Loading...
1// SPDX-License-Identifier: GPL-2.0-only
2/*
3 * Copyright (C) 2006-2010 Red Hat, Inc. All rights reserved.
4 */
5
6#include <linux/miscdevice.h>
7#include <linux/init.h>
8#include <linux/wait.h>
9#include <linux/file.h>
10#include <linux/fs.h>
11#include <linux/poll.h>
12#include <linux/signal.h>
13#include <linux/spinlock.h>
14#include <linux/dlm.h>
15#include <linux/dlm_device.h>
16#include <linux/slab.h>
17#include <linux/sched/signal.h>
18
19#include <trace/events/dlm.h>
20
21#include "dlm_internal.h"
22#include "lockspace.h"
23#include "lock.h"
24#include "lvb_table.h"
25#include "user.h"
26#include "ast.h"
27#include "config.h"
28#include "memory.h"
29
30static const char name_prefix[] = "dlm";
31static const struct file_operations device_fops;
32static atomic_t dlm_monitor_opened;
33static int dlm_monitor_unused = 1;
34
35#ifdef CONFIG_COMPAT
36
37struct dlm_lock_params32 {
38 __u8 mode;
39 __u8 namelen;
40 __u16 unused;
41 __u32 flags;
42 __u32 lkid;
43 __u32 parent;
44 __u64 xid;
45 __u64 timeout;
46 __u32 castparam;
47 __u32 castaddr;
48 __u32 bastparam;
49 __u32 bastaddr;
50 __u32 lksb;
51 char lvb[DLM_USER_LVB_LEN];
52 char name[];
53};
54
55struct dlm_write_request32 {
56 __u32 version[3];
57 __u8 cmd;
58 __u8 is64bit;
59 __u8 unused[2];
60
61 union {
62 struct dlm_lock_params32 lock;
63 struct dlm_lspace_params lspace;
64 struct dlm_purge_params purge;
65 } i;
66};
67
68struct dlm_lksb32 {
69 __u32 sb_status;
70 __u32 sb_lkid;
71 __u8 sb_flags;
72 __u32 sb_lvbptr;
73};
74
75struct dlm_lock_result32 {
76 __u32 version[3];
77 __u32 length;
78 __u32 user_astaddr;
79 __u32 user_astparam;
80 __u32 user_lksb;
81 struct dlm_lksb32 lksb;
82 __u8 bast_mode;
83 __u8 unused[3];
84 /* Offsets may be zero if no data is present */
85 __u32 lvb_offset;
86};
87
88static void compat_input(struct dlm_write_request *kb,
89 struct dlm_write_request32 *kb32,
90 int namelen)
91{
92 kb->version[0] = kb32->version[0];
93 kb->version[1] = kb32->version[1];
94 kb->version[2] = kb32->version[2];
95
96 kb->cmd = kb32->cmd;
97 kb->is64bit = kb32->is64bit;
98 if (kb->cmd == DLM_USER_CREATE_LOCKSPACE ||
99 kb->cmd == DLM_USER_REMOVE_LOCKSPACE) {
100 kb->i.lspace.flags = kb32->i.lspace.flags;
101 kb->i.lspace.minor = kb32->i.lspace.minor;
102 memcpy(kb->i.lspace.name, kb32->i.lspace.name, namelen);
103 } else if (kb->cmd == DLM_USER_PURGE) {
104 kb->i.purge.nodeid = kb32->i.purge.nodeid;
105 kb->i.purge.pid = kb32->i.purge.pid;
106 } else {
107 kb->i.lock.mode = kb32->i.lock.mode;
108 kb->i.lock.namelen = kb32->i.lock.namelen;
109 kb->i.lock.flags = kb32->i.lock.flags;
110 kb->i.lock.lkid = kb32->i.lock.lkid;
111 kb->i.lock.parent = kb32->i.lock.parent;
112 kb->i.lock.xid = kb32->i.lock.xid;
113 kb->i.lock.timeout = kb32->i.lock.timeout;
114 kb->i.lock.castparam = (__user void *)(long)kb32->i.lock.castparam;
115 kb->i.lock.castaddr = (__user void *)(long)kb32->i.lock.castaddr;
116 kb->i.lock.bastparam = (__user void *)(long)kb32->i.lock.bastparam;
117 kb->i.lock.bastaddr = (__user void *)(long)kb32->i.lock.bastaddr;
118 kb->i.lock.lksb = (__user void *)(long)kb32->i.lock.lksb;
119 memcpy(kb->i.lock.lvb, kb32->i.lock.lvb, DLM_USER_LVB_LEN);
120 memcpy(kb->i.lock.name, kb32->i.lock.name, namelen);
121 }
122}
123
124static void compat_output(struct dlm_lock_result *res,
125 struct dlm_lock_result32 *res32)
126{
127 memset(res32, 0, sizeof(*res32));
128
129 res32->version[0] = res->version[0];
130 res32->version[1] = res->version[1];
131 res32->version[2] = res->version[2];
132
133 res32->user_astaddr = (__u32)(__force long)res->user_astaddr;
134 res32->user_astparam = (__u32)(__force long)res->user_astparam;
135 res32->user_lksb = (__u32)(__force long)res->user_lksb;
136 res32->bast_mode = res->bast_mode;
137
138 res32->lvb_offset = res->lvb_offset;
139 res32->length = res->length;
140
141 res32->lksb.sb_status = res->lksb.sb_status;
142 res32->lksb.sb_flags = res->lksb.sb_flags;
143 res32->lksb.sb_lkid = res->lksb.sb_lkid;
144 res32->lksb.sb_lvbptr = (__u32)(long)res->lksb.sb_lvbptr;
145}
146#endif
147
148/* Figure out if this lock is at the end of its life and no longer
149 available for the application to use. The lkb still exists until
150 the final ast is read. A lock becomes EOL in three situations:
151 1. a noqueue request fails with EAGAIN
152 2. an unlock completes with EUNLOCK
153 3. a cancel of a waiting request completes with ECANCEL/EDEADLK
154 An EOL lock needs to be removed from the process's list of locks.
155 And we can't allow any new operation on an EOL lock. This is
156 not related to the lifetime of the lkb struct which is managed
157 entirely by refcount. */
158
159static int lkb_is_endoflife(int mode, int status)
160{
161 switch (status) {
162 case -DLM_EUNLOCK:
163 return 1;
164 case -DLM_ECANCEL:
165 case -ETIMEDOUT:
166 case -EDEADLK:
167 case -EAGAIN:
168 if (mode == DLM_LOCK_IV)
169 return 1;
170 break;
171 }
172 return 0;
173}
174
175/* we could possibly check if the cancel of an orphan has resulted in the lkb
176 being removed and then remove that lkb from the orphans list and free it */
177
178void dlm_user_add_ast(struct dlm_lkb *lkb, uint32_t flags, int mode,
179 int status, uint32_t sbflags)
180{
181 struct dlm_ls *ls;
182 struct dlm_user_args *ua;
183 struct dlm_user_proc *proc;
184 struct dlm_callback *cb;
185 int rv, copy_lvb;
186
187 if (test_bit(DLM_DFL_ORPHAN_BIT, &lkb->lkb_dflags) ||
188 test_bit(DLM_IFL_DEAD_BIT, &lkb->lkb_iflags))
189 return;
190
191 ls = lkb->lkb_resource->res_ls;
192 spin_lock_bh(&ls->ls_clear_proc_locks);
193
194 /* If ORPHAN/DEAD flag is set, it means the process is dead so an ast
195 can't be delivered. For ORPHAN's, dlm_clear_proc_locks() freed
196 lkb->ua so we can't try to use it. This second check is necessary
197 for cases where a completion ast is received for an operation that
198 began before clear_proc_locks did its cancel/unlock. */
199
200 if (test_bit(DLM_DFL_ORPHAN_BIT, &lkb->lkb_dflags) ||
201 test_bit(DLM_IFL_DEAD_BIT, &lkb->lkb_iflags))
202 goto out;
203
204 DLM_ASSERT(lkb->lkb_ua, dlm_print_lkb(lkb););
205 ua = lkb->lkb_ua;
206 proc = ua->proc;
207
208 if ((flags & DLM_CB_BAST) && ua->bastaddr == NULL)
209 goto out;
210
211 if ((flags & DLM_CB_CAST) && lkb_is_endoflife(mode, status))
212 set_bit(DLM_IFL_ENDOFLIFE_BIT, &lkb->lkb_iflags);
213
214 spin_lock_bh(&proc->asts_spin);
215
216 if (!dlm_may_skip_callback(lkb, flags, mode, status, sbflags,
217 ©_lvb)) {
218 rv = dlm_get_cb(lkb, flags, mode, status, sbflags, &cb);
219 if (!rv) {
220 cb->copy_lvb = copy_lvb;
221 cb->ua = *ua;
222 cb->lkb_lksb = &cb->ua.lksb;
223 if (copy_lvb) {
224 memcpy(cb->lvbptr, ua->lksb.sb_lvbptr,
225 DLM_USER_LVB_LEN);
226 cb->lkb_lksb->sb_lvbptr = cb->lvbptr;
