glock.c 44.4 KB
Newer Older
David Teigland's avatar
David Teigland committed
1 2
/*
 * Copyright (C) Sistina Software, Inc.  1997-2003 All rights reserved.
3
 * Copyright (C) 2004-2008 Red Hat, Inc.  All rights reserved.
David Teigland's avatar
David Teigland committed
4 5 6
 *
 * This copyrighted material is made available to anyone wishing to use,
 * modify, copy, or redistribute it subject to the terms and conditions
7
 * of the GNU General Public License version 2.
David Teigland's avatar
David Teigland committed
8 9 10 11 12 13 14 15 16
 */

#include <linux/sched.h>
#include <linux/slab.h>
#include <linux/spinlock.h>
#include <linux/buffer_head.h>
#include <linux/delay.h>
#include <linux/sort.h>
#include <linux/jhash.h>
17
#include <linux/kallsyms.h>
18
#include <linux/gfs2_ondisk.h>
19
#include <linux/list.h>
20
#include <linux/wait.h>
akpm@linux-foundation.org's avatar
akpm@linux-foundation.org committed
21
#include <linux/module.h>
22
#include <linux/rwsem.h>
David Teigland's avatar
David Teigland committed
23
#include <asm/uaccess.h>
24 25
#include <linux/seq_file.h>
#include <linux/debugfs.h>
26 27
#include <linux/kthread.h>
#include <linux/freezer.h>
28 29
#include <linux/workqueue.h>
#include <linux/jiffies.h>
David Teigland's avatar
David Teigland committed
30 31

#include "gfs2.h"
32
#include "incore.h"
David Teigland's avatar
David Teigland committed
33 34 35 36 37 38 39
#include "glock.h"
#include "glops.h"
#include "inode.h"
#include "lops.h"
#include "meta_io.h"
#include "quota.h"
#include "super.h"
40
#include "util.h"
41
#include "bmap.h"
David Teigland's avatar
David Teigland committed
42

43
struct gfs2_gl_hash_bucket {
44
        struct hlist_head hb_list;
45 46
};

47 48 49 50 51
struct gfs2_glock_iter {
	int hash;			/* hash bucket index         */
	struct gfs2_sbd *sdp;		/* incore superblock         */
	struct gfs2_glock *gl;		/* current glock struct      */
	char string[512];		/* scratch space             */
52 53
};

David Teigland's avatar
David Teigland committed
54 55
typedef void (*glock_examiner) (struct gfs2_glock * gl);

56
static int gfs2_dump_lockstate(struct gfs2_sbd *sdp);
57 58 59
static int __dump_glock(struct seq_file *seq, const struct gfs2_glock *gl);
#define GLOCK_BUG_ON(gl,x) do { if (unlikely(x)) { __dump_glock(NULL, gl); BUG(); } } while(0)
static void do_xmote(struct gfs2_glock *gl, struct gfs2_holder *gh, unsigned int target);
60

61
static DECLARE_RWSEM(gfs2_umount_flush_sem);
62
static struct dentry *gfs2_root;
63
static struct workqueue_struct *glock_workqueue;
64 65
static LIST_HEAD(lru_list);
static atomic_t lru_count = ATOMIC_INIT(0);
Julia Lawall's avatar
Julia Lawall committed
66
static DEFINE_SPINLOCK(lru_lock);
67

68
#define GFS2_GL_HASH_SHIFT      15
69 70 71
#define GFS2_GL_HASH_SIZE       (1 << GFS2_GL_HASH_SHIFT)
#define GFS2_GL_HASH_MASK       (GFS2_GL_HASH_SIZE - 1)

72
static struct gfs2_gl_hash_bucket gl_hash_table[GFS2_GL_HASH_SIZE];
73
static struct dentry *gfs2_root;
74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113

/*
 * Despite what you might think, the numbers below are not arbitrary :-)
 * They are taken from the ipv4 routing hash code, which is well tested
 * and thus should be nearly optimal. Later on we might tweek the numbers
 * but for now this should be fine.
 *
 * The reason for putting the locks in a separate array from the list heads
 * is that we can have fewer locks than list heads and save memory. We use
 * the same hash function for both, but with a different hash mask.
 */
#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK) || \
	defined(CONFIG_PROVE_LOCKING)

#ifdef CONFIG_LOCKDEP
# define GL_HASH_LOCK_SZ        256
#else
# if NR_CPUS >= 32
#  define GL_HASH_LOCK_SZ       4096
# elif NR_CPUS >= 16
#  define GL_HASH_LOCK_SZ       2048
# elif NR_CPUS >= 8
#  define GL_HASH_LOCK_SZ       1024
# elif NR_CPUS >= 4
#  define GL_HASH_LOCK_SZ       512
# else
#  define GL_HASH_LOCK_SZ       256
# endif
#endif

/* We never want more locks than chains */
#if GFS2_GL_HASH_SIZE < GL_HASH_LOCK_SZ
# undef GL_HASH_LOCK_SZ
# define GL_HASH_LOCK_SZ GFS2_GL_HASH_SIZE
#endif

static rwlock_t gl_hash_locks[GL_HASH_LOCK_SZ];

static inline rwlock_t *gl_lock_addr(unsigned int x)
{
114
	return &gl_hash_locks[x & (GL_HASH_LOCK_SZ-1)];
115 116
}
#else /* not SMP, so no spinlocks required */
Randy Dunlap's avatar
Randy Dunlap committed
117
static inline rwlock_t *gl_lock_addr(unsigned int x)
118 119 120 121
{
	return NULL;
}
#endif
122

David Teigland's avatar
David Teigland committed
123 124 125 126 127 128 129
/**
 * gl_hash() - Turn glock number into hash bucket number
 * @lock: The glock number
 *
 * Returns: The number of the corresponding hash bucket
 */

130 131
static unsigned int gl_hash(const struct gfs2_sbd *sdp,
			    const struct lm_lockname *name)
David Teigland's avatar
David Teigland committed
132 133 134
{
	unsigned int h;

