Loading...
1
2#include <linux/ceph/ceph_debug.h>
3#include <linux/backing-dev.h>
4#include <linux/ctype.h>
5#include <linux/fs.h>
6#include <linux/inet.h>
7#include <linux/in6.h>
8#include <linux/key.h>
9#include <keys/ceph-type.h>
10#include <linux/module.h>
11#include <linux/mount.h>
12#include <linux/parser.h>
13#include <linux/sched.h>
14#include <linux/seq_file.h>
15#include <linux/slab.h>
16#include <linux/statfs.h>
17#include <linux/string.h>
18
19
20#include <linux/ceph/libceph.h>
21#include <linux/ceph/debugfs.h>
22#include <linux/ceph/decode.h>
23#include <linux/ceph/mon_client.h>
24#include <linux/ceph/auth.h>
25#include "crypto.h"
26
27
28
29/*
30 * find filename portion of a path (/foo/bar/baz -> baz)
31 */
32const char *ceph_file_part(const char *s, int len)
33{
34 const char *e = s + len;
35
36 while (e != s && *(e-1) != '/')
37 e--;
38 return e;
39}
40EXPORT_SYMBOL(ceph_file_part);
41
42const char *ceph_msg_type_name(int type)
43{
44 switch (type) {
45 case CEPH_MSG_SHUTDOWN: return "shutdown";
46 case CEPH_MSG_PING: return "ping";
47 case CEPH_MSG_AUTH: return "auth";
48 case CEPH_MSG_AUTH_REPLY: return "auth_reply";
49 case CEPH_MSG_MON_MAP: return "mon_map";
50 case CEPH_MSG_MON_GET_MAP: return "mon_get_map";
51 case CEPH_MSG_MON_SUBSCRIBE: return "mon_subscribe";
52 case CEPH_MSG_MON_SUBSCRIBE_ACK: return "mon_subscribe_ack";
53 case CEPH_MSG_STATFS: return "statfs";
54 case CEPH_MSG_STATFS_REPLY: return "statfs_reply";
55 case CEPH_MSG_MDS_MAP: return "mds_map";
56 case CEPH_MSG_CLIENT_SESSION: return "client_session";
57 case CEPH_MSG_CLIENT_RECONNECT: return "client_reconnect";
58 case CEPH_MSG_CLIENT_REQUEST: return "client_request";
59 case CEPH_MSG_CLIENT_REQUEST_FORWARD: return "client_request_forward";
60 case CEPH_MSG_CLIENT_REPLY: return "client_reply";
61 case CEPH_MSG_CLIENT_CAPS: return "client_caps";
62 case CEPH_MSG_CLIENT_CAPRELEASE: return "client_cap_release";
63 case CEPH_MSG_CLIENT_SNAP: return "client_snap";
64 case CEPH_MSG_CLIENT_LEASE: return "client_lease";
65 case CEPH_MSG_OSD_MAP: return "osd_map";
66 case CEPH_MSG_OSD_OP: return "osd_op";
67 case CEPH_MSG_OSD_OPREPLY: return "osd_opreply";
68 case CEPH_MSG_WATCH_NOTIFY: return "watch_notify";
69 default: return "unknown";
70 }
71}
72EXPORT_SYMBOL(ceph_msg_type_name);
73
74/*
75 * Initially learn our fsid, or verify an fsid matches.
76 */
77int ceph_check_fsid(struct ceph_client *client, struct ceph_fsid *fsid)
78{
79 if (client->have_fsid) {
80 if (ceph_fsid_compare(&client->fsid, fsid)) {
81 pr_err("bad fsid, had %pU got %pU",
82 &client->fsid, fsid);
83 return -1;
84 }
85 } else {
86 pr_info("client%lld fsid %pU\n", ceph_client_id(client), fsid);
87 memcpy(&client->fsid, fsid, sizeof(*fsid));
88 }
89 return 0;
90}
91EXPORT_SYMBOL(ceph_check_fsid);
92
93static int strcmp_null(const char *s1, const char *s2)
94{
95 if (!s1 && !s2)
96 return 0;
97 if (s1 && !s2)
98 return -1;
99 if (!s1 && s2)
100 return 1;
101 return strcmp(s1, s2);
102}
103
104int ceph_compare_options(struct ceph_options *new_opt,
105 struct ceph_client *client)
106{
107 struct ceph_options *opt1 = new_opt;
108 struct ceph_options *opt2 = client->options;
109 int ofs = offsetof(struct ceph_options, mon_addr);
110 int i;
111 int ret;
112
113 ret = memcmp(opt1, opt2, ofs);
114 if (ret)
115 return ret;
116
117 ret = strcmp_null(opt1->name, opt2->name);
118 if (ret)
119 return ret;
120
121 if (opt1->key && !opt2->key)
122 return -1;
123 if (!opt1->key && opt2->key)
124 return 1;
125 if (opt1->key && opt2->key) {
126 if (opt1->key->type != opt2->key->type)
127 return -1;
128 if (opt1->key->created.tv_sec != opt2->key->created.tv_sec)
129 return -1;
130 if (opt1->key->created.tv_nsec != opt2->key->created.tv_nsec)
131 return -1;
132 if (opt1->key->len != opt2->key->len)
133 return -1;
134 if (opt1->key->key && !opt2->key->key)
135 return -1;
136 if (!opt1->key->key && opt2->key->key)
137 return 1;
138 if (opt1->key->key && opt2->key->key) {
139 ret = memcmp(opt1->key->key, opt2->key->key, opt1->key->len);
140 if (ret)
141 return ret;
142 }
143 }
144
145 /* any matching mon ip implies a match */
146 for (i = 0; i < opt1->num_mon; i++) {
147 if (ceph_monmap_contains(client->monc.monmap,
148 &opt1->mon_addr[i]))
149 return 0;
150 }
151 return -1;
152}
153EXPORT_SYMBOL(ceph_compare_options);
154
155
156static int parse_fsid(const char *str, struct ceph_fsid *fsid)
157{
158 int i = 0;
159 char tmp[3];
160 int err = -EINVAL;
161 int d;
162
