user.c 21.4 KB
Newer Older
D
David Teigland 已提交
1
/*
2
 * Copyright (C) 2006-2007 Red Hat, Inc.  All rights reserved.
D
David Teigland 已提交
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
 *
 * This copyrighted material is made available to anyone wishing to use,
 * modify, copy, or redistribute it subject to the terms and conditions
 * of the GNU General Public License v.2.
 */

#include <linux/miscdevice.h>
#include <linux/init.h>
#include <linux/wait.h>
#include <linux/module.h>
#include <linux/file.h>
#include <linux/fs.h>
#include <linux/poll.h>
#include <linux/signal.h>
#include <linux/spinlock.h>
#include <linux/dlm.h>
#include <linux/dlm_device.h>

#include "dlm_internal.h"
#include "lockspace.h"
#include "lock.h"
#include "lvb_table.h"
25
#include "user.h"
D
David Teigland 已提交
26 27 28

static const char *name_prefix="dlm";
static struct miscdevice ctl_device;
29
static const struct file_operations device_fops;
D
David Teigland 已提交
30 31 32 33 34 35

#ifdef CONFIG_COMPAT

struct dlm_lock_params32 {
	__u8 mode;
	__u8 namelen;
36 37
	__u16 unused;
	__u32 flags;
D
David Teigland 已提交
38 39
	__u32 lkid;
	__u32 parent;
40 41
	__u64 xid;
	__u64 timeout;
D
David Teigland 已提交
42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59
	__u32 castparam;
	__u32 castaddr;
	__u32 bastparam;
	__u32 bastaddr;
	__u32 lksb;
	char lvb[DLM_USER_LVB_LEN];
	char name[0];
};

struct dlm_write_request32 {
	__u32 version[3];
	__u8 cmd;
	__u8 is64bit;
	__u8 unused[2];

	union  {
		struct dlm_lock_params32 lock;
		struct dlm_lspace_params lspace;
D
David Teigland 已提交
60
		struct dlm_purge_params purge;
D
David Teigland 已提交
61 62 63 64 65 66 67 68 69 70 71
	} i;
};

struct dlm_lksb32 {
	__u32 sb_status;
	__u32 sb_lkid;
	__u8 sb_flags;
	__u32 sb_lvbptr;
};

struct dlm_lock_result32 {
72
	__u32 version[3];
D
David Teigland 已提交
73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97
	__u32 length;
	__u32 user_astaddr;
	__u32 user_astparam;
	__u32 user_lksb;
	struct dlm_lksb32 lksb;
	__u8 bast_mode;
	__u8 unused[3];
	/* Offsets may be zero if no data is present */
	__u32 lvb_offset;
};

static void compat_input(struct dlm_write_request *kb,
			 struct dlm_write_request32 *kb32)
{
	kb->version[0] = kb32->version[0];
	kb->version[1] = kb32->version[1];
	kb->version[2] = kb32->version[2];

	kb->cmd = kb32->cmd;
	kb->is64bit = kb32->is64bit;
	if (kb->cmd == DLM_USER_CREATE_LOCKSPACE ||
	    kb->cmd == DLM_USER_REMOVE_LOCKSPACE) {
		kb->i.lspace.flags = kb32->i.lspace.flags;
		kb->i.lspace.minor = kb32->i.lspace.minor;
		strcpy(kb->i.lspace.name, kb32->i.lspace.name);
D
David Teigland 已提交
98 99 100
	} else if (kb->cmd == DLM_USER_PURGE) {
		kb->i.purge.nodeid = kb32->i.purge.nodeid;
		kb->i.purge.pid = kb32->i.purge.pid;
D
David Teigland 已提交
101 102 103 104 105 106
	} else {
		kb->i.lock.mode = kb32->i.lock.mode;
		kb->i.lock.namelen = kb32->i.lock.namelen;
		kb->i.lock.flags = kb32->i.lock.flags;
		kb->i.lock.lkid = kb32->i.lock.lkid;
		kb->i.lock.parent = kb32->i.lock.parent;
107 108
		kb->i.lock.xid = kb32->i.lock.xid;
		kb->i.lock.timeout = kb32->i.lock.timeout;
D
David Teigland 已提交
109 110 111 112 113 114 115 116 117 118 119 120 121
		kb->i.lock.castparam = (void *)(long)kb32->i.lock.castparam;
		kb->i.lock.castaddr = (void *)(long)kb32->i.lock.castaddr;
		kb->i.lock.bastparam = (void *)(long)kb32->i.lock.bastparam;
		kb->i.lock.bastaddr = (void *)(long)kb32->i.lock.bastaddr;
		kb->i.lock.lksb = (void *)(long)kb32->i.lock.lksb;
		memcpy(kb->i.lock.lvb, kb32->i.lock.lvb, DLM_USER_LVB_LEN);
		memcpy(kb->i.lock.name, kb32->i.lock.name, kb->i.lock.namelen);
	}
}