227 }
228
229 list_add_tail(&cb->list, &proc->asts);
230 wake_up_interruptible(&proc->wait);
231 }
232 }
233 spin_unlock_bh(&proc->asts_spin);
234
235 if (test_bit(DLM_IFL_ENDOFLIFE_BIT, &lkb->lkb_iflags)) {
236 /* N.B. spin_lock locks_spin, not asts_spin */
237 spin_lock_bh(&proc->locks_spin);
238 if (!list_empty(&lkb->lkb_ownqueue)) {
239 list_del_init(&lkb->lkb_ownqueue);
240 dlm_put_lkb(lkb);
241 }
242 spin_unlock_bh(&proc->locks_spin);
243 }
244 out:
245 spin_unlock_bh(&ls->ls_clear_proc_locks);
246}
247
248static int device_user_lock(struct dlm_user_proc *proc,
249 struct dlm_lock_params *params)
250{
251 struct dlm_ls *ls;
252 struct dlm_user_args *ua;
253 uint32_t lkid;
254 int error = -ENOMEM;
255
256 ls = dlm_find_lockspace_local(proc->lockspace);
257 if (!ls)
258 return -ENOENT;
259
260 if (!params->castaddr || !params->lksb) {
261 error = -EINVAL;
262 goto out;
263 }
264
265 ua = kzalloc(sizeof(struct dlm_user_args), GFP_NOFS);
266 if (!ua)
267 goto out;
268 ua->proc = proc;
269 ua->user_lksb = params->lksb;
270 ua->castparam = params->castparam;
271 ua->castaddr = params->castaddr;
272 ua->bastparam = params->bastparam;
273 ua->bastaddr = params->bastaddr;
274 ua->xid = params->xid;
275
276 if (params->flags & DLM_LKF_CONVERT) {
277 error = dlm_user_convert(ls, ua,
278 params->mode, params->flags,
279 params->lkid, params->lvb);
280 } else if (params->flags & DLM_LKF_ORPHAN) {
281 error = dlm_user_adopt_orphan(ls, ua,
282 params->mode, params->flags,
283 params->name, params->namelen,
284 &lkid);
285 if (!error)
286 error = lkid;
287 } else {
288 error = dlm_user_request(ls, ua,
289 params->mode, params->flags,
290 params->name, params->namelen);
291 if (!error)
292 error = ua->lksb.sb_lkid;
293 }
294 out:
295 dlm_put_lockspace(ls);
296 return error;
297}
298
299static int device_user_unlock(struct dlm_user_proc *proc,
300 struct dlm_lock_params *params)
301{
302 struct dlm_ls *ls;
303 struct dlm_user_args *ua;
304 int error = -ENOMEM;
305
306 ls = dlm_find_lockspace_local(proc->lockspace);
307 if (!ls)
308 return -ENOENT;
309
310 ua = kzalloc(sizeof(struct dlm_user_args), GFP_NOFS);
311 if (!ua)
312 goto out;
313 ua->proc = proc;
314 ua->user_lksb = params->lksb;
315 ua->castparam = params->castparam;
316 ua->castaddr = params->castaddr;
317
318 if (params->flags & DLM_LKF_CANCEL)
319 error = dlm_user_cancel(ls, ua, params->flags, params->lkid);
320 else
321 error = dlm_user_unlock(ls, ua, params->flags, params->lkid,
322 params->lvb);
323 out:
324 dlm_put_lockspace(ls);
325 return error;
326}
327
328static int device_user_deadlock(struct dlm_user_proc *proc,
329 struct dlm_lock_params *params)
330{
331 struct dlm_ls *ls;
332 int error;
333
334 ls = dlm_find_lockspace_local(proc->lockspace);
335 if (!ls)
336 return -ENOENT;
337
338 error = dlm_user_deadlock(ls, params->flags, params->lkid);
339
340 dlm_put_lockspace(ls);
341 return error;
342}
343
344static int dlm_device_register(struct dlm_ls *ls, char *name)
345{
346 int error, len;
347
348 /* The device is already registered. This happens when the
349 lockspace is created multiple times from userspace. */
350 if (ls->ls_device.name)
351 return 0;
352
353 error = -ENOMEM;
354 len = strlen(name) + strlen(name_prefix) + 2;
355 ls->ls_device.name = kzalloc(len, GFP_NOFS);
356 if (!ls->ls_device.name)
357 goto fail;
358
359 snprintf((char *)ls->ls_device.name, len, "%s_%s", name_prefix,
360 name);
361 ls->ls_device.fops = &device_fops;
362 ls->ls_device.minor = MISC_DYNAMIC_MINOR;
363
364 error = misc_register(&ls->ls_device);
365 if (error) {
366 kfree(ls->ls_device.name);
367 /* this has to be set to NULL
368 * to avoid a double-free in dlm_device_deregister
369 */
370 ls->ls_device.name = NULL;
371 }
372fail:
373 return error;
374}
375
376int dlm_device_deregister(struct dlm_ls *ls)
377{
378 /* The device is not registered. This happens when the lockspace
379 was never used from userspace, or when device_create_lockspace()
380 calls dlm_release_lockspace() after the register fails. */
381 if (!ls->ls_device.name)
382 return 0;
383
384 misc_deregister(&ls->ls_device);
385 kfree(ls->ls_device.name);
386 return 0;
387}
388
389static int device_user_purge(struct dlm_user_proc *proc,
390 struct dlm_purge_params *params)
391{
392 struct dlm_ls *ls;
393 int error;
394
395 ls = dlm_find_lockspace_local(proc->lockspace);
396 if (!ls)
397 return -ENOENT;
398
399 error = dlm_user_purge(ls, proc, params->nodeid, params->pid);
400
401 dlm_put_lockspace(ls);
402 return error;
403}
404
405static int device_create_lockspace(struct dlm_lspace_params *params)
406{
407 dlm_lockspace_t *lockspace;
408 struct dlm_ls *ls;
409 int error;
410
411 if (!capable(CAP_SYS_ADMIN))
412 return -EPERM;
413
414 error = dlm_new_user_lockspace(params->name, dlm_config.ci_cluster_name,
415 params->flags, DLM_USER_LVB_LEN, NULL,
416 NULL, NULL, &lockspace);
417 if (error)
418 return error;
419
420 ls = dlm_find_lockspace_local(lockspace);
421 if (!ls)
422 return -ENOENT;
423
424 error = dlm_device_register(ls, params->name);
425 dlm_put_lockspace(ls);
426
427 if (error)
428 dlm_release_lockspace(lockspace, 0);
429 else
430 error = ls->ls_device.minor;
431
432 return error;
433}
434
435static int device_remove_lockspace(struct dlm_lspace_params *params)
436{
437 dlm_lockspace_t *lockspace;
438 struct dlm_ls *ls;
439 int error, force = 0;
440
441 if (!capable(CAP_SYS_ADMIN))
442 return -EPERM;
443
444 ls = dlm_find_lockspace_device(params->minor);
445 if (!ls)
446 return -ENOENT;
447
448 if (params->flags & DLM_USER_LSFLG_FORCEFREE)
449 force = 2;
450
451 lockspace = ls;
452 dlm_put_lockspace(ls);
453
454 /* The final dlm_release_lockspace waits for references to go to
455 zero, so all processes will need to close their device for the
456 ls before the release will proceed. release also calls the
457 device_deregister above. Converting a positive return value
458 from release to zero means that userspace won't know when its
459 release was the final one, but it shouldn't need to know. */
460
461 error = dlm_release_lockspace(lockspace, force);
462 if (error > 0)
463 error = 0;
464 return error;
465}
466
467/* Check the user's version matches ours */
468static int check_version(struct dlm_write_request *req)
469{
470 if (req->version[0] != DLM_DEVICE_VERSION_MAJOR ||
471 (req->version[0] == DLM_DEVICE_VERSION_MAJOR &&
472 req->version[1] > DLM_DEVICE_VERSION_MINOR)) {
473
474 printk(KERN_DEBUG "dlm: process %s (%d) version mismatch "
475 "user (%d.%d.%d) kernel (%d.%d.%d)\n",
476 current->comm,
477 task_pid_nr(current),
478 req->version[0],
479 req->version[1],
480 req->version[2],
481 DLM_DEVICE_VERSION_MAJOR,
482 DLM_DEVICE_VERSION_MINOR,