135
	h = jhash(&name->ln_number, sizeof(u64), 0);
David Teigland's avatar
David Teigland committed
136
	h = jhash(&name->ln_type, sizeof(unsigned int), h);
137
	h = jhash(&sdp, sizeof(struct gfs2_sbd *), h);
David Teigland's avatar
David Teigland committed
138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158
	h &= GFS2_GL_HASH_MASK;

	return h;
}

/**
 * glock_free() - Perform a few checks and then release struct gfs2_glock
 * @gl: The glock to release
 *
 * Also calls lock module to release its internal structure for this glock.
 *
 */

static void glock_free(struct gfs2_glock *gl)
{
	struct gfs2_sbd *sdp = gl->gl_sbd;
	struct inode *aspace = gl->gl_aspace;

	if (aspace)
		gfs2_aspace_put(aspace);

159
	sdp->sd_lockstruct.ls_ops->lm_put_lock(gfs2_glock_cachep, gl);
David Teigland's avatar
David Teigland committed
160 161 162 163 164 165 166 167
}

/**
 * gfs2_glock_hold() - increment reference count on glock
 * @gl: The glock to hold
 *
 */

168
static void gfs2_glock_hold(struct gfs2_glock *gl)
David Teigland's avatar
David Teigland committed
169
{
170
	GLOCK_BUG_ON(gl, atomic_read(&gl->gl_ref) == 0);
171
	atomic_inc(&gl->gl_ref);
David Teigland's avatar
David Teigland committed
172 173
}

174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189
/**
 * gfs2_glock_schedule_for_reclaim - Add a glock to the reclaim list
 * @gl: the glock
 *
 */

static void gfs2_glock_schedule_for_reclaim(struct gfs2_glock *gl)
{
	spin_lock(&lru_lock);
	if (list_empty(&gl->gl_lru) && gl->gl_state != LM_ST_UNLOCKED) {
		list_add_tail(&gl->gl_lru, &lru_list);
		atomic_inc(&lru_count);
	}
	spin_unlock(&lru_lock);
}

David Teigland's avatar
David Teigland committed
190 191 192 193 194 195 196 197 198 199
/**
 * gfs2_glock_put() - Decrement reference count on glock
 * @gl: The glock to put
 *
 */

int gfs2_glock_put(struct gfs2_glock *gl)
{
	int rv = 0;

200
	write_lock(gl_lock_addr(gl->gl_hash));
201
	if (atomic_dec_and_test(&gl->gl_ref)) {
202
		hlist_del(&gl->gl_list);
203
		write_unlock(gl_lock_addr(gl->gl_hash));
204 205 206 207 208 209
		spin_lock(&lru_lock);
		if (!list_empty(&gl->gl_lru)) {
			list_del_init(&gl->gl_lru);
			atomic_dec(&lru_count);
		}
		spin_unlock(&lru_lock);
210
		GLOCK_BUG_ON(gl, !list_empty(&gl->gl_holders));
David Teigland's avatar
David Teigland committed
211 212 213 214
		glock_free(gl);
		rv = 1;
		goto out;
	}
215 216 217
	/* 1 for being hashed, 1 for having state != LM_ST_UNLOCKED */
	if (atomic_read(&gl->gl_ref) == 2)
		gfs2_glock_schedule_for_reclaim(gl);
218
	write_unlock(gl_lock_addr(gl->gl_hash));
219
out:
David Teigland's avatar
David Teigland committed
220 221 222 223 224 225 226 227 228 229 230
	return rv;
}

/**
 * search_bucket() - Find struct gfs2_glock by lock number
 * @bucket: the bucket to search
 * @name: The lock name
 *
 * Returns: NULL, or the struct gfs2_glock with the requested number
 */

231
static struct gfs2_glock *search_bucket(unsigned int hash,
232
					const struct gfs2_sbd *sdp,
233
					const struct lm_lockname *name)
David Teigland's avatar
David Teigland committed
234 235
{
	struct gfs2_glock *gl;
236
	struct hlist_node *h;
David Teigland's avatar
David Teigland committed
237

238
	hlist_for_each_entry(gl, h, &gl_hash_table[hash].hb_list, gl_list) {
David Teigland's avatar
David Teigland committed
239 240
		if (!lm_name_equal(&gl->gl_name, name))
			continue;
241 242
		if (gl->gl_sbd != sdp)
			continue;
David Teigland's avatar
David Teigland committed
243

244
		atomic_inc(&gl->gl_ref);
David Teigland's avatar
David Teigland committed
245 246 247 248 249 250 251

		return gl;
	}

	return NULL;
}

252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269
/**
 * may_grant - check if its ok to grant a new lock
 * @gl: The glock
 * @gh: The lock request which we wish to grant
 *
 * Returns: true if its ok to grant the lock
 */

static inline int may_grant(const struct gfs2_glock *gl, const struct gfs2_holder *gh)
{
	const struct gfs2_holder *gh_head = list_entry(gl->gl_holders.next, const struct gfs2_holder, gh_list);
	if ((gh->gh_state == LM_ST_EXCLUSIVE ||
	     gh_head->gh_state == LM_ST_EXCLUSIVE) && gh != gh_head)
		return 0;
	if (gl->gl_state == gh->gh_state)
		return 1;
	if (gh->gh_flags & GL_EXACT)
		return 0;
270 271 272 273 274 275
	if (gl->gl_state == LM_ST_EXCLUSIVE) {
		if (gh->gh_state == LM_ST_SHARED && gh_head->gh_state == LM_ST_SHARED)
			return 1;
		if (gh->gh_state == LM_ST_DEFERRED && gh_head->gh_state == LM_ST_DEFERRED)
			return 1;
	}
276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291
	if (gl->gl_state != LM_ST_UNLOCKED && (gh->gh_flags & LM_FLAG_ANY))
		return 1;
	return 0;
}

static void gfs2_holder_wake(struct gfs2_holder *gh)
{
	clear_bit(HIF_WAIT, &gh->gh_iflags);
	smp_mb__after_clear_bit();
	wake_up_bit(&gh->gh_iflags, HIF_WAIT);
}

