]>
Commit | Line | Data |
---|---|---|
602adf40 YS |
1 | /* |
2 | rbd.c -- Export ceph rados objects as a Linux block device | |
3 | ||
4 | ||
5 | based on drivers/block/osdblk.c: | |
6 | ||
7 | Copyright 2009 Red Hat, Inc. | |
8 | ||
9 | This program is free software; you can redistribute it and/or modify | |
10 | it under the terms of the GNU General Public License as published by | |
11 | the Free Software Foundation. | |
12 | ||
13 | This program is distributed in the hope that it will be useful, | |
14 | but WITHOUT ANY WARRANTY; without even the implied warranty of | |
15 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
16 | GNU General Public License for more details. | |
17 | ||
18 | You should have received a copy of the GNU General Public License | |
19 | along with this program; see the file COPYING. If not, write to | |
20 | the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. | |
21 | ||
22 | ||
23 | ||
dfc5606d | 24 | For usage instructions, please refer to: |
602adf40 | 25 | |
dfc5606d | 26 | Documentation/ABI/testing/sysfs-bus-rbd |
602adf40 YS |
27 | |
28 | */ | |
29 | ||
30 | #include <linux/ceph/libceph.h> | |
31 | #include <linux/ceph/osd_client.h> | |
32 | #include <linux/ceph/mon_client.h> | |
33 | #include <linux/ceph/decode.h> | |
59c2be1e | 34 | #include <linux/parser.h> |
602adf40 YS |
35 | |
36 | #include <linux/kernel.h> | |
37 | #include <linux/device.h> | |
38 | #include <linux/module.h> | |
39 | #include <linux/fs.h> | |
40 | #include <linux/blkdev.h> | |
41 | ||
42 | #include "rbd_types.h" | |
43 | ||
593a9e7b AE |
44 | /* |
45 | * The basic unit of block I/O is a sector. It is interpreted in a | |
46 | * number of contexts in Linux (blk, bio, genhd), but the default is | |
47 | * universally 512 bytes. These symbols are just slightly more | |
48 | * meaningful than the bare numbers they represent. | |
49 | */ | |
50 | #define SECTOR_SHIFT 9 | |
51 | #define SECTOR_SIZE (1ULL << SECTOR_SHIFT) | |
52 | ||
f0f8cef5 AE |
53 | #define RBD_DRV_NAME "rbd" |
54 | #define RBD_DRV_NAME_LONG "rbd (rados block device)" | |
602adf40 YS |
55 | |
56 | #define RBD_MINORS_PER_MAJOR 256 /* max minors per blkdev */ | |
57 | ||
602adf40 YS |
58 | #define RBD_MAX_SNAP_NAME_LEN 32 |
59 | #define RBD_MAX_OPT_LEN 1024 | |
60 | ||
61 | #define RBD_SNAP_HEAD_NAME "-" | |
62 | ||
81a89793 AE |
63 | /* |
64 | * An RBD device name will be "rbd#", where the "rbd" comes from | |
65 | * RBD_DRV_NAME above, and # is a unique integer identifier. | |
66 | * MAX_INT_FORMAT_WIDTH is used in ensuring DEV_NAME_LEN is big | |
67 | * enough to hold all possible device names. | |
68 | */ | |
602adf40 | 69 | #define DEV_NAME_LEN 32 |
81a89793 | 70 | #define MAX_INT_FORMAT_WIDTH ((5 * sizeof (int)) / 2 + 1) |
602adf40 | 71 | |
59c2be1e YS |
72 | #define RBD_NOTIFY_TIMEOUT_DEFAULT 10 |
73 | ||
602adf40 YS |
74 | /* |
75 | * block device image metadata (in-memory version) | |
76 | */ | |
77 | struct rbd_image_header { | |
78 | u64 image_size; | |
849b4260 | 79 | char *object_prefix; |
602adf40 YS |
80 | __u8 obj_order; |
81 | __u8 crypt_type; | |
82 | __u8 comp_type; | |
602adf40 YS |
83 | struct ceph_snap_context *snapc; |
84 | size_t snap_names_len; | |
85 | u64 snap_seq; | |
86 | u32 total_snaps; | |
87 | ||
88 | char *snap_names; | |
89 | u64 *snap_sizes; | |
59c2be1e YS |
90 | |
91 | u64 obj_version; | |
92 | }; | |
93 | ||
94 | struct rbd_options { | |
95 | int notify_timeout; | |
602adf40 YS |
96 | }; |
97 | ||
98 | /* | |
f0f8cef5 | 99 | * an instance of the client. multiple devices may share an rbd client. |
602adf40 YS |
100 | */ |
101 | struct rbd_client { | |
102 | struct ceph_client *client; | |
59c2be1e | 103 | struct rbd_options *rbd_opts; |
602adf40 YS |
104 | struct kref kref; |
105 | struct list_head node; | |
106 | }; | |
107 | ||
108 | /* | |
f0f8cef5 | 109 | * a request completion status |
602adf40 | 110 | */ |
1fec7093 YS |
111 | struct rbd_req_status { |
112 | int done; | |
113 | int rc; | |
114 | u64 bytes; | |
115 | }; | |
116 | ||
117 | /* | |
118 | * a collection of requests | |
119 | */ | |
120 | struct rbd_req_coll { | |
121 | int total; | |
122 | int num_done; | |
123 | struct kref kref; | |
124 | struct rbd_req_status status[0]; | |
602adf40 YS |
125 | }; |
126 | ||
f0f8cef5 AE |
127 | /* |
128 | * a single io request | |
129 | */ | |
130 | struct rbd_request { | |
131 | struct request *rq; /* blk layer request */ | |
132 | struct bio *bio; /* cloned bio */ | |
133 | struct page **pages; /* list of used pages */ | |
134 | u64 len; | |
135 | int coll_index; | |
136 | struct rbd_req_coll *coll; | |
137 | }; | |
138 | ||
dfc5606d YS |
139 | struct rbd_snap { |
140 | struct device dev; | |
141 | const char *name; | |
3591538f | 142 | u64 size; |
dfc5606d YS |
143 | struct list_head node; |
144 | u64 id; | |
145 | }; | |
146 | ||
602adf40 YS |
147 | /* |
148 | * a single device | |
149 | */ | |
150 | struct rbd_device { | |
151 | int id; /* blkdev unique id */ | |
152 | ||
153 | int major; /* blkdev assigned major */ | |
154 | struct gendisk *disk; /* blkdev's gendisk and rq */ | |
155 | struct request_queue *q; | |
156 | ||
602adf40 YS |
157 | struct rbd_client *rbd_client; |
158 | ||
159 | char name[DEV_NAME_LEN]; /* blkdev name, e.g. rbd3 */ | |
160 | ||
161 | spinlock_t lock; /* queue lock */ | |
162 | ||
163 | struct rbd_image_header header; | |
bf3e5ae1 AE |
164 | char *obj; /* rbd image name */ |
165 | size_t obj_len; | |
cb8627c7 | 166 | char *obj_md_name; /* hdr nm. */ |
d22f76e7 | 167 | char *pool_name; |
9bb2f334 | 168 | int pool_id; |
602adf40 | 169 | |
59c2be1e YS |
170 | struct ceph_osd_event *watch_event; |
171 | struct ceph_osd_request *watch_request; | |
172 | ||
c666601a JD |
173 | /* protects updating the header */ |
174 | struct rw_semaphore header_rwsem; | |
602adf40 | 175 | char snap_name[RBD_MAX_SNAP_NAME_LEN]; |
77dfe99f | 176 | u64 snap_id; /* current snapshot id */ |
602adf40 YS |
177 | int read_only; |
178 | ||
179 | struct list_head node; | |
dfc5606d YS |
180 | |
181 | /* list of snapshots */ | |
182 | struct list_head snaps; | |
183 | ||
184 | /* sysfs related */ | |
185 | struct device dev; | |
186 | }; | |
187 | ||
602adf40 | 188 | static DEFINE_MUTEX(ctl_mutex); /* Serialize open/close/setup/teardown */ |
e124a82f | 189 | |
602adf40 | 190 | static LIST_HEAD(rbd_dev_list); /* devices */ |
e124a82f AE |
191 | static DEFINE_SPINLOCK(rbd_dev_list_lock); |
192 | ||
432b8587 AE |
193 | static LIST_HEAD(rbd_client_list); /* clients */ |
194 | static DEFINE_SPINLOCK(rbd_client_list_lock); | |
602adf40 | 195 | |
dfc5606d YS |
196 | static int __rbd_init_snaps_header(struct rbd_device *rbd_dev); |
197 | static void rbd_dev_release(struct device *dev); | |
dfc5606d YS |
198 | static ssize_t rbd_snap_add(struct device *dev, |
199 | struct device_attribute *attr, | |
200 | const char *buf, | |
201 | size_t count); | |
202 | static void __rbd_remove_snap_dev(struct rbd_device *rbd_dev, | |
69932487 | 203 | struct rbd_snap *snap); |
dfc5606d | 204 | |
f0f8cef5 AE |
205 | static ssize_t rbd_add(struct bus_type *bus, const char *buf, |
206 | size_t count); | |
207 | static ssize_t rbd_remove(struct bus_type *bus, const char *buf, | |
208 | size_t count); | |
209 | ||
210 | static struct bus_attribute rbd_bus_attrs[] = { | |
211 | __ATTR(add, S_IWUSR, NULL, rbd_add), | |
212 | __ATTR(remove, S_IWUSR, NULL, rbd_remove), | |
213 | __ATTR_NULL | |
214 | }; | |
215 | ||
216 | static struct bus_type rbd_bus_type = { | |
217 | .name = "rbd", | |
218 | .bus_attrs = rbd_bus_attrs, | |
219 | }; | |
220 | ||
221 | static void rbd_root_dev_release(struct device *dev) | |
222 | { | |
223 | } | |
224 | ||
225 | static struct device rbd_root_dev = { | |
226 | .init_name = "rbd", | |
227 | .release = rbd_root_dev_release, | |
228 | }; | |
229 | ||
dfc5606d | 230 | |
dfc5606d YS |
231 | static struct device *rbd_get_dev(struct rbd_device *rbd_dev) |
232 | { | |
233 | return get_device(&rbd_dev->dev); | |
234 | } | |
235 | ||
236 | static void rbd_put_dev(struct rbd_device *rbd_dev) | |
237 | { | |
238 | put_device(&rbd_dev->dev); | |
239 | } | |
602adf40 | 240 | |
263c6ca0 | 241 | static int __rbd_refresh_header(struct rbd_device *rbd_dev); |
59c2be1e | 242 | |
602adf40 YS |
243 | static int rbd_open(struct block_device *bdev, fmode_t mode) |
244 | { | |
f0f8cef5 | 245 | struct rbd_device *rbd_dev = bdev->bd_disk->private_data; |
602adf40 | 246 | |
dfc5606d YS |
247 | rbd_get_dev(rbd_dev); |
248 | ||
602adf40 YS |
249 | set_device_ro(bdev, rbd_dev->read_only); |
250 | ||
251 | if ((mode & FMODE_WRITE) && rbd_dev->read_only) | |
252 | return -EROFS; | |
253 | ||
254 | return 0; | |
255 | } | |
256 | ||
dfc5606d YS |
257 | static int rbd_release(struct gendisk *disk, fmode_t mode) |
258 | { | |
259 | struct rbd_device *rbd_dev = disk->private_data; | |
260 | ||
261 | rbd_put_dev(rbd_dev); | |
262 | ||
263 | return 0; | |
264 | } | |
265 | ||
602adf40 YS |
266 | static const struct block_device_operations rbd_bd_ops = { |
267 | .owner = THIS_MODULE, | |
268 | .open = rbd_open, | |
dfc5606d | 269 | .release = rbd_release, |
602adf40 YS |
270 | }; |
271 | ||
272 | /* | |
273 | * Initialize an rbd client instance. | |
274 | * We own *opt. | |
275 | */ | |
59c2be1e YS |
276 | static struct rbd_client *rbd_client_create(struct ceph_options *opt, |
277 | struct rbd_options *rbd_opts) | |
602adf40 YS |
278 | { |
279 | struct rbd_client *rbdc; | |
280 | int ret = -ENOMEM; | |
281 | ||
282 | dout("rbd_client_create\n"); | |
283 | rbdc = kmalloc(sizeof(struct rbd_client), GFP_KERNEL); | |
284 | if (!rbdc) | |
285 | goto out_opt; | |
286 | ||
287 | kref_init(&rbdc->kref); | |
288 | INIT_LIST_HEAD(&rbdc->node); | |
289 | ||
bc534d86 AE |
290 | mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING); |
291 | ||
6ab00d46 | 292 | rbdc->client = ceph_create_client(opt, rbdc, 0, 0); |
602adf40 | 293 | if (IS_ERR(rbdc->client)) |
bc534d86 | 294 | goto out_mutex; |
28f259b7 | 295 | opt = NULL; /* Now rbdc->client is responsible for opt */ |
602adf40 YS |
296 | |
297 | ret = ceph_open_session(rbdc->client); | |
298 | if (ret < 0) | |
299 | goto out_err; | |
300 | ||
59c2be1e YS |
301 | rbdc->rbd_opts = rbd_opts; |
302 | ||
432b8587 | 303 | spin_lock(&rbd_client_list_lock); |
602adf40 | 304 | list_add_tail(&rbdc->node, &rbd_client_list); |
432b8587 | 305 | spin_unlock(&rbd_client_list_lock); |
602adf40 | 306 | |
bc534d86 AE |
307 | mutex_unlock(&ctl_mutex); |
308 | ||
602adf40 YS |
309 | dout("rbd_client_create created %p\n", rbdc); |
310 | return rbdc; | |
311 | ||
312 | out_err: | |
313 | ceph_destroy_client(rbdc->client); | |
bc534d86 AE |
314 | out_mutex: |
315 | mutex_unlock(&ctl_mutex); | |
602adf40 YS |
316 | kfree(rbdc); |
317 | out_opt: | |
28f259b7 VK |
318 | if (opt) |
319 | ceph_destroy_options(opt); | |
320 | return ERR_PTR(ret); | |
602adf40 YS |
321 | } |
322 | ||
323 | /* | |
324 | * Find a ceph client with specific addr and configuration. | |
325 | */ | |
326 | static struct rbd_client *__rbd_client_find(struct ceph_options *opt) | |
327 | { | |
328 | struct rbd_client *client_node; | |
329 | ||
330 | if (opt->flags & CEPH_OPT_NOSHARE) | |
331 | return NULL; | |
332 | ||
333 | list_for_each_entry(client_node, &rbd_client_list, node) | |
334 | if (ceph_compare_options(opt, client_node->client) == 0) | |
335 | return client_node; | |
336 | return NULL; | |
337 | } | |
338 | ||
59c2be1e YS |
339 | /* |
340 | * mount options | |
341 | */ | |
342 | enum { | |
343 | Opt_notify_timeout, | |
344 | Opt_last_int, | |
