IS_ERR(ptr: req->r_new_inode)) {
1146	ret = ERR_CAST(ptr: req->r_new_inode);
1147	req->r_new_inode = NULL;
1148	goto out_req;
1149	}
1150
1151	req->r_dentry = dget(dentry);
1152	req->r_num_caps = 2;
1153	req->r_parent = dir;
1154	ihold(inode: dir);
1155	set_bit(CEPH_MDS_R_PARENT_LOCKED, addr: &req->r_req_flags);
1156	if (op == CEPH_MDS_OP_MKDIR)
1157	req->r_mnt_idmap = mnt_idmap_get(idmap);
1158	req->r_args.mkdir.mode = cpu_to_le32(mode);
1159	req->r_dentry_drop = CEPH_CAP_FILE_SHARED | CEPH_CAP_AUTH_EXCL |
1160	CEPH_CAP_XATTR_EXCL;
1161	req->r_dentry_unless = CEPH_CAP_FILE_EXCL;
1162
1163	ceph_as_ctx_to_req(req, as_ctx: &as_ctx);
1164
1165	err = ceph_mdsc_do_request(mdsc, dir, req);
1166	if (!err &&
1167	!req->r_reply_info.head->is_target &&
1168	!req->r_reply_info.head->is_dentry)
1169	err = ceph_handle_notrace_create(dir, dentry);
1170	ret = ERR_PTR(error: err);
1171	out_req:
1172	if (!IS_ERR(ptr: ret) && req->r_dentry != dentry)
1173	/* Some other dentry was spliced in */
1174	ret = dget(dentry: req->r_dentry);
1175	ceph_mdsc_put_request(req);
1176	out:
1177	if (!IS_ERR(ptr: ret)) {
1178	if (ret)
1179	dentry = ret;
1180	ceph_init_inode_acls(inode: d_inode(dentry), as_ctx: &as_ctx);
1181	} else {
1182	d_drop(dentry);
1183	}
1184	ceph_release_acl_sec_ctx(as_ctx: &as_ctx);
1185	return ret;
1186	}
1187
1188	static int ceph_link(struct dentry *old_dentry, struct inode *dir,
1189	struct dentry *dentry)
1190	{
1191	struct ceph_mds_client *mdsc = ceph_sb_to_mdsc(sb: dir->i_sb);
1192	struct ceph_client *cl = mdsc->fsc->client;
1193	struct ceph_mds_request *req;
1194	int err;
1195
1196	if (dentry->d_flags & DCACHE_DISCONNECTED)
1197	return -EINVAL;
1198
1199	err = ceph_wait_on_conflict_unlink(dentry);
1200	if (err)
1201	return err;
1202
1203	if (ceph_snap(inode: dir) != CEPH_NOSNAP)
1204	return -EROFS;
1205
1206	err = fscrypt_prepare_link(old_dentry, dir, dentry);
1207	if (err)
1208	return err;
1209
1210	doutc(cl, "%p %llx.%llx/'%pd' to '%pd'\n", dir, ceph_vinop(dir),
1211	old_dentry, dentry);
1212	req = ceph_mdsc_create_request(mdsc, op: CEPH_MDS_OP_LINK, mode: USE_AUTH_MDS);
1213	if (IS_ERR(ptr: req)) {
1214	d_drop(dentry);
1215	return PTR_ERR(ptr: req);
1216	}
1217	req->r_dentry = dget(dentry);
1218	req->r_num_caps = 2;
1219	req->r_old_dentry = dget(dentry: old_dentry);
1220	/*
1221	* The old_dentry maybe a DCACHE_DISCONNECTED dentry, then we
1222	* will just pass the ino# to MDSs.
1223	*/
1224	if (old_dentry->d_flags & DCACHE_DISCONNECTED)
1225	req->r_ino2 = ceph_vino(inode: d_inode(dentry: old_dentry));
1226	req->r_parent = dir;
1227	ihold(inode: dir);
1228	set_bit(CEPH_MDS_R_PARENT_LOCKED, addr: &req->r_req_flags);
1229	req->r_dentry_drop = CEPH_CAP_FILE_SHARED | CEPH_CAP_XATTR_EXCL;
1230	req->r_dentry_unless = CEPH_CAP_FILE_EXCL;
1231	/* release LINK_SHARED on source inode (mds will lock it) */
1232	req->r_old_inode_drop = CEPH_CAP_LINK_SHARED | CEPH_CAP_LINK_EXCL;
1233	err = ceph_mdsc_do_request(mdsc, dir, req);
1234	if (err) {
1235	d_drop(dentry);
1236	} else if (!req->r_reply_info.head->is_dentry) {
1237	ihold(inode: d_inode(dentry: old_dentry));
1238	d_instantiate(dentry, d_inode(dentry: old_dentry));
1239	}
1240	ceph_mdsc_put_request(req);
1241	return err;
1242	}
1243
1244	static void ceph_async_unlink_cb(struct ceph_mds_client *mdsc,
1245	struct ceph_mds_request *req)
1246	{
1247	struct dentry *dentry = req->r_dentry;
1248	struct ceph_fs_client *fsc = ceph_sb_to_fs_client(sb: dentry->d_sb);
1249	struct ceph_client *cl = fsc->client;
1250	struct ceph_dentry_info *di = ceph_dentry(dentry);
1251	int result = req->r_err ? req->r_err :
1252	le32_to_cpu(req->r_reply_info.head->result);
1253
1254	if (!test_bit(CEPH_DENTRY_ASYNC_UNLINK_BIT, &di->flags))
1255	pr_warn_client(cl,
1256	"dentry %p:%pd async unlink bit is not set\n",
1257	dentry, dentry);
1258
1259	spin_lock(lock: &fsc->async_unlink_conflict_lock);
1260	hash_del_rcu(node: &di->hnode);
1261	spin_unlock(lock: &fsc->async_unlink_conflict_lock);
1262
1263	spin_lock(lock: &dentry->d_lock);
1264	clear_and_wake_up_bit(CEPH_DENTRY_ASYNC_UNLINK_BIT, word: &di->flags);
1265	spin_unlock(lock: &dentry->d_lock);
1266
1267	synchronize_rcu();
1268
1269	if (result == -EJUKEBOX)
1270	goto out;
1271
1272	/* If op failed, mark everyone involved for errors */
1273	if (result) {
1274	struct ceph_path_info path_info = {0};
1275	char *path = ceph_mdsc_build_path(mdsc, dentry, path_info: &path_info, for_wire: 0);
1276
1277	/* mark error on parent + clear complete */
1278	mapping_set_error(mapping: req->r_parent->i_mapping, error: result);
1279	ceph_dir_clear_complete(inode: req->r_parent);
1280
1281	/* drop the dentry -- we don't know its status */
