glock.c 45.6 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>
David Teigland's avatar
David Teigland committed
22
#include <asm/uaccess.h>
23 24
#include <linux/seq_file.h>
#include <linux/debugfs.h>
25 26
#include <linux/kthread.h>
#include <linux/freezer.h>
27 28
#include <linux/workqueue.h>
#include <linux/jiffies.h>
29 30 31
#include <linux/rcupdate.h>
#include <linux/rculist_bl.h>
#include <linux/bit_spinlock.h>
David Teigland's avatar
David Teigland committed
32 33

#include "gfs2.h"
34
#include "incore.h"
David Teigland's avatar
David Teigland committed
35 36 37 38 39 40 41
#include "glock.h"
#include "glops.h"
#include "inode.h"
#include "lops.h"
#include "meta_io.h"
#include "quota.h"
#include "super.h"
42
#include "util.h"
43
#include "bmap.h"
Steven Whitehouse's avatar
Steven Whitehouse committed
44 45
#define CREATE_TRACE_POINTS
#include "trace_gfs2.h"
David Teigland's avatar
David Teigland committed
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 57 58
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);
59

60
static struct dentry *gfs2_root;
61
static struct workqueue_struct *glock_workqueue;
62
struct workqueue_struct *gfs2_delete_workqueue;
63 64
static LIST_HEAD(lru_list);
static atomic_t lru_count = ATOMIC_INIT(0);
Julia Lawall's avatar
Julia Lawall committed
65
static DEFINE_SPINLOCK(lru_lock);
66

67
#define GFS2_GL_HASH_SHIFT      15
68 69 70
#define GFS2_GL_HASH_SIZE       (1 << GFS2_GL_HASH_SHIFT)
#define GFS2_GL_HASH_MASK       (GFS2_GL_HASH_SIZE - 1)

71
static struct hlist_bl_head gl_hash_table[GFS2_GL_HASH_SIZE];
72
static struct dentry *gfs2_root;
73

David Teigland's avatar
David Teigland committed
74 75 76 77 78 79 80
/**
 * gl_hash() - Turn glock number into hash bucket number
 * @lock: The glock number
 *
 * Returns: The number of the corresponding hash bucket
 */

81 82
static unsigned int gl_hash(const struct gfs2_sbd *sdp,
			    const struct lm_lockname *name)
David Teigland's avatar
David Teigland committed
83 84 85
{
	unsigned int h;

86
	h = jhash(&name->ln_number, sizeof(u64), 0);
David Teigland's avatar
David Teigland committed
87
	h = jhash(&name->ln_type, sizeof(unsigned int), h);
88
	h = jhash(&sdp, sizeof(struct gfs2_sbd *), h);
David Teigland's avatar
David Teigland committed
89 90 91 92 93
	h &= GFS2_GL_HASH_MASK;

	return h;
}

94 95
static inline void spin_lock_bucket(unsigned int hash)
{
96
	hlist_bl_lock(&gl_hash_table[hash]);
97
}
David Teigland's avatar
David Teigland committed
98

99 100
static inline void spin_unlock_bucket(unsigned int hash)
{
101
	hlist_bl_unlock(&gl_hash_table[hash]);
102
}
David Teigland's avatar
David Teigland committed
103

104
static void gfs2_glock_dealloc(struct rcu_head *rcu)
David Teigland's avatar
David Teigland committed
105
{
106
	struct gfs2_glock *gl = container_of(rcu, struct gfs2_glock, gl_rcu);
David Teigland's avatar
David Teigland committed
107

108 109 110 111
	if (gl->gl_ops->go_flags & GLOF_ASPACE)
		kmem_cache_free(gfs2_glock_aspace_cachep, gl);
	else
		kmem_cache_free(gfs2_glock_cachep, gl);
112 113 114
}

void gfs2_glock_free(struct gfs2_glock *gl)
David Teigland's avatar
David Teigland committed
115 116 117
{
	struct gfs2_sbd *sdp = gl->gl_sbd;