483 DLM_DEVICE_VERSION_PATCH);
484 return -EINVAL;
485 }
486 return 0;
487}
488
489/*
490 * device_write
491 *
492 * device_user_lock
493 * dlm_user_request -> request_lock
494 * dlm_user_convert -> convert_lock
495 *
496 * device_user_unlock
497 * dlm_user_unlock -> unlock_lock
498 * dlm_user_cancel -> cancel_lock
499 *
500 * device_create_lockspace
501 * dlm_new_lockspace
502 *
503 * device_remove_lockspace
504 * dlm_release_lockspace
505 */
506
507/* a write to a lockspace device is a lock or unlock request, a write
508 to the control device is to create/remove a lockspace */
509
510static ssize_t device_write(struct file *file, const char __user *buf,
511 size_t count, loff_t *ppos)
512{
513 struct dlm_user_proc *proc = file->private_data;
514 struct dlm_write_request *kbuf;
515 int error;
516
517#ifdef CONFIG_COMPAT
518 if (count < sizeof(struct dlm_write_request32))
519#else
520 if (count < sizeof(struct dlm_write_request))
521#endif
522 return -EINVAL;
523
524 /*
525 * can't compare against COMPAT/dlm_write_request32 because
526 * we don't yet know if is64bit is zero
527 */
528 if (count > sizeof(struct dlm_write_request) + DLM_RESNAME_MAXLEN)
529 return -EINVAL;
530
531 kbuf = memdup_user_nul(buf, count);
532 if (IS_ERR(kbuf))
533 return PTR_ERR(kbuf);
534
535 if (check_version(kbuf)) {
536 error = -EBADE;
537 goto out_free;
538 }
539
540#ifdef CONFIG_COMPAT
541 if (!kbuf->is64bit) {
542 struct dlm_write_request32 *k32buf;
543 int namelen = 0;
544
545 if (count > sizeof(struct dlm_write_request32))
546 namelen = count - sizeof(struct dlm_write_request32);
547
548 k32buf = (struct dlm_write_request32 *)kbuf;
549
550 /* add 1 after namelen so that the name string is terminated */
551 kbuf = kzalloc(sizeof(struct dlm_write_request) + namelen + 1,
552 GFP_NOFS);
553 if (!kbuf) {
554 kfree(k32buf);
555 return -ENOMEM;
556 }
557
558 if (proc)
559 set_bit(DLM_PROC_FLAGS_COMPAT, &proc->flags);
560
561 compat_input(kbuf, k32buf, namelen);
562 kfree(k32buf);
563 }
564#endif
565
566 /* do we really need this? can a write happen after a close? */
567 if ((kbuf->cmd == DLM_USER_LOCK || kbuf->cmd == DLM_USER_UNLOCK) &&
568 (proc && test_bit(DLM_PROC_FLAGS_CLOSING, &proc->flags))) {
569 error = -EINVAL;
570 goto out_free;
571 }
572
573 error = -EINVAL;
574
575 switch (kbuf->cmd)
576 {
577 case DLM_USER_LOCK:
578 if (!proc) {
579 log_print("no locking on control device");
580 goto out_free;
581 }
582 error = device_user_lock(proc, &kbuf->i.lock);
583 break;
584
585 case DLM_USER_UNLOCK:
586 if (!proc) {
587 log_print("no locking on control device");
588 goto out_free;
589 }
590 error = device_user_unlock(proc, &kbuf->i.lock);
591 break;
592
593 case DLM_USER_DEADLOCK:
594 if (!proc) {
595 log_print("no locking on control device");
596 goto out_free;
597 }
598 error = device_user_deadlock(proc, &kbuf->i.lock);
599 break;
600
601 case DLM_USER_CREATE_LOCKSPACE:
602 if (proc) {
603 log_print("create/remove only on control device");
604 goto out_free;
605 }
606 error = device_create_lockspace(&kbuf->i.lspace);
607 break;
608
609 case DLM_USER_REMOVE_LOCKSPACE:
610 if (proc) {
611 log_print("create/remove only on control device");
612 goto out_free;
613 }
614 error = device_remove_lockspace(&kbuf->i.lspace);
615 break;
616
617 case DLM_USER_PURGE:
618 if (!proc) {
619 log_print("no locking on control device");
620 goto out_free;
621 }
622 error = device_user_purge(proc, &kbuf->i.purge);
623 break;
624
625 default:
626 log_print("Unknown command passed to DLM device : %d\n",
627 kbuf->cmd);
628 }
629
630 out_free:
631 kfree(kbuf);
632 return error;
633}
634
635/* Every process that opens the lockspace device has its own "proc" structure
636 hanging off the open file that's used to keep track of locks owned by the
637 process and asts that need to be delivered to the process. */
638
639static int device_open(struct inode *inode, struct file *file)
640{
641 struct dlm_user_proc *proc;
642 struct dlm_ls *ls;
643
644 ls = dlm_find_lockspace_device(iminor(inode));
645 if (!ls)
646 return -ENOENT;
647
648 proc = kzalloc(sizeof(struct dlm_user_proc), GFP_NOFS);
649 if (!proc) {
650 dlm_put_lockspace(ls);
651 return -ENOMEM;
652 }
653
654 proc->lockspace = ls;
655 INIT_LIST_HEAD(&proc->asts);
656 INIT_LIST_HEAD(&proc->locks);
657 INIT_LIST_HEAD(&proc->unlocking);
658 spin_lock_init(&proc->asts_spin);
659 spin_lock_init(&proc->locks_spin);
660 init_waitqueue_head(&proc->wait);
661 file->private_data = proc;
662
663 return 0;
664}
665
666static int device_close(struct inode *inode, struct file *file)
667{
668 struct dlm_user_proc *proc = file->private_data;
669 struct dlm_ls *ls;
670
671 ls = dlm_find_lockspace_local(proc->lockspace);
672 if (!ls)
673 return -ENOENT;
674
675 set_bit(DLM_PROC_FLAGS_CLOSING, &proc->flags);
676
677 dlm_clear_proc_locks(ls, proc);
678
679 /* at this point no more lkb's should exist for this lockspace,
680 so there's no chance of dlm_user_add_ast() being called and
681 looking for lkb->ua->proc */
682
683 kfree(proc);
684 file->private_data = NULL;
685
686 dlm_put_lockspace(ls);
687 dlm_put_lockspace(ls); /* for the find in device_open() */
688
689 /* FIXME: AUTOFREE: if this ls is no longer used do
690 device_remove_lockspace() */
691
692 return 0;
693}
694
695static int copy_result_to_user(struct dlm_user_args *ua, int compat,
696 uint32_t flags, int mode, int copy_lvb,
697 char __user *buf, size_t count)
698{
699#ifdef CONFIG_COMPAT
700 struct dlm_lock_result32 result32;
701#endif
702 struct dlm_lock_result result;
703 void *resultptr;
704 int error=0;
705 int len;
706 int struct_len;
707
708 memset(&result, 0, sizeof(struct dlm_lock_result));
709 result.version[0] = DLM_DEVICE_VERSION_MAJOR;
710 result.version[1] = DLM_DEVICE_VERSION_MINOR;
711 result.version[2] = DLM_DEVICE_VERSION_PATCH;
712 memcpy(&result.lksb, &ua->lksb, offsetof(struct dlm_lksb, sb_lvbptr));
713 result.user_lksb = ua->user_lksb;
714
715 /* FIXME: dlm1 provides for the user's bastparam/addr to not be updated
716 in a conversion unless the conversion is successful. See code
717 in dlm_user_convert() for updating ua from ua_tmp. OpenVMS, though,
718 notes that a new blocking AST address and parameter are set even if
719 the conversion fails, so maybe we should just do that. */
720
721 if (flags & DLM_CB_BAST) {
722 result.user_astaddr = ua->bastaddr;
723 result.user_astparam = ua->bastparam;
724 result.bast_mode = mode;
725 } else {
726 result.user_astaddr = ua->castaddr;
727 result.user_astparam = ua->castparam;
728 }
729
730#ifdef CONFIG_COMPAT
731 if (compat)
732 len = sizeof(struct dlm_lock_result32);
733 else
734#endif
735 len = sizeof(struct dlm_lock_result);
736 struct_len = len;
737