static void compat_output(struct dlm_lock_result *res,
			  struct dlm_lock_result32 *res32)
{
122 123 124 125
	res32->version[0] = res->version[0];
	res32->version[1] = res->version[1];
	res32->version[2] = res->version[2];

D
David Teigland 已提交
126 127 128 129 130 131 132 133 134 135 136 137 138 139 140
	res32->user_astaddr = (__u32)(long)res->user_astaddr;
	res32->user_astparam = (__u32)(long)res->user_astparam;
	res32->user_lksb = (__u32)(long)res->user_lksb;
	res32->bast_mode = res->bast_mode;

	res32->lvb_offset = res->lvb_offset;
	res32->length = res->length;

	res32->lksb.sb_status = res->lksb.sb_status;
	res32->lksb.sb_flags = res->lksb.sb_flags;
	res32->lksb.sb_lkid = res->lksb.sb_lkid;
	res32->lksb.sb_lvbptr = (__u32)(long)res->lksb.sb_lvbptr;
}
#endif

141 142
/* we could possibly check if the cancel of an orphan has resulted in the lkb
   being removed and then remove that lkb from the orphans list and free it */
D
David Teigland 已提交
143 144 145 146 147 148

void dlm_user_add_ast(struct dlm_lkb *lkb, int type)
{
	struct dlm_ls *ls;
	struct dlm_user_args *ua;
	struct dlm_user_proc *proc;
149
	int eol = 0, ast_type;
D
David Teigland 已提交
150

151
	if (lkb->lkb_flags & (DLM_IFL_ORPHAN | DLM_IFL_DEAD))
D
David Teigland 已提交
152 153 154 155 156 157 158
		return;

	ls = lkb->lkb_resource->res_ls;
	mutex_lock(&ls->ls_clear_proc_locks);

	/* If ORPHAN/DEAD flag is set, it means the process is dead so an ast
	   can't be delivered.  For ORPHAN's, dlm_clear_proc_locks() freed
159 160 161
	   lkb->ua so we can't try to use it.  This second check is necessary
	   for cases where a completion ast is received for an operation that
	   began before clear_proc_locks did its cancel/unlock. */
D
David Teigland 已提交
162

163
	if (lkb->lkb_flags & (DLM_IFL_ORPHAN | DLM_IFL_DEAD))
D
David Teigland 已提交
164 165 166 167 168 169 170 171 172 173
		goto out;

	DLM_ASSERT(lkb->lkb_astparam, dlm_print_lkb(lkb););
	ua = (struct dlm_user_args *)lkb->lkb_astparam;
	proc = ua->proc;

	if (type == AST_BAST && ua->bastaddr == NULL)
		goto out;

	spin_lock(&proc->asts_spin);
174 175 176 177 178

	ast_type = lkb->lkb_ast_type;
	lkb->lkb_ast_type |= type;

	if (!ast_type) {
D
David Teigland 已提交
179 180 181 182
		kref_get(&lkb->lkb_ref);
		list_add_tail(&lkb->lkb_astqueue, &proc->asts);
		wake_up_interruptible(&proc->wait);
	}
183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202
	if (type == AST_COMP && (ast_type & AST_COMP))
		log_debug(ls, "ast overlap %x status %x %x",
			  lkb->lkb_id, ua->lksb.sb_status, lkb->lkb_flags);

