SLUB: Replace __builtin_return_address(0) with _RET_IP_.
[linux-2.6/mini2440.git] / fs / ocfs2 / stack_user.c
blobfaec2d879357dbc62f4d85124e2840d26d439be3
1 /* -*- mode: c; c-basic-offset: 8; -*-
2 * vim: noexpandtab sw=8 ts=8 sts=0:
4 * stack_user.c
6 * Code which interfaces ocfs2 with fs/dlm and a userspace stack.
8 * Copyright (C) 2007 Oracle. All rights reserved.
10 * This program is free software; you can redistribute it and/or
11 * modify it under the terms of the GNU General Public
12 * License as published by the Free Software Foundation, version 2.
14 * This program is distributed in the hope that it will be useful,
15 * but WITHOUT ANY WARRANTY; without even the implied warranty of
16 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
17 * General Public License for more details.
20 #include <linux/module.h>
21 #include <linux/fs.h>
22 #include <linux/miscdevice.h>
23 #include <linux/mutex.h>
24 #include <linux/smp_lock.h>
25 #include <linux/reboot.h>
26 #include <asm/uaccess.h>
28 #include "ocfs2.h" /* For struct ocfs2_lock_res */
29 #include "stackglue.h"
31 #include <linux/dlm_plock.h>
34 * The control protocol starts with a handshake. Until the handshake
35 * is complete, the control device will fail all write(2)s.
37 * The handshake is simple. First, the client reads until EOF. Each line
38 * of output is a supported protocol tag. All protocol tags are a single
39 * character followed by a two hex digit version number. Currently the
40 * only things supported is T01, for "Text-base version 0x01". Next, the
41 * client writes the version they would like to use, including the newline.
42 * Thus, the protocol tag is 'T01\n'. If the version tag written is
43 * unknown, -EINVAL is returned. Once the negotiation is complete, the
44 * client can start sending messages.
46 * The T01 protocol has three messages. First is the "SETN" message.
47 * It has the following syntax:
49 * SETN<space><8-char-hex-nodenum><newline>
51 * This is 14 characters.
53 * The "SETN" message must be the first message following the protocol.
54 * It tells ocfs2_control the local node number.
56 * Next comes the "SETV" message. It has the following syntax:
58 * SETV<space><2-char-hex-major><space><2-char-hex-minor><newline>
60 * This is 11 characters.
62 * The "SETV" message sets the filesystem locking protocol version as
63 * negotiated by the client. The client negotiates based on the maximum
64 * version advertised in /sys/fs/ocfs2/max_locking_protocol. The major
65 * number from the "SETV" message must match
66 * ocfs2_user_plugin.sp_proto->lp_max_version.pv_major, and the minor number
67 * must be less than or equal to ...->lp_max_version.pv_minor.
69 * Once this information has been set, mounts will be allowed. From this
70 * point on, the "DOWN" message can be sent for node down notification.
71 * It has the following syntax:
73 * DOWN<space><32-char-cap-hex-uuid><space><8-char-hex-nodenum><newline>
75 * eg:
77 * DOWN 632A924FDD844190BDA93C0DF6B94899 00000001\n
79 * This is 47 characters.
83 * Whether or not the client has done the handshake.
84 * For now, we have just one protocol version.