1282	if (!d_unhashed(dentry))
1283	d_drop(dentry);
1284
1285	/* mark inode itself for an error (since metadata is bogus) */
1286	mapping_set_error(mapping: req->r_old_inode->i_mapping, error: result);
1287
1288	pr_warn_client(cl, "failure path=(%llx)%s result=%d!\n",
1289	path_info.vino.ino, IS_ERR(path) ? "<<bad>>" : path, result);
1290	ceph_mdsc_free_path_info(path_info: &path_info);
1291	}
1292	out:
1293	iput(req->r_old_inode);
1294	ceph_mdsc_release_dir_caps(req);
1295	}
1296
1297	static int get_caps_for_async_unlink(struct inode *dir, struct dentry *dentry)
1298	{
1299	struct ceph_inode_info *ci = ceph_inode(inode: dir);
1300	struct ceph_dentry_info *di;
1301	int got = 0, want = CEPH_CAP_FILE_EXCL | CEPH_CAP_DIR_UNLINK;
1302
1303	spin_lock(lock: &ci->i_ceph_lock);
1304	if ((__ceph_caps_issued(ci, NULL) & want) == want) {
1305	ceph_take_cap_refs(ci, caps: want, snap_rwsem_locked: false);
1306	got = want;
1307	}
1308	spin_unlock(lock: &ci->i_ceph_lock);
1309
1310	/* If we didn't get anything, return 0 */
1311	if (!got)
1312	return 0;
1313
1314	spin_lock(lock: &dentry->d_lock);
1315	di = ceph_dentry(dentry);
1316	/*
1317	* - We are holding Fx, which implies Fs caps.
1318	* - Only support async unlink for primary linkage
1319	*/
1320	if (atomic_read(v: &ci->i_shared_gen) != di->lease_shared_gen ||
1321	!(di->flags & CEPH_DENTRY_PRIMARY_LINK))
1322	want = 0;
1323	spin_unlock(lock: &dentry->d_lock);
1324
1325	/* Do we still want what we've got? */
1326	if (want == got)
1327	return got;
1328
1329	ceph_put_cap_refs(ci, had: got);
1330	return 0;
1331	}
1332
1333	/*
1334	* rmdir and unlink are differ only by the metadata op code
1335	*/
1336	static int ceph_unlink(struct inode *dir, struct dentry *dentry)
1337	{
1338	struct ceph_fs_client *fsc = ceph_sb_to_fs_client(sb: dir->i_sb);
1339	struct ceph_client *cl = fsc->client;
1340	struct ceph_mds_client *mdsc = fsc->mdsc;
1341	struct inode *inode = d_inode(dentry);
1342	struct ceph_mds_request *req;
1343	bool try_async = ceph_test_mount_opt(fsc, ASYNC_DIROPS);
1344	struct dentry *dn;
1345	int err = -EROFS;
1346	int op;
1347	char *path;
1348
1349	if (ceph_snap(inode: dir) == CEPH_SNAPDIR) {
1350	/* rmdir .snap/foo is RMSNAP */
1351	doutc(cl, "rmsnap %llx.%llx/'%pd' dn\n", ceph_vinop(dir),
1352	dentry);
1353	op = CEPH_MDS_OP_RMSNAP;
1354	} else if (ceph_snap(inode: dir) == CEPH_NOSNAP) {
1355	doutc(cl, "unlink/rmdir %llx.%llx/'%pd' inode %llx.%llx\n",
1356	ceph_vinop(dir), dentry, ceph_vinop(inode));
1357	op = d_is_dir(dentry) ?
1358	CEPH_MDS_OP_RMDIR : CEPH_MDS_OP_UNLINK;
1359	} else
1360	goto out;
1361
1362	dn = d_find_alias(dir);
1363	if (!dn) {
1364	try_async = false;
1365	} else {
1366	struct ceph_path_info path_info;
1367	path = ceph_mdsc_build_path(mdsc, dentry: dn, path_info: &path_info, for_wire: 0);
1368	if (IS_ERR(ptr: path)) {
1369	try_async = false;
1370	err = 0;
1371	} else {
1372	err = ceph_mds_check_access(mdsc, tpath: path, MAY_WRITE);
1373	}
1374	ceph_mdsc_free_path_info(path_info: &path_info);
1375	dput(dn);
1376
1377	/* For none EACCES cases will let the MDS do the mds auth check */
1378	if (err == -EACCES) {
1379	return err;
1380	} else if (err < 0) {
1381	try_async = false;
1382	err = 0;
1383	}
1384	}
1385
1386	retry:
1387	req = ceph_mdsc_create_request(mdsc, op, mode: USE_AUTH_MDS);
1388	if (IS_ERR(ptr: req)) {
1389	err = PTR_ERR(ptr: req);
1390	goto out;
1391	}
1392	req->r_dentry = dget(dentry);
1393	req->r_num_caps = 2;
1394	req->r_parent = dir;
1395	ihold(inode: dir);
1396	req->r_dentry_drop = CEPH_CAP_FILE_SHARED | CEPH_CAP_XATTR_EXCL;
1397	req->r_dentry_unless = CEPH_CAP_FILE_EXCL;
1398	req->r_inode_drop = ceph_drop_caps_for_unlink(inode);
1399
1400	if (try_async && op == CEPH_MDS_OP_UNLINK &&
1401	(req->r_dir_caps = get_caps_for_async_unlink(dir, dentry))) {
1402	struct ceph_dentry_info *di = ceph_dentry(dentry);
1403
1404	doutc(cl, "async unlink on %llx.%llx/'%pd' caps=%s",
1405	ceph_vinop(dir), dentry,
1406	ceph_cap_string(req->r_dir_caps));
1407	set_bit(CEPH_MDS_R_ASYNC, addr: &req->r_req_flags);
1408	req->r_callback = ceph_async_unlink_cb;
1409	req->r_old_inode = d_inode(dentry);
1410	ihold(inode: req->r_old_inode);
1411
1412	spin_lock(lock: &dentry->d_lock);
1413	di->flags |= CEPH_DENTRY_ASYNC_UNLINK;
1414	spin_unlock(lock: &dentry->d_lock);
1415
1416	spin_lock(lock: &fsc->async_unlink_conflict_lock);
1417	hash_add_rcu(fsc->async_unlink_conflict, &di->hnode,
1418	dentry->d_name.hash);
1419	spin_unlock(lock: &fsc->async_unlink_conflict_lock);
1420
1421	err = ceph_mdsc_submit_request(mdsc, dir, req);
1422	if (!err) {
1423	/*
1424	* We have enough caps, so we assume that the unlink
1425	* will succeed. Fix up the target inode and dcache.