738 /* copy lvb to userspace if there is one, it's been updated, and
739 the user buffer has space for it */
740
741 if (copy_lvb && ua->lksb.sb_lvbptr && count >= len + DLM_USER_LVB_LEN) {
742 if (copy_to_user(buf+len, ua->lksb.sb_lvbptr,
743 DLM_USER_LVB_LEN)) {
744 error = -EFAULT;
745 goto out;
746 }
747
748 result.lvb_offset = len;
749 len += DLM_USER_LVB_LEN;
750 }
751
752 result.length = len;
753 resultptr = &result;
754#ifdef CONFIG_COMPAT
755 if (compat) {
756 compat_output(&result, &result32);
757 resultptr = &result32;
758 }
759#endif
760
761 if (copy_to_user(buf, resultptr, struct_len))
762 error = -EFAULT;
763 else
764 error = len;
765 out:
766 return error;
767}
768
769static int copy_version_to_user(char __user *buf, size_t count)
770{
771 struct dlm_device_version ver;
772
773 memset(&ver, 0, sizeof(struct dlm_device_version));
774 ver.version[0] = DLM_DEVICE_VERSION_MAJOR;
775 ver.version[1] = DLM_DEVICE_VERSION_MINOR;
776 ver.version[2] = DLM_DEVICE_VERSION_PATCH;
777
778 if (copy_to_user(buf, &ver, sizeof(struct dlm_device_version)))
779 return -EFAULT;
780 return sizeof(struct dlm_device_version);
781}
782
783/* a read returns a single ast described in a struct dlm_lock_result */
784
785static ssize_t device_read(struct file *file, char __user *buf, size_t count,
786 loff_t *ppos)
787{
788 struct dlm_user_proc *proc = file->private_data;
789 DECLARE_WAITQUEUE(wait, current);
790 struct dlm_callback *cb;
791 int rv, ret;
792
793 if (count == sizeof(struct dlm_device_version)) {
794 rv = copy_version_to_user(buf, count);
795 return rv;
796 }
797
798 if (!proc) {
799 log_print("non-version read from control device %zu", count);
800 return -EINVAL;
801 }
802
803#ifdef CONFIG_COMPAT
804 if (count < sizeof(struct dlm_lock_result32))
805#else
806 if (count < sizeof(struct dlm_lock_result))
807#endif
808 return -EINVAL;
809
810 /* do we really need this? can a read happen after a close? */
811 if (test_bit(DLM_PROC_FLAGS_CLOSING, &proc->flags))
812 return -EINVAL;
813
814 spin_lock_bh(&proc->asts_spin);
815 if (list_empty(&proc->asts)) {
816 if (file->f_flags & O_NONBLOCK) {
817 spin_unlock_bh(&proc->asts_spin);
818 return -EAGAIN;
819 }
820
821 add_wait_queue(&proc->wait, &wait);
822
823 repeat:
824 set_current_state(TASK_INTERRUPTIBLE);
825 if (list_empty(&proc->asts) && !signal_pending(current)) {
826 spin_unlock_bh(&proc->asts_spin);
827 schedule();
828 spin_lock_bh(&proc->asts_spin);
829 goto repeat;
830 }
831 set_current_state(TASK_RUNNING);
832 remove_wait_queue(&proc->wait, &wait);
833
834 if (signal_pending(current)) {
835 spin_unlock_bh(&proc->asts_spin);
836 return -ERESTARTSYS;
837 }
838 }
839
840 /* if we empty lkb_callbacks, we don't want to unlock the spinlock
841 without removing lkb_cb_list; so empty lkb_cb_list is always
842 consistent with empty lkb_callbacks */
843
844 cb = list_first_entry(&proc->asts, struct dlm_callback, list);
845 list_del(&cb->list);
846 spin_unlock_bh(&proc->asts_spin);
847
848 if (cb->flags & DLM_CB_BAST) {
849 trace_dlm_bast(cb->ls_id, cb->lkb_id, cb->mode, cb->res_name,
850 cb->res_length);
851 } else if (cb->flags & DLM_CB_CAST) {
852 cb->lkb_lksb->sb_status = cb->sb_status;
853 cb->lkb_lksb->sb_flags = cb->sb_flags;
854 trace_dlm_ast(cb->ls_id, cb->lkb_id, cb->sb_status,
855 cb->sb_flags, cb->res_name, cb->res_length);
856 }
857
858 ret = copy_result_to_user(&cb->ua,
859 test_bit(DLM_PROC_FLAGS_COMPAT, &proc->flags),
860 cb->flags, cb->mode, cb->copy_lvb, buf, count);
861 dlm_free_cb(cb);
862 return ret;
863}
864
865static __poll_t device_poll(struct file *file, poll_table *wait)
866{
867 struct dlm_user_proc *proc = file->private_data;
868
869 poll_wait(file, &proc->wait, wait);
870
871 spin_lock_bh(&proc->asts_spin);
872 if (!list_empty(&proc->asts)) {
873 spin_unlock_bh(&proc->asts_spin);
874 return EPOLLIN | EPOLLRDNORM;
875 }
876 spin_unlock_bh(&proc->asts_spin);
877 return 0;
878}
879
880int dlm_user_daemon_available(void)
881{
882 /* dlm_controld hasn't started (or, has started, but not
883 properly populated configfs) */
884
885 if (!dlm_our_nodeid())
886 return 0;
887
888 /* This is to deal with versions of dlm_controld that don't
889 know about the monitor device. We assume that if the
890 dlm_controld was started (above), but the monitor device
891 was never opened, that it's an old version. dlm_controld
892 should open the monitor device before populating configfs. */
893
894 if (dlm_monitor_unused)
895 return 1;
896
897 return atomic_read(&dlm_monitor_opened) ? 1 : 0;
898}
899
900static int ctl_device_open(struct inode *inode, struct file *file)
901{
902 file->private_data = NULL;
903 return 0;
904}
905
906static int ctl_device_close(struct inode *inode, struct file *file)
907{
908 return 0;
909}
910
911static int monitor_device_open(struct inode *inode, struct file *file)
912{
913 atomic_inc(&dlm_monitor_opened);
914 dlm_monitor_unused = 0;
915 return 0;
916}
917
918static int monitor_device_close(struct inode *inode, struct file *file)
919{
920 if (atomic_dec_and_test(&dlm_monitor_opened))
921 dlm_stop_lockspaces();
922 return 0;
923}
924
925static const struct file_operations device_fops = {
926 .open = device_open,
927 .release = device_close,
928 .read = device_read,
929 .write = device_write,
930 .poll = device_poll,
931 .owner = THIS_MODULE,
932 .llseek = noop_llseek,
933};
934
935static const struct file_operations ctl_device_fops = {
936 .open = ctl_device_open,
937 .release = ctl_device_close,
938 .read = device_read,
939 .write = device_write,
940 .owner = THIS_MODULE,
941 .llseek = noop_llseek,
942};
943
944static struct miscdevice ctl_device = {
945 .name = "dlm-control",
946 .fops = &ctl_device_fops,
947 .minor = MISC_DYNAMIC_MINOR,
948};
949
950static const struct file_operations monitor_device_fops = {
951 .open = monitor_device_open,
952 .release = monitor_device_close,
953 .owner = THIS_MODULE,
954 .llseek = noop_llseek,
955};
956
957static struct miscdevice monitor_device = {
958 .name = "dlm-monitor",
959 .fops = &monitor_device_fops,
960 .minor = MISC_DYNAMIC_MINOR,
961};
962
963int __init dlm_user_init(void)
964{
965 int error;
966
967 atomic_set(&dlm_monitor_opened, 0);
968
969 error = misc_register(&ctl_device);
970 if (error) {
971 log_print("misc_register failed for control device");
972 goto out;
973 }
974
975 error = misc_register(&monitor_device);
976 if (error) {
977 log_print("misc_register failed for monitor device");
978 misc_deregister(&ctl_device);
979 }
980 out:
981 return error;
982}
983
984void dlm_user_exit(void)
985{
986 misc_deregister(&ctl_device);
987 misc_deregister(&monitor_device);
988}
989
1// SPDX-License-Identifier: GPL-2.0-only
2/*
3 * Copyright (C) 2006-2010 Red Hat, Inc. All rights reserved.