86 #define OCFS2_CONTROL_PROTO "T01\n"
87 #define OCFS2_CONTROL_PROTO_LEN 4
89 /* Handshake states */
90 #define OCFS2_CONTROL_HANDSHAKE_INVALID (0)
91 #define OCFS2_CONTROL_HANDSHAKE_READ (1)
92 #define OCFS2_CONTROL_HANDSHAKE_PROTOCOL (2)
93 #define OCFS2_CONTROL_HANDSHAKE_VALID (3)
95 /* Messages */
96 #define OCFS2_CONTROL_MESSAGE_OP_LEN 4
97 #define OCFS2_CONTROL_MESSAGE_SETNODE_OP "SETN"
98 #define OCFS2_CONTROL_MESSAGE_SETNODE_TOTAL_LEN 14
99 #define OCFS2_CONTROL_MESSAGE_SETVERSION_OP "SETV"
100 #define OCFS2_CONTROL_MESSAGE_SETVERSION_TOTAL_LEN 11
101 #define OCFS2_CONTROL_MESSAGE_DOWN_OP "DOWN"
102 #define OCFS2_CONTROL_MESSAGE_DOWN_TOTAL_LEN 47
103 #define OCFS2_TEXT_UUID_LEN 32
104 #define OCFS2_CONTROL_MESSAGE_VERNUM_LEN 2
105 #define OCFS2_CONTROL_MESSAGE_NODENUM_LEN 8
108 * ocfs2_live_connection is refcounted because the filesystem and
109 * miscdevice sides can detach in different order. Let's just be safe.
111 struct ocfs2_live_connection {
112 struct list_head oc_list;
113 struct ocfs2_cluster_connection *oc_conn;
116 struct ocfs2_control_private {
117 struct list_head op_list;
118 int op_state;
119 int op_this_node;
120 struct ocfs2_protocol_version op_proto;
123 /* SETN<space><8-char-hex-nodenum><newline> */
124 struct ocfs2_control_message_setn {
125 char tag[OCFS2_CONTROL_MESSAGE_OP_LEN];
126 char space;
127 char nodestr[OCFS2_CONTROL_MESSAGE_NODENUM_LEN];
128 char newline;
131 /* SETV<space><2-char-hex-major><space><2-char-hex-minor><newline> */
132 struct ocfs2_control_message_setv {
133 char tag[OCFS2_CONTROL_MESSAGE_OP_LEN];
134 char space1;
135 char major[OCFS2_CONTROL_MESSAGE_VERNUM_LEN];
136 char space2;
137 char minor[OCFS2_CONTROL_MESSAGE_VERNUM_LEN];
138 char newline;
141 /* DOWN<space><32-char-cap-hex-uuid><space><8-char-hex-nodenum><newline> */
142 struct ocfs2_control_message_down {
143 char tag[OCFS2_CONTROL_MESSAGE_OP_LEN];
144 char space1;
145 char uuid[OCFS2_TEXT_UUID_LEN];
146 char space2;
147 char nodestr[OCFS2_CONTROL_MESSAGE_NODENUM_LEN];
148 char newline;
151 union ocfs2_control_message {
152 char tag[OCFS2_CONTROL_MESSAGE_OP_LEN];
153 struct ocfs2_control_message_setn u_setn;
154 struct ocfs2_control_message_setv u_setv;
155 struct ocfs2_control_message_down u_down;
158 static struct ocfs2_stack_plugin ocfs2_user_plugin;
160 static atomic_t ocfs2_control_opened;
161 static int ocfs2_control_this_node = -1;
162 static struct ocfs2_protocol_version running_proto;
164 static LIST_HEAD(ocfs2_live_connection_list);
165 static LIST_HEAD(ocfs2_control_private_list);
166 static DEFINE_MUTEX(ocfs2_control_lock);
168 static inline void ocfs2_control_set_handshake_state(struct file *file,
169 int state)
171 struct ocfs2_control_private *p = file->private_data;
172 p->op_state = state;
175 static inline int ocfs2_control_get_handshake_state(struct file *file)
177 struct ocfs2_control_private *p = file->private_data;
178 return p->op_state;
181 static struct ocfs2_live_connection *ocfs2_connection_find(const char *name)
183 size_t len = strlen(name);
184 struct ocfs2_live_connection *c;
186 BUG_ON(!mutex_is_locked(&ocfs2_control_lock));
188 list_for_each_entry(c, &ocfs2_live_connection_list, oc_list) {
189 if ((c->oc_conn->cc_namelen == len) &&
190 !strncmp(c->oc_conn->cc_name, name, len))
191 return c;
194 return c;
198 * ocfs2_live_connection structures are created underneath the ocfs2
199 * mount path. Since the VFS prevents multiple calls to
200 * fill_super(), we can't get dupes here.