118
	call_rcu(&gl->gl_rcu, gfs2_glock_dealloc);
119 120
	if (atomic_dec_and_test(&sdp->sd_glock_disposal))
		wake_up(&sdp->sd_glock_wait);
David Teigland's avatar
David Teigland committed
121 122 123 124 125 126 127 128
}

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

129
void gfs2_glock_hold(struct gfs2_glock *gl)
David Teigland's avatar
David Teigland committed
130
{
131
	GLOCK_BUG_ON(gl, atomic_read(&gl->gl_ref) == 0);
132
	atomic_inc(&gl->gl_ref);
David Teigland's avatar
David Teigland committed
133 134
}

135 136 137 138 139 140 141 142 143 144 145 146 147
/**
 * demote_ok - Check to see if it's ok to unlock a glock
 * @gl: the glock
 *
 * Returns: 1 if it's ok
 */

static int demote_ok(const struct gfs2_glock *gl)
{
	const struct gfs2_glock_operations *glops = gl->gl_ops;

	if (gl->gl_state == LM_ST_UNLOCKED)
		return 0;
148
	if (!list_empty(&gl->gl_holders))
149 150 151 152 153 154
		return 0;
	if (glops->go_demote_ok)
		return glops->go_demote_ok(gl);
	return 1;
}

155

156 157 158 159 160 161 162 163 164 165
void gfs2_glock_add_to_lru(struct gfs2_glock *gl)
{
	spin_lock(&lru_lock);

	if (!list_empty(&gl->gl_lru))
		list_del_init(&gl->gl_lru);
	else
		atomic_inc(&lru_count);

	list_add_tail(&gl->gl_lru, &lru_list);
166
	set_bit(GLF_LRU, &gl->gl_flags);
167 168 169
	spin_unlock(&lru_lock);
}

170 171 172 173 174 175 176 177 178 179 180
static void gfs2_glock_remove_from_lru(struct gfs2_glock *gl)
{
	spin_lock(&lru_lock);
	if (!list_empty(&gl->gl_lru)) {
		list_del_init(&gl->gl_lru);
		atomic_dec(&lru_count);
		clear_bit(GLF_LRU, &gl->gl_flags);
	}
	spin_unlock(&lru_lock);
}

181
/**
182
 * __gfs2_glock_schedule_for_reclaim - Add a glock to the reclaim list
183 184
 * @gl: the glock
 *
185 186
 * If the glock is demotable, then we add it (or move it) to the end
 * of the glock LRU list.
187 188
 */

189
static void __gfs2_glock_schedule_for_reclaim(struct gfs2_glock *gl)
190
{
191 192
	if (demote_ok(gl))
		gfs2_glock_add_to_lru(gl);
193 194
}

195 196 197 198 199 200 201 202
/**
 * gfs2_glock_put_nolock() - Decrement reference count on glock
 * @gl: The glock to put
 *
 * This function should only be used if the caller has its own reference
 * to the glock, in addition to the one it is dropping.
 */

203
void gfs2_glock_put_nolock(struct gfs2_glock *gl)
204 205 206 207 208
{
	if (atomic_dec_and_test(&gl->gl_ref))
		GLOCK_BUG_ON(gl, 1);
}

David Teigland's avatar
David Teigland committed
209 210 211 212 213 214
/**
 * gfs2_glock_put() - Decrement reference count on glock
 * @gl: The glock to put
 *
 */

215
void gfs2_glock_put(struct gfs2_glock *gl)
David Teigland's avatar
David Teigland committed
216
{
217 218
	struct gfs2_sbd *sdp = gl->gl_sbd;
	struct address_space *mapping = gfs2_glock2aspace(gl);
David Teigland's avatar
David Teigland committed
219

