glock.c 52 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
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

David Teigland's avatar
David Teigland committed
12 13 14 15 16 17 18
#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>
19
#include <linux/kallsyms.h>
20
#include <linux/gfs2_ondisk.h>
21
#include <linux/list.h>
22
#include <linux/wait.h>
akpm@linux-foundation.org's avatar
akpm@linux-foundation.org committed
23
#include <linux/module.h>
David Teigland's avatar
David Teigland committed
24
#include <asm/uaccess.h>
25 26
#include <linux/seq_file.h>
#include <linux/debugfs.h>
27 28
#include <linux/kthread.h>
#include <linux/freezer.h>
29 30
#include <linux/workqueue.h>
#include <linux/jiffies.h>
31 32 33
#include <linux/rcupdate.h>
#include <linux/rculist_bl.h>
#include <linux/bit_spinlock.h>
34
#include <linux/percpu.h>
35
#include <linux/list_sort.h>
36
#include <linux/lockref.h>
David Teigland's avatar
David Teigland committed
37 38

#include "gfs2.h"
39
#include "incore.h"
David Teigland's avatar
David Teigland committed
40 41 42 43 44 45 46
#include "glock.h"
#include "glops.h"
#include "inode.h"
#include "lops.h"
#include "meta_io.h"
#include "quota.h"
#include "super.h"
47
#include "util.h"
48
#include "bmap.h"
Steven Whitehouse's avatar
Steven Whitehouse committed
49 50
#define CREATE_TRACE_POINTS
#include "trace_gfs2.h"
David Teigland's avatar
David Teigland committed
51

52
struct gfs2_glock_iter {
53 54 55 56 57
	int hash;			/* hash bucket index           */
	unsigned nhash;			/* Index within current bucket */
	struct gfs2_sbd *sdp;		/* incore superblock           */
	struct gfs2_glock *gl;		/* current glock struct        */
	loff_t last_pos;		/* last position               */
58 59
};

David Teigland's avatar
David Teigland committed
60 61
typedef void (*glock_examiner) (struct gfs2_glock * gl);

62
static void do_xmote(struct gfs2_glock *gl, struct gfs2_holder *gh, unsigned int target);
63

64
static struct dentry *gfs2_root;
65
static struct workqueue_struct *glock_workqueue;
66
struct workqueue_struct *gfs2_delete_workqueue;
67 68
static LIST_HEAD(lru_list);
static atomic_t lru_count = ATOMIC_INIT(0);
Julia Lawall's avatar
Julia Lawall committed
69
static DEFINE_SPINLOCK(lru_lock);
70

71
#define GFS2_GL_HASH_SHIFT      15
72 73 74
#define GFS2_GL_HASH_SIZE       (1 << GFS2_GL_HASH_SHIFT)
#define GFS2_GL_HASH_MASK       (GFS2_GL_HASH_SIZE - 1)

75
static struct hlist_bl_head gl_hash_table[GFS2_GL_HASH_SIZE];
76
static struct dentry *gfs2_root;
77

David Teigland's avatar
David Teigland committed
78 79 80 81 82 83 84
/**
 * gl_hash() - Turn glock number into hash bucket number
 * @lock: The glock number
 *
 * Returns: The number of the corresponding hash bucket
 */

85 86
static unsigned int gl_hash(const struct gfs2_sbd *sdp,
			    const struct lm_lockname *name)
David Teigland's avatar
David Teigland committed
87 88 89
{
	unsigned int h;

90
	h = jhash(&name->ln_number, sizeof(u64), 0);
David Teigland's avatar
David Teigland committed
91
	h = jhash(&name->ln_type, sizeof(unsigned int), h);
92
	h = jhash(&sdp, sizeof(struct gfs2_sbd *), h);
David Teigland's avatar
David Teigland committed
93 94 95 96 97
	h &= GFS2_GL_HASH_MASK;

	return h;
}

98 99
static inline void spin_lock_bucket(unsigned int hash)
{
100
	hlist_bl_lock(&gl_hash_table[hash]);
101
}
David Teigland's avatar
David Teigland committed
102

103 104
static inline void spin_unlock_bucket(unsigned int hash)
{
105
	hlist_bl_unlock(&gl_hash_table[hash]);
106
}
David Teigland's avatar
David Teigland committed
107

108
static void gfs2_glock_dealloc(struct rcu_head *rcu)
David Teigland's avatar
David Teigland committed
109
{
110
	struct gfs2_glock *gl = container_of(rcu, struct gfs2_glock, gl_rcu);
David Teigland's avatar
David Teigland committed
111

