glock.c 51.5 KB
Newer Older
D
David Teigland 已提交
1 2
/*
 * Copyright (C) Sistina Software, Inc.  1997-2003 All rights reserved.
3
 * Copyright (C) 2004-2008 Red Hat, Inc.  All rights reserved.
D
David Teigland 已提交
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.
D
David Teigland 已提交
8 9
 */

10 11
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

D
David Teigland 已提交
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>
S
Steven Whitehouse 已提交
19
#include <linux/kallsyms.h>
20
#include <linux/gfs2_ondisk.h>
21
#include <linux/list.h>
22
#include <linux/wait.h>
A
akpm@linux-foundation.org 已提交
23
#include <linux/module.h>
D
David Teigland 已提交
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>
S
Steven Whitehouse 已提交
36
#include <linux/lockref.h>
D
David Teigland 已提交
37 38

#include "gfs2.h"
39
#include "incore.h"
D
David Teigland 已提交
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"
S
Steven Whitehouse 已提交
49 50
#define CREATE_TRACE_POINTS
#include "trace_gfs2.h"
D
David Teigland 已提交
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
};

D
David Teigland 已提交
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);
J
Julia Lawall 已提交
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

D
David Teigland 已提交
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)
D
David Teigland 已提交
87 88 89
{
	unsigned int h;

90
	h = jhash(&name->ln_number, sizeof(u64), 0);
D
David Teigland 已提交
91
	h = jhash(&name->ln_type, sizeof(unsigned int), h);
92
	h = jhash(&sdp, sizeof(struct gfs2_sbd *), h);
D
David Teigland 已提交
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
}
D
David Teigland 已提交
102

103 104
static inline void spin_unlock_bucket(unsigned int hash)
{
105
	hlist_bl_unlock(&gl_hash_table[hash]);
106
}
D
David Teigland 已提交
107

108
static void gfs2_glock_dealloc(struct rcu_head *rcu)
D
David Teigland 已提交
109
{
110
	struct gfs2_glock *gl = container_of(rcu, struct gfs2_glock, gl_rcu);
D
David Teigland 已提交
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)
D
David Teigland 已提交
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);
D
David Teigland 已提交
127 128 129 130 131 132 133 134
}

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

S
Steven Whitehouse 已提交
135
static void gfs2_glock_hold(struct gfs2_glock *gl)
D
David Teigland 已提交
136
{
S
Steven Whitehouse 已提交
137 138
	GLOCK_BUG_ON(gl, __lockref_is_dead(&gl->gl_lockref));
	lockref_get(&gl->gl_lockref);
D
David Teigland 已提交
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
	spin_lock(&lru_lock);
179 180 181 182 183 184 185 186
	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);
}

D
David Teigland 已提交
187 188 189 190 191 192
/**
 * gfs2_glock_put() - Decrement reference count on glock
 * @gl: The glock to put
 *
 */

193
void gfs2_glock_put(struct gfs2_glock *gl)
D
David Teigland 已提交
194
{
195 196
	struct gfs2_sbd *sdp = gl->gl_sbd;
	struct address_space *mapping = gfs2_glock2aspace(gl);
D
David Teigland 已提交
197

S
Steven Whitehouse 已提交
198 199 200 201 202
	if (lockref_put_or_lock(&gl->gl_lockref))
		return;

	lockref_mark_dead(&gl->gl_lockref);

203
	gfs2_glock_remove_from_lru(gl);
S
Steven Whitehouse 已提交
204 205 206 207 208 209 210 211
	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);
D
David Teigland 已提交
212 213 214 215 216 217 218 219 220 221
}

/**
 * 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
 */

222
static struct gfs2_glock *search_bucket(unsigned int hash,
223
					const struct gfs2_sbd *sdp,
224
					const struct lm_lockname *name)
D
David Teigland 已提交
225 226
{
	struct gfs2_glock *gl;
227
	struct hlist_bl_node *h;
D
David Teigland 已提交
228

229
	hlist_bl_for_each_entry_rcu(gl, h, &gl_hash_table[hash], gl_list) {
D
David Teigland 已提交
230 231
		if (!lm_name_equal(&gl->gl_name, name))
			continue;
232 233
		if (gl->gl_sbd != sdp)
			continue;
S
Steven Whitehouse 已提交
234
		if (lockref_get_not_dead(&gl->gl_lockref))
235
			return gl;
D
David Teigland 已提交
236 237 238 239 240
	}

	return NULL;
}

241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258
/**
 * 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;
259 260 261 262 263 264
	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;
	}
265 266 267 268 269 270 271 272
	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);
273
	smp_mb__after_atomic();
274 275 276
	wake_up_bit(&gh->gh_iflags, HIF_WAIT);
}

277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300
/**
 * 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);
	}
}

301 302 303 304
/**
 * do_promote - promote as many requests as possible on the current queue
 * @gl: The glock
 * 
305 306
 * Returns: 1 if there is a blocked holder at the head of the list, or 2
 *          if a type specific operation is underway.
307 308 309
 */

static int do_promote(struct gfs2_glock *gl)
310 311
__releases(&gl->gl_spin)
__acquires(&gl->gl_spin)
312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328
{
	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) {
329 330
					if (ret == 1)
						return 2;
331 332
					gh->gh_error = ret;
					list_del_init(&gh->gh_list);
S
Steven Whitehouse 已提交
333
					trace_gfs2_glock_queue(gh, 0);
334 335 336 337
					gfs2_holder_wake(gh);
					goto restart;
				}
				set_bit(HIF_HOLDER, &gh->gh_iflags);
S
Steven Whitehouse 已提交
338
				trace_gfs2_promote(gh, 1);
339 340 341 342
				gfs2_holder_wake(gh);
				goto restart;
			}
			set_bit(HIF_HOLDER, &gh->gh_iflags);
S
Steven Whitehouse 已提交
343
			trace_gfs2_promote(gh, 0);
344 345 346 347 348
			gfs2_holder_wake(gh);
			continue;
		}
		if (gh->gh_list.prev == &gl->gl_holders)
			return 1;
349
		do_error(gl, 0);
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
		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) {
S
Steven Whitehouse 已提交
386
		GLOCK_BUG_ON(gl, __lockref_is_dead(&gl->gl_lockref));
387
		if (held2)
S
Steven Whitehouse 已提交
388
			gl->gl_lockref.count++;
389
		else
S
Steven Whitehouse 已提交
390
			gl->gl_lockref.count--;
391
	}
392 393
	if (held1 && held2 && list_empty(&gl->gl_holders))
		clear_bit(GLF_QUEUED, &gl->gl_flags);
394