220 221 222 223
	if (atomic_dec_and_test(&gl->gl_ref)) {
		spin_lock_bucket(gl->gl_hash);
		hlist_bl_del_rcu(&gl->gl_list);
		spin_unlock_bucket(gl->gl_hash);
224
		gfs2_glock_remove_from_lru(gl);
225
		GLOCK_BUG_ON(gl, !list_empty(&gl->gl_holders));
226 227 228
		GLOCK_BUG_ON(gl, mapping && mapping->nrpages);
		trace_gfs2_glock_put(gl);
		sdp->sd_lockstruct.ls_ops->lm_put_lock(gl);
David Teigland's avatar
David Teigland committed
229 230 231 232 233 234 235 236 237 238 239
	}
}

/**
 * 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
 */

240
static struct gfs2_glock *search_bucket(unsigned int hash,
241
					const struct gfs2_sbd *sdp,
242
					const struct lm_lockname *name)
David Teigland's avatar
David Teigland committed
243 244
{
	struct gfs2_glock *gl;
245
	struct hlist_bl_node *h;
David Teigland's avatar
David Teigland committed
246

247
	hlist_bl_for_each_entry_rcu(gl, h, &gl_hash_table[hash], gl_list) {
David Teigland's avatar
David Teigland committed
248 249
		if (!lm_name_equal(&gl->gl_name, name))
			continue;
250 251
		if (gl->gl_sbd != sdp)
			continue;
252 253
		if (atomic_inc_not_zero(&gl->gl_ref))
			return gl;
David Teigland's avatar
David Teigland committed
254 255 256 257 258
	}

	return NULL;
}

259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276
/**
 * 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;
277 278 279 280 281 282
	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;
	}
283 284 285 286 287 288 289 290 291 292 293 294
	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);
}

295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318
/**
 * 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);
		trace_gfs2_glock_queue(gh, 0);
		gfs2_holder_wake(gh);
	}
}

319 320 321 322
/**
 * do_promote - promote as many requests as possible on the current queue
 * @gl: The glock
 * 
323 324
 * Returns: 1 if there is a blocked holder at the head of the list, or 2
 *          if a type specific operation is underway.
325 326 327
 */

static int do_promote(struct gfs2_glock *gl)
328 329
__releases(&gl->gl_spin)
__acquires(&gl->gl_spin)
330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346
{
	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) {
347 348
					if (ret == 1)
						return 2;
349 350
					gh->gh_error = ret;
					list_del_init(&gh->gh_list);
Steven Whitehouse's avatar
Steven Whitehouse committed
351
					trace_gfs2_glock_queue(gh, 0);
352 353 354 355
					gfs2_holder_wake(gh);
					goto restart;
				}
				set_bit(HIF_HOLDER, &gh->gh_iflags);
Steven Whitehouse's avatar
Steven Whitehouse committed
356
				trace_gfs2_promote(gh, 1);
357 358 359 360
				gfs2_holder_wake(gh);
				goto restart;
			}
			set_bit(HIF_HOLDER, &gh->gh_iflags);
Steven Whitehouse's avatar
Steven Whitehouse committed
361
			trace_gfs2_promote(gh, 0);
362 363 364 365 366
			gfs2_holder_wake(gh);
			continue;
		}
		if (gh->gh_list.prev == &gl->gl_holders)
			return 1;
367
		do_error(gl, 0);
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
		break;
	}
	return 0;
}

/**
 * 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
407
			gfs2_glock_put_nolock(gl);
408
	}
409 410
	if (held1 && held2 && list_empty(&gl->gl_holders))
		clear_bit(GLF_QUEUED, &gl->gl_flags);
411

412 413 414 415
	if (new_state != gl->gl_target)
		/* shorten our minimum hold time */
		gl->gl_hold_time = max(gl->gl_hold_time - GL_GLOCK_HOLD_DECR,
				       GL_GLOCK_MIN_HOLD);
416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439
	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;
440
	int rv;
441 442

	spin_lock(&gl->gl_spin);
Steven Whitehouse's avatar
Steven Whitehouse committed
443
	trace_gfs2_glock_state_change(gl, state);
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
	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);
		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);
			spin_lock(&gl->gl_spin);
			if (rv) {
				do_error(gl, rv);
				goto out;
			}
		}
503 504 505
		rv = do_promote(gl);
		if (rv == 2)
			goto out_locked;
506 507 508
	}
out:
	clear_bit(GLF_LOCK, &gl->gl_flags);
509
out_locked:
510 511 512 513 514 515 516 517 518 519 520 521
	spin_unlock(&gl->gl_spin);
}