112
	if (gl->gl_ops->go_flags & GLOF_ASPACE) {
113
		kmem_cache_free(gfs2_glock_aspace_cachep, gl);
114
	} else {
115
		kfree(gl->gl_lksb.sb_lvbptr);
116
		kmem_cache_free(gfs2_glock_cachep, gl);
117
	}
118 119 120
}

void gfs2_glock_free(struct gfs2_glock *gl)
David Teigland's avatar
David Teigland committed
121 122 123
{
	struct gfs2_sbd *sdp = gl->gl_sbd;

124
	call_rcu(&gl->gl_rcu, gfs2_glock_dealloc);
125 126
	if (atomic_dec_and_test(&sdp->sd_glock_disposal))
		wake_up(&sdp->sd_glock_wait);
David Teigland's avatar
David Teigland committed
127 128 129 130 131 132 133 134
}

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

135
static void gfs2_glock_hold(struct gfs2_glock *gl)
David Teigland's avatar
David Teigland committed
136
{
137 138
	GLOCK_BUG_ON(gl, __lockref_is_dead(&gl->gl_lockref));
	lockref_get(&gl->gl_lockref);
David Teigland's avatar
David Teigland committed
139 140
}

141 142 143 144 145 146 147 148 149 150 151 152 153
/**
 * 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;
154
	if (!list_empty(&gl->gl_holders))
155 156 157 158 159 160
		return 0;
	if (glops->go_demote_ok)
		return glops->go_demote_ok(gl);
	return 1;
}

161

162 163 164 165 166 167 168 169 170 171
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);
172
	set_bit(GLF_LRU, &gl->gl_flags);
173 174 175
	spin_unlock(&lru_lock);
}

176
static void __gfs2_glock_remove_from_lru(struct gfs2_glock *gl)
177 178 179 180 181 182
{
	if (!list_empty(&gl->gl_lru)) {
		list_del_init(&gl->gl_lru);
		atomic_dec(&lru_count);
		clear_bit(GLF_LRU, &gl->gl_flags);
	}
183 184 185 186 187 188
}

static void gfs2_glock_remove_from_lru(struct gfs2_glock *gl)
{
	spin_lock(&lru_lock);
	__gfs2_glock_remove_from_lru(gl);
189 190 191
	spin_unlock(&lru_lock);
}

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

198
void gfs2_glock_put(struct gfs2_glock *gl)
David Teigland's avatar
David Teigland committed
199
{
200 201
	struct gfs2_sbd *sdp = gl->gl_sbd;
	struct address_space *mapping = gfs2_glock2aspace(gl);
David Teigland's avatar
David Teigland committed
202

203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218
	if (lockref_put_or_lock(&gl->gl_lockref))
		return;

	lockref_mark_dead(&gl->gl_lockref);

	spin_lock(&lru_lock);
	__gfs2_glock_remove_from_lru(gl);
	spin_unlock(&lru_lock);
	spin_unlock(&gl->gl_lockref.lock);
	spin_lock_bucket(gl->gl_hash);
	hlist_bl_del_rcu(&gl->gl_list);
	spin_unlock_bucket(gl->gl_hash);
	GLOCK_BUG_ON(gl, !list_empty(&gl->gl_holders));
	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
219 220 221 222 223 224 225 226 227 228
}

/**
 * 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
 */

229
static struct gfs2_glock *search_bucket(unsigned int hash,
230
					const struct gfs2_sbd *sdp,
231
					const struct lm_lockname *name)
David Teigland's avatar
David Teigland committed
232 233
{
	struct gfs2_glock *gl;
234
	struct hlist_bl_node *h;
David Teigland's avatar
David Teigland committed
235

236
	hlist_bl_for_each_entry_rcu(gl, h, &gl_hash_table[hash], gl_list) {
David Teigland's avatar
David Teigland committed
237 238
		if (!lm_name_equal(&gl->gl_name, name))
			continue;
239 240
		if (gl->gl_sbd != sdp)
			continue;
241
		if (lockref_get_not_dead(&gl->gl_lockref))
242
			return gl;
David Teigland's avatar
David Teigland committed
243 244 245 246 247
	}

	return NULL;
}

248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265
/**
 * 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;
266 267 268 269 270 271
	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;
	}
272 273 274 275 276 277 278 279 280 281 282 283
	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);
}

284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307
/**
 * 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);
	}
}