395 396 397 398
	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);
399 400 401 402 403 404 405 406
	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);
407
	smp_mb__after_atomic();
408 409 410 411 412 413 414 415 416 417 418 419 420 421 422
	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;
423
	int rv;
424 425

	spin_lock(&gl->gl_spin);
S
Steven Whitehouse 已提交
426
	trace_gfs2_glock_state_change(gl, state);
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
	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 */
466
			pr_err("wanted %u got %u\n", gl->gl_target, state);
467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485
			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;
			}
		}
486 487 488
		rv = do_promote(gl);
		if (rv == 2)
			goto out_locked;
489 490 491
	}
out:
	clear_bit(GLF_LOCK, &gl->gl_flags);
492
out_locked:
493 494 495 496 497 498 499 500 501 502 503 504
	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)
505 506
__releases(&gl->gl_spin)
__acquires(&gl->gl_spin)
507 508 509 510 511 512 513 514
{
	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);
515 516
	GLOCK_BUG_ON(gl, gl->gl_state == target);
	GLOCK_BUG_ON(gl, gl->gl_state == gl->gl_target);
517 518 519 520 521
	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 */
	}
522
	gl->gl_req = target;
523 524 525 526 527
	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);
528
	spin_unlock(&gl->gl_spin);
529 530
	if (glops->go_sync)
		glops->go_sync(gl);
531 532 533 534 535
	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);
536 537 538
	if (sdp->sd_lockstruct.ls_ops->lm_lock)	{
		/* lock_dlm */
		ret = sdp->sd_lockstruct.ls_ops->lm_lock(gl, target, lck_flags);
539
		if (ret) {
540
			pr_err("lm_lock ret %d\n", ret);
541 542
			GLOCK_BUG_ON(gl, 1);
		}
543 544
	} else { /* lock_nolock */
		finish_xmote(gl, target);
545 546 547
		if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0)
			gfs2_glock_put(gl);
	}
548

549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576
	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)
577 578
__releases(&gl->gl_spin)
__acquires(&gl->gl_spin)
579 580
{
	struct gfs2_holder *gh = NULL;
581
	int ret;
582 583 584 585 586 587 588 589 590

	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))
591
			goto out_unlock;
592 593 594
		if (nonblock)
			goto out_sched;
		set_bit(GLF_DEMOTE_IN_PROGRESS, &gl->gl_flags);
595
		GLOCK_BUG_ON(gl, gl->gl_demote_state == LM_ST_EXCLUSIVE);
596 597 598 599
		gl->gl_target = gl->gl_demote_state;
	} else {
		if (test_bit(GLF_DEMOTE, &gl->gl_flags))
			gfs2_demote_wake(gl);
600 601
		ret = do_promote(gl);
		if (ret == 0)
602
			goto out_unlock;
603
		if (ret == 2)
604
			goto out;
605 606 607 608 609 610
		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);
611
out:
612 613 614
	return;

out_sched:
615
	clear_bit(GLF_LOCK, &gl->gl_flags);
616
	smp_mb__after_atomic();
S
Steven Whitehouse 已提交
617
	gl->gl_lockref.count++;
618
	if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0)
S
Steven Whitehouse 已提交
619
		gl->gl_lockref.count--;
620 621
	return;

622
out_unlock:
623
	clear_bit(GLF_LOCK, &gl->gl_flags);
624
	smp_mb__after_atomic();
625
	return;
626 627
}

628 629 630 631
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;
632
	struct gfs2_inode *ip;
633
	struct inode *inode;
634 635 636 637
	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 */
638 639

	if (ip)
640
		inode = gfs2_ilookup(sdp->sd_vfs, no_addr, 1);
641 642 643 644 645
	else
		inode = gfs2_lookup_by_inum(sdp, no_addr, NULL, GFS2_BLKST_UNLINKED);
	if (inode && !IS_ERR(inode)) {
		d_prune_aliases(inode);
		iput(inode);
646 647 648 649
	}
	gfs2_glock_put(gl);
}

650 651
static void glock_work_func(struct work_struct *work)
{
652
	unsigned long delay = 0;
653
	struct gfs2_glock *gl = container_of(work, struct gfs2_glock, gl_work.work);
654
	int drop_ref = 0;
655

656
	if (test_and_clear_bit(GLF_REPLY_PENDING, &gl->gl_flags)) {
657
		finish_xmote(gl, gl->gl_reply);
658 659
		drop_ref = 1;
	}
660
	spin_lock(&gl->gl_spin);
661
	if (test_bit(GLF_PENDING_DEMOTE, &gl->gl_flags) &&
662 663
	    gl->gl_state != LM_ST_UNLOCKED &&
	    gl->gl_demote_state != LM_ST_EXCLUSIVE) {
664
		unsigned long holdtime, now = jiffies;
665

666
		holdtime = gl->gl_tchange + gl->gl_hold_time;
667 668
		if (time_before(now, holdtime))
			delay = holdtime - now;
669 670 671 672 673

		if (!delay) {
			clear_bit(GLF_PENDING_DEMOTE, &gl->gl_flags);
			set_bit(GLF_DEMOTE, &gl->gl_flags);
		}
674 675
	}
	run_queue(gl, 0);
676
	spin_unlock(&gl->gl_spin);
677
	if (!delay)
678
		gfs2_glock_put(gl);
679 680 681 682 683 684
	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);
	}
685 686
	if (drop_ref)
		gfs2_glock_put(gl);
687 688
}