/**
 * 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
{
	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);
532 533
	GLOCK_BUG_ON(gl, gl->gl_state == target);
	GLOCK_BUG_ON(gl, gl->gl_state == gl->gl_target);
534 535 536 537 538
	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 */
	}
539
	gl->gl_req = target;
540 541 542 543 544 545 546 547
	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);
548 549 550 551 552 553
	if (sdp->sd_lockstruct.ls_ops->lm_lock)	{
		/* lock_dlm */
		ret = sdp->sd_lockstruct.ls_ops->lm_lock(gl, target, lck_flags);
		GLOCK_BUG_ON(gl, ret);
	} else { /* lock_nolock */
		finish_xmote(gl, target);
554 555 556
		if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0)
			gfs2_glock_put(gl);
	}
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
	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)
586 587
__releases(&gl->gl_spin)
__acquires(&gl->gl_spin)
588 589
{
	struct gfs2_holder *gh = NULL;
590
	int ret;
591 592 593 594 595 596 597 598 599

	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))
600
			goto out_unlock;
601 602 603
		if (nonblock)
			goto out_sched;
		set_bit(GLF_DEMOTE_IN_PROGRESS, &gl->gl_flags);
604
		GLOCK_BUG_ON(gl, gl->gl_demote_state == LM_ST_EXCLUSIVE);
605 606 607 608
		gl->gl_target = gl->gl_demote_state;
	} else {
		if (test_bit(GLF_DEMOTE, &gl->gl_flags))
			gfs2_demote_wake(gl);
609 610
		ret = do_promote(gl);
		if (ret == 0)
611
			goto out_unlock;
612
		if (ret == 2)
613
			goto out;
614 615 616 617 618 619
		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);
620
out:
621 622 623
	return;

out_sched:
624 625
	clear_bit(GLF_LOCK, &gl->gl_flags);
	smp_mb__after_clear_bit();
626 627
	gfs2_glock_hold(gl);
	if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0)
628
		gfs2_glock_put_nolock(gl);
629 630
	return;

631
out_unlock:
632
	clear_bit(GLF_LOCK, &gl->gl_flags);
633 634
	smp_mb__after_clear_bit();
	return;
635 636
}

637 638 639 640
static void delete_work_func(struct work_struct *work)
{
	struct gfs2_glock *gl = container_of(work, struct gfs2_glock, gl_delete);
	struct gfs2_sbd *sdp = gl->gl_sbd;
641
	struct gfs2_inode *ip;
642
	struct inode *inode;
643 644 645 646
	u64 no_addr = gl->gl_name.ln_number;

	ip = gl->gl_object;
	/* Note: Unsafe to dereference ip as we don't hold right refs/locks */
647 648

	if (ip)
649
		inode = gfs2_ilookup(sdp->sd_vfs, no_addr, 1);
650 651 652 653 654
	else
		inode = gfs2_lookup_by_inum(sdp, no_addr, NULL, GFS2_BLKST_UNLINKED);
	if (inode && !IS_ERR(inode)) {
		d_prune_aliases(inode);
		iput(inode);
655 656 657 658
	}
	gfs2_glock_put(gl);
}

659 660
static void glock_work_func(struct work_struct *work)
{
661
	unsigned long delay = 0;
662
	struct gfs2_glock *gl = container_of(work, struct gfs2_glock, gl_work.work);
663
	int drop_ref = 0;
664