202 static int ocfs2_live_connection_new(struct ocfs2_cluster_connection *conn,
203 struct ocfs2_live_connection **c_ret)
205 int rc = 0;
206 struct ocfs2_live_connection *c;
208 c = kzalloc(sizeof(struct ocfs2_live_connection), GFP_KERNEL);
209 if (!c)
210 return -ENOMEM;
212 mutex_lock(&ocfs2_control_lock);
213 c->oc_conn = conn;
215 if (atomic_read(&ocfs2_control_opened))
216 list_add(&c->oc_list, &ocfs2_live_connection_list);
217 else {
218 printk(KERN_ERR
219 "ocfs2: Userspace control daemon is not present\n");
220 rc = -ESRCH;
223 mutex_unlock(&ocfs2_control_lock);
225 if (!rc)
226 *c_ret = c;
227 else
228 kfree(c);
230 return rc;
234 * This function disconnects the cluster connection from ocfs2_control.
235 * Afterwards, userspace can't affect the cluster connection.
237 static void ocfs2_live_connection_drop(struct ocfs2_live_connection *c)
239 mutex_lock(&ocfs2_control_lock);
240 list_del_init(&c->oc_list);
241 c->oc_conn = NULL;
242 mutex_unlock(&ocfs2_control_lock);
244 kfree(c);
247 static int ocfs2_control_cfu(void *target, size_t target_len,
248 const char __user *buf, size_t count)
250 /* The T01 expects write(2) calls to have exactly one command */
251 if ((count != target_len) ||
252 (count > sizeof(union ocfs2_control_message)))
253 return -EINVAL;
255 if (copy_from_user(target, buf, target_len))
256 return -EFAULT;
258 return 0;
261 static ssize_t ocfs2_control_validate_protocol(struct file *file,
262 const char __user *buf,
263 size_t count)
265 ssize_t ret;
266 char kbuf[OCFS2_CONTROL_PROTO_LEN];
268 ret = ocfs2_control_cfu(kbuf, OCFS2_CONTROL_PROTO_LEN,
269 buf, count);
270 if (ret)
271 return ret;
273 if (strncmp(kbuf, OCFS2_CONTROL_PROTO, OCFS2_CONTROL_PROTO_LEN))
274 return -EINVAL;
276 ocfs2_control_set_handshake_state(file,
277 OCFS2_CONTROL_HANDSHAKE_PROTOCOL);
279 return count;
282 static void ocfs2_control_send_down(const char *uuid,
283 int nodenum)
285 struct ocfs2_live_connection *c;
287 mutex_lock(&ocfs2_control_lock);
289 c = ocfs2_connection_find(uuid);
290 if (c) {
291 BUG_ON(c->oc_conn == NULL);
292 c->oc_conn->cc_recovery_handler(nodenum,
293 c->oc_conn->cc_recovery_data);
296 mutex_unlock(&ocfs2_control_lock);
300 * Called whenever configuration elements are sent to /dev/ocfs2_control.
301 * If all configuration elements are present, try to set the global
302 * values. If there is a problem, return an error. Skip any missing
303 * elements, and only bump ocfs2_control_opened when we have all elements
304 * and are successful.