D
David Teigland 已提交
689 690 691 692 693 694 695 696 697 698 699 700 701
/**
 * 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
 */

702
int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number,
703
		   const struct gfs2_glock_operations *glops, int create,
D
David Teigland 已提交
704 705
		   struct gfs2_glock **glp)
{
706
	struct super_block *s = sdp->sd_vfs;
707
	struct lm_lockname name = { .ln_number = number, .ln_type = glops->go_type };
D
David Teigland 已提交
708
	struct gfs2_glock *gl, *tmp;
709
	unsigned int hash = gl_hash(sdp, &name);
710
	struct address_space *mapping;
711
	struct kmem_cache *cachep;
D
David Teigland 已提交
712

713
	rcu_read_lock();
714
	gl = search_bucket(hash, sdp, &name);
715
	rcu_read_unlock();
D
David Teigland 已提交
716

717 718
	*glp = gl;
	if (gl)
D
David Teigland 已提交
719
		return 0;
720 721
	if (!create)
		return -ENOENT;
D
David Teigland 已提交
722

723
	if (glops->go_flags & GLOF_ASPACE)
724
		cachep = gfs2_glock_aspace_cachep;
725
	else
726
		cachep = gfs2_glock_cachep;
727
	gl = kmem_cache_alloc(cachep, GFP_NOFS);
D
David Teigland 已提交
728 729 730
	if (!gl)
		return -ENOMEM;

731 732 733
	memset(&gl->gl_lksb, 0, sizeof(struct dlm_lksb));

	if (glops->go_flags & GLOF_LVB) {
734
		gl->gl_lksb.sb_lvbptr = kzalloc(GFS2_MIN_LVB_SIZE, GFP_NOFS);
735
		if (!gl->gl_lksb.sb_lvbptr) {
736 737 738 739 740
			kmem_cache_free(cachep, gl);
			return -ENOMEM;
		}
	}

741
	atomic_inc(&sdp->sd_glock_disposal);
742
	gl->gl_sbd = sdp;
743
	gl->gl_flags = 0;
D
David Teigland 已提交
744
	gl->gl_name = name;
S
Steven Whitehouse 已提交
745
	gl->gl_lockref.count = 1;
D
David Teigland 已提交
746
	gl->gl_state = LM_ST_UNLOCKED;
747
	gl->gl_target = LM_ST_UNLOCKED;
748
	gl->gl_demote_state = LM_ST_EXCLUSIVE;
749
	gl->gl_hash = hash;
D
David Teigland 已提交
750
	gl->gl_ops = glops;
751 752 753 754 755 756 757
	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;
758
	gl->gl_tchange = jiffies;
759
	gl->gl_object = NULL;
760
	gl->gl_hold_time = GL_GLOCK_DFT_HOLD;
761
	INIT_DELAYED_WORK(&gl->gl_work, glock_work_func);
762
	INIT_WORK(&gl->gl_delete, delete_work_func);
D
David Teigland 已提交
763

764 765 766 767 768 769
	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);
770
		mapping->private_data = NULL;
771 772
		mapping->backing_dev_info = s->s_bdi;
		mapping->writeback_index = 0;
D
David Teigland 已提交
773 774
	}

775
	spin_lock_bucket(hash);
776
	tmp = search_bucket(hash, sdp, &name);
D
David Teigland 已提交
777
	if (tmp) {
778
		spin_unlock_bucket(hash);
779
		kfree(gl->gl_lksb.sb_lvbptr);
780
		kmem_cache_free(cachep, gl);
781
		atomic_dec(&sdp->sd_glock_disposal);
D
David Teigland 已提交
782 783
		gl = tmp;
	} else {
784 785
		hlist_bl_add_head_rcu(&gl->gl_list, &gl_hash_table[hash]);
		spin_unlock_bucket(hash);
D
David Teigland 已提交
786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801
	}

	*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
 *
 */

802
void gfs2_holder_init(struct gfs2_glock *gl, unsigned int state, unsigned flags,
D
David Teigland 已提交
803 804 805 806
		      struct gfs2_holder *gh)
{
	INIT_LIST_HEAD(&gh->gh_list);
	gh->gh_gl = gl;
807
	gh->gh_ip = _RET_IP_;
808
	gh->gh_owner_pid = get_pid(task_pid(current));
D
David Teigland 已提交
809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825
	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.
 *
 */

826
void gfs2_holder_reinit(unsigned int state, unsigned flags, struct gfs2_holder *gh)
D
David Teigland 已提交
827 828
{
	gh->gh_state = state;
829
	gh->gh_flags = flags;
830
	gh->gh_iflags = 0;
831
	gh->gh_ip = _RET_IP_;
832
	put_pid(gh->gh_owner_pid);
B
Bob Peterson 已提交
833
	gh->gh_owner_pid = get_pid(task_pid(current));
D
David Teigland 已提交
834 835 836 837 838 839 840 841 842 843
}

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

void gfs2_holder_uninit(struct gfs2_holder *gh)
{
844
	put_pid(gh->gh_owner_pid);
D
David Teigland 已提交
845 846
	gfs2_glock_put(gh->gh_gl);
	gh->gh_gl = NULL;
S
Steven Whitehouse 已提交
847
	gh->gh_ip = 0;
D
David Teigland 已提交
848 849
}

850 851 852 853 854 855 856 857
/**
 * gfs2_glock_wait - wait on a glock acquisition
 * @gh: the glock holder
 *
 * Returns: 0 on success
 */

int gfs2_glock_wait(struct gfs2_holder *gh)
858
{
859 860
	unsigned long time1 = jiffies;

861
	might_sleep();
862
	wait_on_bit(&gh->gh_iflags, HIF_WAIT, TASK_UNINTERRUPTIBLE);
863 864 865 866 867
	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);
868
	return gh->gh_error;
869 870
}