4 */
5
6#include <linux/miscdevice.h>
7#include <linux/init.h>
8#include <linux/wait.h>
9#include <linux/file.h>
10#include <linux/fs.h>
11#include <linux/poll.h>
12#include <linux/signal.h>
13#include <linux/spinlock.h>
14#include <linux/dlm.h>
15#include <linux/dlm_device.h>
16#include <linux/slab.h>
17#include <linux/sched/signal.h>
18
19#include <trace/events/dlm.h>
20
21#include "dlm_internal.h"
22#include "lockspace.h"
23#include "lock.h"
24#include "lvb_table.h"
25#include "user.h"
26#include "ast.h"
27#include "config.h"
28#include "memory.h"
29
30static const char name_prefix[] = "dlm";
31static const struct file_operations device_fops;
32static atomic_t dlm_monitor_opened;
33static int dlm_monitor_unused = 1;
34
35#ifdef CONFIG_COMPAT
36
37struct dlm_lock_params32 {
38 __u8 mode;
39 __u8 namelen;
40 __u16 unused;
41 __u32 flags;
42 __u32 lkid;
43 __u32 parent;
44 __u64 xid;
45 __u64 timeout;
46 __u32 castparam;
47 __u32 castaddr;
48 __u32 bastparam;
49 __u32 bastaddr;
50 __u32 lksb;
51 char lvb[DLM_USER_LVB_LEN];
52 char name[];
53};
54
55struct dlm_write_request32 {
56 __u32 version[3];
57 __u8 cmd;
58 __u8 is64bit;
59 __u8 unused[2];
60
61 union {
62 struct dlm_lock_params32 lock;
63 struct dlm_lspace_params lspace;
64 struct dlm_purge_params purge;
65 } i;
66};
67
68struct dlm_lksb32 {
69 __u32 sb_status;
70 __u32 sb_lkid;
71 __u8 sb_flags;
72 __u32 sb_lvbptr;
73};
74
75struct dlm_lock_result32 {
76 __u32 version[3];
77 __u32 length;
78 __u32 user_astaddr;
79 __u32 user_astparam;
80 __u32 user_lksb;
81 struct dlm_lksb32 lksb;
82 __u8 bast_mode;
83 __u8 unused[3];
84 /* Offsets may be zero if no data is present */
85 __u32 lvb_offset;
86};
87
88static void compat_input(struct dlm_write_request *kb,
89 struct dlm_write_request32 *kb32,
90 int namelen)
91{
92 kb->version[0] = kb32->version[0];
93 kb->version[1] = kb32->version[1];
94 kb->version[2] = kb32->version[2];
95
96 kb->cmd = kb32->cmd;
97 kb->is64bit = kb32->is64bit;
98 if (kb->cmd == DLM_USER_CREATE_LOCKSPACE ||
99 kb->cmd == DLM_USER_REMOVE_LOCKSPACE) {
100 kb->i.lspace.flags = kb32->i.lspace.flags;
101 kb->i.lspace.minor = kb32->i.lspace.minor;
102 memcpy(kb->i.lspace.name, kb32->i.lspace.name, namelen);
103 } else if (kb->cmd == DLM_USER_PURGE) {
104 kb->i.purge.nodeid = kb32->i.purge.nodeid;
105 kb->i.purge.pid = kb32->i.purge.pid;
106 } else {
107 kb->i.lock.mode = kb32->i.lock.mode;
108 kb->i.lock.namelen = kb32->i.lock.namelen;
109 kb->i.lock.flags = kb32->i.lock.flags;
110 kb->i.lock.lkid = kb32->i.lock.lkid;
111 kb->i.lock.parent = kb32->i.lock.parent;
112 kb->i.lock.xid = kb32->i.lock.xid;
113 kb->i.lock.timeout = kb32->i.lock.timeout;
114 kb->i.lock.castparam = (__user void *)(long)kb32->i.lock.castparam;
115 kb->i.lock.castaddr = (__user void *)(long)kb32->i.lock.castaddr;
116 kb->i.lock.bastparam = (__user void *)(long)kb32->i.lock.bastparam;
117 kb->i.lock.bastaddr = (__user void *)(long)kb32->i.lock.bastaddr;
118 kb->i.lock.lksb = (__user void *)(long)kb32->i.lock.lksb;
119 memcpy(kb->i.lock.lvb, kb32->i.lock.lvb, DLM_USER_LVB_LEN);
120 memcpy(kb->i.lock.name, kb32->i.lock.name, namelen);
121 }
122}
123
124static void compat_output(struct dlm_lock_result *res,
125 struct dlm_lock_result32 *res32)
126{
127 memset(res32, 0, sizeof(*res32));
128
129 res32->version[0] = res->version[0];
130 res32->version[1] = res->version[1];
131 res32->version[2] = res->version[2];
132
133 res32->user_astaddr = (__u32)(__force long)res->user_astaddr;
134 res32->user_astparam = (__u32)(__force long)res->user_astparam;
135 res32->user_lksb = (__u32)(__force long)res->user_lksb;
136 res32->bast_mode = res->bast_mode;
137
138 res32->lvb_offset = res->lvb_offset;
139 res32->length = res->length;
140
141 res32->lksb.sb_status = res->lksb.sb_status;
142 res32->lksb.sb_flags = res->lksb.sb_flags;
143 res32->lksb.sb_lkid = res->lksb.sb_lkid;
144 res32->lksb.sb_lvbptr = (__u32)(long)res->lksb.sb_lvbptr;
145}
146#endif
147
148/* should held proc->asts_spin lock */
149void dlm_purge_lkb_callbacks(struct dlm_lkb *lkb)
150{
151 struct dlm_callback *cb, *safe;
152
153 list_for_each_entry_safe(cb, safe, &lkb->lkb_callbacks, list) {
154 list_del(&cb->list);
155 kref_put(&cb->ref, dlm_release_callback);
156 }
157
158 clear_bit(DLM_IFL_CB_PENDING_BIT, &lkb->lkb_iflags);
159
160 /* invalidate */
161 dlm_callback_set_last_ptr(&lkb->lkb_last_cast, NULL);
162 dlm_callback_set_last_ptr(&lkb->lkb_last_cb, NULL);
163 lkb->lkb_last_bast_mode = -1;
164}
165
166/* Figure out if this lock is at the end of its life and no longer
167 available for the application to use. The lkb still exists until
168 the final ast is read. A lock becomes EOL in three situations:
169 1. a noqueue request fails with EAGAIN
170 2. an unlock completes with EUNLOCK
171 3. a cancel of a waiting request completes with ECANCEL/EDEADLK
172 An EOL lock needs to be removed from the process's list of locks.
173 And we can't allow any new operation on an EOL lock. This is
174 not related to the lifetime of the lkb struct which is managed
175 entirely by refcount. */
176
177static int lkb_is_endoflife(int mode, int status)
178{
179 switch (status) {
180 case -DLM_EUNLOCK:
181 return 1;
182 case -DLM_ECANCEL:
183 case -ETIMEDOUT:
184 case -EDEADLK:
185 case -EAGAIN:
186 if (mode == DLM_LOCK_IV)
187 return 1;
188 break;
189 }
190 return 0;
191}
192
193/* we could possibly check if the cancel of an orphan has resulted in the lkb
194 being removed and then remove that lkb from the orphans list and free it */
195
196void dlm_user_add_ast(struct dlm_lkb *lkb, uint32_t flags, int mode,
197 int status, uint32_t sbflags)
198{
199 struct dlm_ls *ls;
200 struct dlm_user_args *ua;
201 struct dlm_user_proc *proc;
202 int rv;
203
204 if (test_bit(DLM_DFL_ORPHAN_BIT, &lkb->lkb_dflags) ||
205 test_bit(DLM_IFL_DEAD_BIT, &lkb->lkb_iflags))
206 return;
207
208 ls = lkb->lkb_resource->res_ls;
209 spin_lock(&ls->ls_clear_proc_locks);
210
211 /* If ORPHAN/DEAD flag is set, it means the process is dead so an ast
212 can't be delivered. For ORPHAN's, dlm_clear_proc_locks() freed
213 lkb->ua so we can't try to use it. This second check is necessary
214 for cases where a completion ast is received for an operation that
215 began before clear_proc_locks did its cancel/unlock. */
216
217 if (test_bit(DLM_DFL_ORPHAN_BIT, &lkb->lkb_dflags) ||
218 test_bit(DLM_IFL_DEAD_BIT, &lkb->lkb_iflags))
219 goto out;
220
221 DLM_ASSERT(lkb->lkb_ua, dlm_print_lkb(lkb););
222 ua = lkb->lkb_ua;
223 proc = ua->proc;
224
225 if ((flags & DLM_CB_BAST) && ua->bastaddr == NULL)
226 goto out;
227
228 if ((flags & DLM_CB_CAST) && lkb_is_endoflife(mode, status))
229 set_bit(DLM_IFL_ENDOFLIFE_BIT, &lkb->lkb_iflags);
230
231 spin_lock(&proc->asts_spin);
232
233 rv = dlm_enqueue_lkb_callback(lkb, flags, mode, status, sbflags);
234 switch (rv) {
235 case DLM_ENQUEUE_CALLBACK_FAILURE:
236 spin_unlock(&proc->asts_spin);
237 WARN_ON_ONCE(1);
238 goto out;
239 case DLM_ENQUEUE_CALLBACK_NEED_SCHED:
240 kref_get(&lkb->lkb_ref);
241 list_add_tail(&lkb->lkb_cb_list, &proc->asts);