665
	if (test_and_clear_bit(GLF_REPLY_PENDING, &gl->gl_flags)) {
666
		finish_xmote(gl, gl->gl_reply);
667 668
		drop_ref = 1;
	}
669
	spin_lock(&gl->gl_spin);
670
	if (test_bit(GLF_PENDING_DEMOTE, &gl->gl_flags) &&
671 672
	    gl->gl_state != LM_ST_UNLOCKED &&
	    gl->gl_demote_state != LM_ST_EXCLUSIVE) {
673
		unsigned long holdtime, now = jiffies;
674

675
		holdtime = gl->gl_tchange + gl->gl_hold_time;
676 677
		if (time_before(now, holdtime))
			delay = holdtime - now;
678 679 680 681 682

		if (!delay) {
			clear_bit(GLF_PENDING_DEMOTE, &gl->gl_flags);
			set_bit(GLF_DEMOTE, &gl->gl_flags);
		}
683 684
	}
	run_queue(gl, 0);
685
	spin_unlock(&gl->gl_spin);
686
	if (!delay)
687
		gfs2_glock_put(gl);
688 689 690 691 692 693
	else {
		if (gl->gl_name.ln_type != LM_TYPE_INODE)
			delay = 0;
		if (queue_delayed_work(glock_workqueue, &gl->gl_work, delay) == 0)
			gfs2_glock_put(gl);
	}
694 695
	if (drop_ref)
		gfs2_glock_put(gl);
696 697
}

David Teigland's avatar
David Teigland committed
698 699 700 701 702 703 704 705 706 707 708 709 710
/**
 * 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
 */

711
int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number,
712
		   const struct gfs2_glock_operations *glops, int create,
David Teigland's avatar
David Teigland committed
713 714
		   struct gfs2_glock **glp)
{
715
	struct super_block *s = sdp->sd_vfs;
716
	struct lm_lockname name = { .ln_number = number, .ln_type = glops->go_type };
David Teigland's avatar
David Teigland committed
717
	struct gfs2_glock *gl, *tmp;
718
	unsigned int hash = gl_hash(sdp, &name);
719
	struct address_space *mapping;
720
	struct kmem_cache *cachep;
David Teigland's avatar
David Teigland committed
721

722
	rcu_read_lock();
723
	gl = search_bucket(hash, sdp, &name);
724
	rcu_read_unlock();
David Teigland's avatar
David Teigland committed
725

726 727
	*glp = gl;
	if (gl)
David Teigland's avatar
David Teigland committed
728
		return 0;
729 730
	if (!create)
		return -ENOENT;
David Teigland's avatar
David Teigland committed
731

732
	if (glops->go_flags & GLOF_ASPACE)
733
		cachep = gfs2_glock_aspace_cachep;
734
	else
735 736
		cachep = gfs2_glock_cachep;
	gl = kmem_cache_alloc(cachep, GFP_KERNEL);
David Teigland's avatar
David Teigland committed
737 738 739
	if (!gl)
		return -ENOMEM;

740
	atomic_inc(&sdp->sd_glock_disposal);
741
	gl->gl_flags = 0;
David Teigland's avatar
David Teigland committed
742
	gl->gl_name = name;
743
	atomic_set(&gl->gl_ref, 1);
David Teigland's avatar
David Teigland committed
744
	gl->gl_state = LM_ST_UNLOCKED;
745
	gl->gl_target = LM_ST_UNLOCKED;
746
	gl->gl_demote_state = LM_ST_EXCLUSIVE;
747
	gl->gl_hash = hash;
David Teigland's avatar
David Teigland committed
748
	gl->gl_ops = glops;
749 750 751
	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;
752
	gl->gl_tchange = jiffies;
753
	gl->gl_object = NULL;
David Teigland's avatar
David Teigland committed
754
	gl->gl_sbd = sdp;
755
	gl->gl_hold_time = GL_GLOCK_DFT_HOLD;
756
	INIT_DELAYED_WORK(&gl->gl_work, glock_work_func);
757
	INIT_WORK(&gl->gl_delete, delete_work_func);
David Teigland's avatar
David Teigland committed
758