163 dout("parse_fsid '%s'\n", str);
164 tmp[2] = 0;
165 while (*str && i < 16) {
166 if (ispunct(*str)) {
167 str++;
168 continue;
169 }
170 if (!isxdigit(str[0]) || !isxdigit(str[1]))
171 break;
172 tmp[0] = str[0];
173 tmp[1] = str[1];
174 if (sscanf(tmp, "%x", &d) < 1)
175 break;
176 fsid->fsid[i] = d & 0xff;
177 i++;
178 str += 2;
179 }
180
181 if (i == 16)
182 err = 0;
183 dout("parse_fsid ret %d got fsid %pU", err, fsid);
184 return err;
185}
186
187/*
188 * ceph options
189 */
190enum {
191 Opt_osdtimeout,
192 Opt_osdkeepalivetimeout,
193 Opt_mount_timeout,
194 Opt_osd_idle_ttl,
195 Opt_last_int,
196 /* int args above */
197 Opt_fsid,
198 Opt_name,
199 Opt_secret,
200 Opt_key,
201 Opt_ip,
202 Opt_last_string,
203 /* string args above */
204 Opt_share,
205 Opt_noshare,
206 Opt_crc,
207 Opt_nocrc,
208};
209
210static match_table_t opt_tokens = {
211 {Opt_osdtimeout, "osdtimeout=%d"},
212 {Opt_osdkeepalivetimeout, "osdkeepalive=%d"},
213 {Opt_mount_timeout, "mount_timeout=%d"},
214 {Opt_osd_idle_ttl, "osd_idle_ttl=%d"},
215 /* int args above */
216 {Opt_fsid, "fsid=%s"},
217 {Opt_name, "name=%s"},
218 {Opt_secret, "secret=%s"},
219 {Opt_key, "key=%s"},
220 {Opt_ip, "ip=%s"},
221 /* string args above */
222 {Opt_share, "share"},
223 {Opt_noshare, "noshare"},
224 {Opt_crc, "crc"},
225 {Opt_nocrc, "nocrc"},
226 {-1, NULL}
227};
228
229void ceph_destroy_options(struct ceph_options *opt)
230{
231 dout("destroy_options %p\n", opt);
232 kfree(opt->name);
233 if (opt->key) {
234 ceph_crypto_key_destroy(opt->key);
235 kfree(opt->key);
236 }
237 kfree(opt->mon_addr);
238 kfree(opt);
239}
240EXPORT_SYMBOL(ceph_destroy_options);
241
242/* get secret from key store */
243static int get_secret(struct ceph_crypto_key *dst, const char *name) {
244 struct key *ukey;
245 int key_err;
246 int err = 0;
247 struct ceph_crypto_key *ckey;
248
249 ukey = request_key(&key_type_ceph, name, NULL);
250 if (!ukey || IS_ERR(ukey)) {
251 /* request_key errors don't map nicely to mount(2)
252 errors; don't even try, but still printk */
253 key_err = PTR_ERR(ukey);
254 switch (key_err) {
255 case -ENOKEY:
256 pr_warning("ceph: Mount failed due to key not found: %s\n", name);
257 break;
258 case -EKEYEXPIRED:
259 pr_warning("ceph: Mount failed due to expired key: %s\n", name);
260 break;
261 case -EKEYREVOKED:
262 pr_warning("ceph: Mount failed due to revoked key: %s\n", name);
263 break;
264 default:
265 pr_warning("ceph: Mount failed due to unknown key error"
266 " %d: %s\n", key_err, name);
267 }
268 err = -EPERM;
269 goto out;
270 }
271
272 ckey = ukey->payload.data;
273 err = ceph_crypto_key_clone(dst, ckey);
274 if (err)
275 goto out_key;
276 /* pass through, err is 0 */
277
278out_key:
279 key_put(ukey);
280out:
281 return err;
282}
283
284struct ceph_options *
285ceph_parse_options(char *options, const char *dev_name,
286 const char *dev_name_end,
287 int (*parse_extra_token)(char *c, void *private),
288 void *private)
289{
290 struct ceph_options *opt;
291 const char *c;
292 int err = -ENOMEM;
293 substring_t argstr[MAX_OPT_ARGS];
294
295 opt = kzalloc(sizeof(*opt), GFP_KERNEL);
296 if (!opt)
297 return ERR_PTR(-ENOMEM);
298 opt->mon_addr = kcalloc(CEPH_MAX_MON, sizeof(*opt->mon_addr),
299 GFP_KERNEL);
300 if (!opt->mon_addr)
301 goto out;
302
303 dout("parse_options %p options '%s' dev_name '%s'\n", opt, options,
304 dev_name);
305
306 /* start with defaults */
307 opt->flags = CEPH_OPT_DEFAULT;
308 opt->osd_timeout = CEPH_OSD_TIMEOUT_DEFAULT;
309 opt->osd_keepalive_timeout = CEPH_OSD_KEEPALIVE_DEFAULT;
310 opt->mount_timeout = CEPH_MOUNT_TIMEOUT_DEFAULT; /* seconds */
311 opt->osd_idle_ttl = CEPH_OSD_IDLE_TTL_DEFAULT; /* seconds */
312
313 /* get mon ip(s) */
314 /* ip1[:port1][,ip2[:port2]...] */
315 err = ceph_parse_ips(dev_name, dev_name_end, opt->mon_addr,
316 CEPH_MAX_MON, &opt->num_mon);
317 if (err < 0)
318 goto out;
319
320 /* parse mount options */
321 while ((c = strsep(&options, ",")) != NULL) {
322 int token, intval, ret;
323 if (!*c)
324 continue;
325 err = -EINVAL;
326 token = match_token((char *)c, opt_tokens, argstr);
327 if (token < 0 && parse_extra_token) {
328 /* extra? */
329 err = parse_extra_token((char *)c, private);
330 if (err < 0) {
331 pr_err("bad option at '%s'\n", c);
332 goto out;
333 }
334 continue;
335 }
336 if (token < Opt_last_int) {
337 ret = match_int(&argstr[0], &intval);
338 if (ret < 0) {
339 pr_err("bad mount option arg (not int) "
340 "at '%s'\n", c);
341 continue;
342 }