345 | /* int args above */ | |
346 | Opt_last_string, | |
347 | /* string args above */ | |
348 | }; | |
349 | ||
350 | static match_table_t rbdopt_tokens = { | |
351 | {Opt_notify_timeout, "notify_timeout=%d"}, | |
352 | /* int args above */ | |
353 | /* string args above */ | |
354 | {-1, NULL} | |
355 | }; | |
356 | ||
357 | static int parse_rbd_opts_token(char *c, void *private) | |
358 | { | |
359 | struct rbd_options *rbdopt = private; | |
360 | substring_t argstr[MAX_OPT_ARGS]; | |
361 | int token, intval, ret; | |
362 | ||
21079786 | 363 | token = match_token(c, rbdopt_tokens, argstr); |
59c2be1e YS |
364 | if (token < 0) |
365 | return -EINVAL; | |
366 | ||
367 | if (token < Opt_last_int) { | |
368 | ret = match_int(&argstr[0], &intval); | |
369 | if (ret < 0) { | |
370 | pr_err("bad mount option arg (not int) " | |
371 | "at '%s'\n", c); | |
372 | return ret; | |
373 | } | |
374 | dout("got int token %d val %d\n", token, intval); | |
375 | } else if (token > Opt_last_int && token < Opt_last_string) { | |
376 | dout("got string token %d val %s\n", token, | |
377 | argstr[0].from); | |
378 | } else { | |
379 | dout("got token %d\n", token); | |
380 | } | |
381 | ||
382 | switch (token) { | |
383 | case Opt_notify_timeout: | |
384 | rbdopt->notify_timeout = intval; | |
385 | break; | |
386 | default: | |
387 | BUG_ON(token); | |
388 | } | |
389 | return 0; | |
390 | } | |
391 | ||
602adf40 YS |
392 | /* |
393 | * Get a ceph client with specific addr and configuration, if one does | |
394 | * not exist create it. | |
395 | */ | |
5214ecc4 AE |
396 | static struct rbd_client *rbd_get_client(const char *mon_addr, |
397 | size_t mon_addr_len, | |
398 | char *options) | |
602adf40 YS |
399 | { |
400 | struct rbd_client *rbdc; | |
401 | struct ceph_options *opt; | |
59c2be1e YS |
402 | struct rbd_options *rbd_opts; |
403 | ||
404 | rbd_opts = kzalloc(sizeof(*rbd_opts), GFP_KERNEL); | |
405 | if (!rbd_opts) | |
d720bcb0 | 406 | return ERR_PTR(-ENOMEM); |
59c2be1e YS |
407 | |
408 | rbd_opts->notify_timeout = RBD_NOTIFY_TIMEOUT_DEFAULT; | |
602adf40 | 409 | |
ee57741c | 410 | opt = ceph_parse_options(options, mon_addr, |
5214ecc4 | 411 | mon_addr + mon_addr_len, |
21079786 | 412 | parse_rbd_opts_token, rbd_opts); |
ee57741c | 413 | if (IS_ERR(opt)) { |
d720bcb0 AE |
414 | kfree(rbd_opts); |
415 | return ERR_CAST(opt); | |
ee57741c | 416 | } |
602adf40 | 417 | |
432b8587 | 418 | spin_lock(&rbd_client_list_lock); |
602adf40 YS |
419 | rbdc = __rbd_client_find(opt); |
420 | if (rbdc) { | |
602adf40 YS |
421 | /* using an existing client */ |
422 | kref_get(&rbdc->kref); | |
432b8587 | 423 | spin_unlock(&rbd_client_list_lock); |
e6994d3d | 424 | |
e6994d3d AE |
425 | ceph_destroy_options(opt); |
426 | kfree(rbd_opts); | |
427 | ||
d720bcb0 | 428 | return rbdc; |
602adf40 | 429 | } |
432b8587 | 430 | spin_unlock(&rbd_client_list_lock); |
602adf40 | 431 | |
59c2be1e | 432 | rbdc = rbd_client_create(opt, rbd_opts); |
d97081b0 | 433 | |
d720bcb0 AE |
434 | if (IS_ERR(rbdc)) |
435 | kfree(rbd_opts); | |
602adf40 | 436 | |
d720bcb0 | 437 | return rbdc; |
602adf40 YS |
438 | } |
439 | ||
440 | /* | |
441 | * Destroy ceph client | |
d23a4b3f | 442 | * |
432b8587 | 443 | * Caller must hold rbd_client_list_lock. |
602adf40 YS |
444 | */ |
445 | static void rbd_client_release(struct kref *kref) | |
446 | { | |
447 | struct rbd_client *rbdc = container_of(kref, struct rbd_client, kref); | |
448 | ||
449 | dout("rbd_release_client %p\n", rbdc); | |
cd9d9f5d | 450 | spin_lock(&rbd_client_list_lock); |
602adf40 | 451 | list_del(&rbdc->node); |
cd9d9f5d | 452 | spin_unlock(&rbd_client_list_lock); |
602adf40 YS |
453 | |
454 | ceph_destroy_client(rbdc->client); | |
59c2be1e | 455 | kfree(rbdc->rbd_opts); |
602adf40 YS |
456 | kfree(rbdc); |
457 | } | |
458 | ||
459 | /* | |
460 | * Drop reference to ceph client node. If it's not referenced anymore, release | |
461 | * it. | |
462 | */ | |
463 | static void rbd_put_client(struct rbd_device *rbd_dev) | |
464 | { | |
465 | kref_put(&rbd_dev->rbd_client->kref, rbd_client_release); | |
466 | rbd_dev->rbd_client = NULL; | |
602adf40 YS |
467 | } |
468 | ||
1fec7093 YS |
469 | /* |
470 | * Destroy requests collection | |
471 | */ | |
472 | static void rbd_coll_release(struct kref *kref) | |
473 | { | |
474 | struct rbd_req_coll *coll = | |
475 | container_of(kref, struct rbd_req_coll, kref); | |
476 | ||
477 | dout("rbd_coll_release %p\n", coll); | |
478 | kfree(coll); | |
479 | } | |
602adf40 YS |
480 | |
481 | /* | |
482 | * Create a new header structure, translate header format from the on-disk | |
483 | * header. | |
484 | */ | |
485 | static int rbd_header_from_disk(struct rbd_image_header *header, | |
486 | struct rbd_image_header_ondisk *ondisk, | |
50f7c4c9 | 487 | u32 allocated_snaps, |
602adf40 YS |
488 | gfp_t gfp_flags) |
489 | { | |
50f7c4c9 | 490 | u32 i, snap_count; |
602adf40 | 491 | |
21079786 | 492 | if (memcmp(ondisk, RBD_HEADER_TEXT, sizeof(RBD_HEADER_TEXT))) |
81e759fb | 493 | return -ENXIO; |
81e759fb | 494 | |
00f1f36f | 495 | snap_count = le32_to_cpu(ondisk->snap_count); |
50f7c4c9 XW |
496 | if (snap_count > (UINT_MAX - sizeof(struct ceph_snap_context)) |
497 | / sizeof (*ondisk)) | |
498 | return -EINVAL; | |
602adf40 | 499 | header->snapc = kmalloc(sizeof(struct ceph_snap_context) + |
f9f9a190 | 500 | snap_count * sizeof(u64), |
602adf40 YS |
501 | gfp_flags); |
502 | if (!header->snapc) | |
503 | return -ENOMEM; | |
00f1f36f | 504 | |
00f1f36f | 505 | header->snap_names_len = le64_to_cpu(ondisk->snap_names_len); |
602adf40 YS |
506 | if (snap_count) { |
507 | header->snap_names = kmalloc(header->snap_names_len, | |
f8ad495a | 508 | gfp_flags); |
602adf40 YS |
509 | if (!header->snap_names) |
510 | goto err_snapc; | |
511 | header->snap_sizes = kmalloc(snap_count * sizeof(u64), | |
f8ad495a | 512 | gfp_flags); |
602adf40 YS |
513 | if (!header->snap_sizes) |
514 | goto err_names; | |
515 | } else { | |
516 | header->snap_names = NULL; | |
517 | header->snap_sizes = NULL; | |
518 | } | |
849b4260 AE |
519 | |
520 | header->object_prefix = kmalloc(sizeof (ondisk->block_name) + 1, | |
521 | gfp_flags); | |
522 | if (!header->object_prefix) | |
523 | goto err_sizes; | |
524 | ||
ca1e49a6 | 525 | memcpy(header->object_prefix, ondisk->block_name, |
602adf40 | 526 | sizeof(ondisk->block_name)); |
849b4260 | 527 | header->object_prefix[sizeof (ondisk->block_name)] = '\0'; |
602adf40 YS |
528 | |
529 | header->image_size = le64_to_cpu(ondisk->image_size); | |
530 | header->obj_order = ondisk->options.order; | |
531 | header->crypt_type = ondisk->options.crypt_type; | |
532 | header->comp_type = ondisk->options.comp_type; | |
533 | ||
534 | atomic_set(&header->snapc->nref, 1); | |
535 | header->snap_seq = le64_to_cpu(ondisk->snap_seq); | |
536 | header->snapc->num_snaps = snap_count; | |
537 | header->total_snaps = snap_count; | |
538 | ||
21079786 | 539 | if (snap_count && allocated_snaps == snap_count) { |
602adf40 YS |
540 | for (i = 0; i < snap_count; i++) { |
541 | header->snapc->snaps[i] = | |
542 | le64_to_cpu(ondisk->snaps[i].id); | |
543 | header->snap_sizes[i] = | |
544 | le64_to_cpu(ondisk->snaps[i].image_size); | |
545 | } | |
546 | ||
547 | /* copy snapshot names */ | |
548 | memcpy(header->snap_names, &ondisk->snaps[i], | |
549 | header->snap_names_len); | |
550 | } | |
551 | ||
552 | return 0; | |
553 | ||
849b4260 AE |
554 | err_sizes: |
555 | kfree(header->snap_sizes); | |
602adf40 YS |
556 | err_names: |
557 | kfree(header->snap_names); | |
558 | err_snapc: | |
559 | kfree(header->snapc); | |
00f1f36f | 560 | return -ENOMEM; |
602adf40 YS |
561 | } |
562 | ||
602adf40 YS |
563 | static int snap_by_name(struct rbd_image_header *header, const char *snap_name, |
564 | u64 *seq, u64 *size) | |
565 | { | |
566 | int i; | |
567 | char *p = header->snap_names; | |
568 | ||
00f1f36f AE |
569 | for (i = 0; i < header->total_snaps; i++) { |
570 | if (!strcmp(snap_name, p)) { | |
602adf40 | 571 | |
00f1f36f | 572 | /* Found it. Pass back its id and/or size */ |
602adf40 | 573 | |
00f1f36f AE |
574 | if (seq) |
575 | *seq = header->snapc->snaps[i]; | |
576 | if (size) | |
577 | *size = header->snap_sizes[i]; | |
578 | return i; | |
579 | } | |
580 | p += strlen(p) + 1; /* Skip ahead to the next name */ | |
581 | } | |
582 | return -ENOENT; | |
602adf40 YS |
583 | } |
584 | ||
cc9d734c | 585 | static int rbd_header_set_snap(struct rbd_device *dev, u64 *size) |
602adf40 YS |
586 | { |
587 | struct rbd_image_header *header = &dev->header; | |
588 | struct ceph_snap_context *snapc = header->snapc; | |
589 | int ret = -ENOENT; | |
590 | ||
cc9d734c JD |
591 | BUILD_BUG_ON(sizeof (dev->snap_name) < sizeof (RBD_SNAP_HEAD_NAME)); |
592 | ||
c666601a | 593 | down_write(&dev->header_rwsem); |
602adf40 | 594 | |
cc9d734c JD |
595 | if (!memcmp(dev->snap_name, RBD_SNAP_HEAD_NAME, |
596 | sizeof (RBD_SNAP_HEAD_NAME))) { | |
602adf40 YS |
597 | if (header->total_snaps) |
598 | snapc->seq = header->snap_seq; | |
599 | else | |
600 | snapc->seq = 0; | |
77dfe99f | 601 | dev->snap_id = CEPH_NOSNAP; |
602adf40 YS |
602 | dev->read_only = 0; |
603 | if (size) | |
604 | *size = header->image_size; | |
605 | } else { | |
cc9d734c | 606 | ret = snap_by_name(header, dev->snap_name, &snapc->seq, size); |
602adf40 YS |
607 | if (ret < 0) |
608 | goto done; | |
77dfe99f | 609 | dev->snap_id = snapc->seq; |
602adf40 YS |
610 | dev->read_only = 1; |
611 | } | |
612 | ||
613 | ret = 0; | |
614 | done: | |
c666601a | 615 | up_write(&dev->header_rwsem); |
602adf40 YS |
616 | return ret; |
617 | } | |
618 | ||
619 | static void rbd_header_free(struct rbd_image_header *header) | |
620 | { | |
849b4260 | 621 | kfree(header->object_prefix); |
602adf40 | 622 | kfree(header->snap_sizes); |
849b4260 AE |
623 | kfree(header->snap_names); |
624 | kfree(header->snapc); | |
602adf40 YS |
625 | } |
626 | ||
627 | /* | |
628 | * get the actual striped segment name, offset and length | |
629 | */ | |
630 | static u64 rbd_get_segment(struct rbd_image_header *header, | |
ca1e49a6 | 631 | const char *object_prefix, |
602adf40 YS |
632 | u64 ofs, u64 len, |
633 | char *seg_name, u64 *segofs) | |
634 | { | |
635 | u64 seg = ofs >> header->obj_order; | |
636 | ||
637 | if (seg_name) | |
638 | snprintf(seg_name, RBD_MAX_SEG_NAME_LEN, | |
ca1e49a6 | 639 | "%s.%012llx", object_prefix, seg); |
602adf40 YS |
640 | |
641 | ofs = ofs & ((1 << header->obj_order) - 1); | |
642 | len = min_t(u64, len, (1 << header->obj_order) - ofs); | |
643 | ||
644 | if (segofs) | |
645 | *segofs = ofs; | |
646 | ||
647 | return len; | |
648 | } | |
649 | ||
1fec7093 YS |
650 | static int rbd_get_num_segments(struct rbd_image_header *header, |
651 | u64 ofs, u64 len) | |
652 | { | |
653 | u64 start_seg = ofs >> header->obj_order; | |
654 | u64 end_seg = (ofs + len - 1) >> header->obj_order; | |
655 | return end_seg - start_seg + 1; | |
656 | } | |
657 | ||
029bcbd8 JD |
658 | /* |
659 | * returns the size of an object in the image | |
660 | */ | |
661 | static u64 rbd_obj_bytes(struct rbd_image_header *header) | |
662 | { | |
663 | return 1 << header->obj_order; | |
664 | } | |
665 | ||
602adf40 YS |
666 | /* |
667 | * bio helpers | |
668 | */ | |
669 | ||
670 | static void bio_chain_put(struct bio *chain) | |
671 | { | |
672 | struct bio *tmp; | |
673 | ||
674 | while (chain) { | |
675 | tmp = chain; | |
676 | chain = chain->bi_next; | |
677 | bio_put(tmp); | |
678 | } | |
679 | } | |
680 | ||
681 | /* | |
682 | * zeros a bio chain, starting at specific offset | |
683 | */ | |