	/* Figure out if this lock is at the end of its life and no longer
	   available for the application to use.  The lkb still exists until
	   the final ast is read.  A lock becomes EOL in three situations:
	     1. a noqueue request fails with EAGAIN
	     2. an unlock completes with EUNLOCK
	     3. a cancel of a waiting request completes with ECANCEL
	   An EOL lock needs to be removed from the process's list of locks.
	   And we can't allow any new operation on an EOL lock.  This is
	   not related to the lifetime of the lkb struct which is managed
	   entirely by refcount. */

	if (type == AST_COMP &&
	    lkb->lkb_grmode == DLM_LOCK_IV &&
	    ua->lksb.sb_status == -EAGAIN)
		eol = 1;
	else if (ua->lksb.sb_status == -DLM_EUNLOCK ||
D
David Teigland 已提交
203 204
	    (ua->lksb.sb_status == -DLM_ECANCEL &&
	     lkb->lkb_grmode == DLM_LOCK_IV))
205 206 207 208 209
		eol = 1;
	if (eol) {
		lkb->lkb_ast_type &= ~AST_BAST;
		lkb->lkb_flags |= DLM_IFL_ENDOFLIFE;
	}
D
David Teigland 已提交
210

D
David Teigland 已提交
211 212 213 214 215 216 217 218 219 220 221 222 223 224
	/* We want to copy the lvb to userspace when the completion
	   ast is read if the status is 0, the lock has an lvb and
	   lvb_ops says we should.  We could probably have set_lvb_lock()
	   set update_user_lvb instead and not need old_mode */

	if ((lkb->lkb_ast_type & AST_COMP) &&
	    (lkb->lkb_lksb->sb_status == 0) &&
	    lkb->lkb_lksb->sb_lvbptr &&
	    dlm_lvb_operations[ua->old_mode + 1][lkb->lkb_grmode + 1])
		ua->update_user_lvb = 1;
	else
		ua->update_user_lvb = 0;

	spin_unlock(&proc->asts_spin);
D
David Teigland 已提交
225

226
	if (eol) {
D
David Teigland 已提交
227
		spin_lock(&ua->proc->locks_spin);
228 229 230 231
		if (!list_empty(&lkb->lkb_ownqueue)) {
			list_del_init(&lkb->lkb_ownqueue);
			dlm_put_lkb(lkb);
		}
D
David Teigland 已提交
232 233
		spin_unlock(&ua->proc->locks_spin);
	}
D
David Teigland 已提交
234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262
 out:
	mutex_unlock(&ls->ls_clear_proc_locks);
}

static int device_user_lock(struct dlm_user_proc *proc,
			    struct dlm_lock_params *params)
{
	struct dlm_ls *ls;
	struct dlm_user_args *ua;
	int error = -ENOMEM;

	ls = dlm_find_lockspace_local(proc->lockspace);
	if (!ls)
		return -ENOENT;

	if (!params->castaddr || !params->lksb) {
		error = -EINVAL;
		goto out;
	}

	ua = kzalloc(sizeof(struct dlm_user_args), GFP_KERNEL);
	if (!ua)
		goto out;
	ua->proc = proc;
	ua->user_lksb = params->lksb;
	ua->castparam = params->castparam;
	ua->castaddr = params->castaddr;
	ua->bastparam = params->bastparam;
	ua->bastaddr = params->bastaddr;
263
	ua->xid = params->xid;
D
David Teigland 已提交
264 265 266 267