759 760 761 762 763 764 765 766 767
	mapping = gfs2_glock2aspace(gl);
	if (mapping) {
                mapping->a_ops = &gfs2_meta_aops;
		mapping->host = s->s_bdev->bd_inode;
		mapping->flags = 0;
		mapping_set_gfp_mask(mapping, GFP_NOFS);
		mapping->assoc_mapping = NULL;
		mapping->backing_dev_info = s->s_bdi;
		mapping->writeback_index = 0;
David Teigland's avatar
David Teigland committed
768 769
	}

770
	spin_lock_bucket(hash);
771
	tmp = search_bucket(hash, sdp, &name);
David Teigland's avatar
David Teigland committed
772
	if (tmp) {
773 774
		spin_unlock_bucket(hash);
		kmem_cache_free(cachep, gl);
775
		atomic_dec(&sdp->sd_glock_disposal);
David Teigland's avatar
David Teigland committed
776 777
		gl = tmp;
	} else {
778 779
		hlist_bl_add_head_rcu(&gl->gl_list, &gl_hash_table[hash]);
		spin_unlock_bucket(hash);
David Teigland's avatar
David Teigland committed
780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795
	}

	*glp = gl;

	return 0;
}

/**
 * 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
 *
 */

796
void gfs2_holder_init(struct gfs2_glock *gl, unsigned int state, unsigned flags,
David Teigland's avatar
David Teigland committed
797 798 799 800
		      struct gfs2_holder *gh)
{
	INIT_LIST_HEAD(&gh->gh_list);
	gh->gh_gl = gl;
801
	gh->gh_ip = (unsigned long)__builtin_return_address(0);
802
	gh->gh_owner_pid = get_pid(task_pid(current));
David Teigland's avatar
David Teigland committed
803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819
	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.
 *
 */

820
void gfs2_holder_reinit(unsigned int state, unsigned flags, struct gfs2_holder *gh)
David Teigland's avatar
David Teigland committed
821 822
{
	gh->gh_state = state;
823
	gh->gh_flags = flags;
824
	gh->gh_iflags = 0;
825
	gh->gh_ip = (unsigned long)__builtin_return_address(0);
Bob Peterson's avatar
Bob Peterson committed
826 827 828
	if (gh->gh_owner_pid)
		put_pid(gh->gh_owner_pid);
	gh->gh_owner_pid = get_pid(task_pid(current));
David Teigland's avatar
David Teigland committed
829 830 831 832 833 834 835 836 837 838
}

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

void gfs2_holder_uninit(struct gfs2_holder *gh)
{
839
	put_pid(gh->gh_owner_pid);
David Teigland's avatar
David Teigland committed
840 841
	gfs2_glock_put(gh->gh_gl);
	gh->gh_gl = NULL;
842
	gh->gh_ip = 0;
David Teigland's avatar
David Teigland committed
843 844
}

845 846 847 848 849 850 851 852 853 854
/**
 * gfs2_glock_holder_wait
 * @word: unused
 *
 * This function and gfs2_glock_demote_wait both show up in the WCHAN
 * field. Thus I've separated these otherwise identical functions in
 * order to be more informative to the user.
 */

static int gfs2_glock_holder_wait(void *word)
855 856 857 858 859
{
        schedule();
        return 0;
}

860 861 862 863 864 865
static int gfs2_glock_demote_wait(void *word)
{
	schedule();
	return 0;
}

866
static void wait_on_holder(struct gfs2_holder *gh)
867
{
868 869
	unsigned long time1 = jiffies;

870
	might_sleep();
871
	wait_on_bit(&gh->gh_iflags, HIF_WAIT, gfs2_glock_holder_wait, TASK_UNINTERRUPTIBLE);
872 873 874 875 876
	if (time_after(jiffies, time1 + HZ)) /* have we waited > a second? */
		/* Lengthen the minimum hold time. */
		gh->gh_gl->gl_hold_time = min(gh->gh_gl->gl_hold_time +
					      GL_GLOCK_HOLD_INCR,
					      GL_GLOCK_MAX_HOLD);
877 878
}