1426	*/
1427	drop_nlink(inode);
1428	d_delete(dentry);
1429	} else {
1430	spin_lock(lock: &fsc->async_unlink_conflict_lock);
1431	hash_del_rcu(node: &di->hnode);
1432	spin_unlock(lock: &fsc->async_unlink_conflict_lock);
1433
1434	spin_lock(lock: &dentry->d_lock);
1435	di->flags &= ~CEPH_DENTRY_ASYNC_UNLINK;
1436	spin_unlock(lock: &dentry->d_lock);
1437
1438	if (err == -EJUKEBOX) {
1439	try_async = false;
1440	ceph_mdsc_put_request(req);
1441	goto retry;
1442	}
1443	}
1444	} else {
1445	set_bit(CEPH_MDS_R_PARENT_LOCKED, addr: &req->r_req_flags);
1446	err = ceph_mdsc_do_request(mdsc, dir, req);
1447	if (!err && !req->r_reply_info.head->is_dentry)
1448	d_delete(dentry);
1449	}
1450
1451	ceph_mdsc_put_request(req);
1452	out:
1453	return err;
1454	}
1455
1456	static int ceph_rename(struct mnt_idmap *idmap, struct inode *old_dir,
1457	struct dentry *old_dentry, struct inode *new_dir,
1458	struct dentry *new_dentry, unsigned int flags)
1459	{
1460	struct ceph_mds_client *mdsc = ceph_sb_to_mdsc(sb: old_dir->i_sb);
1461	struct ceph_client *cl = mdsc->fsc->client;
1462	struct ceph_mds_request *req;
1463	int op = CEPH_MDS_OP_RENAME;
1464	int err;
1465
1466	if (flags)
1467	return -EINVAL;
1468
1469	if (ceph_snap(inode: old_dir) != ceph_snap(inode: new_dir))
1470	return -EXDEV;
1471	if (ceph_snap(inode: old_dir) != CEPH_NOSNAP) {
1472	if (old_dir == new_dir && ceph_snap(inode: old_dir) == CEPH_SNAPDIR)
1473	op = CEPH_MDS_OP_RENAMESNAP;
1474	else
1475	return -EROFS;
1476	}
1477	/* don't allow cross-quota renames */
1478	if ((old_dir != new_dir) &&
1479	(!ceph_quota_is_same_realm(old: old_dir, new: new_dir)))
1480	return -EXDEV;
1481
1482	err = ceph_wait_on_conflict_unlink(dentry: new_dentry);
1483	if (err)
1484	return err;
1485
1486	err = fscrypt_prepare_rename(old_dir, old_dentry, new_dir, new_dentry,
1487	flags);
1488	if (err)
1489	return err;
1490
1491	doutc(cl, "%llx.%llx/'%pd' to %llx.%llx/'%pd'\n",
1492	ceph_vinop(old_dir), old_dentry, ceph_vinop(new_dir),
1493	new_dentry);
1494	req = ceph_mdsc_create_request(mdsc, op, mode: USE_AUTH_MDS);
1495	if (IS_ERR(ptr: req))
1496	return PTR_ERR(ptr: req);
1497	ihold(inode: old_dir);
1498	req->r_dentry = dget(dentry: new_dentry);
1499	req->r_num_caps = 2;
1500	req->r_old_dentry = dget(dentry: old_dentry);
1501	req->r_old_dentry_dir = old_dir;
1502	req->r_parent = new_dir;
1503	ihold(inode: new_dir);
1504	set_bit(CEPH_MDS_R_PARENT_LOCKED, addr: &req->r_req_flags);
1505	req->r_old_dentry_drop = CEPH_CAP_FILE_SHARED | CEPH_CAP_XATTR_EXCL;
1506	req->r_old_dentry_unless = CEPH_CAP_FILE_EXCL;
1507	req->r_dentry_drop = CEPH_CAP_FILE_SHARED | CEPH_CAP_XATTR_EXCL;
1508	req->r_dentry_unless = CEPH_CAP_FILE_EXCL;
1509	/* release LINK_RDCACHE on source inode (mds will lock it) */
1510	req->r_old_inode_drop = CEPH_CAP_LINK_SHARED | CEPH_CAP_LINK_EXCL;
1511	if (d_really_is_positive(dentry: new_dentry)) {
1512	req->r_inode_drop =
1513	ceph_drop_caps_for_unlink(inode: d_inode(dentry: new_dentry));
1514	}
1515	err = ceph_mdsc_do_request(mdsc, dir: old_dir, req);
1516	if (!err && !req->r_reply_info.head->is_dentry) {
1517	/*
1518	* Normally d_move() is done by fill_trace (called by
1519	* do_request, above). If there is no trace, we need
1520	* to do it here.