	if (params->flags & DLM_LKF_CONVERT)
		error = dlm_user_convert(ls, ua,
				         params->mode, params->flags,
268 269
				         params->lkid, params->lvb,
					 (unsigned long) params->timeout);
D
David Teigland 已提交
270 271 272 273
	else {
		error = dlm_user_request(ls, ua,
					 params->mode, params->flags,
					 params->name, params->namelen,
274
					 (unsigned long) params->timeout);
D
David Teigland 已提交
275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311
		if (!error)
			error = ua->lksb.sb_lkid;
	}
 out:
	dlm_put_lockspace(ls);
	return error;
}

static int device_user_unlock(struct dlm_user_proc *proc,
			      struct dlm_lock_params *params)
{
	struct dlm_ls *ls;
	struct dlm_user_args *ua;
	int error = -ENOMEM;

	ls = dlm_find_lockspace_local(proc->lockspace);
	if (!ls)
		return -ENOENT;

	ua = kzalloc(sizeof(struct dlm_user_args), GFP_KERNEL);
	if (!ua)
		goto out;
	ua->proc = proc;
	ua->user_lksb = params->lksb;
	ua->castparam = params->castparam;
	ua->castaddr = params->castaddr;

	if (params->flags & DLM_LKF_CANCEL)
		error = dlm_user_cancel(ls, ua, params->flags, params->lkid);
	else
		error = dlm_user_unlock(ls, ua, params->flags, params->lkid,
					params->lvb);
 out:
	dlm_put_lockspace(ls);
	return error;
}

312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334
static int create_misc_device(struct dlm_ls *ls, char *name)
{
	int error, len;

	error = -ENOMEM;
	len = strlen(name) + strlen(name_prefix) + 2;
	ls->ls_device.name = kzalloc(len, GFP_KERNEL);
	if (!ls->ls_device.name)
		goto fail;

	snprintf((char *)ls->ls_device.name, len, "%s_%s", name_prefix,
		 name);
	ls->ls_device.fops = &device_fops;
	ls->ls_device.minor = MISC_DYNAMIC_MINOR;

	error = misc_register(&ls->ls_device);
	if (error) {
		kfree(ls->ls_device.name);
	}
fail:
	return error;
}

D
David Teigland 已提交
335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350
static int device_user_purge(struct dlm_user_proc *proc,
			     struct dlm_purge_params *params)
{
	struct dlm_ls *ls;
	int error;

	ls = dlm_find_lockspace_local(proc->lockspace);
	if (!ls)
		return -ENOENT;

	error = dlm_user_purge(ls, proc, params->nodeid, params->pid);

	dlm_put_lockspace(ls);
	return error;
}

D
David Teigland 已提交
351 352 353 354
static int device_create_lockspace(struct dlm_lspace_params *params)
{
	dlm_lockspace_t *lockspace;
	struct dlm_ls *ls;
355
	int error;
D
David Teigland 已提交
356 357 358 359 360

	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

	error = dlm_new_lockspace(params->name, strlen(params->name),
361
				  &lockspace, params->flags, DLM_USER_LVB_LEN);
D
David Teigland 已提交
362 363 364 365 366 367 368
	if (error)
		return error;

	ls = dlm_find_lockspace_local(lockspace);
	if (!ls)
		return -ENOENT;

369
	error = create_misc_device(ls, params->name);
D
David Teigland 已提交
370 371
	dlm_put_lockspace(ls);

372 373 374 375 376
	if (error)
		dlm_release_lockspace(lockspace, 0);
	else
		error = ls->ls_device.minor;

D
David Teigland 已提交
377 378 379 380 381 382 383
	return error;
}

static int device_remove_lockspace(struct dlm_lspace_params *params)
{
	dlm_lockspace_t *lockspace;
	struct dlm_ls *ls;
384
	int error, force = 0;
D
David Teigland 已提交
385 386 387 388 389 390 391 392

	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

	ls = dlm_find_lockspace_device(params->minor);
	if (!ls)
		return -ENOENT;