343 dout("got int token %d val %d\n", token, intval);
344 } else if (token > Opt_last_int && token < Opt_last_string) {
345 dout("got string token %d val %s\n", token,
346 argstr[0].from);
347 } else {
348 dout("got token %d\n", token);
349 }
350 switch (token) {
351 case Opt_ip:
352 err = ceph_parse_ips(argstr[0].from,
353 argstr[0].to,
354 &opt->my_addr,
355 1, NULL);
356 if (err < 0)
357 goto out;
358 opt->flags |= CEPH_OPT_MYIP;
359 break;
360
361 case Opt_fsid:
362 err = parse_fsid(argstr[0].from, &opt->fsid);
363 if (err == 0)
364 opt->flags |= CEPH_OPT_FSID;
365 break;
366 case Opt_name:
367 opt->name = kstrndup(argstr[0].from,
368 argstr[0].to-argstr[0].from,
369 GFP_KERNEL);
370 break;
371 case Opt_secret:
372 opt->key = kzalloc(sizeof(*opt->key), GFP_KERNEL);
373 if (!opt->key) {
374 err = -ENOMEM;
375 goto out;
376 }
377 err = ceph_crypto_key_unarmor(opt->key, argstr[0].from);
378 if (err < 0)
379 goto out;
380 break;
381 case Opt_key:
382 opt->key = kzalloc(sizeof(*opt->key), GFP_KERNEL);
383 if (!opt->key) {
384 err = -ENOMEM;
385 goto out;
386 }
387 err = get_secret(opt->key, argstr[0].from);
388 if (err < 0)
389 goto out;
390 break;
391
392 /* misc */
393 case Opt_osdtimeout:
394 opt->osd_timeout = intval;
395 break;
396 case Opt_osdkeepalivetimeout:
397 opt->osd_keepalive_timeout = intval;
398 break;
399 case Opt_osd_idle_ttl:
400 opt->osd_idle_ttl = intval;
401 break;
402 case Opt_mount_timeout:
403 opt->mount_timeout = intval;
404 break;
405
406 case Opt_share:
407 opt->flags &= ~CEPH_OPT_NOSHARE;
408 break;
409 case Opt_noshare:
410 opt->flags |= CEPH_OPT_NOSHARE;
411 break;
412
413 case Opt_crc:
414 opt->flags &= ~CEPH_OPT_NOCRC;
415 break;
416 case Opt_nocrc:
417 opt->flags |= CEPH_OPT_NOCRC;
418 break;
419
420 default:
421 BUG_ON(token);
422 }
423 }
424
425 /* success */
426 return opt;
427
428out:
429 ceph_destroy_options(opt);
430 return ERR_PTR(err);
431}
432EXPORT_SYMBOL(ceph_parse_options);
433
434u64 ceph_client_id(struct ceph_client *client)
435{
436 return client->monc.auth->global_id;
437}
438EXPORT_SYMBOL(ceph_client_id);
439
440/*
441 * create a fresh client instance
442 */
443struct ceph_client *ceph_create_client(struct ceph_options *opt, void *private,
444 unsigned int supported_features,
445 unsigned int required_features)
446{
447 struct ceph_client *client;
448 struct ceph_entity_addr *myaddr = NULL;
449 int err = -ENOMEM;
450
451 client = kzalloc(sizeof(*client), GFP_KERNEL);
452 if (client == NULL)
453 return ERR_PTR(-ENOMEM);
454
455 client->private = private;
456 client->options = opt;
457
458 mutex_init(&client->mount_mutex);
459 init_waitqueue_head(&client->auth_wq);
460 client->auth_err = 0;
461
462 client->extra_mon_dispatch = NULL;
463 client->supported_features = CEPH_FEATURE_SUPPORTED_DEFAULT |
464 supported_features;
465 client->required_features = CEPH_FEATURE_REQUIRED_DEFAULT |
466 required_features;
467
468 /* msgr */
469 if (ceph_test_opt(client, MYIP))
470 myaddr = &client->options->my_addr;
471 client->msgr = ceph_messenger_create(myaddr,
472 client->supported_features,
473 client->required_features);
474 if (IS_ERR(client->msgr)) {
475 err = PTR_ERR(client->msgr);
476 goto fail;
477 }
478 client->msgr->nocrc = ceph_test_opt(client, NOCRC);
479
480 /* subsystems */
481 err = ceph_monc_init(&client->monc, client);
482 if (err < 0)
483 goto fail_msgr;
484 err = ceph_osdc_init(&client->osdc, client);
485 if (err < 0)
486 goto fail_monc;
487
488 return client;
489
490fail_monc:
491 ceph_monc_stop(&client->monc);
492fail_msgr:
493 ceph_messenger_destroy(client->msgr);
494fail:
495 kfree(client);
496 return ERR_PTR(err);
497}
498EXPORT_SYMBOL(ceph_create_client);
499
500void ceph_destroy_client(struct ceph_client *client)
501{
502 dout("destroy_client %p\n", client);
503
504 /* unmount */
505 ceph_osdc_stop(&client->osdc);
506
507 ceph_monc_stop(&client->monc);
508
509 ceph_debugfs_client_cleanup(client);
510
511 ceph_messenger_destroy(client->msgr);
512
513 ceph_destroy_options(client->options);
514
515 kfree(client);
516 dout("destroy_client %p done\n", client);
517}
518EXPORT_SYMBOL(ceph_destroy_client);
519
520/*
521 * true if we have the mon map (and have thus joined the cluster)
522 */
523static int have_mon_and_osd_map(struct ceph_client *client)
524{
525 return client->monc.monmap && client->monc.monmap->epoch &&
526 client->osdc.osdmap && client->osdc.osdmap->epoch;
527}
528
529/*
530 * mount: join the ceph cluster, and open root directory.