306 static int ocfs2_control_install_private(struct file *file)
308 int rc = 0;
309 int set_p = 1;
310 struct ocfs2_control_private *p = file->private_data;
312 BUG_ON(p->op_state != OCFS2_CONTROL_HANDSHAKE_PROTOCOL);
314 mutex_lock(&ocfs2_control_lock);
316 if (p->op_this_node < 0) {
317 set_p = 0;
318 } else if ((ocfs2_control_this_node >= 0) &&
319 (ocfs2_control_this_node != p->op_this_node)) {
320 rc = -EINVAL;
321 goto out_unlock;
324 if (!p->op_proto.pv_major) {
325 set_p = 0;
326 } else if (!list_empty(&ocfs2_live_connection_list) &&
327 ((running_proto.pv_major != p->op_proto.pv_major) ||
328 (running_proto.pv_minor != p->op_proto.pv_minor))) {
329 rc = -EINVAL;
330 goto out_unlock;
333 if (set_p) {
334 ocfs2_control_this_node = p->op_this_node;
335 running_proto.pv_major = p->op_proto.pv_major;
336 running_proto.pv_minor = p->op_proto.pv_minor;
339 out_unlock:
340 mutex_unlock(&ocfs2_control_lock);
342 if (!rc && set_p) {
343 /* We set the global values successfully */
344 atomic_inc(&ocfs2_control_opened);
345 ocfs2_control_set_handshake_state(file,
346 OCFS2_CONTROL_HANDSHAKE_VALID);
349 return rc;
352 static int ocfs2_control_get_this_node(void)
354 int rc;
356 mutex_lock(&ocfs2_control_lock);
357 if (ocfs2_control_this_node < 0)
358 rc = -EINVAL;
359 else
360 rc = ocfs2_control_this_node;
361 mutex_unlock(&ocfs2_control_lock);
363 return rc;
366 static int ocfs2_control_do_setnode_msg(struct file *file,
367 struct ocfs2_control_message_setn *msg)
369 long nodenum;
370 char *ptr = NULL;
371 struct ocfs2_control_private *p = file->private_data;
373 if (ocfs2_control_get_handshake_state(file) !=
374 OCFS2_CONTROL_HANDSHAKE_PROTOCOL)
375 return -EINVAL;
377 if (strncmp(msg->tag, OCFS2_CONTROL_MESSAGE_SETNODE_OP,
378 OCFS2_CONTROL_MESSAGE_OP_LEN))
379 return -EINVAL;
381 if ((msg->space != ' ') || (msg->newline != '\n'))
382 return -EINVAL;
383 msg->space = msg->newline = '\0';
385 nodenum = simple_strtol(msg->nodestr, &ptr, 16);
386 if (!ptr || *ptr)
387 return -EINVAL;
389 if ((nodenum == LONG_MIN) || (nodenum == LONG_MAX) ||
390 (nodenum > INT_MAX) || (nodenum < 0))
391 return -ERANGE;
392 p->op_this_node = nodenum;
394 return ocfs2_control_install_private(file);
397 static int ocfs2_control_do_setversion_msg(struct file *file,
398 struct ocfs2_control_message_setv *msg)
400 long major, minor;
401 char *ptr = NULL;
402 struct ocfs2_control_private *p = file->private_data;
403 struct ocfs2_protocol_version *max =
404 &ocfs2_user_plugin.sp_proto->lp_max_version;
406 if (ocfs2_control_get_handshake_state(file) !=
407 OCFS2_CONTROL_HANDSHAKE_PROTOCOL)
408 return -EINVAL;
410 if (strncmp(msg->tag, OCFS2_CONTROL_MESSAGE_SETVERSION_OP,
411 OCFS2_CONTROL_MESSAGE_OP_LEN))
412 return -EINVAL;
414 if ((msg->space1 != ' ') || (msg->space2 != ' ') ||
415 (msg->newline != '\n'))
416 return -EINVAL;
417 msg->space1 = msg->space2 = msg->newline = '\0';
419 major = simple_strtol(msg->major, &ptr, 16);
420 if (!ptr || *ptr)
421 return -EINVAL;
422 minor = simple_strtol(msg->minor, &ptr, 16);
423 if (!ptr || *ptr)
424 return -EINVAL;
427 * The major must be between 1 and 255, inclusive. The minor
428 * must be between 0 and 255, inclusive. The version passed in
429 * must be within the maximum version supported by the filesystem.