D
David Teigland 已提交
871
/**
872 873 874
 * handle_callback - process a demote request
 * @gl: the glock
 * @state: the state the caller wants us to change to
D
David Teigland 已提交
875
 *
876 877
 * There are only two requests that we are going to see in actual
 * practise: LM_ST_SHARED and LM_ST_UNLOCKED
D
David Teigland 已提交
878 879
 */

880
static void handle_callback(struct gfs2_glock *gl, unsigned int state,
881
			    unsigned long delay, bool remote)
D
David Teigland 已提交
882
{
883
	int bit = delay ? GLF_PENDING_DEMOTE : GLF_DEMOTE;
D
David Teigland 已提交
884

885 886 887 888 889 890 891
	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;
D
David Teigland 已提交
892
	}
893
	if (gl->gl_ops->go_callback)
894
		gl->gl_ops->go_callback(gl, remote);
895
	trace_gfs2_demote_rq(gl, remote);
D
David Teigland 已提交
896 897
}

898
void gfs2_print_dbg(struct seq_file *seq, const char *fmt, ...)
899
{
900
	struct va_format vaf;
901 902 903
	va_list args;

	va_start(args, fmt);
904

905
	if (seq) {
906
		seq_vprintf(seq, fmt, args);
907
	} else {
908 909 910
		vaf.fmt = fmt;
		vaf.va = &args;

911
		pr_err("%pV", &vaf);
912
	}
913

914 915 916
	va_end(args);
}

D
David Teigland 已提交
917 918 919 920
/**
 * add_to_queue - Add a holder to the wait queue (but look for recursion)
 * @gh: the holder structure to add
 *
921 922 923 924
 * 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.
 * 
D
David Teigland 已提交
925 926
 */

927
static inline void add_to_queue(struct gfs2_holder *gh)
928 929
__releases(&gl->gl_spin)
__acquires(&gl->gl_spin)
D
David Teigland 已提交
930 931
{
	struct gfs2_glock *gl = gh->gh_gl;
932 933 934
	struct gfs2_sbd *sdp = gl->gl_sbd;
	struct list_head *insert_pt = NULL;
	struct gfs2_holder *gh2;
935
	int try_futile = 0;
D
David Teigland 已提交
936

937
	BUG_ON(gh->gh_owner_pid == NULL);
938 939
	if (test_and_set_bit(HIF_WAIT, &gh->gh_iflags))
		BUG();
940

941 942
	if (gh->gh_flags & (LM_FLAG_TRY | LM_FLAG_TRY_1CB)) {
		if (test_bit(GLF_LOCK, &gl->gl_flags))
943
			try_futile = !may_grant(gl, gh);
944 945 946 947 948 949 950 951
		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;
952 953
		if (try_futile &&
		    !(gh2->gh_flags & (LM_FLAG_TRY | LM_FLAG_TRY_1CB))) {
954 955 956 957
fail:
			gh->gh_error = GLR_TRYFAILED;
			gfs2_holder_wake(gh);
			return;
958
		}
959 960 961 962 963
		if (test_bit(HIF_HOLDER, &gh2->gh_iflags))
			continue;
		if (unlikely((gh->gh_flags & LM_FLAG_PRIORITY) && !insert_pt))
			insert_pt = &gh2->gh_list;
	}
964
	set_bit(GLF_QUEUED, &gl->gl_flags);
965
	trace_gfs2_glock_queue(gh, 1);
966 967
	gfs2_glstats_inc(gl, GFS2_LKS_QCOUNT);
	gfs2_sbstats_inc(gl, GFS2_LKS_QCOUNT);
968 969 970 971 972 973 974 975 976 977 978
	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);
S
Steven Whitehouse 已提交
979
		if (sdp->sd_lockstruct.ls_ops->lm_cancel)
980
			sdp->sd_lockstruct.ls_ops->lm_cancel(gl);
981
		spin_lock(&gl->gl_spin);
D
David Teigland 已提交
982
	}
983
	return;
D
David Teigland 已提交
984

985
trap_recursive:
986 987 988
	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",
989
	       gh2->gh_gl->gl_name.ln_type, gh2->gh_state);
990 991 992
	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",
993
	       gh->gh_gl->gl_name.ln_type, gh->gh_state);
994
	gfs2_dump_glock(NULL, gl);
995
	BUG();
D
David Teigland 已提交
996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012
}

/**
 * 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;

1013
	if (unlikely(test_bit(SDF_SHUTDOWN, &sdp->sd_flags)))
D
David Teigland 已提交
1014 1015
		return -EIO;

1016 1017 1018
	if (test_bit(GLF_LRU, &gl->gl_flags))
		gfs2_glock_remove_from_lru(gl);

D
David Teigland 已提交
1019 1020
	spin_lock(&gl->gl_spin);
	add_to_queue(gh);
1021 1022
	if (unlikely((LM_FLAG_NOEXP & gh->gh_flags) &&
		     test_and_clear_bit(GLF_FROZEN, &gl->gl_flags))) {
1023
		set_bit(GLF_REPLY_PENDING, &gl->gl_flags);
1024 1025 1026 1027
		gl->gl_lockref.count++;
		if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0)
			gl->gl_lockref.count--;
	}
1028
	run_queue(gl, 1);
D
David Teigland 已提交
1029 1030
	spin_unlock(&gl->gl_spin);

1031 1032
	if (!(gh->gh_flags & GL_ASYNC))
		error = gfs2_glock_wait(gh);
D
David Teigland 已提交
1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045

	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)
{
1046
	return test_bit(HIF_WAIT, &gh->gh_iflags) ? 0 : 1;
D
David Teigland 已提交
1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057
}

/**
 * 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;
1058
	const struct gfs2_glock_operations *glops = gl->gl_ops;
1059
	unsigned delay = 0;
1060
	int fast_path = 0;
D
David Teigland 已提交
1061