242 wake_up_interruptible(&proc->wait);
243 break;
244 case DLM_ENQUEUE_CALLBACK_SUCCESS:
245 break;
246 default:
247 WARN_ON_ONCE(1);
248 break;
249 }
250 spin_unlock(&proc->asts_spin);
251
252 if (test_bit(DLM_IFL_ENDOFLIFE_BIT, &lkb->lkb_iflags)) {
253 /* N.B. spin_lock locks_spin, not asts_spin */
254 spin_lock(&proc->locks_spin);
255 if (!list_empty(&lkb->lkb_ownqueue)) {
256 list_del_init(&lkb->lkb_ownqueue);
257 dlm_put_lkb(lkb);
258 }
259 spin_unlock(&proc->locks_spin);
260 }
261 out:
262 spin_unlock(&ls->ls_clear_proc_locks);
263}
264
265static int device_user_lock(struct dlm_user_proc *proc,
266 struct dlm_lock_params *params)
267{
268 struct dlm_ls *ls;
269 struct dlm_user_args *ua;
270 uint32_t lkid;
271 int error = -ENOMEM;
272
273 ls = dlm_find_lockspace_local(proc->lockspace);
274 if (!ls)
275 return -ENOENT;
276
277 if (!params->castaddr || !params->lksb) {
278 error = -EINVAL;
279 goto out;
280 }
281
282 ua = kzalloc(sizeof(struct dlm_user_args), GFP_NOFS);
283 if (!ua)
284 goto out;
285 ua->proc = proc;
286 ua->user_lksb = params->lksb;
287 ua->castparam = params->castparam;
288 ua->castaddr = params->castaddr;
289 ua->bastparam = params->bastparam;
290 ua->bastaddr = params->bastaddr;
291 ua->xid = params->xid;
292
293 if (params->flags & DLM_LKF_CONVERT) {
294 error = dlm_user_convert(ls, ua,
295 params->mode, params->flags,
296 params->lkid, params->lvb);
297 } else if (params->flags & DLM_LKF_ORPHAN) {
298 error = dlm_user_adopt_orphan(ls, ua,
299 params->mode, params->flags,
300 params->name, params->namelen,
301 &lkid);
302 if (!error)
303 error = lkid;
304 } else {
305 error = dlm_user_request(ls, ua,
306 params->mode, params->flags,
307 params->name, params->namelen);
308 if (!error)
309 error = ua->lksb.sb_lkid;
310 }
311 out:
312 dlm_put_lockspace(ls);
313 return error;
314}
315
316static int device_user_unlock(struct dlm_user_proc *proc,
317 struct dlm_lock_params *params)
318{
319 struct dlm_ls *ls;
320 struct dlm_user_args *ua;
321 int error = -ENOMEM;
322
323 ls = dlm_find_lockspace_local(proc->lockspace);
324 if (!ls)
325 return -ENOENT;
326
327 ua = kzalloc(sizeof(struct dlm_user_args), GFP_NOFS);
328 if (!ua)
329 goto out;
330 ua->proc = proc;
331 ua->user_lksb = params->lksb;
332 ua->castparam = params->castparam;
333 ua->castaddr = params->castaddr;
334
335 if (params->flags & DLM_LKF_CANCEL)
336 error = dlm_user_cancel(ls, ua, params->flags, params->lkid);
337 else
338 error = dlm_user_unlock(ls, ua, params->flags, params->lkid,
339 params->lvb);
340 out:
341 dlm_put_lockspace(ls);
342 return error;
343}
344
345static int device_user_deadlock(struct dlm_user_proc *proc,
346 struct dlm_lock_params *params)
347{
348 struct dlm_ls *ls;
349 int error;
350
351 ls = dlm_find_lockspace_local(proc->lockspace);
352 if (!ls)
353 return -ENOENT;
354
355 error = dlm_user_deadlock(ls, params->flags, params->lkid);
356
357 dlm_put_lockspace(ls);
358 return error;
359}
360
361static int dlm_device_register(struct dlm_ls *ls, char *name)
362{
363 int error, len;
364
365 /* The device is already registered. This happens when the
366 lockspace is created multiple times from userspace. */
367 if (ls->ls_device.name)
368 return 0;
369
370 error = -ENOMEM;
371 len = strlen(name) + strlen(name_prefix) + 2;
372 ls->ls_device.name = kzalloc(len, GFP_NOFS);
373 if (!ls->ls_device.name)
374 goto fail;
375
376 snprintf((char *)ls->ls_device.name, len, "%s_%s", name_prefix,
377 name);
378 ls->ls_device.fops = &device_fops;
379 ls->ls_device.minor = MISC_DYNAMIC_MINOR;
380
381 error = misc_register(&ls->ls_device);
382 if (error) {
383 kfree(ls->ls_device.name);
384 /* this has to be set to NULL
385 * to avoid a double-free in dlm_device_deregister
386 */
387 ls->ls_device.name = NULL;
388 }
389fail:
390 return error;
391}
392
393int dlm_device_deregister(struct dlm_ls *ls)
394{
395 /* The device is not registered. This happens when the lockspace
396 was never used from userspace, or when device_create_lockspace()
397 calls dlm_release_lockspace() after the register fails. */
398 if (!ls->ls_device.name)
399 return 0;
400
401 misc_deregister(&ls->ls_device);
402 kfree(ls->ls_device.name);
403 return 0;
404}
405
406static int device_user_purge(struct dlm_user_proc *proc,
407 struct dlm_purge_params *params)
408{
409 struct dlm_ls *ls;
410 int error;
411
412 ls = dlm_find_lockspace_local(proc->lockspace);
413 if (!ls)
414 return -ENOENT;
415
416 error = dlm_user_purge(ls, proc, params->nodeid, params->pid);
417
418 dlm_put_lockspace(ls);
419 return error;
420}
421
422static int device_create_lockspace(struct dlm_lspace_params *params)
423{
424 dlm_lockspace_t *lockspace;
425 struct dlm_ls *ls;
426 int error;
427
428 if (!capable(CAP_SYS_ADMIN))
429 return -EPERM;
430
431 error = dlm_new_user_lockspace(params->name, dlm_config.ci_cluster_name,
432 params->flags, DLM_USER_LVB_LEN, NULL,
433 NULL, NULL, &lockspace);
434 if (error)
435 return error;
436
437 ls = dlm_find_lockspace_local(lockspace);
438 if (!ls)
439 return -ENOENT;
440
441 error = dlm_device_register(ls, params->name);
442 dlm_put_lockspace(ls);
443
444 if (error)
445 dlm_release_lockspace(lockspace, 0);
446 else
447 error = ls->ls_device.minor;
448
449 return error;
450}
451
452static int device_remove_lockspace(struct dlm_lspace_params *params)
453{
454 dlm_lockspace_t *lockspace;
455 struct dlm_ls *ls;
456 int error, force = 0;
457
458 if (!capable(CAP_SYS_ADMIN))
459 return -EPERM;
460
461 ls = dlm_find_lockspace_device(params->minor);
462 if (!ls)
463 return -ENOENT;
464
465 if (params->flags & DLM_USER_LSFLG_FORCEFREE)
466 force = 2;
467
468 lockspace = ls->ls_local_handle;
469 dlm_put_lockspace(ls);
470
471 /* The final dlm_release_lockspace waits for references to go to
472 zero, so all processes will need to close their device for the
473 ls before the release will proceed. release also calls the
474 device_deregister above. Converting a positive return value
475 from release to zero means that userspace won't know when its
476 release was the final one, but it shouldn't need to know. */
477
478 error = dlm_release_lockspace(lockspace, force);
479 if (error > 0)
480 error = 0;
481 return error;
482}
483
484/* Check the user's version matches ours */
485static int check_version(struct dlm_write_request *req)
486{
487 if (req->version[0] != DLM_DEVICE_VERSION_MAJOR ||
488 (req->version[0] == DLM_DEVICE_VERSION_MAJOR &&
489 req->version[1] > DLM_DEVICE_VERSION_MINOR)) {
490
491 printk(KERN_DEBUG "dlm: process %s (%d) version mismatch "
492 "user (%d.%d.%d) kernel (%d.%d.%d)\n",
493 current->comm,
494 task_pid_nr(current),
495 req->version[0],
496 req->version[1],
497 req->version[2],
498 DLM_DEVICE_VERSION_MAJOR,
499 DLM_DEVICE_VERSION_MINOR,
500 DLM_DEVICE_VERSION_PATCH);
501 return -EINVAL;
502 }
503 return 0;
504}
505
506/*
507 * device_write
508 *
509 * device_user_lock
510 * dlm_user_request -> request_lock
511 * dlm_user_convert -> convert_lock
512 *
513 * device_user_unlock
514 * dlm_user_unlock -> unlock_lock
515 * dlm_user_cancel -> cancel_lock
516 *
517 * device_create_lockspace