393 394 395 396
	/* Deregister the misc device first, so we don't have
	 * a device that's not attached to a lockspace. If
	 * dlm_release_lockspace fails then we can recreate it
	 */
D
David Teigland 已提交
397 398 399 400 401 402 403
	error = misc_deregister(&ls->ls_device);
	if (error) {
		dlm_put_lockspace(ls);
		goto out;
	}
	kfree(ls->ls_device.name);

404 405 406
	if (params->flags & DLM_USER_LSFLG_FORCEFREE)
		force = 2;

D
David Teigland 已提交
407 408 409 410 411 412 413
	lockspace = ls->ls_local_handle;

	/* dlm_release_lockspace waits for references to go to zero,
	   so all processes will need to close their device for the ls
	   before the release will procede */

	dlm_put_lockspace(ls);
414
	error = dlm_release_lockspace(lockspace, force);
415 416
	if (error)
		create_misc_device(ls, ls->ls_name);
417
 out:
D
David Teigland 已提交
418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552
	return error;
}

/* Check the user's version matches ours */
static int check_version(struct dlm_write_request *req)
{
	if (req->version[0] != DLM_DEVICE_VERSION_MAJOR ||
	    (req->version[0] == DLM_DEVICE_VERSION_MAJOR &&
	     req->version[1] > DLM_DEVICE_VERSION_MINOR)) {

		printk(KERN_DEBUG "dlm: process %s (%d) version mismatch "
		       "user (%d.%d.%d) kernel (%d.%d.%d)\n",
		       current->comm,
		       current->pid,
		       req->version[0],
		       req->version[1],
		       req->version[2],
		       DLM_DEVICE_VERSION_MAJOR,
		       DLM_DEVICE_VERSION_MINOR,
		       DLM_DEVICE_VERSION_PATCH);
		return -EINVAL;
	}
	return 0;
}

/*
 * device_write
 *
 *   device_user_lock
 *     dlm_user_request -> request_lock
 *     dlm_user_convert -> convert_lock
 *
 *   device_user_unlock
 *     dlm_user_unlock -> unlock_lock
 *     dlm_user_cancel -> cancel_lock
 *
 *   device_create_lockspace
 *     dlm_new_lockspace
 *
 *   device_remove_lockspace
 *     dlm_release_lockspace
 */

/* a write to a lockspace device is a lock or unlock request, a write
   to the control device is to create/remove a lockspace */

static ssize_t device_write(struct file *file, const char __user *buf,
			    size_t count, loff_t *ppos)
{
	struct dlm_user_proc *proc = file->private_data;
	struct dlm_write_request *kbuf;
	sigset_t tmpsig, allsigs;
	int error;

#ifdef CONFIG_COMPAT
	if (count < sizeof(struct dlm_write_request32))
#else
	if (count < sizeof(struct dlm_write_request))
#endif
		return -EINVAL;

	kbuf = kmalloc(count, GFP_KERNEL);
	if (!kbuf)
		return -ENOMEM;

	if (copy_from_user(kbuf, buf, count)) {
		error = -EFAULT;
		goto out_free;
	}

	if (check_version(kbuf)) {
		error = -EBADE;
		goto out_free;
	}

#ifdef CONFIG_COMPAT
	if (!kbuf->is64bit) {
		struct dlm_write_request32 *k32buf;
		k32buf = (struct dlm_write_request32 *)kbuf;
		kbuf = kmalloc(count + (sizeof(struct dlm_write_request) -
			       sizeof(struct dlm_write_request32)), GFP_KERNEL);
		if (!kbuf)
			return -ENOMEM;

		if (proc)
			set_bit(DLM_PROC_FLAGS_COMPAT, &proc->flags);
		compat_input(kbuf, k32buf);
		kfree(k32buf);
	}
#endif