1062
	spin_lock(&gl->gl_spin);
D
David Teigland 已提交
1063
	if (gh->gh_flags & GL_NOCACHE)
1064
		handle_callback(gl, LM_ST_UNLOCKED, 0, false);
D
David Teigland 已提交
1065 1066

	list_del_init(&gh->gh_list);
1067
	if (find_first_holder(gl) == NULL) {
1068
		if (glops->go_unlock) {
1069
			GLOCK_BUG_ON(gl, test_and_set_bit(GLF_LOCK, &gl->gl_flags));
1070
			spin_unlock(&gl->gl_spin);
D
David Teigland 已提交
1071
			glops->go_unlock(gh);
1072
			spin_lock(&gl->gl_spin);
1073
			clear_bit(GLF_LOCK, &gl->gl_flags);
1074
		}
1075 1076 1077 1078
		if (list_empty(&gl->gl_holders) &&
		    !test_bit(GLF_PENDING_DEMOTE, &gl->gl_flags) &&
		    !test_bit(GLF_DEMOTE, &gl->gl_flags))
			fast_path = 1;
D
David Teigland 已提交
1079
	}
1080 1081 1082
	if (!test_bit(GLF_LFLUSH, &gl->gl_flags) && demote_ok(gl))
		gfs2_glock_add_to_lru(gl);

S
Steven Whitehouse 已提交
1083
	trace_gfs2_glock_queue(gh, 0);
D
David Teigland 已提交
1084
	spin_unlock(&gl->gl_spin);
1085 1086
	if (likely(fast_path))
		return;
1087 1088 1089

	gfs2_glock_hold(gl);
	if (test_bit(GLF_PENDING_DEMOTE, &gl->gl_flags) &&
1090 1091 1092
	    !test_bit(GLF_DEMOTE, &gl->gl_flags) &&
	    gl->gl_name.ln_type == LM_TYPE_INODE)
		delay = gl->gl_hold_time;
1093 1094
	if (queue_delayed_work(glock_workqueue, &gl->gl_work, delay) == 0)
		gfs2_glock_put(gl);
D
David Teigland 已提交
1095 1096
}

A
Abhijith Das 已提交
1097 1098 1099 1100
void gfs2_glock_dq_wait(struct gfs2_holder *gh)
{
	struct gfs2_glock *gl = gh->gh_gl;
	gfs2_glock_dq(gh);
1101
	might_sleep();
1102
	wait_on_bit(&gl->gl_flags, GLF_DEMOTE, TASK_UNINTERRUPTIBLE);
A
Abhijith Das 已提交
1103 1104
}

D
David Teigland 已提交
1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122
/**
 * 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
L
Lucas De Marchi 已提交
1123
 * @flags: modifier flags for the acquisition
D
David Teigland 已提交
1124 1125 1126 1127 1128
 * @gh: the struct gfs2_holder
 *
 * Returns: errno
 */

1129
int gfs2_glock_nq_num(struct gfs2_sbd *sdp, u64 number,
1130 1131
		      const struct gfs2_glock_operations *glops,
		      unsigned int state, int flags, struct gfs2_holder *gh)
D
David Teigland 已提交
1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153
{
	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)
{
1154 1155 1156 1157
	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;
D
David Teigland 已提交
1158 1159

	if (a->ln_number > b->ln_number)
1160 1161 1162
		return 1;
	if (a->ln_number < b->ln_number)
		return -1;
1163
	BUG_ON(gh_a->gh_gl->gl_ops->go_type == gh_b->gh_gl->gl_ops->go_type);
1164
	return 0;
D
David Teigland 已提交
1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212
}

/**
 * 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)
{
1213 1214
	struct gfs2_holder *tmp[4];
	struct gfs2_holder **pph = tmp;
D
David Teigland 已提交
1215 1216
	int error = 0;

1217 1218
	switch(num_gh) {
	case 0:
D
David Teigland 已提交
1219
		return 0;
1220
	case 1:
D
David Teigland 已提交
1221 1222
		ghs->gh_flags &= ~(LM_FLAG_TRY | GL_ASYNC);
		return gfs2_glock_nq(ghs);
1223 1224
	default:
		if (num_gh <= 4)
D
David Teigland 已提交
1225
			break;
1226 1227 1228
		pph = kmalloc(num_gh * sizeof(struct gfs2_holder *), GFP_NOFS);
		if (!pph)
			return -ENOMEM;
D
David Teigland 已提交
1229 1230
	}

1231
	error = nq_m_sync(num_gh, ghs, pph);
D
David Teigland 已提交
1232

1233 1234
	if (pph != tmp)
		kfree(pph);
D
David Teigland 已提交
1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247

	return error;
}

/**
 * gfs2_glock_dq_m - release multiple glocks
 * @num_gh: the number of structures
 * @ghs: an array of struct gfs2_holder structures
 *
 */

void gfs2_glock_dq_m(unsigned int num_gh, struct gfs2_holder *ghs)
{
1248 1249
	while (num_gh--)
		gfs2_glock_dq(&ghs[num_gh]);
D
David Teigland 已提交
1250 1251
}

1252
void gfs2_glock_cb(struct gfs2_glock *gl, unsigned int state)
1253
{
1254 1255 1256
	unsigned long delay = 0;
	unsigned long holdtime;
	unsigned long now = jiffies;
D
David Teigland 已提交
1257

1258
	gfs2_glock_hold(gl);
1259 1260 1261
	holdtime = gl->gl_tchange + gl->gl_hold_time;
	if (test_bit(GLF_QUEUED, &gl->gl_flags) &&
	    gl->gl_name.ln_type == LM_TYPE_INODE) {
1262 1263 1264
		if (time_before(now, holdtime))
			delay = holdtime - now;
		if (test_bit(GLF_REPLY_PENDING, &gl->gl_flags))
1265
			delay = gl->gl_hold_time;
1266
	}
D
David Teigland 已提交
1267

1268
	spin_lock(&gl->gl_spin);
1269
	handle_callback(gl, state, delay, true);
1270
	spin_unlock(&gl->gl_spin);
1271 1272
	if (queue_delayed_work(glock_workqueue, &gl->gl_work, delay) == 0)
		gfs2_glock_put(gl);
D
David Teigland 已提交
1273 1274
}