684 | static void zero_bio_chain(struct bio *chain, int start_ofs) | |
685 | { | |
686 | struct bio_vec *bv; | |
687 | unsigned long flags; | |
688 | void *buf; | |
689 | int i; | |
690 | int pos = 0; | |
691 | ||
692 | while (chain) { | |
693 | bio_for_each_segment(bv, chain, i) { | |
694 | if (pos + bv->bv_len > start_ofs) { | |
695 | int remainder = max(start_ofs - pos, 0); | |
696 | buf = bvec_kmap_irq(bv, &flags); | |
697 | memset(buf + remainder, 0, | |
698 | bv->bv_len - remainder); | |
85b5aaa6 | 699 | bvec_kunmap_irq(buf, &flags); |
602adf40 YS |
700 | } |
701 | pos += bv->bv_len; | |
702 | } | |
703 | ||
704 | chain = chain->bi_next; | |
705 | } | |
706 | } | |
707 | ||
708 | /* | |
709 | * bio_chain_clone - clone a chain of bios up to a certain length. | |
710 | * might return a bio_pair that will need to be released. | |
711 | */ | |
712 | static struct bio *bio_chain_clone(struct bio **old, struct bio **next, | |
713 | struct bio_pair **bp, | |
714 | int len, gfp_t gfpmask) | |
715 | { | |
716 | struct bio *tmp, *old_chain = *old, *new_chain = NULL, *tail = NULL; | |
717 | int total = 0; | |
718 | ||
719 | if (*bp) { | |
720 | bio_pair_release(*bp); | |
721 | *bp = NULL; | |
722 | } | |
723 | ||
724 | while (old_chain && (total < len)) { | |
725 | tmp = bio_kmalloc(gfpmask, old_chain->bi_max_vecs); | |
726 | if (!tmp) | |
727 | goto err_out; | |
728 | ||
729 | if (total + old_chain->bi_size > len) { | |
730 | struct bio_pair *bp; | |
731 | ||
732 | /* | |
733 | * this split can only happen with a single paged bio, | |
734 | * split_bio will BUG_ON if this is not the case | |
735 | */ | |
736 | dout("bio_chain_clone split! total=%d remaining=%d" | |
737 | "bi_size=%d\n", | |
738 | (int)total, (int)len-total, | |
739 | (int)old_chain->bi_size); | |
740 | ||
741 | /* split the bio. We'll release it either in the next | |
742 | call, or it will have to be released outside */ | |
593a9e7b | 743 | bp = bio_split(old_chain, (len - total) / SECTOR_SIZE); |
602adf40 YS |
744 | if (!bp) |
745 | goto err_out; | |
746 | ||
747 | __bio_clone(tmp, &bp->bio1); | |
748 | ||
749 | *next = &bp->bio2; | |
750 | } else { | |
751 | __bio_clone(tmp, old_chain); | |
752 | *next = old_chain->bi_next; | |
753 | } | |
754 | ||
755 | tmp->bi_bdev = NULL; | |
756 | gfpmask &= ~__GFP_WAIT; | |
757 | tmp->bi_next = NULL; | |
758 | ||
759 | if (!new_chain) { | |
760 | new_chain = tail = tmp; | |
761 | } else { | |
762 | tail->bi_next = tmp; | |
763 | tail = tmp; | |
764 | } | |
765 | old_chain = old_chain->bi_next; | |
766 | ||
767 | total += tmp->bi_size; | |
768 | } | |
769 | ||
770 | BUG_ON(total < len); | |
771 | ||
772 | if (tail) | |
773 | tail->bi_next = NULL; | |
774 | ||
775 | *old = old_chain; | |
776 | ||
777 | return new_chain; | |
778 | ||
779 | err_out: | |
780 | dout("bio_chain_clone with err\n"); | |
781 | bio_chain_put(new_chain); | |
782 | return NULL; | |
783 | } | |
784 | ||
785 | /* | |
786 | * helpers for osd request op vectors. | |
787 | */ | |
788 | static int rbd_create_rw_ops(struct ceph_osd_req_op **ops, | |
789 | int num_ops, | |
790 | int opcode, | |
791 | u32 payload_len) | |
792 | { | |
793 | *ops = kzalloc(sizeof(struct ceph_osd_req_op) * (num_ops + 1), | |
794 | GFP_NOIO); | |
795 | if (!*ops) | |
796 | return -ENOMEM; | |
797 | (*ops)[0].op = opcode; | |
798 | /* | |
799 | * op extent offset and length will be set later on | |
800 | * in calc_raw_layout() | |
801 | */ | |
802 | (*ops)[0].payload_len = payload_len; | |
803 | return 0; | |
804 | } | |
805 | ||
806 | static void rbd_destroy_ops(struct ceph_osd_req_op *ops) | |
807 | { | |
808 | kfree(ops); | |
809 | } | |
810 | ||
1fec7093 YS |
811 | static void rbd_coll_end_req_index(struct request *rq, |
812 | struct rbd_req_coll *coll, | |
813 | int index, | |
814 | int ret, u64 len) | |
815 | { | |
816 | struct request_queue *q; | |
817 | int min, max, i; | |
818 | ||
819 | dout("rbd_coll_end_req_index %p index %d ret %d len %lld\n", | |
820 | coll, index, ret, len); | |
821 | ||
822 | if (!rq) | |
823 | return; | |
824 | ||
825 | if (!coll) { | |
826 | blk_end_request(rq, ret, len); | |
827 | return; | |
828 | } | |
829 | ||
830 | q = rq->q; | |
831 | ||
832 | spin_lock_irq(q->queue_lock); | |
833 | coll->status[index].done = 1; | |
834 | coll->status[index].rc = ret; | |
835 | coll->status[index].bytes = len; | |
836 | max = min = coll->num_done; | |
837 | while (max < coll->total && coll->status[max].done) | |
838 | max++; | |
839 | ||
840 | for (i = min; i<max; i++) { | |
841 | __blk_end_request(rq, coll->status[i].rc, | |
842 | coll->status[i].bytes); | |
843 | coll->num_done++; | |
844 | kref_put(&coll->kref, rbd_coll_release); | |
845 | } | |
846 | spin_unlock_irq(q->queue_lock); | |
847 | } | |
848 | ||
849 | static void rbd_coll_end_req(struct rbd_request *req, | |
850 | int ret, u64 len) | |
851 | { | |
852 | rbd_coll_end_req_index(req->rq, req->coll, req->coll_index, ret, len); | |
853 | } | |
854 | ||
602adf40 YS |
855 | /* |
856 | * Send ceph osd request | |
857 | */ | |
858 | static int rbd_do_request(struct request *rq, | |
859 | struct rbd_device *dev, | |
860 | struct ceph_snap_context *snapc, | |
861 | u64 snapid, | |
862 | const char *obj, u64 ofs, u64 len, | |
863 | struct bio *bio, | |
864 | struct page **pages, | |
865 | int num_pages, | |
866 | int flags, | |
867 | struct ceph_osd_req_op *ops, | |
868 | int num_reply, | |
1fec7093 YS |
869 | struct rbd_req_coll *coll, |
870 | int coll_index, | |
602adf40 | 871 | void (*rbd_cb)(struct ceph_osd_request *req, |
59c2be1e YS |
872 | struct ceph_msg *msg), |
873 | struct ceph_osd_request **linger_req, | |
874 | u64 *ver) | |
602adf40 YS |
875 | { |
876 | struct ceph_osd_request *req; | |
877 | struct ceph_file_layout *layout; | |
878 | int ret; | |
879 | u64 bno; | |
880 | struct timespec mtime = CURRENT_TIME; | |
881 | struct rbd_request *req_data; | |
882 | struct ceph_osd_request_head *reqhead; | |
1dbb4399 | 883 | struct ceph_osd_client *osdc; |
602adf40 | 884 | |
602adf40 | 885 | req_data = kzalloc(sizeof(*req_data), GFP_NOIO); |
1fec7093 YS |
886 | if (!req_data) { |
887 | if (coll) | |
888 | rbd_coll_end_req_index(rq, coll, coll_index, | |
889 | -ENOMEM, len); | |
890 | return -ENOMEM; | |
891 | } | |
892 | ||
893 | if (coll) { | |
894 | req_data->coll = coll; | |
895 | req_data->coll_index = coll_index; | |
896 | } | |
602adf40 | 897 | |
1fec7093 | 898 | dout("rbd_do_request obj=%s ofs=%lld len=%lld\n", obj, len, ofs); |
602adf40 | 899 | |
c666601a | 900 | down_read(&dev->header_rwsem); |
602adf40 | 901 | |
1dbb4399 AE |
902 | osdc = &dev->rbd_client->client->osdc; |
903 | req = ceph_osdc_alloc_request(osdc, flags, snapc, ops, | |
904 | false, GFP_NOIO, pages, bio); | |
4ad12621 | 905 | if (!req) { |
c666601a | 906 | up_read(&dev->header_rwsem); |
4ad12621 | 907 | ret = -ENOMEM; |
602adf40 YS |
908 | goto done_pages; |
909 | } | |
910 | ||
911 | req->r_callback = rbd_cb; | |
912 | ||
913 | req_data->rq = rq; | |
914 | req_data->bio = bio; | |
915 | req_data->pages = pages; | |
916 | req_data->len = len; | |
917 | ||
918 | req->r_priv = req_data; | |
919 | ||
920 | reqhead = req->r_request->front.iov_base; | |
921 | reqhead->snapid = cpu_to_le64(CEPH_NOSNAP); | |
922 | ||
923 | strncpy(req->r_oid, obj, sizeof(req->r_oid)); | |
924 | req->r_oid_len = strlen(req->r_oid); | |
925 | ||
926 | layout = &req->r_file_layout; | |
927 | memset(layout, 0, sizeof(*layout)); | |
928 | layout->fl_stripe_unit = cpu_to_le32(1 << RBD_MAX_OBJ_ORDER); | |
929 | layout->fl_stripe_count = cpu_to_le32(1); | |
930 | layout->fl_object_size = cpu_to_le32(1 << RBD_MAX_OBJ_ORDER); | |
9bb2f334 | 931 | layout->fl_pg_pool = cpu_to_le32(dev->pool_id); |
1dbb4399 AE |
932 | ceph_calc_raw_layout(osdc, layout, snapid, ofs, &len, &bno, |
933 | req, ops); | |
602adf40 YS |
934 | |
935 | ceph_osdc_build_request(req, ofs, &len, | |
936 | ops, | |
937 | snapc, | |
938 | &mtime, | |
939 | req->r_oid, req->r_oid_len); | |
c666601a | 940 | up_read(&dev->header_rwsem); |
602adf40 | 941 | |
59c2be1e | 942 | if (linger_req) { |
1dbb4399 | 943 | ceph_osdc_set_request_linger(osdc, req); |
59c2be1e YS |
944 | *linger_req = req; |
945 | } | |
946 | ||
1dbb4399 | 947 | ret = ceph_osdc_start_request(osdc, req, false); |
602adf40 YS |
948 | if (ret < 0) |
949 | goto done_err; | |
950 | ||
951 | if (!rbd_cb) { | |
1dbb4399 | 952 | ret = ceph_osdc_wait_request(osdc, req); |
59c2be1e YS |
953 | if (ver) |
954 | *ver = le64_to_cpu(req->r_reassert_version.version); | |
1fec7093 YS |
955 | dout("reassert_ver=%lld\n", |
956 | le64_to_cpu(req->r_reassert_version.version)); | |
602adf40 YS |
957 | ceph_osdc_put_request(req); |
958 | } | |
959 | return ret; | |
960 | ||
961 | done_err: | |
962 | bio_chain_put(req_data->bio); | |
963 | ceph_osdc_put_request(req); | |
964 | done_pages: | |
1fec7093 | 965 | rbd_coll_end_req(req_data, ret, len); |
602adf40 | 966 | kfree(req_data); |
602adf40 YS |
967 | return ret; |
968 | } | |
969 | ||
970 | /* | |
971 | * Ceph osd op callback | |
972 | */ | |
973 | static void rbd_req_cb(struct ceph_osd_request *req, struct ceph_msg *msg) | |
974 | { | |
975 | struct rbd_request *req_data = req->r_priv; | |
976 | struct ceph_osd_reply_head *replyhead; | |
977 | struct ceph_osd_op *op; | |
978 | __s32 rc; | |
979 | u64 bytes; | |
980 | int read_op; | |
981 | ||
982 | /* parse reply */ | |
983 | replyhead = msg->front.iov_base; | |
984 | WARN_ON(le32_to_cpu(replyhead->num_ops) == 0); | |
985 | op = (void *)(replyhead + 1); | |
986 | rc = le32_to_cpu(replyhead->result); | |
987 | bytes = le64_to_cpu(op->extent.length); | |
895cfcc8 | 988 | read_op = (le16_to_cpu(op->op) == CEPH_OSD_OP_READ); |
602adf40 YS |
989 | |
990 | dout("rbd_req_cb bytes=%lld readop=%d rc=%d\n", bytes, read_op, rc); | |
991 | ||
992 | if (rc == -ENOENT && read_op) { | |
993 | zero_bio_chain(req_data->bio, 0); | |
994 | rc = 0; | |
995 | } else if (rc == 0 && read_op && bytes < req_data->len) { | |
996 | zero_bio_chain(req_data->bio, bytes); | |
997 | bytes = req_data->len; | |
998 | } | |
999 | ||
1fec7093 | 1000 | rbd_coll_end_req(req_data, rc, bytes); |
602adf40 YS |
1001 | |
1002 | if (req_data->bio) | |
1003 | bio_chain_put(req_data->bio); | |
1004 | ||
1005 | ceph_osdc_put_request(req); | |
1006 | kfree(req_data); | |
1007 | } | |
1008 | ||
59c2be1e YS |
1009 | static void rbd_simple_req_cb(struct ceph_osd_request *req, struct ceph_msg *msg) |
1010 | { | |
1011 | ceph_osdc_put_request(req); | |
1012 | } | |
1013 | ||
602adf40 YS |
1014 | /* |
1015 | * Do a synchronous ceph osd operation | |
1016 | */ | |
1017 | static int rbd_req_sync_op(struct rbd_device *dev, | |
1018 | struct ceph_snap_context *snapc, | |
1019 | u64 snapid, | |
1020 | int opcode, | |
1021 | int flags, | |
1022 | struct ceph_osd_req_op *orig_ops, | |
1023 | int num_reply, | |
1024 | const char *obj, | |
1025 | u64 ofs, u64 len, | |
59c2be1e YS |
1026 | char *buf, |
1027 | struct ceph_osd_request **linger_req, | |
1028 | u64 *ver) | |
602adf40 YS |
1029 | { |
1030 | int ret; | |
1031 | struct page **pages; | |
1032 | int num_pages; | |
1033 | struct ceph_osd_req_op *ops = orig_ops; | |
1034 | u32 payload_len; | |
1035 | ||
1036 | num_pages = calc_pages_for(ofs , len); | |
1037 | pages = ceph_alloc_page_vector(num_pages, GFP_KERNEL); | |
b8d0638a DC |
1038 | if (IS_ERR(pages)) |
1039 | return PTR_ERR(pages); | |
602adf40 YS |
1040 | |
1041 | if (!orig_ops) { | |
1042 | payload_len = (flags & CEPH_OSD_FLAG_WRITE ? len : 0); | |
1043 | ret = rbd_create_rw_ops(&ops, 1, opcode, payload_len); | |