/**
 * do_promote - promote as many requests as possible on the current queue
 * @gl: The glock
 * 
292 293
 * Returns: 1 if there is a blocked holder at the head of the list, or 2
 *          if a type specific operation is underway.
294 295 296
 */

static int do_promote(struct gfs2_glock *gl)
297 298
__releases(&gl->gl_spin)
__acquires(&gl->gl_spin)
299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315
{
	const struct gfs2_glock_operations *glops = gl->gl_ops;
	struct gfs2_holder *gh, *tmp;
	int ret;

restart:
	list_for_each_entry_safe(gh, tmp, &gl->gl_holders, gh_list) {
		if (test_bit(HIF_HOLDER, &gh->gh_iflags))
			continue;
		if (may_grant(gl, gh)) {
			if (gh->gh_list.prev == &gl->gl_holders &&
			    glops->go_lock) {
				spin_unlock(&gl->gl_spin);
				/* FIXME: eliminate this eventually */
				ret = glops->go_lock(gh);
				spin_lock(&gl->gl_spin);
				if (ret) {
316 317
					if (ret == 1)
						return 2;
318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421
					gh->gh_error = ret;
					list_del_init(&gh->gh_list);
					gfs2_holder_wake(gh);
					goto restart;
				}
				set_bit(HIF_HOLDER, &gh->gh_iflags);
				gfs2_holder_wake(gh);
				goto restart;
			}
			set_bit(HIF_HOLDER, &gh->gh_iflags);
			gfs2_holder_wake(gh);
			continue;
		}
		if (gh->gh_list.prev == &gl->gl_holders)
			return 1;
		break;
	}
	return 0;
}

/**
 * do_error - Something unexpected has happened during a lock request
 *
 */

static inline void do_error(struct gfs2_glock *gl, const int ret)
{
	struct gfs2_holder *gh, *tmp;

	list_for_each_entry_safe(gh, tmp, &gl->gl_holders, gh_list) {
		if (test_bit(HIF_HOLDER, &gh->gh_iflags))
			continue;
		if (ret & LM_OUT_ERROR)
			gh->gh_error = -EIO;
		else if (gh->gh_flags & (LM_FLAG_TRY | LM_FLAG_TRY_1CB))
			gh->gh_error = GLR_TRYFAILED;
		else
			continue;
		list_del_init(&gh->gh_list);
		gfs2_holder_wake(gh);
	}
}

/**
 * find_first_waiter - find the first gh that's waiting for the glock
 * @gl: the glock
 */

static inline struct gfs2_holder *find_first_waiter(const struct gfs2_glock *gl)
{
	struct gfs2_holder *gh;

	list_for_each_entry(gh, &gl->gl_holders, gh_list) {
		if (!test_bit(HIF_HOLDER, &gh->gh_iflags))
			return gh;
	}
	return NULL;
}

/**
 * state_change - record that the glock is now in a different state
 * @gl: the glock
 * @new_state the new state
 *
 */

static void state_change(struct gfs2_glock *gl, unsigned int new_state)
{
	int held1, held2;

	held1 = (gl->gl_state != LM_ST_UNLOCKED);
	held2 = (new_state != LM_ST_UNLOCKED);

	if (held1 != held2) {
		if (held2)
			gfs2_glock_hold(gl);
		else
			gfs2_glock_put(gl);
	}

	gl->gl_state = new_state;
	gl->gl_tchange = jiffies;
}

static void gfs2_demote_wake(struct gfs2_glock *gl)
{
	gl->gl_demote_state = LM_ST_EXCLUSIVE;
	clear_bit(GLF_DEMOTE, &gl->gl_flags);
	smp_mb__after_clear_bit();
	wake_up_bit(&gl->gl_flags, GLF_DEMOTE);
}

/**
 * finish_xmote - The DLM has replied to one of our lock requests
 * @gl: The glock
 * @ret: The status from the DLM
 *
 */

static void finish_xmote(struct gfs2_glock *gl, unsigned int ret)
{
	const struct gfs2_glock_operations *glops = gl->gl_ops;
	struct gfs2_holder *gh;
	unsigned state = ret & LM_OUT_ST_MASK;
422
	int rv;
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

	spin_lock(&gl->gl_spin);
	state_change(gl, state);
	gh = find_first_waiter(gl);

	/* Demote to UN request arrived during demote to SH or DF */
	if (test_bit(GLF_DEMOTE_IN_PROGRESS, &gl->gl_flags) &&
	    state != LM_ST_UNLOCKED && gl->gl_demote_state == LM_ST_UNLOCKED)
		gl->gl_target = LM_ST_UNLOCKED;

	/* Check for state != intended state */
	if (unlikely(state != gl->gl_target)) {
		if (gh && !test_bit(GLF_DEMOTE_IN_PROGRESS, &gl->gl_flags)) {
			/* move to back of queue and try next entry */
			if (ret & LM_OUT_CANCELED) {
				if ((gh->gh_flags & LM_FLAG_PRIORITY) == 0)
					list_move_tail(&gh->gh_list, &gl->gl_holders);
				gh = find_first_waiter(gl);
				gl->gl_target = gh->gh_state;
				goto retry;
			}
			/* Some error or failed "try lock" - report it */
			if ((ret & LM_OUT_ERROR) ||
			    (gh->gh_flags & (LM_FLAG_TRY | LM_FLAG_TRY_1CB))) {
				gl->gl_target = gl->gl_state;
				do_error(gl, ret);
				goto out;
			}
		}
		switch(state) {
		/* Unlocked due to conversion deadlock, try again */
		case LM_ST_UNLOCKED:
retry:
			do_xmote(gl, gh, gl->gl_target);
			break;
		/* Conversion fails, unlock and try again */
		case LM_ST_SHARED:
		case LM_ST_DEFERRED:
			do_xmote(gl, gh, LM_ST_UNLOCKED);
			break;
		default: /* Everything else */
			printk(KERN_ERR "GFS2: wanted %u got %u\n", gl->gl_target, state);
			GLOCK_BUG_ON(gl, 1);
		}
		spin_unlock(&gl->gl_spin);
		gfs2_glock_put(gl);
		return;
	}