879
static void wait_on_demote(struct gfs2_glock *gl)
David Teigland's avatar
David Teigland committed
880
{
881
	might_sleep();
882
	wait_on_bit(&gl->gl_flags, GLF_DEMOTE, gfs2_glock_demote_wait, TASK_UNINTERRUPTIBLE);
David Teigland's avatar
David Teigland committed
883 884 885
}

/**
886 887 888
 * 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
889
 *
890 891
 * 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
892 893
 */

894
static void handle_callback(struct gfs2_glock *gl, unsigned int state,
895
			    unsigned long delay)
David Teigland's avatar
David Teigland committed
896
{
897
	int bit = delay ? GLF_PENDING_DEMOTE : GLF_DEMOTE;
David Teigland's avatar
David Teigland committed
898

899 900 901 902 903 904 905
	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
906
	}
907 908
	if (gl->gl_ops->go_callback)
		gl->gl_ops->go_callback(gl);
Steven Whitehouse's avatar
Steven Whitehouse committed
909
	trace_gfs2_demote_rq(gl);
David Teigland's avatar
David Teigland committed
910 911 912
}

/**
913
 * gfs2_glock_wait - wait on a glock acquisition
David Teigland's avatar
David Teigland committed
914 915 916 917 918
 * @gh: the glock holder
 *
 * Returns: 0 on success
 */

919
int gfs2_glock_wait(struct gfs2_holder *gh)
David Teigland's avatar
David Teigland committed
920
{
921
	wait_on_holder(gh);
David Teigland's avatar
David Teigland committed
922 923 924
	return gh->gh_error;
}

925
void gfs2_print_dbg(struct seq_file *seq, const char *fmt, ...)
926
{
927
	struct va_format vaf;
928 929 930
	va_list args;

	va_start(args, fmt);
931

932 933
	if (seq) {
		struct gfs2_glock_iter *gi = seq->private;
934
		vsprintf(gi->string, fmt, args);
935 936
		seq_printf(seq, gi->string);
	} else {
937 938 939 940
		vaf.fmt = fmt;
		vaf.va = &args;

		printk(KERN_ERR " %pV", &vaf);
941
	}
942

943 944 945
	va_end(args);
}

David Teigland's avatar
David Teigland committed
946 947 948 949
/**
 * add_to_queue - Add a holder to the wait queue (but look for recursion)
 * @gh: the holder structure to add
 *
950 951 952 953
 * 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
954 955
 */

956
static inline void add_to_queue(struct gfs2_holder *gh)
957 958
__releases(&gl->gl_spin)
__acquires(&gl->gl_spin)
David Teigland's avatar
David Teigland committed
959 960
{
	struct gfs2_glock *gl = gh->gh_gl;
961 962 963 964
	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
965

966
	BUG_ON(gh->gh_owner_pid == NULL);
967 968
	if (test_and_set_bit(HIF_WAIT, &gh->gh_iflags))
		BUG();
969

970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987
	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;
988
		}
989 990 991 992 993
		if (test_bit(HIF_HOLDER, &gh2->gh_iflags))
			continue;
		if (unlikely((gh->gh_flags & LM_FLAG_PRIORITY) && !insert_pt))
			insert_pt = &gh2->gh_list;
	}
994
	set_bit(GLF_QUEUED, &gl->gl_flags);
995
	trace_gfs2_glock_queue(gh, 1);
996 997 998 999 1000 1001 1002 1003 1004 1005 1006
	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
1007
		if (sdp->sd_lockstruct.ls_ops->lm_cancel)
1008
			sdp->sd_lockstruct.ls_ops->lm_cancel(gl);
1009
		spin_lock(&gl->gl_spin);
David Teigland's avatar
David Teigland committed
1010
	}
1011
	return;
David Teigland's avatar
David Teigland committed
1012

1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023
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_s