1044 | if (ret < 0) | |
1045 | goto done; | |
1046 | ||
1047 | if ((flags & CEPH_OSD_FLAG_WRITE) && buf) { | |
1048 | ret = ceph_copy_to_page_vector(pages, buf, ofs, len); | |
1049 | if (ret < 0) | |
1050 | goto done_ops; | |
1051 | } | |
1052 | } | |
1053 | ||
1054 | ret = rbd_do_request(NULL, dev, snapc, snapid, | |
1055 | obj, ofs, len, NULL, | |
1056 | pages, num_pages, | |
1057 | flags, | |
1058 | ops, | |
1059 | 2, | |
1fec7093 | 1060 | NULL, 0, |
59c2be1e YS |
1061 | NULL, |
1062 | linger_req, ver); | |
602adf40 YS |
1063 | if (ret < 0) |
1064 | goto done_ops; | |
1065 | ||
1066 | if ((flags & CEPH_OSD_FLAG_READ) && buf) | |
1067 | ret = ceph_copy_from_page_vector(pages, buf, ofs, ret); | |
1068 | ||
1069 | done_ops: | |
1070 | if (!orig_ops) | |
1071 | rbd_destroy_ops(ops); | |
1072 | done: | |
1073 | ceph_release_page_vector(pages, num_pages); | |
1074 | return ret; | |
1075 | } | |
1076 | ||
1077 | /* | |
1078 | * Do an asynchronous ceph osd operation | |
1079 | */ | |
1080 | static int rbd_do_op(struct request *rq, | |
1081 | struct rbd_device *rbd_dev , | |
1082 | struct ceph_snap_context *snapc, | |
1083 | u64 snapid, | |
1084 | int opcode, int flags, int num_reply, | |
1085 | u64 ofs, u64 len, | |
1fec7093 YS |
1086 | struct bio *bio, |
1087 | struct rbd_req_coll *coll, | |
1088 | int coll_index) | |
602adf40 YS |
1089 | { |
1090 | char *seg_name; | |
1091 | u64 seg_ofs; | |
1092 | u64 seg_len; | |
1093 | int ret; | |
1094 | struct ceph_osd_req_op *ops; | |
1095 | u32 payload_len; | |
1096 | ||
1097 | seg_name = kmalloc(RBD_MAX_SEG_NAME_LEN + 1, GFP_NOIO); | |
1098 | if (!seg_name) | |
1099 | return -ENOMEM; | |
1100 | ||
1101 | seg_len = rbd_get_segment(&rbd_dev->header, | |
ca1e49a6 | 1102 | rbd_dev->header.object_prefix, |
602adf40 YS |
1103 | ofs, len, |
1104 | seg_name, &seg_ofs); | |
602adf40 YS |
1105 | |
1106 | payload_len = (flags & CEPH_OSD_FLAG_WRITE ? seg_len : 0); | |
1107 | ||
1108 | ret = rbd_create_rw_ops(&ops, 1, opcode, payload_len); | |
1109 | if (ret < 0) | |
1110 | goto done; | |
1111 | ||
1112 | /* we've taken care of segment sizes earlier when we | |
1113 | cloned the bios. We should never have a segment | |
1114 | truncated at this point */ | |
1115 | BUG_ON(seg_len < len); | |
1116 | ||
1117 | ret = rbd_do_request(rq, rbd_dev, snapc, snapid, | |
1118 | seg_name, seg_ofs, seg_len, | |
1119 | bio, | |
1120 | NULL, 0, | |
1121 | flags, | |
1122 | ops, | |
1123 | num_reply, | |
1fec7093 | 1124 | coll, coll_index, |
59c2be1e | 1125 | rbd_req_cb, 0, NULL); |
11f77002 SW |
1126 | |
1127 | rbd_destroy_ops(ops); | |
602adf40 YS |
1128 | done: |
1129 | kfree(seg_name); | |
1130 | return ret; | |
1131 | } | |
1132 | ||
1133 | /* | |
1134 | * Request async osd write | |
1135 | */ | |
1136 | static int rbd_req_write(struct request *rq, | |
1137 | struct rbd_device *rbd_dev, | |
1138 | struct ceph_snap_context *snapc, | |
1139 | u64 ofs, u64 len, | |
1fec7093 YS |
1140 | struct bio *bio, |
1141 | struct rbd_req_coll *coll, | |
1142 | int coll_index) | |
602adf40 YS |
1143 | { |
1144 | return rbd_do_op(rq, rbd_dev, snapc, CEPH_NOSNAP, | |
1145 | CEPH_OSD_OP_WRITE, | |
1146 | CEPH_OSD_FLAG_WRITE | CEPH_OSD_FLAG_ONDISK, | |
1147 | 2, | |
1fec7093 | 1148 | ofs, len, bio, coll, coll_index); |
602adf40 YS |
1149 | } |
1150 | ||
1151 | /* | |
1152 | * Request async osd read | |
1153 | */ | |
1154 | static int rbd_req_read(struct request *rq, | |
1155 | struct rbd_device *rbd_dev, | |
1156 | u64 snapid, | |
1157 | u64 ofs, u64 len, | |
1fec7093 YS |
1158 | struct bio *bio, |
1159 | struct rbd_req_coll *coll, | |
1160 | int coll_index) | |
602adf40 YS |
1161 | { |
1162 | return rbd_do_op(rq, rbd_dev, NULL, | |
b06e6a6b | 1163 | snapid, |
602adf40 YS |
1164 | CEPH_OSD_OP_READ, |
1165 | CEPH_OSD_FLAG_READ, | |
1166 | 2, | |
1fec7093 | 1167 | ofs, len, bio, coll, coll_index); |
602adf40 YS |
1168 | } |
1169 | ||
1170 | /* | |
1171 | * Request sync osd read | |
1172 | */ | |
1173 | static int rbd_req_sync_read(struct rbd_device *dev, | |
1174 | struct ceph_snap_context *snapc, | |
1175 | u64 snapid, | |
1176 | const char *obj, | |
1177 | u64 ofs, u64 len, | |
59c2be1e YS |
1178 | char *buf, |
1179 | u64 *ver) | |
602adf40 YS |
1180 | { |
1181 | return rbd_req_sync_op(dev, NULL, | |
b06e6a6b | 1182 | snapid, |
602adf40 YS |
1183 | CEPH_OSD_OP_READ, |
1184 | CEPH_OSD_FLAG_READ, | |
1185 | NULL, | |
59c2be1e | 1186 | 1, obj, ofs, len, buf, NULL, ver); |
602adf40 YS |
1187 | } |
1188 | ||
1189 | /* | |
59c2be1e YS |
1190 | * Request sync osd watch |
1191 | */ | |
1192 | static int rbd_req_sync_notify_ack(struct rbd_device *dev, | |
1193 | u64 ver, | |
1194 | u64 notify_id, | |
1195 | const char *obj) | |
1196 | { | |
1197 | struct ceph_osd_req_op *ops; | |
11f77002 SW |
1198 | int ret; |
1199 | ||
1200 | ret = rbd_create_rw_ops(&ops, 1, CEPH_OSD_OP_NOTIFY_ACK, 0); | |
59c2be1e YS |
1201 | if (ret < 0) |
1202 | return ret; | |
1203 | ||
1204 | ops[0].watch.ver = cpu_to_le64(dev->header.obj_version); | |
1205 | ops[0].watch.cookie = notify_id; | |
1206 | ops[0].watch.flag = 0; | |
1207 | ||
1208 | ret = rbd_do_request(NULL, dev, NULL, CEPH_NOSNAP, | |
1209 | obj, 0, 0, NULL, | |
ad4f232f | 1210 | NULL, 0, |
59c2be1e YS |
1211 | CEPH_OSD_FLAG_READ, |
1212 | ops, | |
1213 | 1, | |
1fec7093 | 1214 | NULL, 0, |
59c2be1e YS |
1215 | rbd_simple_req_cb, 0, NULL); |
1216 | ||
1217 | rbd_destroy_ops(ops); | |
1218 | return ret; | |
1219 | } | |
1220 | ||
1221 | static void rbd_watch_cb(u64 ver, u64 notify_id, u8 opcode, void *data) | |
1222 | { | |
1223 | struct rbd_device *dev = (struct rbd_device *)data; | |
13143d2d SW |
1224 | int rc; |
1225 | ||
59c2be1e YS |
1226 | if (!dev) |
1227 | return; | |
1228 | ||
1229 | dout("rbd_watch_cb %s notify_id=%lld opcode=%d\n", dev->obj_md_name, | |
1230 | notify_id, (int)opcode); | |
1231 | mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING); | |
263c6ca0 | 1232 | rc = __rbd_refresh_header(dev); |
59c2be1e | 1233 | mutex_unlock(&ctl_mutex); |
13143d2d | 1234 | if (rc) |
f0f8cef5 AE |
1235 | pr_warning(RBD_DRV_NAME "%d got notification but failed to " |
1236 | " update snaps: %d\n", dev->major, rc); | |
59c2be1e YS |
1237 | |
1238 | rbd_req_sync_notify_ack(dev, ver, notify_id, dev->obj_md_name); | |
1239 | } | |
1240 | ||
1241 | /* | |
1242 | * Request sync osd watch | |
1243 | */ | |
1244 | static int rbd_req_sync_watch(struct rbd_device *dev, | |
1245 | const char *obj, | |
1246 | u64 ver) | |
1247 | { | |
1248 | struct ceph_osd_req_op *ops; | |
1dbb4399 | 1249 | struct ceph_osd_client *osdc = &dev->rbd_client->client->osdc; |
59c2be1e YS |
1250 | |
1251 | int ret = rbd_create_rw_ops(&ops, 1, CEPH_OSD_OP_WATCH, 0); | |
1252 | if (ret < 0) | |
1253 | return ret; | |
1254 | ||
1255 | ret = ceph_osdc_create_event(osdc, rbd_watch_cb, 0, | |
1256 | (void *)dev, &dev->watch_event); | |
1257 | if (ret < 0) | |
1258 | goto fail; | |
1259 | ||
1260 | ops[0].watch.ver = cpu_to_le64(ver); | |
1261 | ops[0].watch.cookie = cpu_to_le64(dev->watch_event->cookie); | |
1262 | ops[0].watch.flag = 1; | |
1263 | ||
1264 | ret = rbd_req_sync_op(dev, NULL, | |
1265 | CEPH_NOSNAP, | |
1266 | 0, | |
1267 | CEPH_OSD_FLAG_WRITE | CEPH_OSD_FLAG_ONDISK, | |
1268 | ops, | |
1269 | 1, obj, 0, 0, NULL, | |
1270 | &dev->watch_request, NULL); | |
1271 | ||
1272 | if (ret < 0) | |
1273 | goto fail_event; | |
1274 | ||
1275 | rbd_destroy_ops(ops); | |
1276 | return 0; | |
1277 | ||
1278 | fail_event: | |
1279 | ceph_osdc_cancel_event(dev->watch_event); | |
1280 | dev->watch_event = NULL; | |
1281 | fail: | |
1282 | rbd_destroy_ops(ops); | |
1283 | return ret; | |
1284 | } | |
1285 | ||
79e3057c YS |
1286 | /* |
1287 | * Request sync osd unwatch | |
1288 | */ | |
1289 | static int rbd_req_sync_unwatch(struct rbd_device *dev, | |
1290 | const char *obj) | |
1291 | { | |
1292 | struct ceph_osd_req_op *ops; | |
1293 | ||
1294 | int ret = rbd_create_rw_ops(&ops, 1, CEPH_OSD_OP_WATCH, 0); | |
1295 | if (ret < 0) | |
1296 | return ret; | |
1297 | ||
1298 | ops[0].watch.ver = 0; | |
1299 | ops[0].watch.cookie = cpu_to_le64(dev->watch_event->cookie); | |
1300 | ops[0].watch.flag = 0; | |
1301 | ||
1302 | ret = rbd_req_sync_op(dev, NULL, | |
1303 | CEPH_NOSNAP, | |
1304 | 0, | |
1305 | CEPH_OSD_FLAG_WRITE | CEPH_OSD_FLAG_ONDISK, | |
1306 | ops, | |
1307 | 1, obj, 0, 0, NULL, NULL, NULL); | |
1308 | ||
1309 | rbd_destroy_ops(ops); | |
1310 | ceph_osdc_cancel_event(dev->watch_event); | |
1311 | dev->watch_event = NULL; | |
1312 | return ret; | |
1313 | } | |
1314 | ||
59c2be1e YS |
1315 | struct rbd_notify_info { |
1316 | struct rbd_device *dev; | |
1317 | }; | |
1318 | ||
1319 | static void rbd_notify_cb(u64 ver, u64 notify_id, u8 opcode, void *data) | |
1320 | { | |
1321 | struct rbd_device *dev = (struct rbd_device *)data; | |
1322 | if (!dev) | |
1323 | return; | |
1324 | ||
1325 | dout("rbd_notify_cb %s notify_id=%lld opcode=%d\n", dev->obj_md_name, | |
1326 | notify_id, (int)opcode); | |
1327 | } | |
1328 | ||
1329 | /* | |
1330 | * Request sync osd notify | |
1331 | */ | |
1332 | static int rbd_req_sync_notify(struct rbd_device *dev, | |
1333 | const char *obj) | |
1334 | { | |
1335 | struct ceph_osd_req_op *ops; | |
1dbb4399 | 1336 | struct ceph_osd_client *osdc = &dev->rbd_client->client->osdc; |
59c2be1e YS |
1337 | struct ceph_osd_event *event; |
1338 | struct rbd_notify_info info; | |
1339 | int payload_len = sizeof(u32) + sizeof(u32); | |
1340 | int ret; | |
1341 | ||
1342 | ret = rbd_create_rw_ops(&ops, 1, CEPH_OSD_OP_NOTIFY, payload_len); | |
1343 | if (ret < 0) | |
1344 | return ret; | |
1345 | ||
1346 | info.dev = dev; | |
1347 | ||
1348 | ret = ceph_osdc_create_event(osdc, rbd_notify_cb, 1, | |
1349 | (void *)&info, &event); | |
1350 | if (ret < 0) | |
1351 | goto fail; | |
1352 | ||
1353 | ops[0].watch.ver = 1; | |
1354 | ops[0].watch.flag = 1; | |
1355 | ops[0].watch.cookie = event->cookie; | |
1356 | ops[0].watch.prot_ver = RADOS_NOTIFY_VER; | |
1357 | ops[0].watch.timeout = 12; | |
1358 | ||
1359 | ret = rbd_req_sync_op(dev, NULL, | |
1360 | CEPH_NOSNAP, | |
1361 | 0, | |
1362 | CEPH_OSD_FLAG_WRITE | CEPH_OSD_FLAG_ONDISK, | |
1363 | ops, | |
1364 | 1, obj, 0, 0, NULL, NULL, NULL); | |
1365 | if (ret < 0) | |
1366 | goto fail_event; | |
1367 | ||
1368 | ret = ceph_osdc_wait_event(event, CEPH_OSD_TIMEOUT_DEFAULT); | |
1369 | dout("ceph_osdc_wait_event returned %d\n", ret); | |
1370 | rbd_destroy_ops(ops); | |
1371 | return 0; | |
1372 | ||
1373 | fail_event: | |
1374 | ceph_osdc_cancel_event(event); | |
1375 | fail: | |
1376 | rbd_destroy_ops(ops); | |
1377 | return ret; | |
1378 | } | |
1379 | ||
602adf40 YS |
1380 | /* |
1381 | * Request sync osd read | |
1382 | */ | |
1383 | static int rbd_req_sync_exec(struct rbd_device *dev, | |
1384 | const char *obj, | |
1385 | const char *cls, | |
1386 | const char *method, | |
1387 | const char *data, | |
59c2be1e YS |
1388 | int len, |
1389 | u64 *ver) | |
602adf40 YS |
1390 | { |
1391 | struct ceph_osd_req_op *ops; | |
1392 | int cls_len = strlen(cls); | |
1393 | int method_len = strlen(method); | |
1394 | int ret = rbd_create_rw_ops(&ops, 1, CEPH_OSD_OP_CALL, | |
1395 | cls_len + method_len + len); | |
1396 | if (ret < 0) | |
1397 | return ret; | |
1398 | ||
1399 | ops[0].cls.class_name = cls; | |
1400 | ops[0].cls.class_len = (__u8)cls_len; | |
1401 | ops[0].cls.method_name = method; | |
1402 | ops[0].cls.method_len = (__u8)method_len; | |
1403 | ops[0].cls.argc = 0; | |
1404 | ops[0].cls.indata = data; | |
1405 | ops[0].cls.indata_len = len; | |
1406 | ||