518 * dlm_new_lockspace
519 *
520 * device_remove_lockspace
521 * dlm_release_lockspace
522 */
523
524/* a write to a lockspace device is a lock or unlock request, a write
525 to the control device is to create/remove a lockspace */
526
527static ssize_t device_write(struct file *file, const char __user *buf,
528 size_t count, loff_t *ppos)
529{
530 struct dlm_user_proc *proc = file->private_data;
531 struct dlm_write_request *kbuf;
532 int error;
533
534#ifdef CONFIG_COMPAT
535 if (count < sizeof(struct dlm_write_request32))
536#else
537 if (count < sizeof(struct dlm_write_request))
538#endif
539 return -EINVAL;
540
541 /*
542 * can't compare against COMPAT/dlm_write_request32 because
543 * we don't yet know if is64bit is zero
544 */
545 if (count > sizeof(struct dlm_write_request) + DLM_RESNAME_MAXLEN)
546 return -EINVAL;
547
548 kbuf = memdup_user_nul(buf, count);
549 if (IS_ERR(kbuf))
550 return PTR_ERR(kbuf);
551
552 if (check_version(kbuf)) {
553 error = -EBADE;
554 goto out_free;
555 }
556
557#ifdef CONFIG_COMPAT
558 if (!kbuf->is64bit) {
559 struct dlm_write_request32 *k32buf;
560 int namelen = 0;
561
562 if (count > sizeof(struct dlm_write_request32))
563 namelen = count - sizeof(struct dlm_write_request32);
564
565 k32buf = (struct dlm_write_request32 *)kbuf;
566
567 /* add 1 after namelen so that the name string is terminated */
568 kbuf = kzalloc(sizeof(struct dlm_write_request) + namelen + 1,
569 GFP_NOFS);
570 if (!kbuf) {
571 kfree(k32buf);
572 return -ENOMEM;
573 }
574
575 if (proc)
576 set_bit(DLM_PROC_FLAGS_COMPAT, &proc->flags);
577
578 compat_input(kbuf, k32buf, namelen);
579 kfree(k32buf);
580 }
581#endif
582
583 /* do we really need this? can a write happen after a close? */
584 if ((kbuf->cmd == DLM_USER_LOCK || kbuf->cmd == DLM_USER_UNLOCK) &&
585 (proc && test_bit(DLM_PROC_FLAGS_CLOSING, &proc->flags))) {
586 error = -EINVAL;
587 goto out_free;
588 }
589
590 error = -EINVAL;
591
592 switch (kbuf->cmd)
593 {
594 case DLM_USER_LOCK:
595 if (!proc) {
596 log_print("no locking on control device");
597 goto out_free;
598 }
599 error = device_user_lock(proc, &kbuf->i.lock);
600 break;
601
602 case DLM_USER_UNLOCK:
603 if (!proc) {
604 log_print("no locking on control device");
605 goto out_free;
606 }
607 error = device_user_unlock(proc, &kbuf->i.lock);
608 break;
609
610 case DLM_USER_DEADLOCK:
611 if (!proc) {
612 log_print("no locking on control device");
613 goto out_free;
614 }
615 error = device_user_deadlock(proc, &kbuf->i.lock);
616 break;
617
618 case DLM_USER_CREATE_LOCKSPACE:
619 if (proc) {
620 log_print("create/remove only on control device");
621 goto out_free;
622 }
623 error = device_create_lockspace(&kbuf->i.lspace);
624 break;
625
626 case DLM_USER_REMOVE_LOCKSPACE:
627 if (proc) {
628 log_print("create/remove only on control device");
629 goto out_free;
630 }
631 error = device_remove_lockspace(&kbuf->i.lspace);
632 break;
633
634 case DLM_USER_PURGE:
635 if (!proc) {
636 log_print("no locking on control device");
637 goto out_free;
638 }
639 error = device_user_purge(proc, &kbuf->i.purge);
640 break;
641
642 default:
643 log_print("Unknown command passed to DLM device : %d\n",
644 kbuf->cmd);
645 }
646
647 out_free:
648 kfree(kbuf);
649 return error;
650}
651
652/* Every process that opens the lockspace device has its own "proc" structure
653 hanging off the open file that's used to keep track of locks owned by the
654 process and asts that need to be delivered to the process. */
655
656static int device_open(struct inode *inode, struct file *file)
657{
658 struct dlm_user_proc *proc;
659 struct dlm_ls *ls;
660
661 ls = dlm_find_lockspace_device(iminor(inode));
662 if (!ls)
663 return -ENOENT;
664
665 proc = kzalloc(sizeof(struct dlm_user_proc), GFP_NOFS);
666 if (!proc) {
667 dlm_put_lockspace(ls);
668 return -ENOMEM;
669 }
670
671 proc->lockspace = ls->ls_local_handle;
672 INIT_LIST_HEAD(&proc->asts);
673 INIT_LIST_HEAD(&proc->locks);
674 INIT_LIST_HEAD(&proc->unlocking);
675 spin_lock_init(&proc->asts_spin);
676 spin_lock_init(&proc->locks_spin);
677 init_waitqueue_head(&proc->wait);
678 file->private_data = proc;
679
680 return 0;
681}
682
683static int device_close(struct inode *inode, struct file *file)
684{
685 struct dlm_user_proc *proc = file->private_data;
686 struct dlm_ls *ls;
687
688 ls = dlm_find_lockspace_local(proc->lockspace);
689 if (!ls)
690 return -ENOENT;
691
692 set_bit(DLM_PROC_FLAGS_CLOSING, &proc->flags);
693
694 dlm_clear_proc_locks(ls, proc);
695
696 /* at this point no more lkb's should exist for this lockspace,
697 so there's no chance of dlm_user_add_ast() being called and
698 looking for lkb->ua->proc */
699
700 kfree(proc);
701 file->private_data = NULL;
702
703 dlm_put_lockspace(ls);
704 dlm_put_lockspace(ls); /* for the find in device_open() */
705
706 /* FIXME: AUTOFREE: if this ls is no longer used do
707 device_remove_lockspace() */
708
709 return 0;
710}
711
712static int copy_result_to_user(struct dlm_user_args *ua, int compat,
713 uint32_t flags, int mode, int copy_lvb,
714 char __user *buf, size_t count)
715{
716#ifdef CONFIG_COMPAT
717 struct dlm_lock_result32 result32;
718#endif
719 struct dlm_lock_result result;
720 void *resultptr;
721 int error=0;
722 int len;
723 int struct_len;
724
725 memset(&result, 0, sizeof(struct dlm_lock_result));
726 result.version[0] = DLM_DEVICE_VERSION_MAJOR;
727 result.version[1] = DLM_DEVICE_VERSION_MINOR;
728 result.version[2] = DLM_DEVICE_VERSION_PATCH;
729 memcpy(&result.lksb, &ua->lksb, offsetof(struct dlm_lksb, sb_lvbptr));
730 result.user_lksb = ua->user_lksb;
731
732 /* FIXME: dlm1 provides for the user's bastparam/addr to not be updated
733 in a conversion unless the conversion is successful. See code
734 in dlm_user_convert() for updating ua from ua_tmp. OpenVMS, though,
735 notes that a new blocking AST address and parameter are set even if
736 the conversion fails, so maybe we should just do that. */
737
738 if (flags & DLM_CB_BAST) {
739 result.user_astaddr = ua->bastaddr;
740 result.user_astparam = ua->bastparam;
741 result.bast_mode = mode;
742 } else {
743 result.user_astaddr = ua->castaddr;
744 result.user_astparam = ua->castparam;
745 }
746
747#ifdef CONFIG_COMPAT
748 if (compat)
749 len = sizeof(struct dlm_lock_result32);
750 else
751#endif
752 len = sizeof(struct dlm_lock_result);
753 struct_len = len;
754
755 /* copy lvb to userspace if there is one, it's been updated, and
756 the user buffer has space for it */
757
758 if (copy_lvb && ua->lksb.sb_lvbptr && count >= len + DLM_USER_LVB_LEN) {
759 if (copy_to_user(buf+len, ua->lksb.sb_lvbptr,
760 DLM_USER_LVB_LEN)) {
761 error = -EFAULT;
762 goto out;
763 }
764
765 result.lvb_offset = len;
766 len += DLM_USER_LVB_LEN;
767 }
768
769 result.length = len;
770 resultptr = &result;
771#ifdef CONFIG_COMPAT
772 if (compat) {
773 compat_output(&result, &result32);
774 resultptr = &result32;
775 }
776#endif
777
778 if (copy_to_user(buf, resultptr, struct_len))
779 error = -EFAULT;
780 else
781 error = len;
782 out:
783 return error;
784}
785