	/* Fast path - we got what we asked for */
	if (test_and_clear_bit(GLF_DEMOTE_IN_PROGRESS, &gl->gl_flags))
		gfs2_demote_wake(gl);
	if (state != LM_ST_UNLOCKED) {
		if (glops->go_xmote_bh) {
			spin_unlock(&gl->gl_spin);
			rv = glops->go_xmote_bh(gl, gh);
			if (rv == -EAGAIN)
				return;
			spin_lock(&gl->gl_spin);
			if (rv) {
				do_error(gl, rv);
				goto out;
			}
		}
487 488 489
		rv = do_promote(gl);
		if (rv == 2)
			goto out_locked;
490 491 492
	}
out:
	clear_bit(GLF_LOCK, &gl->gl_flags);
493
out_locked:
494 495 496 497 498
	spin_unlock(&gl->gl_spin);
	gfs2_glock_put(gl);
}

static unsigned int gfs2_lm_lock(struct gfs2_sbd *sdp, void *lock,
499
				 unsigned int req_state,
500 501 502
				 unsigned int flags)
{
	int ret = LM_OUT_ERROR;
Steven Whitehouse's avatar
Steven Whitehouse committed
503 504 505 506

	if (!sdp->sd_lockstruct.ls_ops->lm_lock)
		return req_state == LM_ST_UNLOCKED ? 0 : req_state;

507
	if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags)))
508
		ret = sdp->sd_lockstruct.ls_ops->lm_lock(lock,
509 510 511 512 513 514 515 516 517 518 519 520 521
							 req_state, flags);
	return ret;
}

/**
 * do_xmote - Calls the DLM to change the state of a lock
 * @gl: The lock state
 * @gh: The holder (only for promotes)
 * @target: The target lock state
 *
 */

static void do_xmote(struct gfs2_glock *gl, struct gfs2_holder *gh, unsigned int target)
522 523
__releases(&gl->gl_spin)
__acquires(&gl->gl_spin)
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
{
	const struct gfs2_glock_operations *glops = gl->gl_ops;
	struct gfs2_sbd *sdp = gl->gl_sbd;
	unsigned int lck_flags = gh ? gh->gh_flags : 0;
	int ret;

	lck_flags &= (LM_FLAG_TRY | LM_FLAG_TRY_1CB | LM_FLAG_NOEXP |
		      LM_FLAG_PRIORITY);
	BUG_ON(gl->gl_state == target);
	BUG_ON(gl->gl_state == gl->gl_target);
	if ((target == LM_ST_UNLOCKED || target == LM_ST_DEFERRED) &&
	    glops->go_inval) {
		set_bit(GLF_INVALIDATE_IN_PROGRESS, &gl->gl_flags);
		do_error(gl, 0); /* Fail queued try locks */
	}
	spin_unlock(&gl->gl_spin);
	if (glops->go_xmote_th)
		glops->go_xmote_th(gl);
	if (test_bit(GLF_INVALIDATE_IN_PROGRESS, &gl->gl_flags))
		glops->go_inval(gl, target == LM_ST_DEFERRED ? 0 : DIO_METADATA);
	clear_bit(GLF_INVALIDATE_IN_PROGRESS, &gl->gl_flags);

	gfs2_glock_hold(gl);
	if (target != LM_ST_UNLOCKED && (gl->gl_state == LM_ST_SHARED ||
	    gl->gl_state == LM_ST_DEFERRED) &&
	    !(lck_flags & (LM_FLAG_TRY | LM_FLAG_TRY_1CB)))
		lck_flags |= LM_FLAG_TRY_1CB;
551
	ret = gfs2_lm_lock(sdp, gl, target, lck_flags);
552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588

	if (!(ret & LM_OUT_ASYNC)) {
		finish_xmote(gl, ret);
		gfs2_glock_hold(gl);
		if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0)
			gfs2_glock_put(gl);
	} else {
		GLOCK_BUG_ON(gl, ret != LM_OUT_ASYNC);
	}
	spin_lock(&gl->gl_spin);
}

/**
 * find_first_holder - find the first "holder" gh
 * @gl: the glock
 */

static inline struct gfs2_holder *find_first_holder(const struct gfs2_glock *gl)
{
	struct gfs2_holder *gh;

	if (!list_empty(&gl->gl_holders)) {
		gh = list_entry(gl->gl_holders.next, struct gfs2_holder, gh_list);
		if (test_bit(HIF_HOLDER, &gh->gh_iflags))
			return gh;
	}
	return NULL;
}

/**
 * run_queue - do all outstanding tasks related to a glock
 * @gl: The glock in question
 * @nonblock: True if we must not block in run_queue
 *
 */

static void run_queue(struct gfs2_glock *gl, const int nonblock)
589 590
__releases(&gl->gl_spin)
__acquires(&gl->gl_spin)
591 592
{
	struct gfs2_holder *gh = NULL;
593
	int ret;
594 595 596 597 598 599 600 601 602

	if (test_and_set_bit(GLF_LOCK, &gl->gl_flags))
		return;