1407 | ret = rbd_req_sync_op(dev, NULL, | |
1408 | CEPH_NOSNAP, | |
1409 | 0, | |
1410 | CEPH_OSD_FLAG_WRITE | CEPH_OSD_FLAG_ONDISK, | |
1411 | ops, | |
59c2be1e | 1412 | 1, obj, 0, 0, NULL, NULL, ver); |
602adf40 YS |
1413 | |
1414 | rbd_destroy_ops(ops); | |
1415 | ||
1416 | dout("cls_exec returned %d\n", ret); | |
1417 | return ret; | |
1418 | } | |
1419 | ||
1fec7093 YS |
1420 | static struct rbd_req_coll *rbd_alloc_coll(int num_reqs) |
1421 | { | |
1422 | struct rbd_req_coll *coll = | |
1423 | kzalloc(sizeof(struct rbd_req_coll) + | |
1424 | sizeof(struct rbd_req_status) * num_reqs, | |
1425 | GFP_ATOMIC); | |
1426 | ||
1427 | if (!coll) | |
1428 | return NULL; | |
1429 | coll->total = num_reqs; | |
1430 | kref_init(&coll->kref); | |
1431 | return coll; | |
1432 | } | |
1433 | ||
602adf40 YS |
1434 | /* |
1435 | * block device queue callback | |
1436 | */ | |
1437 | static void rbd_rq_fn(struct request_queue *q) | |
1438 | { | |
1439 | struct rbd_device *rbd_dev = q->queuedata; | |
1440 | struct request *rq; | |
1441 | struct bio_pair *bp = NULL; | |
1442 | ||
00f1f36f | 1443 | while ((rq = blk_fetch_request(q))) { |
602adf40 YS |
1444 | struct bio *bio; |
1445 | struct bio *rq_bio, *next_bio = NULL; | |
1446 | bool do_write; | |
1447 | int size, op_size = 0; | |
1448 | u64 ofs; | |
1fec7093 YS |
1449 | int num_segs, cur_seg = 0; |
1450 | struct rbd_req_coll *coll; | |
602adf40 YS |
1451 | |
1452 | /* peek at request from block layer */ | |
1453 | if (!rq) | |
1454 | break; | |
1455 | ||
1456 | dout("fetched request\n"); | |
1457 | ||
1458 | /* filter out block requests we don't understand */ | |
1459 | if ((rq->cmd_type != REQ_TYPE_FS)) { | |
1460 | __blk_end_request_all(rq, 0); | |
00f1f36f | 1461 | continue; |
602adf40 YS |
1462 | } |
1463 | ||
1464 | /* deduce our operation (read, write) */ | |
1465 | do_write = (rq_data_dir(rq) == WRITE); | |
1466 | ||
1467 | size = blk_rq_bytes(rq); | |
593a9e7b | 1468 | ofs = blk_rq_pos(rq) * SECTOR_SIZE; |
602adf40 YS |
1469 | rq_bio = rq->bio; |
1470 | if (do_write && rbd_dev->read_only) { | |
1471 | __blk_end_request_all(rq, -EROFS); | |
00f1f36f | 1472 | continue; |
602adf40 YS |
1473 | } |
1474 | ||
1475 | spin_unlock_irq(q->queue_lock); | |
1476 | ||
1477 | dout("%s 0x%x bytes at 0x%llx\n", | |
1478 | do_write ? "write" : "read", | |
593a9e7b | 1479 | size, blk_rq_pos(rq) * SECTOR_SIZE); |
602adf40 | 1480 | |
1fec7093 YS |
1481 | num_segs = rbd_get_num_segments(&rbd_dev->header, ofs, size); |
1482 | coll = rbd_alloc_coll(num_segs); | |
1483 | if (!coll) { | |
1484 | spin_lock_irq(q->queue_lock); | |
1485 | __blk_end_request_all(rq, -ENOMEM); | |
00f1f36f | 1486 | continue; |
1fec7093 YS |
1487 | } |
1488 | ||
602adf40 YS |
1489 | do { |
1490 | /* a bio clone to be passed down to OSD req */ | |
1491 | dout("rq->bio->bi_vcnt=%d\n", rq->bio->bi_vcnt); | |
1492 | op_size = rbd_get_segment(&rbd_dev->header, | |
ca1e49a6 | 1493 | rbd_dev->header.object_prefix, |
602adf40 YS |
1494 | ofs, size, |
1495 | NULL, NULL); | |
1fec7093 | 1496 | kref_get(&coll->kref); |
602adf40 YS |
1497 | bio = bio_chain_clone(&rq_bio, &next_bio, &bp, |
1498 | op_size, GFP_ATOMIC); | |
1499 | if (!bio) { | |
1fec7093 YS |
1500 | rbd_coll_end_req_index(rq, coll, cur_seg, |
1501 | -ENOMEM, op_size); | |
1502 | goto next_seg; | |
602adf40 YS |
1503 | } |
1504 | ||
1fec7093 | 1505 | |
602adf40 YS |
1506 | /* init OSD command: write or read */ |
1507 | if (do_write) | |
1508 | rbd_req_write(rq, rbd_dev, | |
1509 | rbd_dev->header.snapc, | |
1510 | ofs, | |
1fec7093 YS |
1511 | op_size, bio, |
1512 | coll, cur_seg); | |
602adf40 YS |
1513 | else |
1514 | rbd_req_read(rq, rbd_dev, | |
77dfe99f | 1515 | rbd_dev->snap_id, |
602adf40 | 1516 | ofs, |
1fec7093 YS |
1517 | op_size, bio, |
1518 | coll, cur_seg); | |
602adf40 | 1519 | |
1fec7093 | 1520 | next_seg: |
602adf40 YS |
1521 | size -= op_size; |
1522 | ofs += op_size; | |
1523 | ||
1fec7093 | 1524 | cur_seg++; |
602adf40 YS |
1525 | rq_bio = next_bio; |
1526 | } while (size > 0); | |
1fec7093 | 1527 | kref_put(&coll->kref, rbd_coll_release); |
602adf40 YS |
1528 | |
1529 | if (bp) | |
1530 | bio_pair_release(bp); | |
602adf40 | 1531 | spin_lock_irq(q->queue_lock); |
602adf40 YS |
1532 | } |
1533 | } | |
1534 | ||
1535 | /* | |
1536 | * a queue callback. Makes sure that we don't create a bio that spans across | |
1537 | * multiple osd objects. One exception would be with a single page bios, | |
1538 | * which we handle later at bio_chain_clone | |
1539 | */ | |
1540 | static int rbd_merge_bvec(struct request_queue *q, struct bvec_merge_data *bmd, | |
1541 | struct bio_vec *bvec) | |
1542 | { | |
1543 | struct rbd_device *rbd_dev = q->queuedata; | |
593a9e7b AE |
1544 | unsigned int chunk_sectors; |
1545 | sector_t sector; | |
1546 | unsigned int bio_sectors; | |
602adf40 YS |
1547 | int max; |
1548 | ||
593a9e7b AE |
1549 | chunk_sectors = 1 << (rbd_dev->header.obj_order - SECTOR_SHIFT); |
1550 | sector = bmd->bi_sector + get_start_sect(bmd->bi_bdev); | |
1551 | bio_sectors = bmd->bi_size >> SECTOR_SHIFT; | |
1552 | ||
602adf40 | 1553 | max = (chunk_sectors - ((sector & (chunk_sectors - 1)) |
593a9e7b | 1554 | + bio_sectors)) << SECTOR_SHIFT; |
602adf40 YS |
1555 | if (max < 0) |
1556 | max = 0; /* bio_add cannot handle a negative return */ | |
1557 | if (max <= bvec->bv_len && bio_sectors == 0) | |
1558 | return bvec->bv_len; | |
1559 | return max; | |
1560 | } | |
1561 | ||
1562 | static void rbd_free_disk(struct rbd_device *rbd_dev) | |
1563 | { | |
1564 | struct gendisk *disk = rbd_dev->disk; | |
1565 | ||
1566 | if (!disk) | |
1567 | return; | |
1568 | ||
1569 | rbd_header_free(&rbd_dev->header); | |
1570 | ||
1571 | if (disk->flags & GENHD_FL_UP) | |
1572 | del_gendisk(disk); | |
1573 | if (disk->queue) | |
1574 | blk_cleanup_queue(disk->queue); | |
1575 | put_disk(disk); | |
1576 | } | |
1577 | ||
1578 | /* | |
1579 | * reload the ondisk the header | |
1580 | */ | |
1581 | static int rbd_read_header(struct rbd_device *rbd_dev, | |
1582 | struct rbd_image_header *header) | |
1583 | { | |
1584 | ssize_t rc; | |
1585 | struct rbd_image_header_ondisk *dh; | |
50f7c4c9 | 1586 | u32 snap_count = 0; |
59c2be1e | 1587 | u64 ver; |
00f1f36f | 1588 | size_t len; |
602adf40 | 1589 | |
00f1f36f AE |
1590 | /* |
1591 | * First reads the fixed-size header to determine the number | |
1592 | * of snapshots, then re-reads it, along with all snapshot | |
1593 | * records as well as their stored names. | |
1594 | */ | |
1595 | len = sizeof (*dh); | |
602adf40 | 1596 | while (1) { |
602adf40 YS |
1597 | dh = kmalloc(len, GFP_KERNEL); |
1598 | if (!dh) | |
1599 | return -ENOMEM; | |
1600 | ||
1601 | rc = rbd_req_sync_read(rbd_dev, | |
1602 | NULL, CEPH_NOSNAP, | |
1603 | rbd_dev->obj_md_name, | |
1604 | 0, len, | |
59c2be1e | 1605 | (char *)dh, &ver); |
602adf40 YS |
1606 | if (rc < 0) |
1607 | goto out_dh; | |
1608 | ||
1609 | rc = rbd_header_from_disk(header, dh, snap_count, GFP_KERNEL); | |
81e759fb | 1610 | if (rc < 0) { |
00f1f36f | 1611 | if (rc == -ENXIO) |
81e759fb JD |
1612 | pr_warning("unrecognized header format" |
1613 | " for image %s", rbd_dev->obj); | |
602adf40 | 1614 | goto out_dh; |
81e759fb | 1615 | } |
602adf40 | 1616 | |
00f1f36f AE |
1617 | if (snap_count == header->total_snaps) |
1618 | break; | |
1619 | ||
1620 | snap_count = header->total_snaps; | |
1621 | len = sizeof (*dh) + | |
1622 | snap_count * sizeof(struct rbd_image_snap_ondisk) + | |
1623 | header->snap_names_len; | |
1624 | ||
1625 | rbd_header_free(header); | |
1626 | kfree(dh); | |
602adf40 | 1627 | } |
59c2be1e | 1628 | header->obj_version = ver; |
602adf40 YS |
1629 | |
1630 | out_dh: | |
1631 | kfree(dh); | |
1632 | return rc; | |
1633 | } | |
1634 | ||
1635 | /* | |
1636 | * create a snapshot | |
1637 | */ | |
1638 | static int rbd_header_add_snap(struct rbd_device *dev, | |
1639 | const char *snap_name, | |
1640 | gfp_t gfp_flags) | |
1641 | { | |
1642 | int name_len = strlen(snap_name); | |
1643 | u64 new_snapid; | |
1644 | int ret; | |
916d4d67 | 1645 | void *data, *p, *e; |
59c2be1e | 1646 | u64 ver; |
1dbb4399 | 1647 | struct ceph_mon_client *monc; |
602adf40 YS |
1648 | |
1649 | /* we should create a snapshot only if we're pointing at the head */ | |
77dfe99f | 1650 | if (dev->snap_id != CEPH_NOSNAP) |
602adf40 YS |
1651 | return -EINVAL; |
1652 | ||
1dbb4399 | 1653 | monc = &dev->rbd_client->client->monc; |
9bb2f334 | 1654 | ret = ceph_monc_create_snapid(monc, dev->pool_id, &new_snapid); |
602adf40 YS |
1655 | dout("created snapid=%lld\n", new_snapid); |
1656 | if (ret < 0) | |
1657 | return ret; | |
1658 | ||
1659 | data = kmalloc(name_len + 16, gfp_flags); | |
1660 | if (!data) | |
1661 | return -ENOMEM; | |
1662 | ||
916d4d67 SW |
1663 | p = data; |
1664 | e = data + name_len + 16; | |
602adf40 | 1665 | |
916d4d67 SW |
1666 | ceph_encode_string_safe(&p, e, snap_name, name_len, bad); |
1667 | ceph_encode_64_safe(&p, e, new_snapid, bad); | |
602adf40 YS |
1668 | |
1669 | ret = rbd_req_sync_exec(dev, dev->obj_md_name, "rbd", "snap_add", | |
916d4d67 | 1670 | data, p - data, &ver); |
602adf40 | 1671 | |
916d4d67 | 1672 | kfree(data); |
602adf40 YS |
1673 | |
1674 | if (ret < 0) | |
1675 | return ret; | |
1676 | ||
403f24d3 JD |
1677 | down_write(&dev->header_rwsem); |
1678 | dev->header.snapc->seq = new_snapid; | |
1679 | up_write(&dev->header_rwsem); | |
602adf40 YS |
1680 | |
1681 | return 0; | |
1682 | bad: | |
1683 | return -ERANGE; | |
1684 | } | |
1685 | ||
dfc5606d YS |
1686 | static void __rbd_remove_all_snaps(struct rbd_device *rbd_dev) |
1687 | { | |
1688 | struct rbd_snap *snap; | |
1689 | ||
1690 | while (!list_empty(&rbd_dev->snaps)) { | |
1691 | snap = list_first_entry(&rbd_dev->snaps, struct rbd_snap, node); | |
1692 | __rbd_remove_snap_dev(rbd_dev, snap); | |
1693 | } | |
1694 | } | |
1695 | ||
602adf40 YS |
1696 | /* |
1697 | * only read the first part of the ondisk header, without the snaps info | |
1698 | */ | |
263c6ca0 | 1699 | static int __rbd_refresh_header(struct rbd_device *rbd_dev) |
602adf40 YS |
1700 | { |
1701 | int ret; | |
1702 | struct rbd_image_header h; | |
1703 | u64 snap_seq; | |
59c2be1e | 1704 | int follow_seq = 0; |
602adf40 YS |
1705 | |
1706 | ret = rbd_read_header(rbd_dev, &h); | |
1707 | if (ret < 0) | |
1708 | return ret; | |
1709 | ||
9db4b3e3 | 1710 | /* resized? */ |
593a9e7b | 1711 | set_capacity(rbd_dev->disk, h.image_size / SECTOR_SIZE); |
9db4b3e3 | 1712 | |
c666601a | 1713 | down_write(&rbd_dev->header_rwsem); |
602adf40 YS |
1714 | |
1715 | snap_seq = rbd_dev->header.snapc->seq; | |
59c2be1e YS |
1716 | if (rbd_dev->header.total_snaps && |
1717 | rbd_dev->header.snapc->snaps[0] == snap_seq) | |
1718 | /* pointing at the head, will need to follow that | |
1719 | if head moves */ | |
1720 | follow_seq = 1; | |
602adf40 | 1721 | |
849b4260 | 1722 | /* rbd_dev->header.object_prefix shouldn't change */ |
602adf40 | 1723 | kfree(rbd_dev->header.snap_sizes); |
849b4260 AE |
1724 | kfree(rbd_dev->header.snap_names); |
1725 | kfree(rbd_dev->header.snapc); | |
602adf40 YS |
1726 | |
1727 | rbd_dev->header.total_snaps = h.total_snaps; | |
1728 | rbd_dev->header.snapc = h.snapc; | |
1729 | rbd_dev->header.snap_names = h.snap_names; | |
dfc5606d | 1730 | rbd_dev->header.snap_names_len = h.snap_names_len; |
602adf40 | 1731 | rbd_dev->header.snap_sizes = h.snap_sizes; |
849b4260 AE |
1732 | /* Free the extra copy of the object prefix */ |
1733 | WARN_ON(strcmp(rbd_dev->header.object_prefix, h.object_prefix)); | |
1734 | kfree(h.object_prefix); | |
1735 | ||
59c2be1e YS |