1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304
/**
 * gfs2_should_freeze - Figure out if glock should be frozen
 * @gl: The glock in question
 *
 * Glocks are not frozen if (a) the result of the dlm operation is
 * an error, (b) the locking operation was an unlock operation or
 * (c) if there is a "noexp" flagged request anywhere in the queue
 *
 * Returns: 1 if freezing should occur, 0 otherwise
 */

static int gfs2_should_freeze(const struct gfs2_glock *gl)
{
	const struct gfs2_holder *gh;

	if (gl->gl_reply & ~LM_OUT_ST_MASK)
		return 0;
	if (gl->gl_target == LM_ST_UNLOCKED)
		return 0;

	list_for_each_entry(gh, &gl->gl_holders, gh_list) {
		if (test_bit(HIF_HOLDER, &gh->gh_iflags))
			continue;
		if (LM_FLAG_NOEXP & gh->gh_flags)
			return 0;
	}

	return 1;
}

D
David Teigland 已提交
1305
/**
1306 1307 1308
 * gfs2_glock_complete - Callback used by locking
 * @gl: Pointer to the glock
 * @ret: The return value from the dlm
D
David Teigland 已提交
1309
 *
1310 1311
 * The gl_reply field is under the gl_spin lock so that it is ok
 * to use a bitfield shared with other glock state fields.
D
David Teigland 已提交
1312 1313
 */

1314
void gfs2_glock_complete(struct gfs2_glock *gl, int ret)
D
David Teigland 已提交
1315
{
1316
	struct lm_lockstruct *ls = &gl->gl_sbd->sd_lockstruct;
1317

1318
	spin_lock(&gl->gl_spin);
1319
	gl->gl_reply = ret;
1320

1321
	if (unlikely(test_bit(DFL_BLOCK_LOCKS, &ls->ls_recover_flags))) {
1322
		if (gfs2_should_freeze(gl)) {
1323
			set_bit(GLF_FROZEN, &gl->gl_flags);
1324
			spin_unlock(&gl->gl_spin);
D
David Teigland 已提交
1325
			return;
1326
		}
D
David Teigland 已提交
1327
	}
1328

S
Steven Whitehouse 已提交
1329
	gl->gl_lockref.count++;
1330
	set_bit(GLF_REPLY_PENDING, &gl->gl_flags);
S
Steven Whitehouse 已提交
1331 1332
	spin_unlock(&gl->gl_spin);

1333 1334
	if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0)
		gfs2_glock_put(gl);
D
David Teigland 已提交
1335 1336
}

1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376
static int glock_cmp(void *priv, struct list_head *a, struct list_head *b)
{
	struct gfs2_glock *gla, *glb;

	gla = list_entry(a, struct gfs2_glock, gl_lru);
	glb = list_entry(b, struct gfs2_glock, gl_lru);

	if (gla->gl_name.ln_number > glb->gl_name.ln_number)
		return 1;
	if (gla->gl_name.ln_number < glb->gl_name.ln_number)
		return -1;

	return 0;
}

/**
 * gfs2_dispose_glock_lru - Demote a list of glocks
 * @list: The list to dispose of
 *
 * Disposing of glocks may involve disk accesses, so that here we sort
 * the glocks by number (i.e. disk location of the inodes) so that if
 * there are any such accesses, they'll be sent in order (mostly).
 *
 * Must be called under the lru_lock, but may drop and retake this
 * lock. While the lru_lock is dropped, entries may vanish from the
 * list, but no new entries will appear on the list (since it is
 * private)
 */

static void gfs2_dispose_glock_lru(struct list_head *list)
__releases(&lru_lock)
__acquires(&lru_lock)
{
	struct gfs2_glock *gl;

	list_sort(NULL, list, glock_cmp);

	while(!list_empty(list)) {
		gl = list_entry(list->next, struct gfs2_glock, gl_lru);
		list_del_init(&gl->gl_lru);
S
Steven Whitehouse 已提交
1377
		if (!spin_trylock(&gl->gl_spin)) {
1378
add_back_to_lru:
S
Steven Whitehouse 已提交
1379 1380 1381 1382
			list_add(&gl->gl_lru, &lru_list);
			atomic_inc(&lru_count);
			continue;
		}
1383 1384 1385 1386
		if (test_and_set_bit(GLF_LOCK, &gl->gl_flags)) {
			spin_unlock(&gl->gl_spin);
			goto add_back_to_lru;
		}
1387
		clear_bit(GLF_LRU, &gl->gl_flags);
S
Steven Whitehouse 已提交
1388
		gl->gl_lockref.count++;
1389
		if (demote_ok(gl))
1390
			handle_callback(gl, LM_ST_UNLOCKED, 0, false);
1391 1392
		WARN_ON(!test_and_clear_bit(GLF_LOCK, &gl->gl_flags));
		if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0)
S
Steven Whitehouse 已提交
1393
			gl->gl_lockref.count--;
1394
		spin_unlock(&gl->gl_spin);
1395
		cond_resched_lock(&lru_lock);
1396 1397 1398
	}
}

1399 1400 1401 1402
/**
 * gfs2_scan_glock_lru - Scan the LRU looking for locks to demote
 * @nr: The number of entries to scan
 *
1403 1404 1405
 * This function selects the entries on the LRU which are able to
 * be demoted, and then kicks off the process by calling
 * gfs2_dispose_glock_lru() above.
1406
 */
D
David Teigland 已提交
1407

1408
static long gfs2_scan_glock_lru(int nr)
D
David Teigland 已提交
1409 1410
{
	struct gfs2_glock *gl;
1411
	LIST_HEAD(skipped);
1412
	LIST_HEAD(dispose);
1413
	long freed = 0;
D
David Teigland 已提交
1414

1415
	spin_lock(&lru_lock);
1416
	while ((nr-- >= 0) && !list_empty(&lru_list)) {
1417 1418 1419
		gl = list_entry(lru_list.next, struct gfs2_glock, gl_lru);