308 309 310 311
/**
 * do_promote - promote as many requests as possible on the current queue
 * @gl: The glock
 * 
312 313
 * Returns: 1 if there is a blocked holder at the head of the list, or 2
 *          if a type specific operation is underway.
314 315 316
 */

static int do_promote(struct gfs2_glock *gl)
317 318
__releases(&gl->gl_spin)
__acquires(&gl->gl_spin)
319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335
{
	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) {
336 337
					if (ret == 1)
						return 2;
338 339
					gh->gh_error = ret;
					list_del_init(&gh->gh_list);
Steven Whitehouse's avatar
Steven Whitehouse committed
340
					trace_gfs2_glock_queue(gh, 0);
341 342 343 344
					gfs2_holder_wake(gh);
					goto restart;
				}
				set_bit(HIF_HOLDER, &gh->gh_iflags);
Steven Whitehouse's avatar
Steven Whitehouse committed
345
				trace_gfs2_promote(gh, 1);
346 347 348 349
				gfs2_holder_wake(gh);
				goto restart;
			}
			set_bit(HIF_HOLDER, &gh->gh_iflags);
Steven Whitehouse's avatar
Steven Whitehouse committed
350
			trace_gfs2_promote(gh, 0);
351 352 353 354 355
			gfs2_holder_wake(gh);
			continue;
		}
		if (gh->gh_list.prev == &gl->gl_holders)
			return 1;
356
		do_error(gl, 0);
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
		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) {
393
		GLOCK_BUG_ON(gl, __lockref_is_dead(&gl->gl_lockref));
394
		if (held2)
395
			gl->gl_lockref.count++;
396
		else
397
			gl->gl_lockref.count--;
398
	}
399 400
	if (held1 && held2 && list_empty(&gl->gl_holders))
		clear_bit(GLF_QUEUED, &gl->gl_flags);
401

402 403 404 405
	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);
406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429
	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;
430
	int rv;
431 432

	spin_lock(&gl->gl_spin);
Steven Whitehouse's avatar
Steven Whitehouse committed
433
	trace_gfs2_glock_state_change(gl, state);
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
	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 */
473
			pr_err("wanted %u got %u\n", gl->gl_target, state);
474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492
			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;
			}
		}
493 494 495
		rv = do_promote(gl);
		if (rv == 2)
			goto out_locked;
496 497 498
	}
out:
	clear_bit(GLF_LOCK, &gl->gl_flags);
499
out_locked:
500 501 502 503 504 505 506 507 508 509 510 511
	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)
512 513
__releases(&gl->gl_spin)
__acquires(&gl->gl_spin)
514 515 516 517 518 519 520 521
{
	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);
522 523
	GLOCK_BUG_ON(gl, gl->gl_state == target);
	GLOCK_BUG_ON(gl, gl->gl_state == gl->gl_target);
524 525 526 527 528
	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 */
	}
529
	gl->gl_req = target;
530 531 532 533 534
	set_bit(GLF_BLOCKING, &gl->gl_flags);
	if ((gl->gl_req == LM_ST_UNLOCKED) ||
	    (gl->gl_state == LM_ST_EXCLUSIVE) ||
	    (lck_flags & (LM_FLAG_TRY|LM_FLAG_TRY_1CB)))
		clear_bit(GLF_BLOCKING, &gl->gl_flags);
535
	spin_unlock(&gl->gl_spin);
536 537
	if (glops->go_sync)
		glops->go_sync(gl);
538 539 540 541 542
	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);
543 544 545
	if (sdp->sd_lockstruct.ls_ops->lm_lock)	{
		/* lock_dlm */
		ret = sdp->sd_lockstruct.ls_ops->lm_lock(gl, target, lck_flags);
546
		if (ret) {
547
			pr_err("lm_lock ret %d\n", ret);
548 549
			GLOCK_BUG_ON(gl, 1);
		}
550 551
	} else { /* lock_nolock */
		finish_xmote(gl, target);
552 553 554
		if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0)
			gfs2_glock_put(gl);
	}
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
	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)
584 585
__releases(&gl->gl_spin)
__acquires(&gl->gl_spin)
586 587
{
	struct gfs2_holder *gh = NULL;
588
	int ret;
589 590 591 592 593 594 595 596 597