	/* do we really need this? can a write happen after a close? */
	if ((kbuf->cmd == DLM_USER_LOCK || kbuf->cmd == DLM_USER_UNLOCK) &&
	    test_bit(DLM_PROC_FLAGS_CLOSING, &proc->flags))
		return -EINVAL;

	sigfillset(&allsigs);
	sigprocmask(SIG_BLOCK, &allsigs, &tmpsig);

	error = -EINVAL;

	switch (kbuf->cmd)
	{
	case DLM_USER_LOCK:
		if (!proc) {
			log_print("no locking on control device");
			goto out_sig;
		}
		error = device_user_lock(proc, &kbuf->i.lock);
		break;

	case DLM_USER_UNLOCK:
		if (!proc) {
			log_print("no locking on control device");
			goto out_sig;
		}
		error = device_user_unlock(proc, &kbuf->i.lock);
		break;

	case DLM_USER_CREATE_LOCKSPACE:
		if (proc) {
			log_print("create/remove only on control device");
			goto out_sig;
		}
		error = device_create_lockspace(&kbuf->i.lspace);
		break;

	case DLM_USER_REMOVE_LOCKSPACE:
		if (proc) {
			log_print("create/remove only on control device");
			goto out_sig;
		}
		error = device_remove_lockspace(&kbuf->i.lspace);
		break;

D
David Teigland 已提交
553 554 555 556 557 558 559 560
	case DLM_USER_PURGE:
		if (!proc) {
			log_print("no locking on control device");
			goto out_sig;
		}
		error = device_user_purge(proc, &kbuf->i.purge);
		break;

D
David Teigland 已提交
561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595
	default:
		log_print("Unknown command passed to DLM device : %d\n",
			  kbuf->cmd);
	}

 out_sig:
	sigprocmask(SIG_SETMASK, &tmpsig, NULL);
	recalc_sigpending();
 out_free:
	kfree(kbuf);
	return error;
}

/* Every process that opens the lockspace device has its own "proc" structure
   hanging off the open file that's used to keep track of locks owned by the
   process and asts that need to be delivered to the process. */

static int device_open(struct inode *inode, struct file *file)
{
	struct dlm_user_proc *proc;
	struct dlm_ls *ls;

	ls = dlm_find_lockspace_device(iminor(inode));
	if (!ls)
		return -ENOENT;

	proc = kzalloc(sizeof(struct dlm_user_proc), GFP_KERNEL);
	if (!proc) {
		dlm_put_lockspace(ls);
		return -ENOMEM;
	}

	proc->lockspace = ls->ls_local_handle;
	INIT_LIST_HEAD(&proc->asts);
	INIT_LIST_HEAD(&proc->locks);
D
David Teigland 已提交
596
	INIT_LIST_HEAD(&proc->unlocking);
D
David Teigland 已提交
597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653
	spin_lock_init(&proc->asts_spin);
	spin_lock_init(&proc->locks_spin);
	init_waitqueue_head(&proc->wait);
	file->private_data = proc;

	return 0;
}

static int device_close(struct inode *inode, struct file *file)
{
	struct dlm_user_proc *proc = file->private_data;
	struct dlm_ls *ls;
	sigset_t tmpsig, allsigs;

	ls = dlm_find_lockspace_local(proc->lockspace);
	if (!ls)
		return -ENOENT;

	sigfillset(&allsigs);
	sigprocmask(SIG_BLOCK, &allsigs, &tmpsig);

	set_bit(DLM_PROC_FLAGS_CLOSING, &proc->flags);

	dlm_clear_proc_locks(ls, proc);

	/* at this point no more lkb's should exist for this lockspace,
	   so there's no chance of dlm_user_add_ast() being called and
	   looking for lkb->ua->proc */

	kfree(proc);
	file->private_data = NULL;

	dlm_put_lockspace(ls);
	dlm_put_lockspace(ls);  /* for the find in device_open() */