		/* Test for being demotable */
1420
		if (!test_bit(GLF_LOCK, &gl->gl_flags)) {
1421 1422
			list_move(&gl->gl_lru, &dispose);
			atomic_dec(&lru_count);
1423
			freed++;
S
Steven Whitehouse 已提交
1424
			continue;
1425
		}
1426 1427

		list_move(&gl->gl_lru, &skipped);
D
David Teigland 已提交
1428
	}
1429
	list_splice(&skipped, &lru_list);
1430 1431
	if (!list_empty(&dispose))
		gfs2_dispose_glock_lru(&dispose);
1432
	spin_unlock(&lru_lock);
1433 1434

	return freed;
1435 1436
}

1437 1438
static unsigned long gfs2_glock_shrink_scan(struct shrinker *shrink,
					    struct shrink_control *sc)
1439
{
1440 1441 1442 1443
	if (!(sc->gfp_mask & __GFP_FS))
		return SHRINK_STOP;
	return gfs2_scan_glock_lru(sc->nr_to_scan);
}
1444

1445 1446 1447
static unsigned long gfs2_glock_shrink_count(struct shrinker *shrink,
					     struct shrink_control *sc)
{
1448
	return vfs_pressure_ratio(atomic_read(&lru_count));
D
David Teigland 已提交
1449 1450
}

1451 1452
static struct shrinker glock_shrinker = {
	.seeks = DEFAULT_SEEKS,
1453 1454
	.count_objects = gfs2_glock_shrink_count,
	.scan_objects = gfs2_glock_shrink_scan,
1455 1456
};

D
David Teigland 已提交
1457 1458 1459 1460 1461 1462 1463 1464
/**
 * examine_bucket - Call a function for glock in a hash bucket
 * @examiner: the function
 * @sdp: the filesystem
 * @bucket: the bucket
 *
 */

1465
static void examine_bucket(glock_examiner examiner, const struct gfs2_sbd *sdp,
1466
			  unsigned int hash)
D
David Teigland 已提交
1467
{
1468 1469 1470
	struct gfs2_glock *gl;
	struct hlist_bl_head *head = &gl_hash_table[hash];
	struct hlist_bl_node *pos;
D
David Teigland 已提交
1471

1472 1473
	rcu_read_lock();
	hlist_bl_for_each_entry_rcu(gl, pos, head, gl_list) {
S
Steven Whitehouse 已提交
1474
		if ((gl->gl_sbd == sdp) && lockref_get_not_dead(&gl->gl_lockref))
1475
			examiner(gl);
D
David Teigland 已提交
1476
	}
1477
	rcu_read_unlock();
1478
	cond_resched();
1479 1480 1481 1482 1483 1484 1485 1486
}

static void glock_hash_walk(glock_examiner examiner, const struct gfs2_sbd *sdp)
{
	unsigned x;

	for (x = 0; x < GFS2_GL_HASH_SIZE; x++)
		examine_bucket(examiner, sdp, x);
D
David Teigland 已提交
1487 1488
}

1489 1490 1491 1492 1493 1494 1495 1496 1497 1498

/**
 * thaw_glock - thaw out a glock which has an unprocessed reply waiting
 * @gl: The glock to thaw
 *
 */

static void thaw_glock(struct gfs2_glock *gl)
{
	if (!test_and_clear_bit(GLF_FROZEN, &gl->gl_flags))
1499
		goto out;
1500
	set_bit(GLF_REPLY_PENDING, &gl->gl_flags);
1501 1502
	if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0) {
out:
1503
		gfs2_glock_put(gl);
1504
	}
1505 1506
}

D
David Teigland 已提交
1507 1508 1509 1510 1511 1512 1513 1514
/**
 * clear_glock - look at a glock and see if we can free it from glock cache
 * @gl: the glock to look at
 *
 */

static void clear_glock(struct gfs2_glock *gl)
{
1515
	gfs2_glock_remove_from_lru(gl);
D
David Teigland 已提交
1516

1517
	spin_lock(&gl->gl_spin);
1518
	if (gl->gl_state != LM_ST_UNLOCKED)
1519
		handle_callback(gl, LM_ST_UNLOCKED, 0, false);
1520 1521 1522
	spin_unlock(&gl->gl_spin);
	if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0)
		gfs2_glock_put(gl);
D
David Teigland 已提交
1523 1524
}

1525 1526 1527 1528 1529 1530 1531 1532
/**
 * gfs2_glock_thaw - Thaw any frozen glocks
 * @sdp: The super block
 *
 */

void gfs2_glock_thaw(struct gfs2_sbd *sdp)
{
1533 1534
	glock_hash_walk(thaw_glock, sdp);
}
1535

1536
static void dump_glock(struct seq_file *seq, struct gfs2_glock *gl)
1537 1538
{
	spin_lock(&gl->gl_spin);
1539
	gfs2_dump_glock(seq, gl);
1540 1541 1542 1543 1544 1545
	spin_unlock(&gl->gl_spin);
}

static void dump_glock_func(struct gfs2_glock *gl)
{
	dump_glock(NULL, gl);
1546 1547
}

D
David Teigland 已提交
1548 1549 1550 1551 1552
/**
 * gfs2_gl_hash_clear - Empty out the glock hash table
 * @sdp: the filesystem
 * @wait: wait until it's all gone
 *
1553
 * Called when unmounting the filesystem.
D
David Teigland 已提交
1554 1555
 */

1556
void gfs2_gl_hash_clear(struct gfs2_sbd *sdp)
D
David Teigland 已提交
1557
{
1558
	set_bit(SDF_SKIP_DLM_UNLOCK, &sdp->sd_flags);
1559
	flush_workqueue(glock_workqueue);
1560
	glock_hash_walk(clear_glock, sdp);
1561 1562
	flush_workqueue(glock_workqueue);
	wait_event(sdp->sd_glock_wait, atomic_read(&sdp->sd_glock_disposal) == 0);
1563
	glock_hash_walk(dump_glock_func, sdp);
D
David Teigland 已提交
1564 1565
}