	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))
598
			goto out_unlock;
599 600 601
		if (nonblock)
			goto out_sched;
		set_bit(GLF_DEMOTE_IN_PROGRESS, &gl->gl_flags);
602
		GLOCK_BUG_ON(gl, gl->gl_demote_state == LM_ST_EXCLUSIVE);
603 604 605 606
		gl->gl_target = gl->gl_demote_state;
	} else {
		if (test_bit(GLF_DEMOTE, &gl->gl_flags))
			gfs2_demote_wake(gl);
607 608
		ret = do_promote(gl);
		if (ret == 0)
609
			goto out_unlock;
610
		if (ret == 2)
611
			goto out;
612 613 614 615 616 617
		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);
618
out:
619 620 621
	return;

out_sched:
622 623
	clear_bit(GLF_LOCK, &gl->gl_flags);
	smp_mb__after_clear_bit();
624
	gl->gl_lockref.count++;
625
	if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0)
626
		gl->gl_lockref.count--;
627 628
	return;

629
out_unlock:
630
	clear_bit(GLF_LOCK, &gl->gl_flags);
631 632
	smp_mb__after_clear_bit();
	return;
633 634
}

635 636 637 638
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;
639
	struct gfs2_inode *ip;
640
	struct inode *inode;
641 642 643 644
	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 */
645 646

	if (ip)
647
		inode = gfs2_ilookup(sdp->sd_vfs, no_addr, 1);
648 649 650 651 652
	else
		inode = gfs2_lookup_by_inum(sdp, no_addr, NULL, GFS2_BLKST_UNLINKED);
	if (inode && !IS_ERR(inode)) {
		d_prune_aliases(inode);
		iput(inode);
653 654 655 656
	}
	gfs2_glock_put(gl);
}

657 658
static void glock_work_func(struct work_struct *work)
{
659
	unsigned long delay = 0;
660
	struct gfs2_glock *gl = container_of(work, struct gfs2_glock, gl_work.work);
661
	int drop_ref = 0;
662

663
	if (test_and_clear_bit(GLF_REPLY_PENDING, &gl->gl_flags)) {
664
		finish_xmote(gl, gl->gl_reply);
665 666
		drop_ref = 1;
	}
667
	spin_lock(&gl->gl_spin);
668
	if (test_bit(GLF_PENDING_DEMOTE, &gl->gl_flags) &&
669 670
	    gl->gl_state != LM_ST_UNLOCKED &&
	    gl->gl_demote_state != LM_ST_EXCLUSIVE) {
671
		unsigned long holdtime, now = jiffies;
672

673
		holdtime = gl->gl_tchange + gl->gl_hold_time;
674 675
		if (time_before(now, holdtime))
			delay = holdtime - now;
676 677 678 679 680

		if (!delay) {
			clear_bit(GLF_PENDING_DEMOTE, &gl->gl_flags);
			set_bit(GLF_DEMOTE, &gl->gl_flags);
		}
681 682
	}
	run_queue(gl, 0);
683
	spin_unlock(&gl->gl_spin);
684
	if (!delay)
685
		gfs2_glock_put(gl);
686 687 688 689 690 691
	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);
	}
692 693
	if (drop_ref)
		gfs2_glock_put(gl);
694 695
}

David Teigland's avatar
David Teigland committed
696 697 698 699 700 701 702 703 704 705 706 707 708
/**
 * 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
 */

709
int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number,
710
		   const struct gfs2_glock_operations *glops, int create,
David Teigland's avatar
David Teigland committed
711 712
		   struct gfs2_glock **glp)
{
713
	struct super_block *s = sdp->sd_vfs;
714
	struct lm_lockname name = { .ln_number = number, .ln_type = glops->go_type };
David Teigland's avatar
David Teigland committed
715
	struct gfs2_glock *gl, *tmp;
716
	unsigned int hash = gl_hash(sdp, &name);
717
	struct address_space *mapping;
718
	struct kmem_cache *cachep;
David Teigland's avatar
David Teigland committed
719

720
	rcu_read_lock();
721
	gl = search_bucket(hash, sdp, &name);
722
	rcu_read_unlock();
David Teigland's avatar
David Teigland committed
723

724 725
	*glp = gl;
	if (gl)
David Teigland's avatar
David Teigland committed
726
		return 0;
727 728
	if (!create)
		return -ENOENT;
David Teigland's avatar
David Teigland committed
729

730
	if (glops->go_flags & GLOF_ASPACE)
731
		cachep = gfs2_glock_aspace_cachep;
732
	else
733 734
		cachep = gfs2_glock_cachep;
	gl = kmem_cache_alloc(cachep, GFP_KERNEL);
David Teigland's avatar
David Teigland committed
735 736 737
	if (!gl)
		return -ENOMEM;