531 */
532int __ceph_open_session(struct ceph_client *client, unsigned long started)
533{
534 int err;
535 unsigned long timeout = client->options->mount_timeout * HZ;
536
537 /* open session, and wait for mon and osd maps */
538 err = ceph_monc_open_session(&client->monc);
539 if (err < 0)
540 return err;
541
542 while (!have_mon_and_osd_map(client)) {
543 err = -EIO;
544 if (timeout && time_after_eq(jiffies, started + timeout))
545 return err;
546
547 /* wait */
548 dout("mount waiting for mon_map\n");
549 err = wait_event_interruptible_timeout(client->auth_wq,
550 have_mon_and_osd_map(client) || (client->auth_err < 0),
551 timeout);
552 if (err == -EINTR || err == -ERESTARTSYS)
553 return err;
554 if (client->auth_err < 0)
555 return client->auth_err;
556 }
557
558 return 0;
559}
560EXPORT_SYMBOL(__ceph_open_session);
561
562
563int ceph_open_session(struct ceph_client *client)
564{
565 int ret;
566 unsigned long started = jiffies; /* note the start time */
567
568 dout("open_session start\n");
569 mutex_lock(&client->mount_mutex);
570
571 ret = __ceph_open_session(client, started);
572
573 mutex_unlock(&client->mount_mutex);
574 return ret;
575}
576EXPORT_SYMBOL(ceph_open_session);
577
578
579static int __init init_ceph_lib(void)
580{
581 int ret = 0;
582
583 ret = ceph_debugfs_init();
584 if (ret < 0)
585 goto out;
586
587 ret = ceph_crypto_init();
588 if (ret < 0)
589 goto out_debugfs;
590
591 ret = ceph_msgr_init();
592 if (ret < 0)
593 goto out_crypto;
594
595 pr_info("loaded (mon/osd proto %d/%d, osdmap %d/%d %d/%d)\n",
596 CEPH_MONC_PROTOCOL, CEPH_OSDC_PROTOCOL,
597 CEPH_OSDMAP_VERSION, CEPH_OSDMAP_VERSION_EXT,
598 CEPH_OSDMAP_INC_VERSION, CEPH_OSDMAP_INC_VERSION_EXT);
599
600 return 0;
601
602out_crypto:
603 ceph_crypto_shutdown();
604out_debugfs:
605 ceph_debugfs_cleanup();
606out:
607 return ret;
608}
609
610static void __exit exit_ceph_lib(void)
611{
612 dout("exit_ceph_lib\n");
613 ceph_msgr_exit();
614 ceph_crypto_shutdown();
615 ceph_debugfs_cleanup();
616}
617
618module_init(init_ceph_lib);
619module_exit(exit_ceph_lib);
620
621MODULE_AUTHOR("Sage Weil <sage@newdream.net>");
622MODULE_AUTHOR("Yehuda Sadeh <yehuda@hq.newdream.net>");
623MODULE_AUTHOR("Patience Warnick <patience@newdream.net>");
624MODULE_DESCRIPTION("Ceph filesystem for Linux");
625MODULE_LICENSE("GPL");
1
2#include <linux/ceph/ceph_debug.h>
3#include <linux/backing-dev.h>
4#include <linux/ctype.h>
5#include <linux/fs.h>
6#include <linux/inet.h>
7#include <linux/in6.h>
8#include <linux/key.h>
9#include <keys/ceph-type.h>
10#include <linux/module.h>
11#include <linux/mount.h>
12#include <linux/nsproxy.h>
13#include <linux/parser.h>
14#include <linux/sched.h>
15#include <linux/seq_file.h>
16#include <linux/slab.h>
17#include <linux/statfs.h>
18#include <linux/string.h>
19#include <linux/vmalloc.h>
20
21
22#include <linux/ceph/ceph_features.h>
23#include <linux/ceph/libceph.h>
24#include <linux/ceph/debugfs.h>
25#include <linux/ceph/decode.h>
26#include <linux/ceph/mon_client.h>
27#include <linux/ceph/auth.h>
28#include "crypto.h"
29
30
31/*
32 * Module compatibility interface. For now it doesn't do anything,
33 * but its existence signals a certain level of functionality.
34 *
35 * The data buffer is used to pass information both to and from
36 * libceph. The return value indicates whether libceph determines
37 * it is compatible with the caller (from another kernel module),
38 * given the provided data.
39 *
40 * The data pointer can be null.
41 */
42bool libceph_compatible(void *data)
43{
44 return true;
45}
46EXPORT_SYMBOL(libceph_compatible);
47
48/*
49 * find filename portion of a path (/foo/bar/baz -> baz)
50 */
51const char *ceph_file_part(const char *s, int len)
52{
53 const char *e = s + len;
54
55 while (e != s && *(e-1) != '/')
56 e--;
57 return e;
58}
59EXPORT_SYMBOL(ceph_file_part);
60
61const char *ceph_msg_type_name(int type)
62{
63 switch (type) {
64 case CEPH_MSG_SHUTDOWN: return "shutdown";
65 case CEPH_MSG_PING: return "ping";
66 case CEPH_MSG_AUTH: return "auth";
67 case CEPH_MSG_AUTH_REPLY: return "auth_reply";
68 case CEPH_MSG_MON_MAP: return "mon_map";
69 case CEPH_MSG_MON_GET_MAP: return "mon_get_map";
70 case CEPH_MSG_MON_SUBSCRIBE: return "mon_subscribe";
71 case CEPH_MSG_MON_SUBSCRIBE_ACK: return "mon_subscribe_ack";
72 case CEPH_MSG_STATFS: return "statfs";
73 case CEPH_MSG_STATFS_REPLY: return "statfs_reply";
74 case CEPH_MSG_MON_GET_VERSION: return "mon_get_version";
75 case CEPH_MSG_MON_GET_VERSION_REPLY: return "mon_get_version_reply";
76 case CEPH_MSG_MDS_MAP: return "mds_map";
77 case CEPH_MSG_CLIENT_SESSION: return "client_session";
78 case CEPH_MSG_CLIENT_RECONNECT: return "client_reconnect";
79 case CEPH_MSG_CLIENT_REQUEST: return "client_request";
80 case CEPH_MSG_CLIENT_REQUEST_FORWARD: return "client_request_forward";
81 case CEPH_MSG_CLIENT_REPLY: return "client_reply";
82 case CEPH_MSG_CLIENT_CAPS: return "client_caps";
83 case CEPH_MSG_CLIENT_CAPRELEASE: return "client_cap_release";
84 case CEPH_MSG_CLIENT_SNAP: return "client_snap";
85 case CEPH_MSG_CLIENT_LEASE: return "client_lease";
86 case CEPH_MSG_OSD_MAP: return "osd_map";
87 case CEPH_MSG_OSD_OP: return "osd_op";
88 case CEPH_MSG_OSD_OPREPLY: return "osd_opreply";
89 case CEPH_MSG_WATCH_NOTIFY: return "watch_notify";
90 default: return "unknown";
91 }
92}
93EXPORT_SYMBOL(ceph_msg_type_name);
94
95/*
96 * Initially learn our fsid, or verify an fsid matches.