1521	*/
1522	d_move(old_dentry, new_dentry);
1523	}
1524	ceph_mdsc_put_request(req);
1525	return err;
1526	}
1527
1528	/*
1529	* Move dentry to tail of mdsc->dentry_leases list when lease is updated.
1530	* Leases at front of the list will expire first. (Assume all leases have
1531	* similar duration)
1532	*
1533	* Called under dentry->d_lock.
1534	*/
1535	void __ceph_dentry_lease_touch(struct ceph_dentry_info *di)
1536	{
1537	struct dentry *dn = di->dentry;
1538	struct ceph_mds_client *mdsc = ceph_sb_to_fs_client(sb: dn->d_sb)->mdsc;
1539	struct ceph_client *cl = mdsc->fsc->client;
1540
1541	doutc(cl, "%p %p '%pd'\n", di, dn, dn);
1542
1543	di->flags |= CEPH_DENTRY_LEASE_LIST;
1544	if (di->flags & CEPH_DENTRY_SHRINK_LIST) {
1545	di->flags |= CEPH_DENTRY_REFERENCED;
1546	return;
1547	}
1548
1549	spin_lock(lock: &mdsc->dentry_list_lock);
1550	list_move_tail(list: &di->lease_list, head: &mdsc->dentry_leases);
1551	spin_unlock(lock: &mdsc->dentry_list_lock);
1552	}
1553
1554	static void __dentry_dir_lease_touch(struct ceph_mds_client* mdsc,
1555	struct ceph_dentry_info *di)
1556	{
1557	di->flags &= ~(CEPH_DENTRY_LEASE_LIST | CEPH_DENTRY_REFERENCED);
1558	di->lease_gen = 0;
1559	di->time = jiffies;
1560	list_move_tail(list: &di->lease_list, head: &mdsc->dentry_dir_leases);
1561	}
1562
1563	/*
1564	* When dir lease is used, add dentry to tail of mdsc->dentry_dir_leases
1565	* list if it's not in the list, otherwise set 'referenced' flag.
1566	*
1567	* Called under dentry->d_lock.
1568	*/
1569	void __ceph_dentry_dir_lease_touch(struct ceph_dentry_info *di)
1570	{
1571	struct dentry *dn = di->dentry;
1572	struct ceph_mds_client *mdsc = ceph_sb_to_fs_client(sb: dn->d_sb)->mdsc;
1573	struct ceph_client *cl = mdsc->fsc->client;
1574
1575	doutc(cl, "%p %p '%pd' (offset 0x%llx)\n", di, dn, dn, di->offset);
1576
1577	if (!list_empty(head: &di->lease_list)) {
1578	if (di->flags & CEPH_DENTRY_LEASE_LIST) {
1579	/* don't remove dentry from dentry lease list
1580	* if its lease is valid */
1581	if (__dentry_lease_is_valid(di))
1582	return;
1583	} else {
1584	di->flags |= CEPH_DENTRY_REFERENCED;
1585	return;
1586	}
1587	}
1588
1589	if (di->flags & CEPH_DENTRY_SHRINK_LIST) {
1590	di->flags |= CEPH_DENTRY_REFERENCED;
1591	di->flags &= ~CEPH_DENTRY_LEASE_LIST;
1592	return;
1593	}
1594
1595	spin_lock(lock: &mdsc->dentry_list_lock);
1596	__dentry_dir_lease_touch(mdsc, di);
1597	spin_unlock(lock: &mdsc->dentry_list_lock);
1598	}
1599
1600	static void __dentry_lease_unlist(struct ceph_dentry_info *di)
1601	{
1602	struct ceph_mds_client *mdsc;
1603	if (di->flags & CEPH_DENTRY_SHRINK_LIST)
1604	return;
1605	if (list_empty(head: &di->lease_list))
1606	return;
1607
1608	mdsc = ceph_sb_to_fs_client(sb: di->dentry->d_sb)->mdsc;
1609	spin_lock(lock: &mdsc->dentry_list_lock);
1610	list_del_init(entry: &di->lease_list);
1611	spin_unlock(lock: &mdsc->dentry_list_lock);
1612	}
1613
1614	enum {
1615	KEEP = 0,
1616	DELETE = 1,
1617	TOUCH = 2,
1618	STOP = 4,
1619	};
1620
1621	struct ceph_lease_walk_control {
1622	bool dir_lease;
1623	bool expire_dir_lease;
1624	unsigned long nr_to_scan;
1625	unsigned long dir_lease_ttl;
1626	};
1627
1628	static int __dir_lease_check(const struct dentry *, struct ceph_lease_walk_control *);
1629	static int __dentry_lease_check(const struct dentry *);
1630
1631	static unsigned long
1632	__dentry_leases_walk(struct ceph_mds_client *mdsc,
1633	struct ceph_lease_walk_control *lwc)
1634	{
1635	struct ceph_dentry_info *di, *tmp;
1636	struct dentry *dentry, *last = NULL;
1637	struct list_head* list;
1638	LIST_HEAD(dispose);
1639	unsigned long freed = 0;
1640	int ret = 0;
1641
1642	list = lwc->dir_lease ? &mdsc->dentry_dir_leases : &mdsc->dentry_leases;
1643	spin_lock(lock: &mdsc->dentry_list_lock);
1644	list_for_each_entry_safe(di, tmp, list, lease_list) {
1645	if (!lwc->nr_to_scan)
1646	break;
1647	--lwc->nr_to_scan;
1648
1649	dentry = di->dentry;
1650	if (last == dentry)
1651	break;
1652
1653	if (!spin_trylock(lock: &dentry->d_lock))
1654	continue;
1655
1656	if (__lockref_is_dead(l: &dentry->d_lockref)) {
1657	list_del_init(entry: &di->lease_list);
1658	goto next;
1659	}
1660
1661	if (lwc->dir_lease)
1662	ret = __dir_lease_check(dentry, lwc);
1663	else
1664	ret = __dentry_lease_check(dentry);
1665	if (ret & TOUCH) {