1736 | if (follow_seq) |
1737 | rbd_dev->header.snapc->seq = rbd_dev->header.snapc->snaps[0]; | |
1738 | else | |
1739 | rbd_dev->header.snapc->seq = snap_seq; | |
602adf40 | 1740 | |
dfc5606d YS |
1741 | ret = __rbd_init_snaps_header(rbd_dev); |
1742 | ||
c666601a | 1743 | up_write(&rbd_dev->header_rwsem); |
602adf40 | 1744 | |
dfc5606d | 1745 | return ret; |
602adf40 YS |
1746 | } |
1747 | ||
1748 | static int rbd_init_disk(struct rbd_device *rbd_dev) | |
1749 | { | |
1750 | struct gendisk *disk; | |
1751 | struct request_queue *q; | |
1752 | int rc; | |
593a9e7b | 1753 | u64 segment_size; |
602adf40 YS |
1754 | u64 total_size = 0; |
1755 | ||
1756 | /* contact OSD, request size info about the object being mapped */ | |
1757 | rc = rbd_read_header(rbd_dev, &rbd_dev->header); | |
1758 | if (rc) | |
1759 | return rc; | |
1760 | ||
dfc5606d YS |
1761 | /* no need to lock here, as rbd_dev is not registered yet */ |
1762 | rc = __rbd_init_snaps_header(rbd_dev); | |
1763 | if (rc) | |
1764 | return rc; | |
1765 | ||
cc9d734c | 1766 | rc = rbd_header_set_snap(rbd_dev, &total_size); |
602adf40 YS |
1767 | if (rc) |
1768 | return rc; | |
1769 | ||
1770 | /* create gendisk info */ | |
1771 | rc = -ENOMEM; | |
1772 | disk = alloc_disk(RBD_MINORS_PER_MAJOR); | |
1773 | if (!disk) | |
1774 | goto out; | |
1775 | ||
f0f8cef5 | 1776 | snprintf(disk->disk_name, sizeof(disk->disk_name), RBD_DRV_NAME "%d", |
aedfec59 | 1777 | rbd_dev->id); |
602adf40 YS |
1778 | disk->major = rbd_dev->major; |
1779 | disk->first_minor = 0; | |
1780 | disk->fops = &rbd_bd_ops; | |
1781 | disk->private_data = rbd_dev; | |
1782 | ||
1783 | /* init rq */ | |
1784 | rc = -ENOMEM; | |
1785 | q = blk_init_queue(rbd_rq_fn, &rbd_dev->lock); | |
1786 | if (!q) | |
1787 | goto out_disk; | |
029bcbd8 | 1788 | |
593a9e7b AE |
1789 | /* We use the default size, but let's be explicit about it. */ |
1790 | blk_queue_physical_block_size(q, SECTOR_SIZE); | |
1791 | ||
029bcbd8 | 1792 | /* set io sizes to object size */ |
593a9e7b AE |
1793 | segment_size = rbd_obj_bytes(&rbd_dev->header); |
1794 | blk_queue_max_hw_sectors(q, segment_size / SECTOR_SIZE); | |
1795 | blk_queue_max_segment_size(q, segment_size); | |
1796 | blk_queue_io_min(q, segment_size); | |
1797 | blk_queue_io_opt(q, segment_size); | |
029bcbd8 | 1798 | |
602adf40 YS |
1799 | blk_queue_merge_bvec(q, rbd_merge_bvec); |
1800 | disk->queue = q; | |
1801 | ||
1802 | q->queuedata = rbd_dev; | |
1803 | ||
1804 | rbd_dev->disk = disk; | |
1805 | rbd_dev->q = q; | |
1806 | ||
1807 | /* finally, announce the disk to the world */ | |
593a9e7b | 1808 | set_capacity(disk, total_size / SECTOR_SIZE); |
602adf40 YS |
1809 | add_disk(disk); |
1810 | ||
1811 | pr_info("%s: added with size 0x%llx\n", | |
1812 | disk->disk_name, (unsigned long long)total_size); | |
1813 | return 0; | |
1814 | ||
1815 | out_disk: | |
1816 | put_disk(disk); | |
1817 | out: | |
1818 | return rc; | |
1819 | } | |
1820 | ||
dfc5606d YS |
1821 | /* |
1822 | sysfs | |
1823 | */ | |
1824 | ||
593a9e7b AE |
1825 | static struct rbd_device *dev_to_rbd_dev(struct device *dev) |
1826 | { | |
1827 | return container_of(dev, struct rbd_device, dev); | |
1828 | } | |
1829 | ||
dfc5606d YS |
1830 | static ssize_t rbd_size_show(struct device *dev, |
1831 | struct device_attribute *attr, char *buf) | |
1832 | { | |
593a9e7b | 1833 | struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); |
dfc5606d YS |
1834 | |
1835 | return sprintf(buf, "%llu\n", (unsigned long long)rbd_dev->header.image_size); | |
1836 | } | |
1837 | ||
1838 | static ssize_t rbd_major_show(struct device *dev, | |
1839 | struct device_attribute *attr, char *buf) | |
1840 | { | |
593a9e7b | 1841 | struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); |
602adf40 | 1842 | |
dfc5606d YS |
1843 | return sprintf(buf, "%d\n", rbd_dev->major); |
1844 | } | |
1845 | ||
1846 | static ssize_t rbd_client_id_show(struct device *dev, | |
1847 | struct device_attribute *attr, char *buf) | |
602adf40 | 1848 | { |
593a9e7b | 1849 | struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); |
dfc5606d | 1850 | |
1dbb4399 AE |
1851 | return sprintf(buf, "client%lld\n", |
1852 | ceph_client_id(rbd_dev->rbd_client->client)); | |
602adf40 YS |
1853 | } |
1854 | ||
dfc5606d YS |
1855 | static ssize_t rbd_pool_show(struct device *dev, |
1856 | struct device_attribute *attr, char *buf) | |
602adf40 | 1857 | { |
593a9e7b | 1858 | struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); |
dfc5606d YS |
1859 | |
1860 | return sprintf(buf, "%s\n", rbd_dev->pool_name); | |
1861 | } | |
1862 | ||
9bb2f334 AE |
1863 | static ssize_t rbd_pool_id_show(struct device *dev, |
1864 | struct device_attribute *attr, char *buf) | |
1865 | { | |
1866 | struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); | |
1867 | ||
1868 | return sprintf(buf, "%d\n", rbd_dev->pool_id); | |
1869 | } | |
1870 | ||
dfc5606d YS |
1871 | static ssize_t rbd_name_show(struct device *dev, |
1872 | struct device_attribute *attr, char *buf) | |
1873 | { | |
593a9e7b | 1874 | struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); |
dfc5606d YS |
1875 | |
1876 | return sprintf(buf, "%s\n", rbd_dev->obj); | |
1877 | } | |
1878 | ||
1879 | static ssize_t rbd_snap_show(struct device *dev, | |
1880 | struct device_attribute *attr, | |
1881 | char *buf) | |
1882 | { | |
593a9e7b | 1883 | struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); |
dfc5606d YS |
1884 | |
1885 | return sprintf(buf, "%s\n", rbd_dev->snap_name); | |
1886 | } | |
1887 | ||
1888 | static ssize_t rbd_image_refresh(struct device *dev, | |
1889 | struct device_attribute *attr, | |
1890 | const char *buf, | |
1891 | size_t size) | |
1892 | { | |
593a9e7b | 1893 | struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); |
dfc5606d YS |
1894 | int rc; |
1895 | int ret = size; | |
602adf40 YS |
1896 | |
1897 | mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING); | |
1898 | ||
263c6ca0 | 1899 | rc = __rbd_refresh_header(rbd_dev); |
dfc5606d YS |
1900 | if (rc < 0) |
1901 | ret = rc; | |
602adf40 | 1902 | |
dfc5606d YS |
1903 | mutex_unlock(&ctl_mutex); |
1904 | return ret; | |
1905 | } | |
602adf40 | 1906 | |
dfc5606d YS |
1907 | static DEVICE_ATTR(size, S_IRUGO, rbd_size_show, NULL); |
1908 | static DEVICE_ATTR(major, S_IRUGO, rbd_major_show, NULL); | |
1909 | static DEVICE_ATTR(client_id, S_IRUGO, rbd_client_id_show, NULL); | |
1910 | static DEVICE_ATTR(pool, S_IRUGO, rbd_pool_show, NULL); | |
9bb2f334 | 1911 | static DEVICE_ATTR(pool_id, S_IRUGO, rbd_pool_id_show, NULL); |
dfc5606d YS |
1912 | static DEVICE_ATTR(name, S_IRUGO, rbd_name_show, NULL); |
1913 | static DEVICE_ATTR(refresh, S_IWUSR, NULL, rbd_image_refresh); | |
1914 | static DEVICE_ATTR(current_snap, S_IRUGO, rbd_snap_show, NULL); | |
1915 | static DEVICE_ATTR(create_snap, S_IWUSR, NULL, rbd_snap_add); | |
dfc5606d YS |
1916 | |
1917 | static struct attribute *rbd_attrs[] = { | |
1918 | &dev_attr_size.attr, | |
1919 | &dev_attr_major.attr, | |
1920 | &dev_attr_client_id.attr, | |
1921 | &dev_attr_pool.attr, | |
9bb2f334 | 1922 | &dev_attr_pool_id.attr, |
dfc5606d YS |
1923 | &dev_attr_name.attr, |
1924 | &dev_attr_current_snap.attr, | |
1925 | &dev_attr_refresh.attr, | |
1926 | &dev_attr_create_snap.attr, | |
dfc5606d YS |
1927 | NULL |
1928 | }; | |
1929 | ||
1930 | static struct attribute_group rbd_attr_group = { | |
1931 | .attrs = rbd_attrs, | |
1932 | }; | |
1933 | ||
1934 | static const struct attribute_group *rbd_attr_groups[] = { | |
1935 | &rbd_attr_group, | |
1936 | NULL | |
1937 | }; | |
1938 | ||
1939 | static void rbd_sysfs_dev_release(struct device *dev) | |
1940 | { | |
1941 | } | |
1942 | ||
1943 | static struct device_type rbd_device_type = { | |
1944 | .name = "rbd", | |
1945 | .groups = rbd_attr_groups, | |
1946 | .release = rbd_sysfs_dev_release, | |
1947 | }; | |
1948 | ||
1949 | ||
1950 | /* | |
1951 | sysfs - snapshots | |
1952 | */ | |
1953 | ||
1954 | static ssize_t rbd_snap_size_show(struct device *dev, | |
1955 | struct device_attribute *attr, | |
1956 | char *buf) | |
1957 | { | |
1958 | struct rbd_snap *snap = container_of(dev, struct rbd_snap, dev); | |
1959 | ||
3591538f | 1960 | return sprintf(buf, "%llu\n", (unsigned long long)snap->size); |
dfc5606d YS |
1961 | } |
1962 | ||
1963 | static ssize_t rbd_snap_id_show(struct device *dev, | |
1964 | struct device_attribute *attr, | |
1965 | char *buf) | |
1966 | { | |
1967 | struct rbd_snap *snap = container_of(dev, struct rbd_snap, dev); | |
1968 | ||
3591538f | 1969 | return sprintf(buf, "%llu\n", (unsigned long long)snap->id); |
dfc5606d YS |
1970 | } |
1971 | ||
1972 | static DEVICE_ATTR(snap_size, S_IRUGO, rbd_snap_size_show, NULL); | |
1973 | static DEVICE_ATTR(snap_id, S_IRUGO, rbd_snap_id_show, NULL); | |
1974 | ||
1975 | static struct attribute *rbd_snap_attrs[] = { | |
1976 | &dev_attr_snap_size.attr, | |
1977 | &dev_attr_snap_id.attr, | |
1978 | NULL, | |
1979 | }; | |
1980 | ||
1981 | static struct attribute_group rbd_snap_attr_group = { | |
1982 | .attrs = rbd_snap_attrs, | |
1983 | }; | |
1984 | ||
1985 | static void rbd_snap_dev_release(struct device *dev) | |
1986 | { | |
1987 | struct rbd_snap *snap = container_of(dev, struct rbd_snap, dev); | |
1988 | kfree(snap->name); | |
1989 | kfree(snap); | |
1990 | } | |
1991 | ||
1992 | static const struct attribute_group *rbd_snap_attr_groups[] = { | |
1993 | &rbd_snap_attr_group, | |
1994 | NULL | |
1995 | }; | |
1996 | ||
1997 | static struct device_type rbd_snap_device_type = { | |
1998 | .groups = rbd_snap_attr_groups, | |
1999 | .release = rbd_snap_dev_release, | |
2000 | }; | |
2001 | ||
2002 | static void __rbd_remove_snap_dev(struct rbd_device *rbd_dev, | |
2003 | struct rbd_snap *snap) | |
2004 | { | |
2005 | list_del(&snap->node); | |
2006 | device_unregister(&snap->dev); | |
2007 | } | |
2008 | ||
2009 | static int rbd_register_snap_dev(struct rbd_device *rbd_dev, | |
2010 | struct rbd_snap *snap, | |
2011 | struct device *parent) | |
2012 | { | |
2013 | struct device *dev = &snap->dev; | |
2014 | int ret; | |
2015 | ||
2016 | dev->type = &rbd_snap_device_type; | |
2017 | dev->parent = parent; | |
2018 | dev->release = rbd_snap_dev_release; | |
2019 | dev_set_name(dev, "snap_%s", snap->name); | |
2020 | ret = device_register(dev); | |
2021 | ||
2022 | return ret; | |
2023 | } | |
2024 | ||
2025 | static int __rbd_add_snap_dev(struct rbd_device *rbd_dev, | |
2026 | int i, const char *name, | |
2027 | struct rbd_snap **snapp) | |
2028 | { | |
2029 | int ret; | |
2030 | struct rbd_snap *snap = kzalloc(sizeof(*snap), GFP_KERNEL); | |
2031 | if (!snap) | |
2032 | return -ENOMEM; | |
2033 | snap->name = kstrdup(name, GFP_KERNEL); | |
2034 | snap->size = rbd_dev->header.snap_sizes[i]; | |
2035 | snap->id = rbd_dev->header.snapc->snaps[i]; | |
2036 | if (device_is_registered(&rbd_dev->dev)) { | |
2037 | ret = rbd_register_snap_dev(rbd_dev, snap, | |
2038 | &rbd_dev->dev); | |
2039 | if (ret < 0) | |
2040 | goto err; | |
2041 | } | |
2042 | *snapp = snap; | |
2043 | return 0; | |
2044 | err: | |
2045 | kfree(snap->name); | |
2046 | kfree(snap); | |
2047 | return ret; | |
2048 | } | |
2049 | ||
2050 | /* | |
2051 | * search for the previous snap in a null delimited string list | |
2052 | */ | |
2053 | const char *rbd_prev_snap_name(const char *name, const char *start) | |
2054 | { | |
2055 | if (name < start + 2) | |
2056 | return NULL; | |
2057 | ||
2058 | name -= 2; | |
2059 | while (*name) { | |
2060 | if (name == start) | |
2061 | return start; | |
2062 | name--; | |