97 */
98int ceph_check_fsid(struct ceph_client *client, struct ceph_fsid *fsid)
99{
100 if (client->have_fsid) {
101 if (ceph_fsid_compare(&client->fsid, fsid)) {
102 pr_err("bad fsid, had %pU got %pU",
103 &client->fsid, fsid);
104 return -1;
105 }
106 } else {
107 memcpy(&client->fsid, fsid, sizeof(*fsid));
108 }
109 return 0;
110}
111EXPORT_SYMBOL(ceph_check_fsid);
112
113static int strcmp_null(const char *s1, const char *s2)
114{
115 if (!s1 && !s2)
116 return 0;
117 if (s1 && !s2)
118 return -1;
119 if (!s1 && s2)
120 return 1;
121 return strcmp(s1, s2);
122}
123
124int ceph_compare_options(struct ceph_options *new_opt,
125 struct ceph_client *client)
126{
127 struct ceph_options *opt1 = new_opt;
128 struct ceph_options *opt2 = client->options;
129 int ofs = offsetof(struct ceph_options, mon_addr);
130 int i;
131 int ret;
132
133 /*
134 * Don't bother comparing options if network namespaces don't
135 * match.
136 */
137 if (!net_eq(current->nsproxy->net_ns, read_pnet(&client->msgr.net)))
138 return -1;
139
140 ret = memcmp(opt1, opt2, ofs);
141 if (ret)
142 return ret;
143
144 ret = strcmp_null(opt1->name, opt2->name);
145 if (ret)
146 return ret;
147
148 if (opt1->key && !opt2->key)
149 return -1;
150 if (!opt1->key && opt2->key)
151 return 1;
152 if (opt1->key && opt2->key) {
153 if (opt1->key->type != opt2->key->type)
154 return -1;
155 if (opt1->key->created.tv_sec != opt2->key->created.tv_sec)
156 return -1;
157 if (opt1->key->created.tv_nsec != opt2->key->created.tv_nsec)
158 return -1;
159 if (opt1->key->len != opt2->key->len)
160 return -1;
161 if (opt1->key->key && !opt2->key->key)
162 return -1;
163 if (!opt1->key->key && opt2->key->key)
164 return 1;
165 if (opt1->key->key && opt2->key->key) {
166 ret = memcmp(opt1->key->key, opt2->key->key, opt1->key->len);
167 if (ret)
168 return ret;
169 }
170 }
171
172 /* any matching mon ip implies a match */
173 for (i = 0; i < opt1->num_mon; i++) {
174 if (ceph_monmap_contains(client->monc.monmap,
175 &opt1->mon_addr[i]))
176 return 0;
177 }
178 return -1;
179}
180EXPORT_SYMBOL(ceph_compare_options);
181
182void *ceph_kvmalloc(size_t size, gfp_t flags)
183{
184 if (size <= (PAGE_SIZE << PAGE_ALLOC_COSTLY_ORDER)) {
185 void *ptr = kmalloc(size, flags | __GFP_NOWARN);
186 if (ptr)
187 return ptr;
188 }
189
190 return __vmalloc(size, flags | __GFP_HIGHMEM, PAGE_KERNEL);
191}
192
193
194static int parse_fsid(const char *str, struct ceph_fsid *fsid)
195{
196 int i = 0;
197 char tmp[3];
198 int err = -EINVAL;
199 int d;
200
201 dout("parse_fsid '%s'\n", str);
202 tmp[2] = 0;
203 while (*str && i < 16) {
204 if (ispunct(*str)) {
205 str++;
206 continue;
207 }
208 if (!isxdigit(str[0]) || !isxdigit(str[1]))
209 break;
210 tmp[0] = str[0];
211 tmp[1] = str[1];
212 if (sscanf(tmp, "%x", &d) < 1)
213 break;
214 fsid->fsid[i] = d & 0xff;
215 i++;
216 str += 2;
217 }
218
219 if (i == 16)
220 err = 0;
221 dout("parse_fsid ret %d got fsid %pU", err, fsid);
222 return err;
223}
224
225/*
226 * ceph options
227 */
228enum {
229 Opt_osdtimeout,
230 Opt_osdkeepalivetimeout,
231 Opt_mount_timeout,
232 Opt_osd_idle_ttl,
233 Opt_last_int,
234 /* int args above */
235 Opt_fsid,
236 Opt_name,
237 Opt_secret,
238 Opt_key,
239 Opt_ip,
240 Opt_last_string,
241 /* string args above */
242 Opt_share,
243 Opt_noshare,
244 Opt_crc,
245 Opt_nocrc,
246 Opt_cephx_require_signatures,
247 Opt_nocephx_require_signatures,
248 Opt_cephx_sign_messages,
249 Opt_nocephx_sign_messages,
250 Opt_tcp_nodelay,
251 Opt_notcp_nodelay,
252};
253
254static match_table_t opt_tokens = {
255 {Opt_osdtimeout, "osdtimeout=%d"},
256 {Opt_osdkeepalivetimeout, "osdkeepalive=%d"},
257 {Opt_mount_timeout, "mount_timeout=%d"},
258 {Opt_osd_idle_ttl, "osd_idle_ttl=%d"},
259 /* int args above */
260 {Opt_fsid, "fsid=%s"},
261 {Opt_name, "name=%s"},
262 {Opt_secret, "secret=%s"},
263 {Opt_key, "key=%s"},
264 {Opt_ip, "ip=%s"},
265 /* string args above */
266 {Opt_share, "share"},
267 {Opt_noshare, "noshare"},
268 {Opt_crc, "crc"},
269 {Opt_nocrc, "nocrc"},
270 {Opt_cephx_require_signatures, "cephx_require_signatures"},
271 {Opt_nocephx_require_signatures, "nocephx_require_signatures"},
272 {Opt_cephx_sign_messages, "cephx_sign_messages"},