1666	/* move it into tail of dir lease list */
1667	__dentry_dir_lease_touch(mdsc, di);
1668	if (!last)
1669	last = dentry;
1670	}
1671	if (ret & DELETE) {
1672	/* stale lease */
1673	di->flags &= ~CEPH_DENTRY_REFERENCED;
1674	if (dentry->d_lockref.count > 0) {
1675	/* update_dentry_lease() will re-add
1676	* it to lease list, or
1677	* ceph_d_delete() will return 1 when
1678	* last reference is dropped */
1679	list_del_init(entry: &di->lease_list);
1680	} else {
1681	di->flags |= CEPH_DENTRY_SHRINK_LIST;
1682	list_move_tail(list: &di->lease_list, head: &dispose);
1683	dget_dlock(dentry);
1684	}
1685	}
1686	next:
1687	spin_unlock(lock: &dentry->d_lock);
1688	if (ret & STOP)
1689	break;
1690	}
1691	spin_unlock(lock: &mdsc->dentry_list_lock);
1692
1693	while (!list_empty(head: &dispose)) {
1694	di = list_first_entry(&dispose, struct ceph_dentry_info,
1695	lease_list);
1696	dentry = di->dentry;
1697	spin_lock(lock: &dentry->d_lock);
1698
1699	list_del_init(entry: &di->lease_list);
1700	di->flags &= ~CEPH_DENTRY_SHRINK_LIST;
1701	if (di->flags & CEPH_DENTRY_REFERENCED) {
1702	spin_lock(lock: &mdsc->dentry_list_lock);
1703	if (di->flags & CEPH_DENTRY_LEASE_LIST) {
1704	list_add_tail(new: &di->lease_list,
1705	head: &mdsc->dentry_leases);
1706	} else {
1707	__dentry_dir_lease_touch(mdsc, di);
1708	}
1709	spin_unlock(lock: &mdsc->dentry_list_lock);
1710	} else {
1711	freed++;
1712	}
1713
1714	spin_unlock(lock: &dentry->d_lock);
1715	/* ceph_d_delete() does the trick */
1716	dput(dentry);
1717	}
1718	return freed;
1719	}
1720
1721	static int __dentry_lease_check(const struct dentry *dentry)
1722	{
1723	struct ceph_dentry_info *di = ceph_dentry(dentry);
1724	int ret;
1725
1726	if (__dentry_lease_is_valid(di))
1727	return STOP;
1728	ret = __dir_lease_try_check(dentry);
1729	if (ret == -EBUSY)
1730	return KEEP;
1731	if (ret > 0)
1732	return TOUCH;
1733	return DELETE;
1734	}
1735
1736	static int __dir_lease_check(const struct dentry *dentry,
1737	struct ceph_lease_walk_control *lwc)
1738	{
1739	struct ceph_dentry_info *di = ceph_dentry(dentry);
1740
1741	int ret = __dir_lease_try_check(dentry);
1742	if (ret == -EBUSY)
1743	return KEEP;
1744	if (ret > 0) {
1745	if (time_before(jiffies, di->time + lwc->dir_lease_ttl))
1746	return STOP;
1747	/* Move dentry to tail of dir lease list if we don't want
1748	* to delete it. So dentries in the list are checked in a
1749	* round robin manner */
1750	if (!lwc->expire_dir_lease)
1751	return TOUCH;
1752	if (dentry->d_lockref.count > 0 ||
1753	(di->flags & CEPH_DENTRY_REFERENCED))
1754	return TOUCH;
1755	/* invalidate dir lease */
1756	di->lease_shared_gen = 0;
1757	}
1758	return DELETE;
1759	}
1760
1761	int ceph_trim_dentries(struct ceph_mds_client *mdsc)
1762	{
1763	struct ceph_lease_walk_control lwc;
1764	unsigned long count;
1765	unsigned long freed;
1766
1767	spin_lock(lock: &mdsc->caps_list_lock);
1768	if (mdsc->caps_use_max > 0 &&
1769	mdsc->caps_use_count > mdsc->caps_use_max)
1770	count = mdsc->caps_use_count - mdsc->caps_use_max;
1771	else
1772	count = 0;
1773	spin_unlock(lock: &mdsc->caps_list_lock);
1774
1775	lwc.dir_lease = false;
1776	lwc.nr_to_scan = CEPH_CAPS_PER_RELEASE * 2;
1777	freed = __dentry_leases_walk(mdsc, lwc: &lwc);
1778	if (!lwc.nr_to_scan) /* more invalid leases */
1779	return -EAGAIN;
1780
1781	if (lwc.nr_to_scan < CEPH_CAPS_PER_RELEASE)
1782	lwc.nr_to_scan = CEPH_CAPS_PER_RELEASE;
1783
1784	lwc.dir_lease = true;
1785	lwc.expire_dir_lease = freed < count;
1786	lwc.dir_lease_ttl = mdsc->fsc->mount_options->caps_wanted_delay_max * HZ;
1787	freed +=__dentry_leases_walk(mdsc, lwc: &lwc);
1788	if (!lwc.nr_to_scan) /* more to check */
1789	return -EAGAIN;
1790
1791	return freed > 0 ? 1 : 0;
1792	}
1793
1794	/*
1795	* Ensure a dentry lease will no longer revalidate.
1796	*/
1797	void ceph_invalidate_dentry_lease(struct dentry *dentry)
1798	{
1799	struct ceph_dentry_info *di = ceph_dentry(dentry);
1800	spin_lock(lock: &dentry->d_lock);
1801	di->time = jiffies;
1802	di->lease_shared_gen = 0;
1803	di->flags &= ~CEPH_DENTRY_PRIMARY_LINK;
1804	__dentry_lease_unlist(di);
1805	spin_unlock(lock: &dentry->d_lock);
1806	}
1807
1808	/*
1809	* Check if dentry lease is valid. If not, delete the lease. Try to
1810	* renew if the least is more than half up.