	/* FIXME: AUTOFREE: if this ls is no longer used do
	   device_remove_lockspace() */

	sigprocmask(SIG_SETMASK, &tmpsig, NULL);
	recalc_sigpending();

	return 0;
}

static int copy_result_to_user(struct dlm_user_args *ua, int compat, int type,
			       int bmode, char __user *buf, size_t count)
{
#ifdef CONFIG_COMPAT
	struct dlm_lock_result32 result32;
#endif
	struct dlm_lock_result result;
	void *resultptr;
	int error=0;
	int len;
	int struct_len;

	memset(&result, 0, sizeof(struct dlm_lock_result));
654 655 656
	result.version[0] = DLM_DEVICE_VERSION_MAJOR;
	result.version[1] = DLM_DEVICE_VERSION_MINOR;
	result.version[2] = DLM_DEVICE_VERSION_PATCH;
D
David Teigland 已提交
657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714
	memcpy(&result.lksb, &ua->lksb, sizeof(struct dlm_lksb));
	result.user_lksb = ua->user_lksb;

	/* FIXME: dlm1 provides for the user's bastparam/addr to not be updated
	   in a conversion unless the conversion is successful.  See code
	   in dlm_user_convert() for updating ua from ua_tmp.  OpenVMS, though,
	   notes that a new blocking AST address and parameter are set even if
	   the conversion fails, so maybe we should just do that. */

	if (type == AST_BAST) {
		result.user_astaddr = ua->bastaddr;
		result.user_astparam = ua->bastparam;
		result.bast_mode = bmode;
	} else {
		result.user_astaddr = ua->castaddr;
		result.user_astparam = ua->castparam;
	}

#ifdef CONFIG_COMPAT
	if (compat)
		len = sizeof(struct dlm_lock_result32);
	else
#endif
		len = sizeof(struct dlm_lock_result);
	struct_len = len;

	/* copy lvb to userspace if there is one, it's been updated, and
	   the user buffer has space for it */

	if (ua->update_user_lvb && ua->lksb.sb_lvbptr &&
	    count >= len + DLM_USER_LVB_LEN) {
		if (copy_to_user(buf+len, ua->lksb.sb_lvbptr,
				 DLM_USER_LVB_LEN)) {
			error = -EFAULT;
			goto out;
		}

		result.lvb_offset = len;
		len += DLM_USER_LVB_LEN;
	}

	result.length = len;
	resultptr = &result;
#ifdef CONFIG_COMPAT
	if (compat) {
		compat_output(&result, &result32);
		resultptr = &result32;
	}
#endif

	if (copy_to_user(buf, resultptr, struct_len))
		error = -EFAULT;
	else
		error = len;
 out:
	return error;
}

715 716 717 718 719 720 721 722 723 724 725 726 727 728
static int copy_version_to_user(char __user *buf, size_t count)
{
	struct dlm_device_version ver;

	memset(&ver, 0, sizeof(struct dlm_device_version));
	ver.version[0] = DLM_DEVICE_VERSION_MAJOR;
	ver.version[1] = DLM_DEVICE_VERSION_MINOR;
	ver.version[2] = DLM_DEVICE_VERSION_PATCH;

	if (copy_to_user(buf, &ver, sizeof(struct dlm_device_version)))
		return -EFAULT;
	return sizeof(struct dlm_device_version);
}

D
David Teigland 已提交
729 730 731 732 733 734 735 736 737 738 739
/* a read returns a single ast described in a struct dlm_lock_result */

static ssize_t device_read(struct file *file, char __user *buf, size_t count,
			   loff_t *ppos)
{
	struct dlm_user_proc *proc = file->private_data;
	struct dlm_lkb *lkb;
	struct dlm_user_args *ua;
	DECLARE_WAITQUEUE(wait, current);
	int error, type=0, bmode=0, removed = 0;

740 741 742 743 744 745 746 747 748 749
	if (count == sizeof(struct dlm_device_version)) {
		error = copy_version_to_user(buf, count);
		return error;
	}

	if (!proc) {
		log_print("non-version read from control device %zu", count);
		return -EINVAL;
	}