273 {Opt_nocephx_sign_messages, "nocephx_sign_messages"},
274 {Opt_tcp_nodelay, "tcp_nodelay"},
275 {Opt_notcp_nodelay, "notcp_nodelay"},
276 {-1, NULL}
277};
278
279void ceph_destroy_options(struct ceph_options *opt)
280{
281 dout("destroy_options %p\n", opt);
282 kfree(opt->name);
283 if (opt->key) {
284 ceph_crypto_key_destroy(opt->key);
285 kfree(opt->key);
286 }
287 kfree(opt->mon_addr);
288 kfree(opt);
289}
290EXPORT_SYMBOL(ceph_destroy_options);
291
292/* get secret from key store */
293static int get_secret(struct ceph_crypto_key *dst, const char *name) {
294 struct key *ukey;
295 int key_err;
296 int err = 0;
297 struct ceph_crypto_key *ckey;
298
299 ukey = request_key(&key_type_ceph, name, NULL);
300 if (!ukey || IS_ERR(ukey)) {
301 /* request_key errors don't map nicely to mount(2)
302 errors; don't even try, but still printk */
303 key_err = PTR_ERR(ukey);
304 switch (key_err) {
305 case -ENOKEY:
306 pr_warn("ceph: Mount failed due to key not found: %s\n",
307 name);
308 break;
309 case -EKEYEXPIRED:
310 pr_warn("ceph: Mount failed due to expired key: %s\n",
311 name);
312 break;
313 case -EKEYREVOKED:
314 pr_warn("ceph: Mount failed due to revoked key: %s\n",
315 name);
316 break;
317 default:
318 pr_warn("ceph: Mount failed due to unknown key error %d: %s\n",
319 key_err, name);
320 }
321 err = -EPERM;
322 goto out;
323 }
324
325 ckey = ukey->payload.data[0];
326 err = ceph_crypto_key_clone(dst, ckey);
327 if (err)
328 goto out_key;
329 /* pass through, err is 0 */
330
331out_key:
332 key_put(ukey);
333out:
334 return err;
335}
336
337struct ceph_options *
338ceph_parse_options(char *options, const char *dev_name,
339 const char *dev_name_end,
340 int (*parse_extra_token)(char *c, void *private),
341 void *private)
342{
343 struct ceph_options *opt;
344 const char *c;
345 int err = -ENOMEM;
346 substring_t argstr[MAX_OPT_ARGS];
347
348 opt = kzalloc(sizeof(*opt), GFP_KERNEL);
349 if (!opt)
350 return ERR_PTR(-ENOMEM);
351 opt->mon_addr = kcalloc(CEPH_MAX_MON, sizeof(*opt->mon_addr),
352 GFP_KERNEL);
353 if (!opt->mon_addr)
354 goto out;
355
356 dout("parse_options %p options '%s' dev_name '%s'\n", opt, options,
357 dev_name);
358
359 /* start with defaults */
360 opt->flags = CEPH_OPT_DEFAULT;
361 opt->osd_keepalive_timeout = CEPH_OSD_KEEPALIVE_DEFAULT;
362 opt->mount_timeout = CEPH_MOUNT_TIMEOUT_DEFAULT;
363 opt->osd_idle_ttl = CEPH_OSD_IDLE_TTL_DEFAULT;
364
365 /* get mon ip(s) */
366 /* ip1[:port1][,ip2[:port2]...] */
367 err = ceph_parse_ips(dev_name, dev_name_end, opt->mon_addr,
368 CEPH_MAX_MON, &opt->num_mon);
369 if (err < 0)
370 goto out;
371
372 /* parse mount options */
373 while ((c = strsep(&options, ",")) != NULL) {
374 int token, intval, ret;
375 if (!*c)
376 continue;
377 err = -EINVAL;
378 token = match_token((char *)c, opt_tokens, argstr);
379 if (token < 0 && parse_extra_token) {
380 /* extra? */
381 err = parse_extra_token((char *)c, private);
382 if (err < 0) {
383 pr_err("bad option at '%s'\n", c);
384 goto out;
385 }
386 continue;
387 }
388 if (token < Opt_last_int) {
389 ret = match_int(&argstr[0], &intval);
390 if (ret < 0) {
391 pr_err("bad mount option arg (not int) "
392 "at '%s'\n", c);
393 continue;
394 }
395 dout("got int token %d val %d\n", token, intval);
396 } else if (token > Opt_last_int && token < Opt_last_string) {
397 dout("got string token %d val %s\n", token,
398 argstr[0].from);
399 } else {
400 dout("got token %d\n", token);
401 }
402 switch (token) {
403 case Opt_ip:
404 err = ceph_parse_ips(argstr[0].from,
405 argstr[0].to,
406 &opt->my_addr,
407 1, NULL);
408 if (err < 0)
409 goto out;
410 opt->flags |= CEPH_OPT_MYIP;
411 break;
412
413 case Opt_fsid:
414 err = parse_fsid(argstr[0].from, &opt->fsid);
415 if (err == 0)
416 opt->flags |= CEPH_OPT_FSID;
417 break;
418 case Opt_name:
419 opt->name = kstrndup(argstr[0].from,
420 argstr[0].to-argstr[0].from,
421 GFP_KERNEL);
422 break;
423 case Opt_secret:
424 opt->key = kzalloc(sizeof(*opt->key), GFP_KERNEL);
425 if (!opt->key) {
426 err = -ENOMEM;
427 goto out;
428 }
429 err = ceph_crypto_key_unarmor(opt->key, argstr[0].from);
430 if (err < 0)
431 goto out;
432 break;
433 case Opt_key:
434 opt->key = kzalloc(sizeof(*opt->key), GFP_KERNEL);
435 if (!opt->key) {
436 err = -ENOMEM;
437 goto out;
438 }
439 err = get_secret(opt->key, argstr[0].from);
440 if (err < 0)
441 goto out;
442 break;
443
444 /* misc */
445 case Opt_osdtimeout:
446 pr_warn("ignoring deprecated osdtimeout option\n");
447 break;
448 case Opt_osdkeepalivetimeout:
449 /* 0 isn't well defined right now, reject it */
450 if (intval < 1 || intval > INT_MAX / 1000) {
451 pr_err("osdkeepalive out of range\n");
452 err = -EINVAL;
453 goto out;
454 }
455 opt->osd_keepalive_timeout =
456 msecs_to_jiffies(intval * 1000);
457 break;
458 case Opt_osd_idle_ttl:
459 /* 0 isn't well defined right now, reject it */
460 if (intval < 1 || intval > INT_MAX / 1000) {
461 pr_err("osd_idle_ttl out of range\n");
462 err = -EINVAL;
463 goto out;
464 }
465 opt->osd_idle_ttl = msecs_to_jiffies(intval * 1000);
466 break;
467 case Opt_mount_timeout:
468 /* 0 is "wait forever" (i.e. infinite timeout) */
469 if (intval < 0 || intval > INT_MAX / 1000) {
470 pr_err("mount_timeout out of range\n");
471 err = -EINVAL;
472 goto out;
473 }
474 opt->mount_timeout = msecs_to_jiffies(intval * 1000);
475 break;
476
477 case Opt_share:
478 opt->flags &= ~CEPH_OPT_NOSHARE;
479 break;
480 case Opt_noshare:
481 opt->flags |= CEPH_OPT_NOSHARE;
482 break;
483
484 case Opt_crc:
485 opt->flags &= ~CEPH_OPT_NOCRC;
486 break;
487 case Opt_nocrc:
488 opt->flags |= CEPH_OPT_NOCRC;
489 break;
490
491 case Opt_cephx_require_signatures:
492 opt->flags &= ~CEPH_OPT_NOMSGAUTH;
493 break;
494 case Opt_nocephx_require_signatures:
495 opt->flags |= CEPH_OPT_NOMSGAUTH;
496 break;
497 case Opt_cephx_sign_messages:
498 opt->flags &= ~CEPH_OPT_NOMSGSIGN;
499 break;
500 case Opt_nocephx_sign_messages:
501 opt->flags |= CEPH_OPT_NOMSGSIGN;
502 break;
503
504 case Opt_tcp_nodelay:
505 opt->flags |= CEPH_OPT_TCP_NODELAY;
506 break;
507 case Opt_notcp_nodelay:
508 opt->flags &= ~CEPH_OPT_TCP_NODELAY;
509 break;
510
511 default:
512 BUG_ON(token);
513 }
514 }
515
516 /* success */
517 return opt;
518
519out:
520 ceph_destroy_options(opt);
521 return ERR_PTR(err);
522}
523EXPORT_SYMBOL(ceph_parse_options);
524
525int ceph_print_client_options(struct seq_file *m, struct ceph_client *client)
526{
527 struct ceph_options *opt = client->options;
528 size_t pos = m->count;
529
530 if (opt->name) {
531 seq_puts(m, "name=");
532 seq_escape(m, opt->name, ", \t\n\\");
533 seq_putc(m, ',');
534 }
535 if (opt->key)
536 seq_puts(m, "secret=<hidden>,");
537
538 if (opt->flags & CEPH_OPT_FSID)
539 seq_printf(m, "fsid=%pU,", &opt->fsid);
540 if (opt->flags & CEPH_OPT_NOSHARE)
541 seq_puts(m, "noshare,");
542 if (opt->flags & CEPH_OPT_NOCRC)
543 seq_puts(m, "nocrc,");
544 if (opt->flags & CEPH_OPT_NOMSGAUTH)
545 seq_puts(m, "nocephx_require_signatures,");
546 if (opt->flags & CEPH_OPT_NOMSGSIGN)
547 seq_puts(m, "nocephx_sign_messages,");
548 if ((opt->flags & CEPH_OPT_TCP_NODELAY) == 0)
549 seq_puts(m, "notcp_nodelay,");
550
551 if (opt->mount_timeout != CEPH_MOUNT_TIMEOUT_DEFAULT)
552 seq_printf(m, "mount_timeout=%d,",
553 jiffies_to_msecs(opt->mount_timeout) / 1000);
554 if (opt->osd_idle_ttl != CEPH_OSD_IDLE_TTL_DEFAULT)
555 seq_printf(m, "osd_idle_ttl=%d,",
556 jiffies_to_msecs(opt->osd_idle_ttl) / 1000);
557 if (opt->osd_keepalive_timeout != CEPH_OSD_KEEPALIVE_DEFAULT)
558 seq_printf(m, "osdkeepalivetimeout=%d,",
559 jiffies_to_msecs(opt->osd_keepalive_timeout) / 1000);
560
561 /* drop redundant comma */
562 if (m->count != pos)
563 m->count--;
564
565 return 0;
566}
567EXPORT_SYMBOL(ceph_print_client_options);
568
569u64 ceph_client_id(struct ceph_client *client)
570{
571 return client->monc.auth->global_id;
572}
573EXPORT_SYMBOL(ceph_client_id);
574
575/*
576 * create a fresh client instance
577 */
578struct ceph_client *ceph_create_client(struct ceph_options *opt, void *private,
579 u64 supported_features,
580 u64 required_features)
581{
582 struct ceph_client *client;
583 struct ceph_entity_addr *myaddr = NULL;
584 int err = -ENOMEM;
585
586 client = kzalloc(sizeof(*client), GFP_KERNEL);
587 if (client == NULL)
588 return ERR_PTR(-ENOMEM);
589
590 client->private = private;
591 client->options = opt;
592
593 mutex_init(&client->mount_mutex);