1811	*/
1812	static bool __dentry_lease_is_valid(struct ceph_dentry_info *di)
1813	{
1814	struct ceph_mds_session *session;
1815
1816	if (!di->lease_gen)
1817	return false;
1818
1819	session = di->lease_session;
1820	if (session) {
1821	u32 gen;
1822	unsigned long ttl;
1823
1824	gen = atomic_read(v: &session->s_cap_gen);
1825	ttl = session->s_cap_ttl;
1826
1827	if (di->lease_gen == gen &&
1828	time_before(jiffies, ttl) &&
1829	time_before(jiffies, di->time))
1830	return true;
1831	}
1832	di->lease_gen = 0;
1833	return false;
1834	}
1835
1836	static int dentry_lease_is_valid(struct dentry *dentry, unsigned int flags)
1837	{
1838	struct ceph_dentry_info *di;
1839	struct ceph_mds_session *session = NULL;
1840	struct ceph_mds_client *mdsc = ceph_sb_to_fs_client(sb: dentry->d_sb)->mdsc;
1841	struct ceph_client *cl = mdsc->fsc->client;
1842	u32 seq = 0;
1843	int valid = 0;
1844
1845	spin_lock(lock: &dentry->d_lock);
1846	di = ceph_dentry(dentry);
1847	if (di && __dentry_lease_is_valid(di)) {
1848	valid = 1;
1849
1850	if (di->lease_renew_after &&
1851	time_after(jiffies, di->lease_renew_after)) {
1852	/*
1853	* We should renew. If we're in RCU walk mode
1854	* though, we can't do that so just return
1855	* -ECHILD.
1856	*/
1857	if (flags & LOOKUP_RCU) {
1858	valid = -ECHILD;
1859	} else {
1860	session = ceph_get_mds_session(s: di->lease_session);
1861	seq = di->lease_seq;
1862	di->lease_renew_after = 0;
1863	di->lease_renew_from = jiffies;
1864	}
1865	}
1866	}
1867	spin_unlock(lock: &dentry->d_lock);
1868
1869	if (session) {
1870	ceph_mdsc_lease_send_msg(session, dentry,
1871	CEPH_MDS_LEASE_RENEW, seq);
1872	ceph_put_mds_session(s: session);
1873	}
1874	doutc(cl, "dentry %p = %d\n", dentry, valid);
1875	return valid;
1876	}
1877
1878	/*
1879	* Called under dentry->d_lock.
1880	*/
1881	static int __dir_lease_try_check(const struct dentry *dentry)
1882	{
1883	struct ceph_dentry_info *di = ceph_dentry(dentry);
1884	struct inode *dir;
1885	struct ceph_inode_info *ci;
1886	int valid = 0;
1887
1888	if (!di->lease_shared_gen)
1889	return 0;
1890	if (IS_ROOT(dentry))
1891	return 0;
1892
1893	dir = d_inode(dentry: dentry->d_parent);
1894	ci = ceph_inode(inode: dir);
1895
1896	if (spin_trylock(lock: &ci->i_ceph_lock)) {
1897	if (atomic_read(v: &ci->i_shared_gen) == di->lease_shared_gen &&
1898	__ceph_caps_issued_mask(ci, CEPH_CAP_FILE_SHARED, t: 0))
1899	valid = 1;
1900	spin_unlock(lock: &ci->i_ceph_lock);
1901	} else {
1902	valid = -EBUSY;
1903	}
1904
1905	if (!valid)
1906	di->lease_shared_gen = 0;
1907	return valid;
1908	}
1909
1910	/*
1911	* Check if directory-wide content lease/cap is valid.
1912	*/
1913	static int dir_lease_is_valid(struct inode *dir, struct dentry *dentry,
1914	struct ceph_mds_client *mdsc)
1915	{
1916	struct ceph_inode_info *ci = ceph_inode(inode: dir);
1917	struct ceph_client *cl = mdsc->fsc->client;
1918	int valid;
1919	int shared_gen;
1920
1921	spin_lock(lock: &ci->i_ceph_lock);
1922	valid = __ceph_caps_issued_mask(ci, CEPH_CAP_FILE_SHARED, t: 1);
1923	if (valid) {
1924	__ceph_touch_fmode(ci, mdsc, CEPH_FILE_MODE_RD);
1925	shared_gen = atomic_read(v: &ci->i_shared_gen);
1926	}
1927	spin_unlock(lock: &ci->i_ceph_lock);
1928	if (valid) {
1929	struct ceph_dentry_info *di;
1930	spin_lock(lock: &dentry->d_lock);
1931	di = ceph_dentry(dentry);
1932	if (dir == d_inode(dentry: dentry->d_parent) &&
1933	di && di->lease_shared_gen == shared_gen)
1934	__ceph_dentry_dir_lease_touch(di);
1935	else
1936	valid = 0;
1937	spin_unlock(lock: &dentry->d_lock);
1938	}
1939	doutc(cl, "dir %p %llx.%llx v%u dentry %p '%pd' = %d\n", dir,
1940	ceph_vinop(dir), (unsigned)atomic_read(&ci->i_shared_gen),
1941	dentry, dentry, valid);
1942	return valid;
1943	}
1944
1945	/*
1946	* Check if cached dentry can be trusted.