	GLOCK_BUG_ON(gl, test_bit(GLF_DEMOTE_IN_PROGRESS, &gl->gl_flags));

	if (test_bit(GLF_DEMOTE, &gl->gl_flags) &&
	    gl->gl_demote_state != gl->gl_state) {
		if (find_first_holder(gl))
603
			goto out_unlock;
604 605 606
		if (nonblock)
			goto out_sched;
		set_bit(GLF_DEMOTE_IN_PROGRESS, &gl->gl_flags);
607
		GLOCK_BUG_ON(gl, gl->gl_demote_state == LM_ST_EXCLUSIVE);
608 609 610 611
		gl->gl_target = gl->gl_demote_state;
	} else {
		if (test_bit(GLF_DEMOTE, &gl->gl_flags))
			gfs2_demote_wake(gl);
612 613
		ret = do_promote(gl);
		if (ret == 0)
614
			goto out_unlock;
615
		if (ret == 2)
616
			goto out;
617 618 619 620 621 622
		gh = find_first_waiter(gl);
		gl->gl_target = gh->gh_state;
		if (!(gh->gh_flags & (LM_FLAG_TRY | LM_FLAG_TRY_1CB)))
			do_error(gl, 0); /* Fail queued try locks */
	}
	do_xmote(gl, gh, gl->gl_target);
623
out:
624 625 626 627 628 629
	return;

out_sched:
	gfs2_glock_hold(gl);
	if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0)
		gfs2_glock_put(gl);
630
out_unlock:
631
	clear_bit(GLF_LOCK, &gl->gl_flags);
632
	goto out;
633 634
}

635 636
static void glock_work_func(struct work_struct *work)
{
637
	unsigned long delay = 0;
638 639
	struct gfs2_glock *gl = container_of(work, struct gfs2_glock, gl_work.work);

640 641
	if (test_and_clear_bit(GLF_REPLY_PENDING, &gl->gl_flags))
		finish_xmote(gl, gl->gl_reply);
642
	down_read(&gfs2_umount_flush_sem);
643
	spin_lock(&gl->gl_spin);
644 645 646
	if (test_and_clear_bit(GLF_PENDING_DEMOTE, &gl->gl_flags) &&
	    gl->gl_state != LM_ST_UNLOCKED &&
	    gl->gl_demote_state != LM_ST_EXCLUSIVE) {
647 648 649 650 651 652 653
		unsigned long holdtime, now = jiffies;
		holdtime = gl->gl_tchange + gl->gl_ops->go_min_hold_time;
		if (time_before(now, holdtime))
			delay = holdtime - now;
		set_bit(delay ? GLF_PENDING_DEMOTE : GLF_DEMOTE, &gl->gl_flags);
	}
	run_queue(gl, 0);
654
	spin_unlock(&gl->gl_spin);
655
	up_read(&gfs2_umount_flush_sem);
656 657 658
	if (!delay ||
	    queue_delayed_work(glock_workqueue, &gl->gl_work, delay) == 0)
		gfs2_glock_put(gl);
659 660
}

David Teigland's avatar
David Teigland committed
661 662 663 664 665 666 667 668 669 670 671 672 673
/**
 * gfs2_glock_get() - Get a glock, or create one if one doesn't exist
 * @sdp: The GFS2 superblock
 * @number: the lock number
 * @glops: The glock_operations to use
 * @create: If 0, don't create the glock if it doesn't exist
 * @glp: the glock is returned here
 *
 * This does not lock a glock, just finds/creates structures for one.
 *
 * Returns: errno
 */

674
int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number,
675
		   const struct gfs2_glock_operations *glops, int create,
David Teigland's avatar
David Teigland committed
676 677
		   struct gfs2_glock **glp)
{
678
	struct lm_lockname name = { .ln_number = number, .ln_type = glops->go_type };
David Teigland's avatar
David Teigland committed
679
	struct gfs2_glock *gl, *tmp;
680
	unsigned int hash = gl_hash(sdp, &name);
David Teigland's avatar
David Teigland committed
681 682
	int error;

683
	read_lock(gl_lock_addr(hash));
684
	gl = search_bucket(hash, sdp, &name);
685
	read_unlock(gl_lock_addr(hash));
David Teigland's avatar
David Teigland committed
686

687 688
	*glp = gl;
	if (gl)
David Teigland's avatar
David Teigland committed
689
		return 0;
690 691
	if (!create)
		return -ENOENT;
David Teigland's avatar
David Teigland committed
692 693 694 695 696

	gl = kmem_cache_alloc(gfs2_glock_cachep, GFP_KERNEL);
	if (!gl)
		return -ENOMEM;

697
	gl->gl_flags = 0;
David Teigland's avatar
David Teigland committed
698
	gl->gl_name = name;
699
	atomic_set(&gl->gl_ref, 1);
David Teigland's avatar
David Teigland committed
700
	gl->gl_state = LM_ST_UNLOCKED;
701
	gl->gl_target = LM_ST_UNLOCKED;
702
	gl->gl_demote_state = LM_ST_EXCLUSIVE;
703
	gl->gl_hash = hash;
David Teigland's avatar
David Teigland committed
704
	gl->gl_ops = glops;
705 706 707
	snprintf(gl->gl_strname, GDLM_STRNAME_BYTES, "%8x%16llx", name.ln_type, (unsigned long long)number);
	memset(&gl->gl_lksb, 0, sizeof(struct dlm_lksb));
	gl->gl_lksb.sb_lvbptr = gl->gl_lvb;
708
	gl->gl_tchange = jiffies;
709
	gl->gl_object = NULL;
David Teigland's avatar
David Teigland committed
710
	gl->gl_sbd = sdp;
711
	gl->gl_aspace = NULL;
712
	INIT_DELAYED_WORK(&gl->gl_work, glock_work_func);
David Teigland's avatar
David Teigland committed
713 714 715

	/* If this glock protects actual on-disk data or metadata blocks,
	   create a VFS inode to manage the pages/buffers holding them. */
716
	if (glops == &gfs2_inode_glops || glops == &gfs2_rgrp_glops) {
David Teigland's avatar
David Teigland committed
717 718 719 720 721 722 723
		gl->gl_aspace = gfs2_aspace_get(sdp);
		if (!gl->gl_aspace) {
			error = -ENOMEM;
			goto fail;
		}
	}

724
	write_lock(gl_lock_addr(hash));
725
	tmp = search_bucket(hash, sdp, &name);
David Teigland's avatar
David Teigland committed
726
	if (tmp) {
727
		write_unlock(gl_lock_addr(hash));
David Teigland's avatar
David Teigland committed
728 729 730
		glock_free(gl);
		gl = tmp;
	} else {
731
		hlist_add_head(&gl->gl_list, &gl_hash_table[hash].hb_list);
732
		write_unlock(gl_lock_addr(hash));
David Teigland's avatar
David Teigland committed
733 734 735 736 737 738
	}