594 init_waitqueue_head(&client->auth_wq);
595 client->auth_err = 0;
596
597 if (!ceph_test_opt(client, NOMSGAUTH))
598 required_features |= CEPH_FEATURE_MSG_AUTH;
599
600 client->extra_mon_dispatch = NULL;
601 client->supported_features = CEPH_FEATURES_SUPPORTED_DEFAULT |
602 supported_features;
603 client->required_features = CEPH_FEATURES_REQUIRED_DEFAULT |
604 required_features;
605
606 /* msgr */
607 if (ceph_test_opt(client, MYIP))
608 myaddr = &client->options->my_addr;
609
610 ceph_messenger_init(&client->msgr, myaddr);
611
612 /* subsystems */
613 err = ceph_monc_init(&client->monc, client);
614 if (err < 0)
615 goto fail;
616 err = ceph_osdc_init(&client->osdc, client);
617 if (err < 0)
618 goto fail_monc;
619
620 return client;
621
622fail_monc:
623 ceph_monc_stop(&client->monc);
624fail:
625 ceph_messenger_fini(&client->msgr);
626 kfree(client);
627 return ERR_PTR(err);
628}
629EXPORT_SYMBOL(ceph_create_client);
630
631void ceph_destroy_client(struct ceph_client *client)
632{
633 dout("destroy_client %p\n", client);
634
635 atomic_set(&client->msgr.stopping, 1);
636
637 /* unmount */
638 ceph_osdc_stop(&client->osdc);
639 ceph_monc_stop(&client->monc);
640 ceph_messenger_fini(&client->msgr);
641
642 ceph_debugfs_client_cleanup(client);
643
644 ceph_destroy_options(client->options);
645
646 kfree(client);
647 dout("destroy_client %p done\n", client);
648}
649EXPORT_SYMBOL(ceph_destroy_client);
650
651/*
652 * true if we have the mon map (and have thus joined the cluster)
653 */
654static int have_mon_and_osd_map(struct ceph_client *client)
655{
656 return client->monc.monmap && client->monc.monmap->epoch &&
657 client->osdc.osdmap && client->osdc.osdmap->epoch;
658}
659
660/*
661 * mount: join the ceph cluster, and open root directory.
662 */
663int __ceph_open_session(struct ceph_client *client, unsigned long started)
664{
665 unsigned long timeout = client->options->mount_timeout;
666 long err;
667
668 /* open session, and wait for mon and osd maps */
669 err = ceph_monc_open_session(&client->monc);
670 if (err < 0)
671 return err;
672
673 while (!have_mon_and_osd_map(client)) {
674 if (timeout && time_after_eq(jiffies, started + timeout))
675 return -ETIMEDOUT;
676
677 /* wait */
678 dout("mount waiting for mon_map\n");
679 err = wait_event_interruptible_timeout(client->auth_wq,
680 have_mon_and_osd_map(client) || (client->auth_err < 0),
681 ceph_timeout_jiffies(timeout));
682 if (err < 0)
683 return err;
684 if (client->auth_err < 0)
685 return client->auth_err;
686 }
687
688 pr_info("client%llu fsid %pU\n", ceph_client_id(client), &client->fsid);
689 ceph_debugfs_client_init(client);
690
691 return 0;
692}
693EXPORT_SYMBOL(__ceph_open_session);
694
695
696int ceph_open_session(struct ceph_client *client)
697{
698 int ret;
699 unsigned long started = jiffies; /* note the start time */
700
701 dout("open_session start\n");
702 mutex_lock(&client->mount_mutex);
703
704 ret = __ceph_open_session(client, started);
705
706 mutex_unlock(&client->mount_mutex);
707 return ret;
708}
709EXPORT_SYMBOL(ceph_open_session);
710
711
712static int __init init_ceph_lib(void)
713{
714 int ret = 0;
715
716 ret = ceph_debugfs_init();
717 if (ret < 0)
718 goto out;
719
720 ret = ceph_crypto_init();
721 if (ret < 0)
722 goto out_debugfs;
723
724 ret = ceph_msgr_init();
725 if (ret < 0)
726 goto out_crypto;
727
728 ret = ceph_osdc_setup();
729 if (ret < 0)
730 goto out_msgr;
731
732 pr_info("loaded (mon/osd proto %d/%d)\n",
733 CEPH_MONC_PROTOCOL, CEPH_OSDC_PROTOCOL);
734
735 return 0;
736
737out_msgr:
738 ceph_msgr_exit();
739out_crypto:
740 ceph_crypto_shutdown();
741out_debugfs:
742 ceph_debugfs_cleanup();
743out:
744 return ret;
745}
746
747static void __exit exit_ceph_lib(void)
748{
749 dout("exit_ceph_lib\n");
750 ceph_osdc_cleanup();
751 ceph_msgr_exit();
752 ceph_crypto_shutdown();
753 ceph_debugfs_cleanup();
754}
755
756module_init(init_ceph_lib);
757module_exit(exit_ceph_lib);
758
759MODULE_AUTHOR("Sage Weil <sage@newdream.net>");
760MODULE_AUTHOR("Yehuda Sadeh <yehuda@hq.newdream.net>");
761MODULE_AUTHOR("Patience Warnick <patience@newdream.net>");
762MODULE_DESCRIPTION("Ceph core library");
763MODULE_LICENSE("GPL");