1947	*/
1948	static int ceph_d_revalidate(struct inode *dir, const struct qstr *name,
1949	struct dentry *dentry, unsigned int flags)
1950	{
1951	struct ceph_mds_client *mdsc = ceph_sb_to_fs_client(sb: dentry->d_sb)->mdsc;
1952	struct ceph_client *cl = mdsc->fsc->client;
1953	int valid = 0;
1954	struct inode *inode;
1955
1956	valid = fscrypt_d_revalidate(dir, name, dentry, flags);
1957	if (valid <= 0)
1958	return valid;
1959
1960	inode = d_inode_rcu(dentry);
1961
1962	doutc(cl, "%p '%pd' inode %p offset 0x%llx nokey %d\n",
1963	dentry, dentry, inode, ceph_dentry(dentry)->offset,
1964	!!(dentry->d_flags & DCACHE_NOKEY_NAME));
1965
1966	mdsc = ceph_sb_to_fs_client(sb: dir->i_sb)->mdsc;
1967
1968	/* always trust cached snapped dentries, snapdir dentry */
1969	if (ceph_snap(inode: dir) != CEPH_NOSNAP) {
1970	doutc(cl, "%p '%pd' inode %p is SNAPPED\n", dentry,
1971	dentry, inode);
1972	valid = 1;
1973	} else if (inode && ceph_snap(inode) == CEPH_SNAPDIR) {
1974	valid = 1;
1975	} else {
1976	valid = dentry_lease_is_valid(dentry, flags);
1977	if (valid == -ECHILD)
1978	return valid;
1979	if (valid || dir_lease_is_valid(dir, dentry, mdsc)) {
1980	if (inode)
1981	valid = ceph_is_any_caps(inode);
1982	else
1983	valid = 1;
1984	}
1985	}
1986
1987	if (!valid) {
1988	struct ceph_mds_request *req;
1989	int op, err;
1990	u32 mask;
1991
1992	if (flags & LOOKUP_RCU)
1993	return -ECHILD;
1994
1995	percpu_counter_inc(fbc: &mdsc->metric.d_lease_mis);
1996
1997	op = ceph_snap(inode: dir) == CEPH_SNAPDIR ?
1998	CEPH_MDS_OP_LOOKUPSNAP : CEPH_MDS_OP_LOOKUP;
1999	req = ceph_mdsc_create_request(mdsc, op, mode: USE_ANY_MDS);
2000	if (!IS_ERR(ptr: req)) {
2001	req->r_dentry = dget(dentry);
2002	req->r_num_caps = 2;
2003	req->r_parent = dir;
2004	ihold(inode: dir);
2005
2006	req->r_dname = name;
2007
2008	mask = CEPH_STAT_CAP_INODE | CEPH_CAP_AUTH_SHARED;
2009	if (ceph_security_xattr_wanted(in: dir))
2010	mask |= CEPH_CAP_XATTR_SHARED;
2011	req->r_args.getattr.mask = cpu_to_le32(mask);
2012
2013	err = ceph_mdsc_do_request(mdsc, NULL, req);
2014	switch (err) {
2015	case 0:
2016	if (d_really_is_positive(dentry) &&
2017	d_inode(dentry) == req->r_target_inode)
2018	valid = 1;
2019	break;
2020	case -ENOENT:
2021	if (d_really_is_negative(dentry))
2022	valid = 1;
2023	fallthrough;
2024	default:
2025	break;
2026	}
2027	ceph_mdsc_put_request(req);
2028	doutc(cl, "%p '%pd', lookup result=%d\n", dentry,
2029	dentry, err);
2030	}
2031	} else {
2032	percpu_counter_inc(fbc: &mdsc->metric.d_lease_hit);
2033	}
2034
2035	doutc(cl, "%p '%pd' %s\n", dentry, dentry, valid ? "valid" : "invalid");
2036	if (!valid)
2037	ceph_dir_clear_complete(inode: dir);
2038	return valid;
2039	}
2040
2041	/*
2042	* Delete unused dentry that doesn't have valid lease
2043	*
2044	* Called under dentry->d_lock.
2045	*/
2046	static int ceph_d_delete(const struct dentry *dentry)
2047	{
2048	struct ceph_dentry_info *di;
2049
2050	/* won't release caps */
2051	if (d_really_is_negative(dentry))
2052	return 0;
2053	if (ceph_snap(inode: d_inode(dentry)) != CEPH_NOSNAP)
2054	return 0;
2055	/* valid lease? */
2056	di = ceph_dentry(dentry);
2057	if (di) {
2058	if (__dentry_lease_is_valid(di))
2059	return 0;
2060	if (__dir_lease_try_check(dentry))
2061	return 0;
2062	}
2063	return 1;
2064	}
2065
2066	/*
2067	* Release our ceph_dentry_info.
2068	*/
2069	static void ceph_d_release(struct dentry *dentry)
2070	{
2071	struct ceph_dentry_info *di = ceph_dentry(dentry);
2072	struct ceph_fs_client *fsc = ceph_sb_to_fs_client(sb: dentry->d_sb);
2073
2074	doutc(fsc->client, "dentry %p '%pd'\n", dentry, dentry);
2075
2076	atomic64_dec(v: &fsc->mdsc->metric.total_dentries);
2077
2078	spin_lock(lock: &dentry->d_lock);
2079	__dentry_lease_unlist(di);
2080	dentry->d_fsdata = NULL;
2081	spin_unlock(lock: &dentry->d_lock);
2082
2083	ceph_put_mds_session(s: di->lease_session);
2084	kmem_cache_free(s: ceph_dentry_cachep, objp: di);
2085	}
2086
2087	/*
2088	* When the VFS prunes a dentry from the cache, we need to clear the
2089	* complete flag on the parent directory.
2090	*
2091	* Called under dentry->d_lock.
2092	*/
2093	static void ceph_d_prune(struct dentry *dentry)
2094	{
2095	struct ceph_mds_client *mdsc = ceph_sb_to_mdsc(sb: dentry->d_sb);
2096	struct ceph_client *cl = mdsc->fsc->client;
2097	struct ceph_inode_info *dir_ci;
2098	struct ceph_dentry_info *di;
2099
2100	doutc(cl, "dentry %p '%pd'\n", dentry, dentry);
2101
2102	/* do we have a valid parent? */
2103	if (IS_ROOT(dentry))
2104	return;
2105
2106	/* we hold d_lock, so d_parent is stable */
2107	dir_ci = ceph_inode(inode: d_inode(dentry: dentry->d_parent));
2108	if (dir_ci->i_vino.snap == CEPH_SNAPDIR)
2109	return;
2110
2111	/* who calls d_delete() should also disable dcache readdir */
2112	if (d_really_is_negative(dentry))
2113	return;
2114
2115	/* d_fsdata does not get cleared until d_release */
2116	if (!d_unhashed(dentry)) {
2117	__ceph_dir_clear_complete(ci: dir_ci);
2118	return;
2119	}
2120
2121	/* Disable dcache readdir just in case that someone called d_drop()
2122	* or d_invalidate(), but MDS didn't revoke CEPH_CAP_FILE_SHARED
2123	* properly (dcache readdir is still enabled) */
2124	di = ceph_dentry(dentry);
2125	if (di->offset > 0 &&
2126	di->lease_shared_gen == atomic_read(v: &dir_ci->i_shared_gen))
2127	__ceph_dir_clear_ordered(ci: dir_ci);
2128	}
2129
2130	/*
2131	* read() on a dir. This weird interface hack only works if mounted
2132	* with '-o dirstat'.