786static int copy_version_to_user(char __user *buf, size_t count)
787{
788 struct dlm_device_version ver;
789
790 memset(&ver, 0, sizeof(struct dlm_device_version));
791 ver.version[0] = DLM_DEVICE_VERSION_MAJOR;
792 ver.version[1] = DLM_DEVICE_VERSION_MINOR;
793 ver.version[2] = DLM_DEVICE_VERSION_PATCH;
794
795 if (copy_to_user(buf, &ver, sizeof(struct dlm_device_version)))
796 return -EFAULT;
797 return sizeof(struct dlm_device_version);
798}
799
800/* a read returns a single ast described in a struct dlm_lock_result */
801
802static ssize_t device_read(struct file *file, char __user *buf, size_t count,
803 loff_t *ppos)
804{
805 struct dlm_user_proc *proc = file->private_data;
806 struct dlm_lkb *lkb;
807 DECLARE_WAITQUEUE(wait, current);
808 struct dlm_callback *cb;
809 int rv, copy_lvb = 0;
810 int old_mode, new_mode;
811
812 if (count == sizeof(struct dlm_device_version)) {
813 rv = copy_version_to_user(buf, count);
814 return rv;
815 }
816
817 if (!proc) {
818 log_print("non-version read from control device %zu", count);
819 return -EINVAL;
820 }
821
822#ifdef CONFIG_COMPAT
823 if (count < sizeof(struct dlm_lock_result32))
824#else
825 if (count < sizeof(struct dlm_lock_result))
826#endif
827 return -EINVAL;
828
829 try_another:
830
831 /* do we really need this? can a read happen after a close? */
832 if (test_bit(DLM_PROC_FLAGS_CLOSING, &proc->flags))
833 return -EINVAL;
834
835 spin_lock(&proc->asts_spin);
836 if (list_empty(&proc->asts)) {
837 if (file->f_flags & O_NONBLOCK) {
838 spin_unlock(&proc->asts_spin);
839 return -EAGAIN;
840 }
841
842 add_wait_queue(&proc->wait, &wait);
843
844 repeat:
845 set_current_state(TASK_INTERRUPTIBLE);
846 if (list_empty(&proc->asts) && !signal_pending(current)) {
847 spin_unlock(&proc->asts_spin);
848 schedule();
849 spin_lock(&proc->asts_spin);
850 goto repeat;
851 }
852 set_current_state(TASK_RUNNING);
853 remove_wait_queue(&proc->wait, &wait);
854
855 if (signal_pending(current)) {
856 spin_unlock(&proc->asts_spin);
857 return -ERESTARTSYS;
858 }
859 }
860
861 /* if we empty lkb_callbacks, we don't want to unlock the spinlock
862 without removing lkb_cb_list; so empty lkb_cb_list is always
863 consistent with empty lkb_callbacks */
864
865 lkb = list_first_entry(&proc->asts, struct dlm_lkb, lkb_cb_list);
866
867 /* rem_lkb_callback sets a new lkb_last_cast */
868 old_mode = lkb->lkb_last_cast->mode;
869
870 rv = dlm_dequeue_lkb_callback(lkb, &cb);
871 switch (rv) {
872 case DLM_DEQUEUE_CALLBACK_EMPTY:
873 /* this shouldn't happen; lkb should have been removed from
874 * list when last item was dequeued
875 */
876 log_print("dlm_rem_lkb_callback empty %x", lkb->lkb_id);
877 list_del_init(&lkb->lkb_cb_list);
878 spin_unlock(&proc->asts_spin);
879 /* removes ref for proc->asts, may cause lkb to be freed */
880 dlm_put_lkb(lkb);
881 WARN_ON_ONCE(1);
882 goto try_another;
883 case DLM_DEQUEUE_CALLBACK_LAST:
884 list_del_init(&lkb->lkb_cb_list);
885 clear_bit(DLM_IFL_CB_PENDING_BIT, &lkb->lkb_iflags);
886 break;
887 case DLM_DEQUEUE_CALLBACK_SUCCESS:
888 break;
889 default:
890 WARN_ON_ONCE(1);
891 break;
892 }
893 spin_unlock(&proc->asts_spin);
894
895 if (cb->flags & DLM_CB_BAST) {
896 trace_dlm_bast(lkb->lkb_resource->res_ls, lkb, cb->mode);
897 } else if (cb->flags & DLM_CB_CAST) {
898 new_mode = cb->mode;
899
900 if (!cb->sb_status && lkb->lkb_lksb->sb_lvbptr &&
901 dlm_lvb_operations[old_mode + 1][new_mode + 1])
902 copy_lvb = 1;
903
904 lkb->lkb_lksb->sb_status = cb->sb_status;
905 lkb->lkb_lksb->sb_flags = cb->sb_flags;
906 trace_dlm_ast(lkb->lkb_resource->res_ls, lkb);
907 }
908
909 rv = copy_result_to_user(lkb->lkb_ua,
910 test_bit(DLM_PROC_FLAGS_COMPAT, &proc->flags),
911 cb->flags, cb->mode, copy_lvb, buf, count);
912
913 kref_put(&cb->ref, dlm_release_callback);
914
915 /* removes ref for proc->asts, may cause lkb to be freed */
916 if (rv == DLM_DEQUEUE_CALLBACK_LAST)
917 dlm_put_lkb(lkb);
918
919 return rv;
920}
921
922static __poll_t device_poll(struct file *file, poll_table *wait)
923{
924 struct dlm_user_proc *proc = file->private_data;
925
926 poll_wait(file, &proc->wait, wait);
927
928 spin_lock(&proc->asts_spin);
929 if (!list_empty(&proc->asts)) {
930 spin_unlock(&proc->asts_spin);
931 return EPOLLIN | EPOLLRDNORM;
932 }
933 spin_unlock(&proc->asts_spin);
934 return 0;
935}
936
937int dlm_user_daemon_available(void)
938{
939 /* dlm_controld hasn't started (or, has started, but not
940 properly populated configfs) */
941
942 if (!dlm_our_nodeid())
943 return 0;
944
945 /* This is to deal with versions of dlm_controld that don't
946 know about the monitor device. We assume that if the
947 dlm_controld was started (above), but the monitor device
948 was never opened, that it's an old version. dlm_controld
949 should open the monitor device before populating configfs. */
950
951 if (dlm_monitor_unused)
952 return 1;
953
954 return atomic_read(&dlm_monitor_opened) ? 1 : 0;
955}
956
957static int ctl_device_open(struct inode *inode, struct file *file)
958{
959 file->private_data = NULL;
960 return 0;
961}
962
963static int ctl_device_close(struct inode *inode, struct file *file)
964{
965 return 0;
966}
967
968static int monitor_device_open(struct inode *inode, struct file *file)
969{
970 atomic_inc(&dlm_monitor_opened);
971 dlm_monitor_unused = 0;
972 return 0;
973}
974
975static int monitor_device_close(struct inode *inode, struct file *file)
976{
977 if (atomic_dec_and_test(&dlm_monitor_opened))
978 dlm_stop_lockspaces();
979 return 0;
980}
981
982static const struct file_operations device_fops = {
983 .open = device_open,
984 .release = device_close,
985 .read = device_read,
986 .write = device_write,
987 .poll = device_poll,
988 .owner = THIS_MODULE,
989 .llseek = noop_llseek,
990};
991
992static const struct file_operations ctl_device_fops = {
993 .open = ctl_device_open,
994 .release = ctl_device_close,
995 .read = device_read,
996 .write = device_write,
997 .owner = THIS_MODULE,
998 .llseek = noop_llseek,
999};
1000
1001static struct miscdevice ctl_device = {
1002 .name = "dlm-control",
1003 .fops = &ctl_device_fops,
1004 .minor = MISC_DYNAMIC_MINOR,
1005};
1006
1007static const struct file_operations monitor_device_fops = {
1008 .open = monitor_device_open,
1009 .release = monitor_device_close,
1010 .owner = THIS_MODULE,
1011 .llseek = noop_llseek,
1012};
1013
1014static struct miscdevice monitor_device = {
1015 .name = "dlm-monitor",
1016 .fops = &monitor_device_fops,
1017 .minor = MISC_DYNAMIC_MINOR,
1018};
1019
1020int __init dlm_user_init(void)
1021{
1022 int error;
1023
1024 atomic_set(&dlm_monitor_opened, 0);
1025
1026 error = misc_register(&ctl_device);
1027 if (error) {
1028 log_print("misc_register failed for control device");
1029 goto out;
1030 }
1031
1032 error = misc_register(&monitor_device);
1033 if (error) {
1034 log_print("misc_register failed for monitor device");
1035 misc_deregister(&ctl_device);
1036 }
1037 out:
1038 return error;
1039}
1040
1041void dlm_user_exit(void)
1042{
1043 misc_deregister(&ctl_device);
1044 misc_deregister(&monitor_device);
1045}
1046