431 if ((major == LONG_MIN) || (major == LONG_MAX) ||
432 (major > (u8)-1) || (major < 1))
433 return -ERANGE;
434 if ((minor == LONG_MIN) || (minor == LONG_MAX) ||
435 (minor > (u8)-1) || (minor < 0))
436 return -ERANGE;
437 if ((major != max->pv_major) ||
438 (minor > max->pv_minor))
439 return -EINVAL;
441 p->op_proto.pv_major = major;
442 p->op_proto.pv_minor = minor;
444 return ocfs2_control_install_private(file);
447 static int ocfs2_control_do_down_msg(struct file *file,
448 struct ocfs2_control_message_down *msg)
450 long nodenum;
451 char *p = NULL;
453 if (ocfs2_control_get_handshake_state(file) !=
454 OCFS2_CONTROL_HANDSHAKE_VALID)
455 return -EINVAL;
457 if (strncmp(msg->tag, OCFS2_CONTROL_MESSAGE_DOWN_OP,
458 OCFS2_CONTROL_MESSAGE_OP_LEN))
459 return -EINVAL;
461 if ((msg->space1 != ' ') || (msg->space2 != ' ') ||
462 (msg->newline != '\n'))
463 return -EINVAL;
464 msg->space1 = msg->space2 = msg->newline = '\0';
466 nodenum = simple_strtol(msg->nodestr, &p, 16);
467 if (!p || *p)
468 return -EINVAL;
470 if ((nodenum == LONG_MIN) || (nodenum == LONG_MAX) ||
471 (nodenum > INT_MAX) || (nodenum < 0))
472 return -ERANGE;
474 ocfs2_control_send_down(msg->uuid, nodenum);
476 return 0;
479 static ssize_t ocfs2_control_message(struct file *file,
480 const char __user *buf,
481 size_t count)
483 ssize_t ret;
484 union ocfs2_control_message msg;
486 /* Try to catch padding issues */
487 WARN_ON(offsetof(struct ocfs2_control_message_down, uuid) !=
488 (sizeof(msg.u_down.tag) + sizeof(msg.u_down.space1)));
490 memset(&msg, 0, sizeof(union ocfs2_control_message));
491 ret = ocfs2_control_cfu(&msg, count, buf, count);
492 if (ret)
493 goto out;
495 if ((count == OCFS2_CONTROL_MESSAGE_SETNODE_TOTAL_LEN) &&
496 !strncmp(msg.tag, OCFS2_CONTROL_MESSAGE_SETNODE_OP,
497 OCFS2_CONTROL_MESSAGE_OP_LEN))
498 ret = ocfs2_control_do_setnode_msg(file, &msg.u_setn);
499 else if ((count == OCFS2_CONTROL_MESSAGE_SETVERSION_TOTAL_LEN) &&
500 !strncmp(msg.tag, OCFS2_CONTROL_MESSAGE_SETVERSION_OP,
501 OCFS2_CONTROL_MESSAGE_OP_LEN))
502 ret = ocfs2_control_do_setversion_msg(file, &msg.u_setv);
503 else if ((count == OCFS2_CONTROL_MESSAGE_DOWN_TOTAL_LEN) &&
504 !strncmp(msg.tag, OCFS2_CONTROL_MESSAGE_DOWN_OP,
505 OCFS2_CONTROL_MESSAGE_OP_LEN))
506 ret = ocfs2_control_do_down_msg(file, &msg.u_down);
507 else
508 ret = -EINVAL;
510 out:
511 return ret ? ret : count;
514 static ssize_t ocfs2_control_write(struct file *file,
515 const char __user *buf,
516 size_t count,
517 loff_t *ppos)
519 ssize_t ret;
521 switch (ocfs2_control_get_handshake_state(file)) {
522 case OCFS2_CONTROL_HANDSHAKE_INVALID:
523 ret = -EINVAL;
524 break;
526 case OCFS2_CONTROL_HANDSHAKE_READ:
527 ret = ocfs2_control_validate_protocol(file, buf,
528 count);
529 break;
531 case OCFS2_CONTROL_HANDSHAKE_PROTOCOL:
532 case OCFS2_CONTROL_HANDSHAKE_VALID:
533 ret = ocfs2_control_message(file, buf, count);
534 break;
536 default:
537 BUG();
538 ret = -EIO;
539 break;
542 return ret;
546 * This is a naive version. If we ever have a new protocol, we'll expand
547 * it. Probably using seq_file.