	*glp = gl;

	return 0;

739
fail:
740
	kmem_cache_free(gfs2_glock_cachep, gl);
David Teigland's avatar
David Teigland committed
741 742 743 744 745 746 747 748 749 750 751 752
	return error;
}

/**
 * gfs2_holder_init - initialize a struct gfs2_holder in the default way
 * @gl: the glock
 * @state: the state we're requesting
 * @flags: the modifier flags
 * @gh: the holder structure
 *
 */

753
void gfs2_holder_init(struct gfs2_glock *gl, unsigned int state, unsigned flags,
David Teigland's avatar
David Teigland committed
754 755 756 757
		      struct gfs2_holder *gh)
{
	INIT_LIST_HEAD(&gh->gh_list);
	gh->gh_gl = gl;
758
	gh->gh_ip = (unsigned long)__builtin_return_address(0);
759
	gh->gh_owner_pid = get_pid(task_pid(current));
David Teigland's avatar
David Teigland committed
760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776
	gh->gh_state = state;
	gh->gh_flags = flags;
	gh->gh_error = 0;
	gh->gh_iflags = 0;
	gfs2_glock_hold(gl);
}

/**
 * gfs2_holder_reinit - reinitialize a struct gfs2_holder so we can requeue it
 * @state: the state we're requesting
 * @flags: the modifier flags
 * @gh: the holder structure
 *
 * Don't mess with the glock.
 *
 */

777
void gfs2_holder_reinit(unsigned int state, unsigned flags, struct gfs2_holder *gh)
David Teigland's avatar
David Teigland committed
778 779
{
	gh->gh_state = state;
780
	gh->gh_flags = flags;
781
	gh->gh_iflags = 0;
782
	gh->gh_ip = (unsigned long)__builtin_return_address(0);
David Teigland's avatar
David Teigland committed
783 784 785 786 787 788 789 790 791 792
}

/**
 * gfs2_holder_uninit - uninitialize a holder structure (drop glock reference)
 * @gh: the holder structure
 *
 */

void gfs2_holder_uninit(struct gfs2_holder *gh)
{
793
	put_pid(gh->gh_owner_pid);
David Teigland's avatar
David Teigland committed
794 795
	gfs2_glock_put(gh->gh_gl);
	gh->gh_gl = NULL;
796
	gh->gh_ip = 0;
David Teigland's avatar
David Teigland committed
797 798
}

Abhijith Das's avatar
Abhijith Das committed
799
static int just_schedule(void *word)
800 801 802 803 804
{
        schedule();
        return 0;
}

805
static void wait_on_holder(struct gfs2_holder *gh)
806
{
807 808
	might_sleep();
	wait_on_bit(&gh->gh_iflags, HIF_WAIT, just_schedule, TASK_UNINTERRUPTIBLE);
809 810
}

811
static void wait_on_demote(struct gfs2_glock *gl)
David Teigland's avatar
David Teigland committed
812
{
813 814
	might_sleep();
	wait_on_bit(&gl->gl_flags, GLF_DEMOTE, just_schedule, TASK_UNINTERRUPTIBLE);
David Teigland's avatar
David Teigland committed
815 816 817
}

/**
818 819 820
 * handle_callback - process a demote request
 * @gl: the glock
 * @state: the state the caller wants us to change to
David Teigland's avatar
David Teigland committed
821
 *
822 823
 * There are only two requests that we are going to see in actual
 * practise: LM_ST_SHARED and LM_ST_UNLOCKED
David Teigland's avatar
David Teigland committed
824 825
 */

826
static void handle_callback(struct gfs2_glock *gl, unsigned int state,
827
			    unsigned long delay)
David Teigland's avatar
David Teigland committed
828
{
829
	int bit = delay ? GLF_PENDING_DEMOTE : GLF_DEMOTE;
David Teigland's avatar
David Teigland committed
830

831 832 833 834 835 836 837
	set_bit(bit, &gl->gl_flags);
	if (gl->gl_demote_state == LM_ST_EXCLUSIVE) {
		gl->gl_demote_state = state;
		gl->gl_demote_time = jiffies;
	} else if (gl->gl_demote_state != LM_ST_UNLOCKED &&
			gl->gl_demote_state != state) {
		gl->gl_demote_state = LM_ST_UNLOCKED;
David Teigland's avatar
David Teigland committed
838 839 840 841
	}
}

/**
842
 * gfs2_glock_wait - wait on a glock acquisition
David Teigland's avatar
David Teigland committed
843 844 845 846 847
 * @gh: the glock holder
 *
 * Returns: 0 on success
 */

848
int gfs2_glock_wait(struct gfs2_holder *gh)
David Teigland's avatar
David Teigland committed
849
{
850
	wait_on_holder(gh);
David Teigland's avatar
David Teigland committed
851 852 853
	return gh->gh_error;
}

854
void gfs2_print_dbg(struct seq_file *seq, const char *fmt, ...)
855 856 857 858
{
	va_list args;

	va_start(args, fmt);
859 860
	if (seq) {
		struct gfs2_glock_iter *gi = seq->private;
861
		vsprintf(gi->string, fmt, args);
862 863 864
		seq_printf(seq, gi->string);
	} else {
		printk(KERN_ERR " ");
865
		vprintk(fmt, args);
866
	}
867 868 869
	va_end(args);
}

David Teigland's avatar
David Teigland committed
870 871 872 873
/**
 * add_to_queue - Add a holder to the wait queue (but look for recursion)
 * @gh: the holder structure to add
 *
874 875 876 877
 * Eventually we should move the recursive locking trap to a
 * debugging option or something like that. This is the fast
 * path and needs to have the minimum number of distractions.
 * 
David Teigland's avatar
David Teigland committed
878 879
 */

