forked from torvalds/linux
-
Notifications
You must be signed in to change notification settings - Fork 0
/
volume.c
413 lines (354 loc) · 10.3 KB
/
volume.c
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
// SPDX-License-Identifier: GPL-2.0-or-later
/* AFS volume management
*
* Copyright (C) 2002, 2007 Red Hat, Inc. All Rights Reserved.
* Written by David Howells ([email protected])
*/
#include <linux/kernel.h>
#include <linux/slab.h>
#include "internal.h"
unsigned __read_mostly afs_volume_gc_delay = 10;
unsigned __read_mostly afs_volume_record_life = 60 * 60;
/*
* Insert a volume into a cell. If there's an existing volume record, that is
* returned instead with a ref held.
*/
static struct afs_volume *afs_insert_volume_into_cell(struct afs_cell *cell,
struct afs_volume *volume)
{
struct afs_volume *p;
struct rb_node *parent = NULL, **pp;
write_seqlock(&cell->volume_lock);
pp = &cell->volumes.rb_node;
while (*pp) {
parent = *pp;
p = rb_entry(parent, struct afs_volume, cell_node);
if (p->vid < volume->vid) {
pp = &(*pp)->rb_left;
} else if (p->vid > volume->vid) {
pp = &(*pp)->rb_right;
} else {
volume = afs_get_volume(p, afs_volume_trace_get_cell_insert);
goto found;
}
}
rb_link_node_rcu(&volume->cell_node, parent, pp);
rb_insert_color(&volume->cell_node, &cell->volumes);
hlist_add_head_rcu(&volume->proc_link, &cell->proc_volumes);
found:
write_sequnlock(&cell->volume_lock);
return volume;
}
static void afs_remove_volume_from_cell(struct afs_volume *volume)
{
struct afs_cell *cell = volume->cell;
if (!hlist_unhashed(&volume->proc_link)) {
trace_afs_volume(volume->vid, atomic_read(&volume->usage),
afs_volume_trace_remove);
write_seqlock(&cell->volume_lock);
hlist_del_rcu(&volume->proc_link);
rb_erase(&volume->cell_node, &cell->volumes);
write_sequnlock(&cell->volume_lock);
}
}
/*
* Allocate a volume record and load it up from a vldb record.
*/
static struct afs_volume *afs_alloc_volume(struct afs_fs_context *params,
struct afs_vldb_entry *vldb,
unsigned long type_mask)
{
struct afs_server_list *slist;
struct afs_volume *volume;
int ret = -ENOMEM, nr_servers = 0, i;
for (i = 0; i < vldb->nr_servers; i++)
if (vldb->fs_mask[i] & type_mask)
nr_servers++;
volume = kzalloc(sizeof(struct afs_volume), GFP_KERNEL);
if (!volume)
goto error_0;
volume->vid = vldb->vid[params->type];
volume->update_at = ktime_get_real_seconds() + afs_volume_record_life;
volume->cell = afs_get_cell(params->cell, afs_cell_trace_get_vol);
volume->type = params->type;
volume->type_force = params->force;
volume->name_len = vldb->name_len;
atomic_set(&volume->usage, 1);
INIT_HLIST_NODE(&volume->proc_link);
rwlock_init(&volume->servers_lock);
rwlock_init(&volume->cb_v_break_lock);
memcpy(volume->name, vldb->name, vldb->name_len + 1);
slist = afs_alloc_server_list(params->cell, params->key, vldb, type_mask);
if (IS_ERR(slist)) {
ret = PTR_ERR(slist);
goto error_1;
}
refcount_set(&slist->usage, 1);
rcu_assign_pointer(volume->servers, slist);
trace_afs_volume(volume->vid, 1, afs_volume_trace_alloc);
return volume;
error_1:
afs_put_cell(volume->cell, afs_cell_trace_put_vol);
kfree(volume);
error_0:
return ERR_PTR(ret);
}
/*
* Look up or allocate a volume record.
*/
static struct afs_volume *afs_lookup_volume(struct afs_fs_context *params,
struct afs_vldb_entry *vldb,
unsigned long type_mask)
{
struct afs_volume *candidate, *volume;
candidate = afs_alloc_volume(params, vldb, type_mask);
if (IS_ERR(candidate))
return candidate;
volume = afs_insert_volume_into_cell(params->cell, candidate);
if (volume != candidate)
afs_put_volume(params->net, candidate, afs_volume_trace_put_cell_dup);
return volume;
}
/*
* Look up a VLDB record for a volume.