549 static ssize_t ocfs2_control_read(struct file *file,
550 char __user *buf,
551 size_t count,
552 loff_t *ppos)
554 ssize_t ret;
556 ret = simple_read_from_buffer(buf, count, ppos,
557 OCFS2_CONTROL_PROTO, OCFS2_CONTROL_PROTO_LEN);
559 /* Have we read the whole protocol list? */
560 if (ret > 0 && *ppos >= OCFS2_CONTROL_PROTO_LEN)
561 ocfs2_control_set_handshake_state(file,
562 OCFS2_CONTROL_HANDSHAKE_READ);
564 return ret;
567 static int ocfs2_control_release(struct inode *inode, struct file *file)
569 struct ocfs2_control_private *p = file->private_data;
571 mutex_lock(&ocfs2_control_lock);
573 if (ocfs2_control_get_handshake_state(file) !=
574 OCFS2_CONTROL_HANDSHAKE_VALID)
575 goto out;
577 if (atomic_dec_and_test(&ocfs2_control_opened)) {
578 if (!list_empty(&ocfs2_live_connection_list)) {
579 /* XXX: Do bad things! */
580 printk(KERN_ERR
581 "ocfs2: Unexpected release of ocfs2_control!\n"
582 " Loss of cluster connection requires "
583 "an emergency restart!\n");
584 emergency_restart();
587 * Last valid close clears the node number and resets
588 * the locking protocol version
590 ocfs2_control_this_node = -1;
591 running_proto.pv_major = 0;
592 running_proto.pv_major = 0;
595 out:
596 list_del_init(&p->op_list);
597 file->private_data = NULL;
599 mutex_unlock(&ocfs2_control_lock);
601 kfree(p);
603 return 0;
606 static int ocfs2_control_open(struct inode *inode, struct file *file)
608 struct ocfs2_control_private *p;
610 p = kzalloc(sizeof(struct ocfs2_control_private), GFP_KERNEL);
611 if (!p)
612 return -ENOMEM;
613 p->op_this_node = -1;
615 lock_kernel();
616 mutex_lock(&ocfs2_control_lock);
617 file->private_data = p;
618 list_add(&p->op_list, &ocfs2_control_private_list);
619 mutex_unlock(&ocfs2_control_lock);
620 unlock_kernel();
622 return 0;
625 static const struct file_operations ocfs2_control_fops = {
626 .open = ocfs2_control_open,
627 .release = ocfs2_control_release,
628 .read = ocfs2_control_read,
629 .write = ocfs2_control_write,
630 .owner = THIS_MODULE,
633 static struct miscdevice ocfs2_control_device = {
634 .minor = MISC_DYNAMIC_MINOR,
635 .name = "ocfs2_control",
636 .fops = &ocfs2_control_fops,
639 static int ocfs2_control_init(void)
641 int rc;
643 atomic_set(&ocfs2_control_opened, 0);
645 rc = misc_register(&ocfs2_control_device);
646 if (rc)
647 printk(KERN_ERR
648 "ocfs2: Unable to register ocfs2_control device "
649 "(errno %d)\n",
650 -rc);
652 return rc;
655 static void ocfs2_control_exit(void)
657 int rc;
659 rc = misc_deregister(&ocfs2_control_device);
660 if (rc)
661 printk(KERN_ERR
662 "ocfs2: Unable to deregister ocfs2_control device "
663 "(errno %d)\n",
664 -rc);
667 static struct dlm_lksb *fsdlm_astarg_to_lksb(void *astarg)
669 struct ocfs2_lock_res *res = astarg;
670 return &res->l_lksb.lksb_fsdlm;
673 static void fsdlm_lock_ast_wrapper(void *astarg)
675 struct dlm_lksb *lksb = fsdlm_astarg_to_lksb(astarg);
676 int status = lksb->sb_status;
678 BUG_ON(ocfs2_user_plugin.sp_proto == NULL);
681 * For now we're punting on the issue of other non-standard errors
682 * where we can't tell if the unlock_ast or lock_ast should be called.