2063 | } | |
2064 | return name + 1; | |
2065 | } | |
2066 | ||
2067 | /* | |
2068 | * compare the old list of snapshots that we have to what's in the header | |
2069 | * and update it accordingly. Note that the header holds the snapshots | |
2070 | * in a reverse order (from newest to oldest) and we need to go from | |
2071 | * older to new so that we don't get a duplicate snap name when | |
2072 | * doing the process (e.g., removed snapshot and recreated a new | |
2073 | * one with the same name. | |
2074 | */ | |
2075 | static int __rbd_init_snaps_header(struct rbd_device *rbd_dev) | |
2076 | { | |
2077 | const char *name, *first_name; | |
2078 | int i = rbd_dev->header.total_snaps; | |
2079 | struct rbd_snap *snap, *old_snap = NULL; | |
2080 | int ret; | |
2081 | struct list_head *p, *n; | |
2082 | ||
2083 | first_name = rbd_dev->header.snap_names; | |
2084 | name = first_name + rbd_dev->header.snap_names_len; | |
2085 | ||
2086 | list_for_each_prev_safe(p, n, &rbd_dev->snaps) { | |
2087 | u64 cur_id; | |
2088 | ||
2089 | old_snap = list_entry(p, struct rbd_snap, node); | |
2090 | ||
2091 | if (i) | |
2092 | cur_id = rbd_dev->header.snapc->snaps[i - 1]; | |
2093 | ||
2094 | if (!i || old_snap->id < cur_id) { | |
2095 | /* old_snap->id was skipped, thus was removed */ | |
2096 | __rbd_remove_snap_dev(rbd_dev, old_snap); | |
2097 | continue; | |
2098 | } | |
2099 | if (old_snap->id == cur_id) { | |
2100 | /* we have this snapshot already */ | |
2101 | i--; | |
2102 | name = rbd_prev_snap_name(name, first_name); | |
2103 | continue; | |
2104 | } | |
2105 | for (; i > 0; | |
2106 | i--, name = rbd_prev_snap_name(name, first_name)) { | |
2107 | if (!name) { | |
2108 | WARN_ON(1); | |
2109 | return -EINVAL; | |
2110 | } | |
2111 | cur_id = rbd_dev->header.snapc->snaps[i]; | |
2112 | /* snapshot removal? handle it above */ | |
2113 | if (cur_id >= old_snap->id) | |
2114 | break; | |
2115 | /* a new snapshot */ | |
2116 | ret = __rbd_add_snap_dev(rbd_dev, i - 1, name, &snap); | |
2117 | if (ret < 0) | |
2118 | return ret; | |
2119 | ||
2120 | /* note that we add it backward so using n and not p */ | |
2121 | list_add(&snap->node, n); | |
2122 | p = &snap->node; | |
2123 | } | |
2124 | } | |
2125 | /* we're done going over the old snap list, just add what's left */ | |
2126 | for (; i > 0; i--) { | |
2127 | name = rbd_prev_snap_name(name, first_name); | |
2128 | if (!name) { | |
2129 | WARN_ON(1); | |
2130 | return -EINVAL; | |
2131 | } | |
2132 | ret = __rbd_add_snap_dev(rbd_dev, i - 1, name, &snap); | |
2133 | if (ret < 0) | |
2134 | return ret; | |
2135 | list_add(&snap->node, &rbd_dev->snaps); | |
2136 | } | |
2137 | ||
2138 | return 0; | |
2139 | } | |
2140 | ||
dfc5606d YS |
2141 | static int rbd_bus_add_dev(struct rbd_device *rbd_dev) |
2142 | { | |
f0f8cef5 | 2143 | int ret; |
dfc5606d YS |
2144 | struct device *dev; |
2145 | struct rbd_snap *snap; | |
2146 | ||
2147 | mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING); | |
2148 | dev = &rbd_dev->dev; | |
2149 | ||
2150 | dev->bus = &rbd_bus_type; | |
2151 | dev->type = &rbd_device_type; | |
2152 | dev->parent = &rbd_root_dev; | |
2153 | dev->release = rbd_dev_release; | |
2154 | dev_set_name(dev, "%d", rbd_dev->id); | |
2155 | ret = device_register(dev); | |
2156 | if (ret < 0) | |
f0f8cef5 | 2157 | goto out; |
dfc5606d YS |
2158 | |
2159 | list_for_each_entry(snap, &rbd_dev->snaps, node) { | |
2160 | ret = rbd_register_snap_dev(rbd_dev, snap, | |
2161 | &rbd_dev->dev); | |
2162 | if (ret < 0) | |
602adf40 YS |
2163 | break; |
2164 | } | |
f0f8cef5 | 2165 | out: |
dfc5606d YS |
2166 | mutex_unlock(&ctl_mutex); |
2167 | return ret; | |
602adf40 YS |
2168 | } |
2169 | ||
dfc5606d YS |
2170 | static void rbd_bus_del_dev(struct rbd_device *rbd_dev) |
2171 | { | |
2172 | device_unregister(&rbd_dev->dev); | |
2173 | } | |
2174 | ||
59c2be1e YS |
2175 | static int rbd_init_watch_dev(struct rbd_device *rbd_dev) |
2176 | { | |
2177 | int ret, rc; | |
2178 | ||
2179 | do { | |
2180 | ret = rbd_req_sync_watch(rbd_dev, rbd_dev->obj_md_name, | |
2181 | rbd_dev->header.obj_version); | |
2182 | if (ret == -ERANGE) { | |
2183 | mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING); | |
263c6ca0 | 2184 | rc = __rbd_refresh_header(rbd_dev); |
59c2be1e YS |
2185 | mutex_unlock(&ctl_mutex); |
2186 | if (rc < 0) | |
2187 | return rc; | |
2188 | } | |
2189 | } while (ret == -ERANGE); | |
2190 | ||
2191 | return ret; | |
2192 | } | |
2193 | ||
1ddbe94e AE |
2194 | static atomic64_t rbd_id_max = ATOMIC64_INIT(0); |
2195 | ||
2196 | /* | |
499afd5b AE |
2197 | * Get a unique rbd identifier for the given new rbd_dev, and add |
2198 | * the rbd_dev to the global list. The minimum rbd id is 1. | |
1ddbe94e | 2199 | */ |
499afd5b | 2200 | static void rbd_id_get(struct rbd_device *rbd_dev) |
b7f23c36 | 2201 | { |
499afd5b AE |
2202 | rbd_dev->id = atomic64_inc_return(&rbd_id_max); |
2203 | ||
2204 | spin_lock(&rbd_dev_list_lock); | |
2205 | list_add_tail(&rbd_dev->node, &rbd_dev_list); | |
2206 | spin_unlock(&rbd_dev_list_lock); | |
1ddbe94e | 2207 | } |
b7f23c36 | 2208 | |
1ddbe94e | 2209 | /* |
499afd5b AE |
2210 | * Remove an rbd_dev from the global list, and record that its |
2211 | * identifier is no longer in use. | |
1ddbe94e | 2212 | */ |
499afd5b | 2213 | static void rbd_id_put(struct rbd_device *rbd_dev) |
1ddbe94e | 2214 | { |
d184f6bf AE |
2215 | struct list_head *tmp; |
2216 | int rbd_id = rbd_dev->id; | |
2217 | int max_id; | |
2218 | ||
2219 | BUG_ON(rbd_id < 1); | |
499afd5b AE |
2220 | |
2221 | spin_lock(&rbd_dev_list_lock); | |
2222 | list_del_init(&rbd_dev->node); | |
d184f6bf AE |
2223 | |
2224 | /* | |
2225 | * If the id being "put" is not the current maximum, there | |
2226 | * is nothing special we need to do. | |
2227 | */ | |
2228 | if (rbd_id != atomic64_read(&rbd_id_max)) { | |
2229 | spin_unlock(&rbd_dev_list_lock); | |
2230 | return; | |
2231 | } | |
2232 | ||
2233 | /* | |
2234 | * We need to update the current maximum id. Search the | |
2235 | * list to find out what it is. We're more likely to find | |
2236 | * the maximum at the end, so search the list backward. | |
2237 | */ | |
2238 | max_id = 0; | |
2239 | list_for_each_prev(tmp, &rbd_dev_list) { | |
2240 | struct rbd_device *rbd_dev; | |
2241 | ||
2242 | rbd_dev = list_entry(tmp, struct rbd_device, node); | |
2243 | if (rbd_id > max_id) | |
2244 | max_id = rbd_id; | |
2245 | } | |
499afd5b | 2246 | spin_unlock(&rbd_dev_list_lock); |
b7f23c36 | 2247 | |
1ddbe94e | 2248 | /* |
d184f6bf AE |
2249 | * The max id could have been updated by rbd_id_get(), in |
2250 | * which case it now accurately reflects the new maximum. | |
2251 | * Be careful not to overwrite the maximum value in that | |
2252 | * case. | |
1ddbe94e | 2253 | */ |
d184f6bf | 2254 | atomic64_cmpxchg(&rbd_id_max, rbd_id, max_id); |
b7f23c36 AE |
2255 | } |
2256 | ||
e28fff26 AE |
2257 | /* |
2258 | * Skips over white space at *buf, and updates *buf to point to the | |
2259 | * first found non-space character (if any). Returns the length of | |
593a9e7b AE |
2260 | * the token (string of non-white space characters) found. Note |
2261 | * that *buf must be terminated with '\0'. | |
e28fff26 AE |
2262 | */ |
2263 | static inline size_t next_token(const char **buf) | |
2264 | { | |
2265 | /* | |
2266 | * These are the characters that produce nonzero for | |
2267 | * isspace() in the "C" and "POSIX" locales. | |
2268 | */ | |
2269 | const char *spaces = " \f\n\r\t\v"; | |
2270 | ||
2271 | *buf += strspn(*buf, spaces); /* Find start of token */ | |
2272 | ||
2273 | return strcspn(*buf, spaces); /* Return token length */ | |
2274 | } | |
2275 | ||
2276 | /* | |
2277 | * Finds the next token in *buf, and if the provided token buffer is | |
2278 | * big enough, copies the found token into it. The result, if | |
593a9e7b AE |
2279 | * copied, is guaranteed to be terminated with '\0'. Note that *buf |
2280 | * must be terminated with '\0' on entry. | |
e28fff26 AE |
2281 | * |
2282 | * Returns the length of the token found (not including the '\0'). | |
2283 | * Return value will be 0 if no token is found, and it will be >= | |
2284 | * token_size if the token would not fit. | |
2285 | * | |
593a9e7b | 2286 | * The *buf pointer will be updated to point beyond the end of the |
e28fff26 AE |
2287 | * found token. Note that this occurs even if the token buffer is |
2288 | * too small to hold it. | |
2289 | */ | |
2290 | static inline size_t copy_token(const char **buf, | |
2291 | char *token, | |
2292 | size_t token_size) | |
2293 | { | |
2294 | size_t len; | |
2295 | ||
2296 | len = next_token(buf); | |
2297 | if (len < token_size) { | |
2298 | memcpy(token, *buf, len); | |
2299 | *(token + len) = '\0'; | |
2300 | } | |
2301 | *buf += len; | |
2302 | ||
2303 | return len; | |
2304 | } | |
2305 | ||
ea3352f4 AE |
2306 | /* |
2307 | * Finds the next token in *buf, dynamically allocates a buffer big | |
2308 | * enough to hold a copy of it, and copies the token into the new | |
2309 | * buffer. The copy is guaranteed to be terminated with '\0'. Note | |
2310 | * that a duplicate buffer is created even for a zero-length token. | |
2311 | * | |
2312 | * Returns a pointer to the newly-allocated duplicate, or a null | |
2313 | * pointer if memory for the duplicate was not available. If | |
2314 | * the lenp argument is a non-null pointer, the length of the token | |
2315 | * (not including the '\0') is returned in *lenp. | |
2316 | * | |
2317 | * If successful, the *buf pointer will be updated to point beyond | |
2318 | * the end of the found token. | |
2319 | * | |
2320 | * Note: uses GFP_KERNEL for allocation. | |
2321 | */ | |
2322 | static inline char *dup_token(const char **buf, size_t *lenp) | |
2323 | { | |
2324 | char *dup; | |
2325 | size_t len; | |
2326 | ||
2327 | len = next_token(buf); | |
2328 | dup = kmalloc(len + 1, GFP_KERNEL); | |
2329 | if (!dup) | |
2330 | return NULL; | |
2331 | ||
2332 | memcpy(dup, *buf, len); | |
2333 | *(dup + len) = '\0'; | |
2334 | *buf += len; | |
2335 | ||
2336 | if (lenp) | |
2337 | *lenp = len; | |
2338 | ||
2339 | return dup; | |
2340 | } | |
2341 | ||
a725f65e AE |
2342 | /* |
2343 | * This fills in the pool_name, obj, obj_len, snap_name, obj_len, | |
2344 | * rbd_dev, rbd_md_name, and name fields of the given rbd_dev, based | |
2345 | * on the list of monitor addresses and other options provided via | |
2346 | * /sys/bus/rbd/add. | |
d22f76e7 AE |
2347 | * |
2348 | * Note: rbd_dev is assumed to have been initially zero-filled. | |
a725f65e AE |
2349 | */ |
2350 | static int rbd_add_parse_args(struct rbd_device *rbd_dev, | |
2351 | const char *buf, | |
7ef3214a | 2352 | const char **mon_addrs, |
5214ecc4 | 2353 | size_t *mon_addrs_size, |
e28fff26 AE |
2354 | char *options, |
2355 | size_t options_size) | |
2356 | { | |
d22f76e7 AE |
2357 | size_t len; |
2358 | int ret; | |
e28fff26 AE |
2359 | |
2360 | /* The first four tokens are required */ | |
2361 | ||
7ef3214a AE |
2362 | len = next_token(&buf); |
2363 | if (!len) | |
a725f65e | 2364 | return -EINVAL; |
5214ecc4 | 2365 | *mon_addrs_size = len + 1; |
7ef3214a AE |
2366 | *mon_addrs = buf; |
2367 | ||
2368 | buf += len; | |
a725f65e | 2369 | |
e28fff26 AE |
2370 | len = copy_token(&buf, options, options_size); |
2371 | if (!len || len >= options_size) | |
2372 | return -EINVAL; | |
2373 | ||
bf3e5ae1 | 2374 | ret = -ENOMEM; |
d22f76e7 AE |
2375 | rbd_dev->pool_name = dup_token(&buf, NULL); |
2376 | if (!rbd_dev->pool_name) | |
d22f76e7 | 2377 | goto out_err; |
e28fff26 | 2378 | |