880
static inline void add_to_queue(struct gfs2_holder *gh)
881 882
__releases(&gl->gl_spin)
__acquires(&gl->gl_spin)
David Teigland's avatar
David Teigland committed
883 884
{
	struct gfs2_glock *gl = gh->gh_gl;
885 886 887 888
	struct gfs2_sbd *sdp = gl->gl_sbd;
	struct list_head *insert_pt = NULL;
	struct gfs2_holder *gh2;
	int try_lock = 0;
David Teigland's avatar
David Teigland committed
889

890
	BUG_ON(gh->gh_owner_pid == NULL);
891 892
	if (test_and_set_bit(HIF_WAIT, &gh->gh_iflags))
		BUG();
893

894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911
	if (gh->gh_flags & (LM_FLAG_TRY | LM_FLAG_TRY_1CB)) {
		if (test_bit(GLF_LOCK, &gl->gl_flags))
			try_lock = 1;
		if (test_bit(GLF_INVALIDATE_IN_PROGRESS, &gl->gl_flags))
			goto fail;
	}

	list_for_each_entry(gh2, &gl->gl_holders, gh_list) {
		if (unlikely(gh2->gh_owner_pid == gh->gh_owner_pid &&
		    (gh->gh_gl->gl_ops->go_type != LM_TYPE_FLOCK)))
			goto trap_recursive;
		if (try_lock &&
		    !(gh2->gh_flags & (LM_FLAG_TRY | LM_FLAG_TRY_1CB)) &&
		    !may_grant(gl, gh)) {
fail:
			gh->gh_error = GLR_TRYFAILED;
			gfs2_holder_wake(gh);
			return;
912
		}
913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928
		if (test_bit(HIF_HOLDER, &gh2->gh_iflags))
			continue;
		if (unlikely((gh->gh_flags & LM_FLAG_PRIORITY) && !insert_pt))
			insert_pt = &gh2->gh_list;
	}
	if (likely(insert_pt == NULL)) {
		list_add_tail(&gh->gh_list, &gl->gl_holders);
		if (unlikely(gh->gh_flags & LM_FLAG_PRIORITY))
			goto do_cancel;
		return;
	}
	list_add_tail(&gh->gh_list, insert_pt);
do_cancel:
	gh = list_entry(gl->gl_holders.next, struct gfs2_holder, gh_list);
	if (!(gh->gh_flags & LM_FLAG_PRIORITY)) {
		spin_unlock(&gl->gl_spin);
Steven Whitehouse's avatar
Steven Whitehouse committed
929
		if (sdp->sd_lockstruct.ls_ops->lm_cancel)
930
			sdp->sd_lockstruct.ls_ops->lm_cancel(gl);
931
		spin_lock(&gl->gl_spin);
David Teigland's avatar
David Teigland committed
932
	}
933
	return;
David Teigland's avatar
David Teigland committed
934

935 936 937 938 939 940 941 942 943 944 945
trap_recursive:
	print_symbol(KERN_ERR "original: %s\n", gh2->gh_ip);
	printk(KERN_ERR "pid: %d\n", pid_nr(gh2->gh_owner_pid));
	printk(KERN_ERR "lock type: %d req lock state : %d\n",
	       gh2->gh_gl->gl_name.ln_type, gh2->gh_state);
	print_symbol(KERN_ERR "new: %s\n", gh->gh_ip);
	printk(KERN_ERR "pid: %d\n", pid_nr(gh->gh_owner_pid));
	printk(KERN_ERR "lock type: %d req lock state : %d\n",
	       gh->gh_gl->gl_name.ln_type, gh->gh_state);
	__dump_glock(NULL, gl);
	BUG();
David Teigland's avatar
David Teigland committed
946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962
}

/**
 * gfs2_glock_nq - enqueue a struct gfs2_holder onto a glock (acquire a glock)
 * @gh: the holder structure
 *
 * if (gh->gh_flags & GL_ASYNC), this never returns an error
 *
 * Returns: 0, GLR_TRYFAILED, or errno on failure
 */

int gfs2_glock_nq(struct gfs2_holder *gh)
{
	struct gfs2_glock *gl = gh->gh_gl;
	struct gfs2_sbd *sdp = gl->gl_sbd;
	int error = 0;

963
	if (unlikely(test_bit(SDF_SHUTDOWN, &sdp->sd_flags)))
David Teigland's avatar
David Teigland committed
964 965 966 967
		return -EIO;

	spin_lock(&gl->gl_spin);
	add_to_queue(gh);
968
	run_queue(gl, 1);
David Teigland's avatar
David Teigland committed
969 970
	spin_unlock(&gl->gl_spin);

971 972
	if (!(gh->gh_flags & GL_ASYNC))
		error = gfs2_glock_wait(gh);
David Teigland's avatar
David Teigland committed
973 974 975 976 977 978 979 980 981 982 983 984 985

	return error;
}

/**
 * gfs2_glock_poll - poll to see if an async request has been completed
 * @gh: the holder
 *
 * Returns: 1 if the request is ready to be gfs2_glock_wait()ed on
 */

int gfs2_glock_poll(struct gfs2_holder *gh)
{
986
	return test_bit(HIF_WAIT, &gh->gh_iflags) ? 0 : 1;
David Teigland's avatar
David Teigland committed
987 988 989 990 991 992 993 994 995 996 997
}

/**
 * gfs2_glock_dq - dequeue a struct gfs2_holder from a glock (release a glock)
 * @gh: the glock holder
 *
 */

void gfs2_glock_dq(struct gfs2_holder *gh)
{
	struct gfs2_glock *gl = gh->gh_gl;
998
	const struct gfs2_glock_operations *glops = gl->gl_ops;
999
	unsigned delay = 0;
1000
	int fast_path = 0;
David Teigland's avatar
David Teigland committed
1001