683 * The main "other error" that's possible is EINVAL which means the
684 * function was called with invalid args, which shouldn't be possible
685 * since the caller here is under our control. Other non-standard
686 * errors probably fall into the same category, or otherwise are fatal
687 * which means we can't carry on anyway.
690 if (status == -DLM_EUNLOCK || status == -DLM_ECANCEL)
691 ocfs2_user_plugin.sp_proto->lp_unlock_ast(astarg, 0);
692 else
693 ocfs2_user_plugin.sp_proto->lp_lock_ast(astarg);
696 static void fsdlm_blocking_ast_wrapper(void *astarg, int level)
698 BUG_ON(ocfs2_user_plugin.sp_proto == NULL);
700 ocfs2_user_plugin.sp_proto->lp_blocking_ast(astarg, level);
703 static int user_dlm_lock(struct ocfs2_cluster_connection *conn,
704 int mode,
705 union ocfs2_dlm_lksb *lksb,
706 u32 flags,
707 void *name,
708 unsigned int namelen,
709 void *astarg)
711 int ret;
713 if (!lksb->lksb_fsdlm.sb_lvbptr)
714 lksb->lksb_fsdlm.sb_lvbptr = (char *)lksb +
715 sizeof(struct dlm_lksb);
717 ret = dlm_lock(conn->cc_lockspace, mode, &lksb->lksb_fsdlm,
718 flags|DLM_LKF_NODLCKWT, name, namelen, 0,
719 fsdlm_lock_ast_wrapper, astarg,
720 fsdlm_blocking_ast_wrapper);
721 return ret;
724 static int user_dlm_unlock(struct ocfs2_cluster_connection *conn,
725 union ocfs2_dlm_lksb *lksb,
726 u32 flags,
727 void *astarg)
729 int ret;
731 ret = dlm_unlock(conn->cc_lockspace, lksb->lksb_fsdlm.sb_lkid,
732 flags, &lksb->lksb_fsdlm, astarg);
733 return ret;
736 static int user_dlm_lock_status(union ocfs2_dlm_lksb *lksb)
738 return lksb->lksb_fsdlm.sb_status;
741 static void *user_dlm_lvb(union ocfs2_dlm_lksb *lksb)
743 return (void *)(lksb->lksb_fsdlm.sb_lvbptr);
746 static void user_dlm_dump_lksb(union ocfs2_dlm_lksb *lksb)
750 static int user_plock(struct ocfs2_cluster_connection *conn,
751 u64 ino,
752 struct file *file,
753 int cmd,
754 struct file_lock *fl)
757 * This more or less just demuxes the plock request into any
758 * one of three dlm calls.
760 * Internally, fs/dlm will pass these to a misc device, which
761 * a userspace daemon will read and write to.
763 * For now, cancel requests (which happen internally only),
764 * are turned into unlocks. Most of this function taken from
765 * gfs2_lock.
768 if (cmd == F_CANCELLK) {
769 cmd = F_SETLK;
770 fl->fl_type = F_UNLCK;
773 if (IS_GETLK(cmd))
774 return dlm_posix_get(conn->cc_lockspace, ino, file, fl);
775 else if (fl->fl_type == F_UNLCK)
776 return dlm_posix_unlock(conn->cc_lockspace, ino, file, fl);
777 else
778 return dlm_posix_lock(conn->cc_lockspace, ino, file, cmd, fl);
782 * Compare a requested locking protocol version against the current one.