bf3e5ae1 AE |
2379 | rbd_dev->obj = dup_token(&buf, &rbd_dev->obj_len); |
2380 | if (!rbd_dev->obj) | |
2381 | goto out_err; | |
a725f65e | 2382 | |
cb8627c7 AE |
2383 | /* Create the name of the header object */ |
2384 | ||
bf3e5ae1 AE |
2385 | rbd_dev->obj_md_name = kmalloc(rbd_dev->obj_len |
2386 | + sizeof (RBD_SUFFIX), | |
2387 | GFP_KERNEL); | |
2388 | if (!rbd_dev->obj_md_name) | |
cb8627c7 | 2389 | goto out_err; |
81a89793 | 2390 | sprintf(rbd_dev->obj_md_name, "%s%s", rbd_dev->obj, RBD_SUFFIX); |
a725f65e | 2391 | |
e28fff26 AE |
2392 | /* |
2393 | * The snapshot name is optional, but it's an error if it's | |
2394 | * too long. If no snapshot is supplied, fill in the default. | |
2395 | */ | |
2396 | len = copy_token(&buf, rbd_dev->snap_name, sizeof (rbd_dev->snap_name)); | |
2397 | if (!len) | |
2398 | memcpy(rbd_dev->snap_name, RBD_SNAP_HEAD_NAME, | |
2399 | sizeof (RBD_SNAP_HEAD_NAME)); | |
849b4260 AE |
2400 | else if (len >= sizeof (rbd_dev->snap_name)) { |
2401 | ret = -EINVAL; | |
d22f76e7 | 2402 | goto out_err; |
849b4260 | 2403 | } |
e28fff26 | 2404 | |
a725f65e | 2405 | return 0; |
d22f76e7 AE |
2406 | |
2407 | out_err: | |
cb8627c7 | 2408 | kfree(rbd_dev->obj_md_name); |
bf3e5ae1 | 2409 | kfree(rbd_dev->obj); |
d22f76e7 AE |
2410 | kfree(rbd_dev->pool_name); |
2411 | rbd_dev->pool_name = NULL; | |
2412 | ||
2413 | return ret; | |
a725f65e AE |
2414 | } |
2415 | ||
59c2be1e YS |
2416 | static ssize_t rbd_add(struct bus_type *bus, |
2417 | const char *buf, | |
2418 | size_t count) | |
602adf40 | 2419 | { |
cb8627c7 AE |
2420 | char *options; |
2421 | struct rbd_device *rbd_dev = NULL; | |
7ef3214a AE |
2422 | const char *mon_addrs = NULL; |
2423 | size_t mon_addrs_size = 0; | |
27cc2594 AE |
2424 | struct ceph_osd_client *osdc; |
2425 | int rc = -ENOMEM; | |
602adf40 YS |
2426 | |
2427 | if (!try_module_get(THIS_MODULE)) | |
2428 | return -ENODEV; | |
2429 | ||
60571c7d | 2430 | options = kmalloc(count, GFP_KERNEL); |
602adf40 | 2431 | if (!options) |
27cc2594 | 2432 | goto err_nomem; |
cb8627c7 AE |
2433 | rbd_dev = kzalloc(sizeof(*rbd_dev), GFP_KERNEL); |
2434 | if (!rbd_dev) | |
2435 | goto err_nomem; | |
602adf40 YS |
2436 | |
2437 | /* static rbd_device initialization */ | |
2438 | spin_lock_init(&rbd_dev->lock); | |
2439 | INIT_LIST_HEAD(&rbd_dev->node); | |
dfc5606d | 2440 | INIT_LIST_HEAD(&rbd_dev->snaps); |
c666601a | 2441 | init_rwsem(&rbd_dev->header_rwsem); |
602adf40 | 2442 | |
c666601a | 2443 | init_rwsem(&rbd_dev->header_rwsem); |
0e805a1d | 2444 | |
d184f6bf | 2445 | /* generate unique id: find highest unique id, add one */ |
499afd5b | 2446 | rbd_id_get(rbd_dev); |
602adf40 | 2447 | |
a725f65e | 2448 | /* Fill in the device name, now that we have its id. */ |
81a89793 AE |
2449 | BUILD_BUG_ON(DEV_NAME_LEN |
2450 | < sizeof (RBD_DRV_NAME) + MAX_INT_FORMAT_WIDTH); | |
2451 | sprintf(rbd_dev->name, "%s%d", RBD_DRV_NAME, rbd_dev->id); | |
a725f65e | 2452 | |
602adf40 | 2453 | /* parse add command */ |
7ef3214a | 2454 | rc = rbd_add_parse_args(rbd_dev, buf, &mon_addrs, &mon_addrs_size, |
e28fff26 | 2455 | options, count); |
a725f65e | 2456 | if (rc) |
f0f8cef5 | 2457 | goto err_put_id; |
e124a82f | 2458 | |
5214ecc4 AE |
2459 | rbd_dev->rbd_client = rbd_get_client(mon_addrs, mon_addrs_size - 1, |
2460 | options); | |
d720bcb0 AE |
2461 | if (IS_ERR(rbd_dev->rbd_client)) { |
2462 | rc = PTR_ERR(rbd_dev->rbd_client); | |
f0f8cef5 | 2463 | goto err_put_id; |
d720bcb0 | 2464 | } |
602adf40 | 2465 | |
602adf40 | 2466 | /* pick the pool */ |
1dbb4399 | 2467 | osdc = &rbd_dev->rbd_client->client->osdc; |
602adf40 YS |
2468 | rc = ceph_pg_poolid_by_name(osdc->osdmap, rbd_dev->pool_name); |
2469 | if (rc < 0) | |
2470 | goto err_out_client; | |
9bb2f334 | 2471 | rbd_dev->pool_id = rc; |
602adf40 YS |
2472 | |
2473 | /* register our block device */ | |
27cc2594 AE |
2474 | rc = register_blkdev(0, rbd_dev->name); |
2475 | if (rc < 0) | |
602adf40 | 2476 | goto err_out_client; |
27cc2594 | 2477 | rbd_dev->major = rc; |
602adf40 | 2478 | |
dfc5606d YS |
2479 | rc = rbd_bus_add_dev(rbd_dev); |
2480 | if (rc) | |
766fc439 YS |
2481 | goto err_out_blkdev; |
2482 | ||
32eec68d AE |
2483 | /* |
2484 | * At this point cleanup in the event of an error is the job | |
2485 | * of the sysfs code (initiated by rbd_bus_del_dev()). | |
2486 | * | |
2487 | * Set up and announce blkdev mapping. | |
2488 | */ | |
602adf40 YS |
2489 | rc = rbd_init_disk(rbd_dev); |
2490 | if (rc) | |
766fc439 | 2491 | goto err_out_bus; |
602adf40 | 2492 | |
59c2be1e YS |
2493 | rc = rbd_init_watch_dev(rbd_dev); |
2494 | if (rc) | |
2495 | goto err_out_bus; | |
2496 | ||
602adf40 YS |
2497 | return count; |
2498 | ||
766fc439 | 2499 | err_out_bus: |
766fc439 YS |
2500 | /* this will also clean up rest of rbd_dev stuff */ |
2501 | ||
2502 | rbd_bus_del_dev(rbd_dev); | |
2503 | kfree(options); | |
766fc439 YS |
2504 | return rc; |
2505 | ||
602adf40 YS |
2506 | err_out_blkdev: |
2507 | unregister_blkdev(rbd_dev->major, rbd_dev->name); | |
2508 | err_out_client: | |
2509 | rbd_put_client(rbd_dev); | |
f0f8cef5 | 2510 | err_put_id: |
cb8627c7 AE |
2511 | if (rbd_dev->pool_name) { |
2512 | kfree(rbd_dev->obj_md_name); | |
bf3e5ae1 | 2513 | kfree(rbd_dev->obj); |
cb8627c7 AE |
2514 | kfree(rbd_dev->pool_name); |
2515 | } | |
499afd5b | 2516 | rbd_id_put(rbd_dev); |
27cc2594 | 2517 | err_nomem: |
27cc2594 | 2518 | kfree(rbd_dev); |
cb8627c7 | 2519 | kfree(options); |
27cc2594 | 2520 | |
602adf40 YS |
2521 | dout("Error adding device %s\n", buf); |
2522 | module_put(THIS_MODULE); | |
27cc2594 AE |
2523 | |
2524 | return (ssize_t) rc; | |
602adf40 YS |
2525 | } |
2526 | ||
2527 | static struct rbd_device *__rbd_get_dev(unsigned long id) | |
2528 | { | |
2529 | struct list_head *tmp; | |
2530 | struct rbd_device *rbd_dev; | |
2531 | ||
e124a82f | 2532 | spin_lock(&rbd_dev_list_lock); |
602adf40 YS |
2533 | list_for_each(tmp, &rbd_dev_list) { |
2534 | rbd_dev = list_entry(tmp, struct rbd_device, node); | |
e124a82f AE |
2535 | if (rbd_dev->id == id) { |
2536 | spin_unlock(&rbd_dev_list_lock); | |
602adf40 | 2537 | return rbd_dev; |
e124a82f | 2538 | } |
602adf40 | 2539 | } |
e124a82f | 2540 | spin_unlock(&rbd_dev_list_lock); |
602adf40 YS |
2541 | return NULL; |
2542 | } | |
2543 | ||
dfc5606d | 2544 | static void rbd_dev_release(struct device *dev) |
602adf40 | 2545 | { |
593a9e7b | 2546 | struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); |
602adf40 | 2547 | |
1dbb4399 AE |
2548 | if (rbd_dev->watch_request) { |
2549 | struct ceph_client *client = rbd_dev->rbd_client->client; | |
2550 | ||
2551 | ceph_osdc_unregister_linger_request(&client->osdc, | |
59c2be1e | 2552 | rbd_dev->watch_request); |
1dbb4399 | 2553 | } |
59c2be1e | 2554 | if (rbd_dev->watch_event) |
79e3057c | 2555 | rbd_req_sync_unwatch(rbd_dev, rbd_dev->obj_md_name); |
59c2be1e | 2556 | |
602adf40 YS |
2557 | rbd_put_client(rbd_dev); |
2558 | ||
2559 | /* clean up and free blkdev */ | |
2560 | rbd_free_disk(rbd_dev); | |
2561 | unregister_blkdev(rbd_dev->major, rbd_dev->name); | |
32eec68d AE |
2562 | |
2563 | /* done with the id, and with the rbd_dev */ | |
cb8627c7 | 2564 | kfree(rbd_dev->obj_md_name); |
d22f76e7 | 2565 | kfree(rbd_dev->pool_name); |
bf3e5ae1 | 2566 | kfree(rbd_dev->obj); |
32eec68d | 2567 | rbd_id_put(rbd_dev); |
602adf40 YS |
2568 | kfree(rbd_dev); |
2569 | ||
2570 | /* release module ref */ | |
2571 | module_put(THIS_MODULE); | |
602adf40 YS |
2572 | } |
2573 | ||
dfc5606d YS |
2574 | static ssize_t rbd_remove(struct bus_type *bus, |
2575 | const char *buf, | |
2576 | size_t count) | |
602adf40 YS |
2577 | { |
2578 | struct rbd_device *rbd_dev = NULL; | |
2579 | int target_id, rc; | |
2580 | unsigned long ul; | |
2581 | int ret = count; | |
2582 | ||
2583 | rc = strict_strtoul(buf, 10, &ul); | |
2584 | if (rc) | |
2585 | return rc; | |
2586 | ||
2587 | /* convert to int; abort if we lost anything in the conversion */ | |
2588 | target_id = (int) ul; | |
2589 | if (target_id != ul) | |
2590 | return -EINVAL; | |
2591 | ||
2592 | mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING); | |
2593 | ||
2594 | rbd_dev = __rbd_get_dev(target_id); | |
2595 | if (!rbd_dev) { | |
2596 | ret = -ENOENT; | |
2597 | goto done; | |
2598 | } | |
2599 | ||
dfc5606d YS |
2600 | __rbd_remove_all_snaps(rbd_dev); |
2601 | rbd_bus_del_dev(rbd_dev); | |
602adf40 YS |
2602 | |
2603 | done: | |
2604 | mutex_unlock(&ctl_mutex); | |
2605 | return ret; | |
2606 | } | |
2607 | ||
dfc5606d YS |
2608 | static ssize_t rbd_snap_add(struct device *dev, |
2609 | struct device_attribute *attr, | |
2610 | const char *buf, | |
2611 | size_t count) | |
602adf40 | 2612 | { |
593a9e7b | 2613 | struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); |
dfc5606d YS |
2614 | int ret; |
2615 | char *name = kmalloc(count + 1, GFP_KERNEL); | |
602adf40 YS |
2616 | if (!name) |
2617 | return -ENOMEM; | |
2618 | ||
dfc5606d | 2619 | snprintf(name, count, "%s", buf); |
602adf40 YS |
2620 | |
2621 | mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING); | |
2622 | ||
602adf40 YS |
2623 | ret = rbd_header_add_snap(rbd_dev, |
2624 | name, GFP_KERNEL); | |
2625 | if (ret < 0) | |
59c2be1e | 2626 | goto err_unlock; |
602adf40 | 2627 | |
263c6ca0 | 2628 | ret = __rbd_refresh_header(rbd_dev); |
602adf40 | 2629 | if (ret < 0) |
59c2be1e YS |
2630 | goto err_unlock; |
2631 | ||
2632 | /* shouldn't hold ctl_mutex when notifying.. notify might | |
2633 | trigger a watch callback that would need to get that mutex */ | |
2634 | mutex_unlock(&ctl_mutex); | |
2635 | ||
2636 | /* make a best effort, don't error if failed */ | |
2637 | rbd_req_sync_notify(rbd_dev, rbd_dev->obj_md_name); | |
602adf40 YS |
2638 | |
2639 | ret = count; | |
59c2be1e YS |
2640 | kfree(name); |
2641 | return ret; | |
2642 | ||
2643 | err_unlock: | |
602adf40 | 2644 | mutex_unlock(&ctl_mutex); |
602adf40 YS |
2645 | kfree(name); |
2646 | return ret; | |
2647 | } | |
2648 | ||
602adf40 YS |
2649 | /* |
2650 | * create control files in sysfs | |
dfc5606d | 2651 | * /sys/bus/rbd/... |
602adf40 YS |
2652 | */ |
2653 | static int rbd_sysfs_init(void) | |
2654 | { | |
dfc5606d | 2655 | int ret; |
602adf40 | 2656 | |
fed4c143 | 2657 | ret = device_register(&rbd_root_dev); |
21079786 | 2658 | if (ret < 0) |
dfc5606d | 2659 | return ret; |
602adf40 | 2660 | |
fed4c143 AE |
2661 | ret = bus_register(&rbd_bus_type); |
2662 | if (ret < 0) | |
2663 | device_unregister(&rbd_root_dev); | |
602adf40 | 2664 | |
602adf40 YS |
2665 | return ret; |
2666 | } | |
2667 | ||
2668 | static void rbd_sysfs_cleanup(void) | |
2669 | { | |
dfc5606d | 2670 | bus_unregister(&rbd_bus_type); |
fed4c143 | 2671 | device_unregister(&rbd_root_dev); |
602adf40 YS |
2672 | } |
2673 | ||
2674 | int __init rbd_init(void) | |
2675 | { | |
2676 | int rc; | |
2677 | ||
2678 | rc = rbd_sysfs_init(); | |
2679 | if (rc) | |
2680 | return rc; | |
f0f8cef5 | 2681 | pr_info("loaded " RBD_DRV_NAME_LONG "\n"); |
602adf40 YS |
2682 | return 0; |
2683 | } | |
2684 | ||
2685 | void __exit rbd_exit(void) | |
2686 | { | |
2687 | rbd_sysfs_cleanup(); | |
2688 | } | |
2689 | ||
2690 | module_init(rbd_init); | |
2691 | module_exit(rbd_exit); | |
2692 | ||
2693 | MODULE_AUTHOR("Sage Weil <sage@newdream.net>"); | |
2694 | MODULE_AUTHOR("Yehuda Sadeh <yehuda@hq.newdream.net>"); | |
2695 | MODULE_DESCRIPTION("rados block device"); | |
2696 | ||
2697 | /* following authorship retained from original osdblk.c */ | |
2698 | MODULE_AUTHOR("Jeff Garzik <jeff@garzik.org>"); | |
2699 | ||
2700 | MODULE_LICENSE("GPL"); |