738 739 740
	memset(&gl->gl_lksb, 0, sizeof(struct dlm_lksb));

	if (glops->go_flags & GLOF_LVB) {
741 742
		gl->gl_lksb.sb_lvbptr = kzalloc(GFS2_MIN_LVB_SIZE, GFP_KERNEL);
		if (!gl->gl_lksb.sb_lvbptr) {
743 744 745 746 747
			kmem_cache_free(cachep, gl);
			return -ENOMEM;
		}
	}

748
	atomic_inc(&sdp->sd_glock_disposal);
749
	gl->gl_sbd = sdp;
750
	gl->gl_flags = 0;
David Teigland's avatar
David Teigland committed
751
	gl->gl_name = name;
752
	gl->gl_lockref.count = 1;
David Teigland's avatar
David Teigland committed
753
	gl->gl_state = LM_ST_UNLOCKED;
754
	gl->gl_target = LM_ST_UNLOCKED;
755
	gl->gl_demote_state = LM_ST_EXCLUSIVE;
756
	gl->gl_hash = hash;
David Teigland's avatar
David Teigland committed
757
	gl->gl_ops = glops;
758 759 760 761 762 763 764
	gl->gl_dstamp = ktime_set(0, 0);
	preempt_disable();
	/* We use the global stats to estimate the initial per-glock stats */
	gl->gl_stats = this_cpu_ptr(sdp->sd_lkstats)->lkstats[glops->go_type];
	preempt_enable();
	gl->gl_stats.stats[GFS2_LKS_DCOUNT] = 0;
	gl->gl_stats.stats[GFS2_LKS_QCOUNT] = 0;
765
	gl->gl_tchange = jiffies;
766
	gl->gl_object = NULL;
767
	gl->gl_hold_time = GL_GLOCK_DFT_HOLD;
768
	INIT_DELAYED_WORK(&gl->gl_work, glock_work_func);
769
	INIT_WORK(&gl->gl_delete, delete_work_func);
David Teigland's avatar
David Teigland committed
770

771 772 773 774 775 776
	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);
777
		mapping->private_data = NULL;
778 779
		mapping->backing_dev_info = s->s_bdi;
		mapping->writeback_index = 0;
David Teigland's avatar
David Teigland committed
780 781
	}

782
	spin_lock_bucket(hash);
783
	tmp = search_bucket(hash, sdp, &name);
David Teigland's avatar
David Teigland committed
784
	if (tmp) {
785
		spin_unlock_bucket(hash);
786
		kfree(gl->gl_lksb.sb_lvbptr);
787
		kmem_cache_free(cachep, gl);
788
		atomic_dec(&sdp->sd_glock_disposal);
David Teigland's avatar
David Teigland committed
789 790
		gl = tmp;
	} else {
791 792
		hlist_bl_add_head_rcu(&gl->gl_list, &gl_hash_table[hash]);
		spin_unlock_bucket(hash);
David Teigland's avatar
David Teigland committed
793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808
	}

	*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
 *
 */

809
void gfs2_holder_init(struct gfs2_glock *gl, unsigned int state, unsigned flags,
David Teigland's avatar
David Teigland committed
810 811 812 813
		      struct gfs2_holder *gh)
{
	INIT_LIST_HEAD(&gh->gh_list);
	gh->gh_gl = gl;
814
	gh->gh_ip = (unsigned long)__builtin_return_address(0);
815
	gh->gh_owner_pid = get_pid(task_pid(current));
David Teigland's avatar
David Teigland committed
816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832
	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.
 *
 */

833
void gfs2_holder_reinit(unsigned int state, unsigned flags, struct gfs2_holder *gh)
David Teigland's avatar
David Teigland committed
834 835
{
	gh->gh_state = state;
836
	gh->gh_flags = flags;
837
	gh->gh_iflags = 0;
838
	gh->gh_ip = (unsigned long)__builtin_return_address(0);
Bob Peterson's avatar
Bob Peterson committed
839 840 841
	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
842 843 844 845 846 847 848 849 850 851
}

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

void gfs2_holder_uninit(struct gfs2_holder *gh)
{
852
	put_pid(gh->gh_owner_pid);
David Teigland's avatar
David Teigland committed
853 854
	gfs2_glock_put(gh->gh_gl);
	gh->gh_gl = NULL;
855
	gh->gh_ip = 0;
David Teigland's avatar
David Teigland committed
856 857
}