784 * If the major numbers are different, they are incompatible.
785 * If the current minor is greater than the request, they are incompatible.
786 * If the current minor is less than or equal to the request, they are
787 * compatible, and the requester should run at the current minor version.
789 static int fs_protocol_compare(struct ocfs2_protocol_version *existing,
790 struct ocfs2_protocol_version *request)
792 if (existing->pv_major != request->pv_major)
793 return 1;
795 if (existing->pv_minor > request->pv_minor)
796 return 1;
798 if (existing->pv_minor < request->pv_minor)
799 request->pv_minor = existing->pv_minor;
801 return 0;
804 static int user_cluster_connect(struct ocfs2_cluster_connection *conn)
806 dlm_lockspace_t *fsdlm;
807 struct ocfs2_live_connection *control;
808 int rc = 0;
810 BUG_ON(conn == NULL);
812 rc = ocfs2_live_connection_new(conn, &control);
813 if (rc)
814 goto out;
817 * running_proto must have been set before we allowed any mounts
818 * to proceed.
820 if (fs_protocol_compare(&running_proto, &conn->cc_version)) {
821 printk(KERN_ERR
822 "Unable to mount with fs locking protocol version "
823 "%u.%u because the userspace control daemon has "
824 "negotiated %u.%u\n",
825 conn->cc_version.pv_major, conn->cc_version.pv_minor,
826 running_proto.pv_major, running_proto.pv_minor);
827 rc = -EPROTO;
828 ocfs2_live_connection_drop(control);
829 goto out;
832 rc = dlm_new_lockspace(conn->cc_name, strlen(conn->cc_name),
833 &fsdlm, DLM_LSFL_FS, DLM_LVB_LEN);
834 if (rc) {
835 ocfs2_live_connection_drop(control);
836 goto out;
839 conn->cc_private = control;
840 conn->cc_lockspace = fsdlm;
841 out:
842 return rc;
845 static int user_cluster_disconnect(struct ocfs2_cluster_connection *conn)
847 dlm_release_lockspace(conn->cc_lockspace, 2);
848 conn->cc_lockspace = NULL;
849 ocfs2_live_connection_drop(conn->cc_private);
850 conn->cc_private = NULL;
851 return 0;
854 static int user_cluster_this_node(unsigned int *this_node)
856 int rc;
858 rc = ocfs2_control_get_this_node();
859 if (rc < 0)
860 return rc;
862 *this_node = rc;
863 return 0;
866 static struct ocfs2_stack_operations ocfs2_user_plugin_ops = {
867 .connect = user_cluster_connect,
868 .disconnect = user_cluster_disconnect,
869 .this_node = user_cluster_this_node,
870 .dlm_lock = user_dlm_lock,
871 .dlm_unlock = user_dlm_unlock,
872 .lock_status = user_dlm_lock_status,
873 .lock_lvb = user_dlm_lvb,
874 .plock = user_plock,
875 .dump_lksb = user_dlm_dump_lksb,
878 static struct ocfs2_stack_plugin ocfs2_user_plugin = {
879 .sp_name = "user",
880 .sp_ops = &ocfs2_user_plugin_ops,
881 .sp_owner = THIS_MODULE,
885 static int __init ocfs2_user_plugin_init(void)
887 int rc;
889 rc = ocfs2_control_init();
890 if (!rc) {
891 rc = ocfs2_stack_glue_register(&ocfs2_user_plugin);
892 if (rc)
893 ocfs2_control_exit();
896 return rc;
899 static void __exit ocfs2_user_plugin_exit(void)
901 ocfs2_stack_glue_unregister(&ocfs2_user_plugin);
902 ocfs2_control_exit();
905 MODULE_AUTHOR("Oracle");
906 MODULE_DESCRIPTION("ocfs2 driver for userspace cluster stacks");
907 MODULE_LICENSE("GPL");
908 module_init(ocfs2_user_plugin_init);
909 module_exit(ocfs2_user_plugin_exit);