2133	*/
2134	static ssize_t ceph_read_dir(struct file *file, char __user *buf, size_t size,
2135	loff_t *ppos)
2136	{
2137	struct ceph_dir_file_info *dfi = file->private_data;
2138	struct inode *inode = file_inode(f: file);
2139	struct ceph_inode_info *ci = ceph_inode(inode);
2140	int left;
2141	const int bufsize = 1024;
2142
2143	if (!ceph_test_mount_opt(ceph_sb_to_fs_client(inode->i_sb), DIRSTAT))
2144	return -EISDIR;
2145
2146	if (!dfi->dir_info) {
2147	dfi->dir_info = kmalloc(bufsize, GFP_KERNEL);
2148	if (!dfi->dir_info)
2149	return -ENOMEM;
2150	dfi->dir_info_len =
2151	snprintf(buf: dfi->dir_info, size: bufsize,
2152	fmt: "entries: %20lld\n"
2153	" files: %20lld\n"
2154	" subdirs: %20lld\n"
2155	"rentries: %20lld\n"
2156	" rfiles: %20lld\n"
2157	" rsubdirs: %20lld\n"
2158	"rbytes: %20lld\n"
2159	"rctime: %ptSp\n",
2160	ci->i_files + ci->i_subdirs,
2161	ci->i_files,
2162	ci->i_subdirs,
2163	ci->i_rfiles + ci->i_rsubdirs,
2164	ci->i_rfiles,
2165	ci->i_rsubdirs,
2166	ci->i_rbytes,
2167	&ci->i_rctime);
2168	}
2169
2170	if (*ppos >= dfi->dir_info_len)
2171	return 0;
2172	size = min_t(unsigned, size, dfi->dir_info_len-*ppos);
2173	left = copy_to_user(to: buf, from: dfi->dir_info + *ppos, n: size);
2174	if (left == size)
2175	return -EFAULT;
2176	*ppos += (size - left);
2177	return size - left;
2178	}
2179
2180
2181
2182	/*
2183	* Return name hash for a given dentry. This is dependent on
2184	* the parent directory's hash function.
2185	*/
2186	unsigned ceph_dentry_hash(struct inode *dir, struct dentry *dn)
2187	{
2188	struct ceph_inode_info *dci = ceph_inode(inode: dir);
2189	unsigned hash;
2190
2191	switch (dci->i_dir_layout.dl_dir_hash) {
2192	case 0: /* for backward compat */
2193	case CEPH_STR_HASH_LINUX:
2194	return dn->d_name.hash;
2195
2196	default:
2197	spin_lock(lock: &dn->d_lock);
2198	hash = ceph_str_hash(type: dci->i_dir_layout.dl_dir_hash,
2199	s: dn->d_name.name, len: dn->d_name.len);
2200	spin_unlock(lock: &dn->d_lock);
2201	return hash;
2202	}
2203	}
2204
2205	WRAP_DIR_ITER(ceph_readdir) // FIXME!
2206	const struct file_operations ceph_dir_fops = {
2207	.read = ceph_read_dir,
2208	.iterate_shared = shared_ceph_readdir,
2209	.llseek = ceph_dir_llseek,
2210	.open = ceph_open,
2211	.release = ceph_release,
2212	.unlocked_ioctl = ceph_ioctl,
2213	.compat_ioctl = compat_ptr_ioctl,
2214	.fsync = ceph_fsync,
2215	.lock = ceph_lock,
2216	.flock = ceph_flock,
2217	.setlease = simple_nosetlease,
2218	};
2219
2220	const struct file_operations ceph_snapdir_fops = {
2221	.iterate_shared = shared_ceph_readdir,
2222	.llseek = ceph_dir_llseek,
2223	.open = ceph_open,
2224	.release = ceph_release,
2225	.setlease = simple_nosetlease,
2226	};
2227
2228	const struct inode_operations ceph_dir_iops = {
2229	.lookup = ceph_lookup,
2230	.permission = ceph_permission,
2231	.getattr = ceph_getattr,
2232	.setattr = ceph_setattr,
2233	.listxattr = ceph_listxattr,
2234	.get_inode_acl = ceph_get_acl,
2235	.set_acl = ceph_set_acl,
2236	.mknod = ceph_mknod,
2237	.symlink = ceph_symlink,
2238	.mkdir = ceph_mkdir,
2239	.link = ceph_link,
2240	.unlink = ceph_unlink,
2241	.rmdir = ceph_unlink,
2242	.rename = ceph_rename,
2243	.create = ceph_create,
2244	.atomic_open = ceph_atomic_open,
2245	};
2246
2247	const struct inode_operations ceph_snapdir_iops = {
2248	.lookup = ceph_lookup,
2249	.permission = ceph_permission,
2250	.getattr = ceph_getattr,
2251	.mkdir = ceph_mkdir,
2252	.rmdir = ceph_unlink,
2253	.rename = ceph_rename,
2254	};
2255
2256	const struct dentry_operations ceph_dentry_ops = {
2257	.d_revalidate = ceph_d_revalidate,
2258	.d_delete = ceph_d_delete,
2259	.d_release = ceph_d_release,
2260	.d_prune = ceph_d_prune,
2261	.d_init = ceph_d_init,
2262	};
2263