858 859 860 861 862 863 864 865 866 867
/**
 * 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)
868 869 870 871 872
{
        schedule();
        return 0;
}

873 874 875 876 877 878
static int gfs2_glock_demote_wait(void *word)
{
	schedule();
	return 0;
}

879 880 881 882 883 884 885 886
/**
 * gfs2_glock_wait - wait on a glock acquisition
 * @gh: the glock holder
 *
 * Returns: 0 on success
 */

int gfs2_glock_wait(struct gfs2_holder *gh)
887
{
888 889
	unsigned long time1 = jiffies;

890
	might_sleep();
891
	wait_on_bit(&gh->gh_iflags, HIF_WAIT, gfs2_glock_holder_wait, TASK_UNINTERRUPTIBLE);
892 893 894 895 896
	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);
897
	return gh->gh_error;
898 899
}

David Teigland's avatar
David Teigland committed
900
/**
901 902 903
 * 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
904
 *
905 906
 * 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
907 908
 */

909
static void handle_callback(struct gfs2_glock *gl, unsigned int state,
910
			    unsigned long delay, bool remote)
David Teigland's avatar
David Teigland committed
911
{
912
	int bit = delay ? GLF_PENDING_DEMOTE : GLF_DEMOTE;
David Teigland's avatar
David Teigland committed
913

914 915 916 917 918 919 920
	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
921
	}
922
	if (gl->gl_ops->go_callback)
923
		gl->gl_ops->go_callback(gl, remote);
924
	trace_gfs2_demote_rq(gl, remote);
David Teigland's avatar
David Teigland committed
925 926
}

927
void gfs2_print_dbg(struct seq_file *seq, const char *fmt, ...)
928
{
929
	struct va_format vaf;
930 931 932
	va_list args;

	va_start(args, fmt);
933

934
	if (seq) {
935
		seq_vprintf(seq, fmt, args);
936
	} else {
937 938 939
		vaf.fmt = fmt;
		vaf.va = &args;

940
		pr_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
	struct gfs2_sbd *sdp = gl->gl_sbd;
	struct list_head *insert_pt = NULL;
	struct gfs2_holder *gh2;
964
	int try_futile = 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
	if (gh->gh_flags & (LM_FLAG_TRY | LM_FLAG_TRY_1CB)) {
		if (test_bit(GLF_LOCK, &gl->gl_flags))
972
			try_futile = !may_grant(gl, gh);
973 974 975 976 977 978 979 980
		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;
981 982
		if (try_futile &&
		    !(gh2->gh_flags & (LM_FLAG_TRY | LM_FLAG_TRY_1CB))) {
983 984 985 986
fail:
			gh->gh_error = GLR_TRYFAILED;
			gfs2_holder_wake(gh);
			return;
987
		}
988 989 990 991 992
		if (test_bit(HIF_HOLDER, &gh2->gh_iflags))
			continue;
		if (unlikely((gh->gh_flags & LM_FLAG_PRIORITY) && !insert_pt))
			insert_pt = &gh2->gh_list;
	}
993
	set_bit(GLF_QUEUED, &gl->gl_flags);
994
	trace_gfs2_glock_queue(gh, 1);
995 996
	gfs2_glstats_inc(gl, GFS2_LKS_QCOUNT);
	gfs2_sbstats_inc(gl, GFS2_LKS_QCOUNT);
997 998 999 1000 1001 1002 1003 1004 1005 1006 1007
	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
1008
		if (sdp->sd_lockstruct.ls_ops->lm_cancel)
1009
			sdp->sd_lockstruct.ls_ops->lm_cancel(gl);
1010
		spin_lock(&gl->gl_spin);
David Teigland's avatar
David Teigland committed
1011
	}
1012
	return;
David Teigland's avatar
David Teigland committed
1013

1014
trap_recursive:
1015 1016 1017
	pr_err("original: %pSR\n", (void *)gh2->gh_ip);
	pr_err("pid: %d\n", pid_nr(gh2->gh_owner_pid));
	pr_err("lock type: %d req lock state : %d\n",
1018
	       gh2->gh_gl->gl_name.ln_type, gh2->gh_state);
1019 1020 1021
	pr_err("new: %pSR\n", (void *)gh->gh_ip);
	pr_err("pid: %d\n", pid_nr(gh->gh_owner_pid));
	pr_err("lock type: %d req lock state : %d\n",
1022
	       gh->gh_gl->gl_name.ln_type, gh->gh_state);
1023
	gfs2_dump_glock(NULL, gl);