*/
static struct afs_vldb_entry *afs_vl_lookup_vldb(struct afs_cell *cell,
struct key *key,
const char *volname,
size_t volnamesz)
{
struct afs_vldb_entry *vldb = ERR_PTR(-EDESTADDRREQ);
struct afs_vl_cursor vc;
int ret;
if (!afs_begin_vlserver_operation(&vc, cell, key))
return ERR_PTR(-ERESTARTSYS);
while (afs_select_vlserver(&vc)) {
vldb = afs_vl_get_entry_by_name_u(&vc, volname, volnamesz);
}
ret = afs_end_vlserver_operation(&vc);
return ret < 0 ? ERR_PTR(ret) : vldb;
}
/*
* Look up a volume in the VL server and create a candidate volume record for
* it.
*
* The volume name can be one of the following:
* "%[cell:]volume[.]" R/W volume
* "#[cell:]volume[.]" R/O or R/W volume (rwparent=0),
* or R/W (rwparent=1) volume
* "%[cell:]volume.readonly" R/O volume
* "#[cell:]volume.readonly" R/O volume
* "%[cell:]volume.backup" Backup volume
* "#[cell:]volume.backup" Backup volume
*
* The cell name is optional, and defaults to the current cell.
*
* See "The Rules of Mount Point Traversal" in Chapter 5 of the AFS SysAdmin
* Guide
* - Rule 1: Explicit type suffix forces access of that type or nothing
* (no suffix, then use Rule 2 & 3)
* - Rule 2: If parent volume is R/O, then mount R/O volume by preference, R/W
* if not available
* - Rule 3: If parent volume is R/W, then only mount R/W volume unless
* explicitly told otherwise
*/
struct afs_volume *afs_create_volume(struct afs_fs_context *params)
{
struct afs_vldb_entry *vldb;
struct afs_volume *volume;
unsigned long type_mask = 1UL << params->type;
vldb = afs_vl_lookup_vldb(params->cell, params->key,
params->volname, params->volnamesz);
if (IS_ERR(vldb))
return ERR_CAST(vldb);
if (test_bit(AFS_VLDB_QUERY_ERROR, &vldb->flags)) {
volume = ERR_PTR(vldb->error);
goto error;
}
/* Make the final decision on the type we want */
volume = ERR_PTR(-ENOMEDIUM);
if (params->force) {
if (!(vldb->flags & type_mask))
goto error;
} else if (test_bit(AFS_VLDB_HAS_RO, &vldb->flags)) {
params->type = AFSVL_ROVOL;
} else if (test_bit(AFS_VLDB_HAS_RW, &vldb->flags)) {
params->type = AFSVL_RWVOL;
} else {
goto error;
}
type_mask = 1UL << params->type;
volume = afs_lookup_volume(params, vldb, type_mask);
error:
kfree(vldb);
return volume;
}
/*
* Destroy a volume record
*/
static void afs_destroy_volume(struct afs_net *net, struct afs_volume *volume)
{
_enter("%p", volume);
#ifdef CONFIG_AFS_FSCACHE
ASSERTCMP(volume->cache, ==, NULL);
#endif
afs_remove_volume_from_cell(volume);
afs_put_serverlist(net, rcu_access_pointer(volume->servers));
afs_put_cell(volume->cell, afs_cell_trace_put_vol);
trace_afs_volume(volume->vid, atomic_read(&volume->usage),
afs_volume_trace_free);
kfree_rcu(volume, rcu);
_leave(" [destroyed]");
}
/*
* Get a reference on a volume record.
*/
struct afs_volume *afs_get_volume(struct afs_volume *volume,
enum afs_volume_trace reason)
{
if (volume) {
int u = atomic_inc_return(&volume->usage);
trace_afs_volume(volume->vid, u, reason);
}
return volume;
}
/*
* Drop a reference on a volume record.
*/
void afs_put_volume(struct afs_net *net, struct afs_volume *volume,
enum afs_volume_trace reason)
{
if (volume) {
afs_volid_t vid = volume->vid;
int u = atomic_dec_return(&volume->usage);
trace_afs_volume(vid, u, reason);
if (u == 0)
afs_destroy_volume(net, volume);
}
}
/*
* Activate a volume.