D
David Teigland 已提交
750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845
#ifdef CONFIG_COMPAT
	if (count < sizeof(struct dlm_lock_result32))
#else
	if (count < sizeof(struct dlm_lock_result))
#endif
		return -EINVAL;

	/* do we really need this? can a read happen after a close? */
	if (test_bit(DLM_PROC_FLAGS_CLOSING, &proc->flags))
		return -EINVAL;

	spin_lock(&proc->asts_spin);
	if (list_empty(&proc->asts)) {
		if (file->f_flags & O_NONBLOCK) {
			spin_unlock(&proc->asts_spin);
			return -EAGAIN;
		}

		add_wait_queue(&proc->wait, &wait);

	repeat:
		set_current_state(TASK_INTERRUPTIBLE);
		if (list_empty(&proc->asts) && !signal_pending(current)) {
			spin_unlock(&proc->asts_spin);
			schedule();
			spin_lock(&proc->asts_spin);
			goto repeat;
		}
		set_current_state(TASK_RUNNING);
		remove_wait_queue(&proc->wait, &wait);

		if (signal_pending(current)) {
			spin_unlock(&proc->asts_spin);
			return -ERESTARTSYS;
		}
	}

	/* there may be both completion and blocking asts to return for
	   the lkb, don't remove lkb from asts list unless no asts remain */

	lkb = list_entry(proc->asts.next, struct dlm_lkb, lkb_astqueue);

	if (lkb->lkb_ast_type & AST_COMP) {
		lkb->lkb_ast_type &= ~AST_COMP;
		type = AST_COMP;
	} else if (lkb->lkb_ast_type & AST_BAST) {
		lkb->lkb_ast_type &= ~AST_BAST;
		type = AST_BAST;
		bmode = lkb->lkb_bastmode;
	}

	if (!lkb->lkb_ast_type) {
		list_del(&lkb->lkb_astqueue);
		removed = 1;
	}
	spin_unlock(&proc->asts_spin);

	ua = (struct dlm_user_args *)lkb->lkb_astparam;
	error = copy_result_to_user(ua,
			 	test_bit(DLM_PROC_FLAGS_COMPAT, &proc->flags),
				type, bmode, buf, count);

	/* removes reference for the proc->asts lists added by
	   dlm_user_add_ast() and may result in the lkb being freed */
	if (removed)
		dlm_put_lkb(lkb);

	return error;
}

static unsigned int device_poll(struct file *file, poll_table *wait)
{
	struct dlm_user_proc *proc = file->private_data;

	poll_wait(file, &proc->wait, wait);

	spin_lock(&proc->asts_spin);
	if (!list_empty(&proc->asts)) {
		spin_unlock(&proc->asts_spin);
		return POLLIN | POLLRDNORM;
	}
	spin_unlock(&proc->asts_spin);
	return 0;
}

static int ctl_device_open(struct inode *inode, struct file *file)
{
	file->private_data = NULL;
	return 0;
}

static int ctl_device_close(struct inode *inode, struct file *file)
{
	return 0;
}

846
static const struct file_operations device_fops = {
D
David Teigland 已提交
847 848 849 850 851 852 853 854
	.open    = device_open,
	.release = device_close,
	.read    = device_read,
	.write   = device_write,
	.poll    = device_poll,
	.owner   = THIS_MODULE,
};

855
static const struct file_operations ctl_device_fops = {
D
David Teigland 已提交
856 857
	.open    = ctl_device_open,
	.release = ctl_device_close,
858
	.read    = device_read,
D
David Teigland 已提交
859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882
	.write   = device_write,
	.owner   = THIS_MODULE,
};

int dlm_user_init(void)
{
	int error;

	ctl_device.name = "dlm-control";
	ctl_device.fops = &ctl_device_fops;
	ctl_device.minor = MISC_DYNAMIC_MINOR;

	error = misc_register(&ctl_device);
	if (error)
		log_print("misc_register failed for control device");

	return error;
}

void dlm_user_exit(void)
{
	misc_deregister(&ctl_device);
}