1002
	spin_lock(&gl->gl_spin);
David Teigland's avatar
David Teigland committed
1003
	if (gh->gh_flags & GL_NOCACHE)
1004
		handle_callback(gl, LM_ST_UNLOCKED, 0);
David Teigland's avatar
David Teigland committed
1005 1006

	list_del_init(&gh->gh_list);
1007
	if (find_first_holder(gl) == NULL) {
1008
		if (glops->go_unlock) {
1009
			GLOCK_BUG_ON(gl, test_and_set_bit(GLF_LOCK, &gl->gl_flags));
1010
			spin_unlock(&gl->gl_spin);
David Teigland's avatar
David Teigland committed
1011
			glops->go_unlock(gh);
1012
			spin_lock(&gl->gl_spin);
1013
			clear_bit(GLF_LOCK, &gl->gl_flags);
1014
		}
1015 1016 1017 1018
		if (list_empty(&gl->gl_holders) &&
		    !test_bit(GLF_PENDING_DEMOTE, &gl->gl_flags) &&
		    !test_bit(GLF_DEMOTE, &gl->gl_flags))
			fast_path = 1;
David Teigland's avatar
David Teigland committed
1019 1020
	}
	spin_unlock(&gl->gl_spin);
1021 1022
	if (likely(fast_path))
		return;
1023 1024 1025 1026 1027 1028 1029

	gfs2_glock_hold(gl);
	if (test_bit(GLF_PENDING_DEMOTE, &gl->gl_flags) &&
	    !test_bit(GLF_DEMOTE, &gl->gl_flags))
		delay = gl->gl_ops->go_min_hold_time;
	if (queue_delayed_work(glock_workqueue, &gl->gl_work, delay) == 0)
		gfs2_glock_put(gl);
David Teigland's avatar
David Teigland committed
1030 1031
}

Abhijith Das's avatar
Abhijith Das committed
1032 1033 1034 1035 1036 1037 1038
void gfs2_glock_dq_wait(struct gfs2_holder *gh)
{
	struct gfs2_glock *gl = gh->gh_gl;
	gfs2_glock_dq(gh);
	wait_on_demote(gl);
}

David Teigland's avatar
David Teigland committed
1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062
/**
 * gfs2_glock_dq_uninit - dequeue a holder from a glock and initialize it
 * @gh: the holder structure
 *
 */

void gfs2_glock_dq_uninit(struct gfs2_holder *gh)
{
	gfs2_glock_dq(gh);
	gfs2_holder_uninit(gh);
}

/**
 * gfs2_glock_nq_num - acquire a glock based on lock number
 * @sdp: the filesystem
 * @number: the lock number
 * @glops: the glock operations for the type of glock
 * @state: the state to acquire the glock in
 * @flags: modifier flags for the aquisition
 * @gh: the struct gfs2_holder
 *
 * Returns: errno
 */

1063
int gfs2_glock_nq_num(struct gfs2_sbd *sdp, u64 number,
1064 1065
		      const struct gfs2_glock_operations *glops,
		      unsigned int state, int flags, struct gfs2_holder *gh)
David Teigland's avatar
David Teigland committed
1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087
{
	struct gfs2_glock *gl;
	int error;

	error = gfs2_glock_get(sdp, number, glops, CREATE, &gl);
	if (!error) {
		error = gfs2_glock_nq_init(gl, state, flags, gh);
		gfs2_glock_put(gl);
	}

	return error;
}

/**
 * glock_compare - Compare two struct gfs2_glock structures for sorting
 * @arg_a: the first structure
 * @arg_b: the second structure
 *
 */

static int glock_compare(const void *arg_a, const void *arg_b)
{
1088 1089 1090 1091
	const struct gfs2_holder *gh_a = *(const struct gfs2_holder **)arg_a;
	const struct gfs2_holder *gh_b = *(const struct gfs2_holder **)arg_b;
	const struct lm_lockname *a = &gh_a->gh_gl->gl_name;
	const struct lm_lockname *b = &gh_b->gh_gl->gl_name;
David Teigland's avatar
David Teigland committed
1092 1093

	if (a->ln_number > b->ln_number)
1094 1095 1096
		return 1;
	if (a->ln_number < b->ln_number)
		return -1;
1097
	BUG_ON(gh_a->gh_gl->gl_ops->go_type == gh_b->gh_gl->gl_ops->go_type);
1098
	return 0;
David Teigland's avatar
David Teigland committed
1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146
}

/**
 * nq_m_sync - synchonously acquire more than one glock in deadlock free order
 * @num_gh: the number of structures
 * @ghs: an array of struct gfs2_holder structures
 *
 * Returns: 0 on success (all glocks acquired),
 *          errno on failure (no glocks acquired)
 */

static int nq_m_sync(unsigned int num_gh, struct gfs2_holder *ghs,
		     struct gfs2_holder **p)
{
	unsigned int x;
	int error = 0;

	for (x = 0; x < num_gh; x++)
		p[x] = &ghs[x];

	sort(p, num_gh, sizeof(struct gfs2_holder *), glock_compare, NULL);

	for (x = 0; x < num_gh; x++) {
		p[x]->gh_flags &= ~(LM_FLAG_TRY | GL_ASYNC);

		error = gfs2_glock_nq(p[x]);
		if (error) {
			while (x--)
				gfs2_glock_dq(p[x]);
			break;
		}
	}

	return error;
}

/**
 * gfs2_glock_nq_m - acquire multiple glocks
 * @num_gh: the number of structures
 * @ghs: an array of struct gfs2_holder structures
 *
 *
 * Returns: 0 on success (all glocks acquired),
 *          errno on failure (no glocks acquired)
 */

int gfs2_glock_nq_m(unsigned int num_gh, struct gfs2_holder *ghs)
{
1147 1148
	struct gfs2_holder *tmp[4];
	struct gfs2_holder **pph = tmp;
David Teigland's avatar
David Teigland committed
1149 1150
	int error = 0;

1151 1152
	switch(num_gh) {
	case 0:
David Teigland's avatar
David Teigland committed
1153
		return 0;
1154
	case 1:
David Teigland's avatar
David Teigland committed
1155 1156
		ghs->gh_flags &= ~(LM_FLAG_TRY | GL_ASYNC);
		return gfs2_glock_nq(ghs);
1157 1158
	default:
		if (num_gh <= 4)
David Teigland's avatar
David Teigland committed