*/
void afs_activate_volume(struct afs_volume *volume)
{
#ifdef CONFIG_AFS_FSCACHE
volume->cache = fscache_acquire_cookie(volume->cell->cache,
&afs_volume_cache_index_def,
&volume->vid, sizeof(volume->vid),
NULL, 0,
volume, 0, true);
#endif
}
/*
* Deactivate a volume.
*/
void afs_deactivate_volume(struct afs_volume *volume)
{
_enter("%s", volume->name);
#ifdef CONFIG_AFS_FSCACHE
fscache_relinquish_cookie(volume->cache, NULL,
test_bit(AFS_VOLUME_DELETED, &volume->flags));
volume->cache = NULL;
#endif
_leave("");
}
/*
* Query the VL service to update the volume status.
*/
static int afs_update_volume_status(struct afs_volume *volume, struct key *key)
{
struct afs_server_list *new, *old, *discard;
struct afs_vldb_entry *vldb;
char idbuf[16];
int ret, idsz;
_enter("");
/* We look up an ID by passing it as a decimal string in the
* operation's name parameter.
*/
idsz = sprintf(idbuf, "%llu", volume->vid);
vldb = afs_vl_lookup_vldb(volume->cell, key, idbuf, idsz);
if (IS_ERR(vldb)) {
ret = PTR_ERR(vldb);
goto error;
}
/* See if the volume got renamed. */
if (vldb->name_len != volume->name_len ||
memcmp(vldb->name, volume->name, vldb->name_len) != 0) {
/* TODO: Use RCU'd string. */
memcpy(volume->name, vldb->name, AFS_MAXVOLNAME);
volume->name_len = vldb->name_len;
}
/* See if the volume's server list got updated. */
new = afs_alloc_server_list(volume->cell, key,
vldb, (1 << volume->type));
if (IS_ERR(new)) {
ret = PTR_ERR(new);
goto error_vldb;
}
write_lock(&volume->servers_lock);
discard = new;
old = rcu_dereference_protected(volume->servers,
lockdep_is_held(&volume->servers_lock));
if (afs_annotate_server_list(new, old)) {
new->seq = volume->servers_seq + 1;
rcu_assign_pointer(volume->servers, new);
smp_wmb();
volume->servers_seq++;
discard = old;
}
volume->update_at = ktime_get_real_seconds() + afs_volume_record_life;
write_unlock(&volume->servers_lock);
ret = 0;
afs_put_serverlist(volume->cell->net, discard);
error_vldb:
kfree(vldb);
error:
_leave(" = %d", ret);
return ret;
}
/*
* Make sure the volume record is up to date.
*/
int afs_check_volume_status(struct afs_volume *volume, struct afs_operation *op)
{
int ret, retries = 0;
_enter("");
retry:
if (test_bit(AFS_VOLUME_WAIT, &volume->flags))
goto wait;
if (volume->update_at <= ktime_get_real_seconds() ||
test_bit(AFS_VOLUME_NEEDS_UPDATE, &volume->flags))
goto update;
_leave(" = 0");
return 0;
update:
if (!test_and_set_bit_lock(AFS_VOLUME_UPDATING, &volume->flags)) {
clear_bit(AFS_VOLUME_NEEDS_UPDATE, &volume->flags);
ret = afs_update_volume_status(volume, op->key);
if (ret < 0)
set_bit(AFS_VOLUME_NEEDS_UPDATE, &volume->flags);
clear_bit_unlock(AFS_VOLUME_WAIT, &volume->flags);
clear_bit_unlock(AFS_VOLUME_UPDATING, &volume->flags);
wake_up_bit(&volume->flags, AFS_VOLUME_WAIT);
_leave(" = %d", ret);
return ret;
}
wait:
if (!test_bit(AFS_VOLUME_WAIT, &volume->flags)) {
_leave(" = 0 [no wait]");
return 0;
}
ret = wait_on_bit(&volume->flags, AFS_VOLUME_WAIT,
(op->flags & AFS_OPERATION_UNINTR) ?
TASK_UNINTERRUPTIBLE : TASK_INTERRUPTIBLE);
if (ret == -ERESTARTSYS) {
_leave(" = %d", ret);
return ret;
}
retries++;
if (retries == 4) {
_leave(" = -ESTALE");
return -ESTALE;
}
goto retry;
}