1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579
void gfs2_glock_finish_truncate(struct gfs2_inode *ip)
{
	struct gfs2_glock *gl = ip->i_gl;
	int ret;

	ret = gfs2_truncatei_resume(ip);
	gfs2_assert_withdraw(gl->gl_sbd, ret == 0);

	spin_lock(&gl->gl_spin);
	clear_bit(GLF_LOCK, &gl->gl_flags);
	run_queue(gl, 1);
	spin_unlock(&gl->gl_spin);
}

1580
static const char *state2str(unsigned state)
1581
{
1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604
	switch(state) {
	case LM_ST_UNLOCKED:
		return "UN";
	case LM_ST_SHARED:
		return "SH";
	case LM_ST_DEFERRED:
		return "DF";
	case LM_ST_EXCLUSIVE:
		return "EX";
	}
	return "??";
}

static const char *hflags2str(char *buf, unsigned flags, unsigned long iflags)
{
	char *p = buf;
	if (flags & LM_FLAG_TRY)
		*p++ = 't';
	if (flags & LM_FLAG_TRY_1CB)
		*p++ = 'T';
	if (flags & LM_FLAG_NOEXP)
		*p++ = 'e';
	if (flags & LM_FLAG_ANY)
1605
		*p++ = 'A';
1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621
	if (flags & LM_FLAG_PRIORITY)
		*p++ = 'p';
	if (flags & GL_ASYNC)
		*p++ = 'a';
	if (flags & GL_EXACT)
		*p++ = 'E';
	if (flags & GL_NOCACHE)
		*p++ = 'c';
	if (test_bit(HIF_HOLDER, &iflags))
		*p++ = 'H';
	if (test_bit(HIF_WAIT, &iflags))
		*p++ = 'W';
	if (test_bit(HIF_FIRST, &iflags))
		*p++ = 'F';
	*p = 0;
	return buf;
1622 1623
}

D
David Teigland 已提交
1624 1625
/**
 * dump_holder - print information about a glock holder
1626
 * @seq: the seq_file struct
D
David Teigland 已提交
1627 1628 1629 1630
 * @gh: the glock holder
 *
 */

1631
static void dump_holder(struct seq_file *seq, const struct gfs2_holder *gh)
D
David Teigland 已提交
1632
{
1633 1634
	struct task_struct *gh_owner = NULL;
	char flags_buf[32];
D
David Teigland 已提交
1635

1636
	rcu_read_lock();
1637
	if (gh->gh_owner_pid)
1638
		gh_owner = pid_task(gh->gh_owner_pid, PIDTYPE_PID);
1639 1640 1641 1642 1643 1644 1645
	gfs2_print_dbg(seq, " H: s:%s f:%s e:%d p:%ld [%s] %pS\n",
		       state2str(gh->gh_state),
		       hflags2str(flags_buf, gh->gh_flags, gh->gh_iflags),
		       gh->gh_error,
		       gh->gh_owner_pid ? (long)pid_nr(gh->gh_owner_pid) : -1,
		       gh_owner ? gh_owner->comm : "(ended)",
		       (void *)gh->gh_ip);
1646
	rcu_read_unlock();
D
David Teigland 已提交
1647 1648
}

1649
static const char *gflags2str(char *buf, const struct gfs2_glock *gl)
1650
{
1651
	const unsigned long *gflags = &gl->gl_flags;
1652
	char *p = buf;
1653

1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669
	if (test_bit(GLF_LOCK, gflags))
		*p++ = 'l';
	if (test_bit(GLF_DEMOTE, gflags))
		*p++ = 'D';
	if (test_bit(GLF_PENDING_DEMOTE, gflags))
		*p++ = 'd';
	if (test_bit(GLF_DEMOTE_IN_PROGRESS, gflags))
		*p++ = 'p';
	if (test_bit(GLF_DIRTY, gflags))
		*p++ = 'y';
	if (test_bit(GLF_LFLUSH, gflags))
		*p++ = 'f';
	if (test_bit(GLF_INVALIDATE_IN_PROGRESS, gflags))
		*p++ = 'i';
	if (test_bit(GLF_REPLY_PENDING, gflags))
		*p++ = 'r';
1670
	if (test_bit(GLF_INITIAL, gflags))
1671
		*p++ = 'I';
1672 1673
	if (test_bit(GLF_FROZEN, gflags))
		*p++ = 'F';
1674 1675
	if (test_bit(GLF_QUEUED, gflags))
		*p++ = 'q';
1676 1677 1678 1679
	if (test_bit(GLF_LRU, gflags))
		*p++ = 'L';
	if (gl->gl_object)
		*p++ = 'o';
1680 1681
	if (test_bit(GLF_BLOCKING, gflags))
		*p++ = 'b';
1682 1683
	*p = 0;
	return buf;
D
David Teigland 已提交
1684 1685 1686
}

/**
1687
 * gfs2_dump_glock - print information about a glock
1688
 * @seq: The seq_file struct
D
David Teigland 已提交
1689
 * @gl: the glock
1690 1691 1692 1693 1694 1695 1696 1697 1698 1699
 *
 * The file format is as follows:
 * One line per object, capital letters are used to indicate objects
 * G = glock, I = Inode, R = rgrp, H = holder. Glocks are not indented,
 * other objects are indented by a single space and follow the glock to
 * which they are related. Fields are indicated by lower case letters
 * followed by a colon and the field value, except for strings which are in
 * [] so that its possible to see if they are composed of spaces for
 * example. The field's are n = number (id of the object), f = flags,
 * t = type, s = state, r = refcount, e = error, p = pid.
D
David Teigland 已提交
1700 1701 1702
 *
 */

1703
void gfs2_dump_glock(struct seq_file *seq, const struct gfs2_glock *gl)
D
David Teigland 已提交
1704
{
1705 1706 1707 1708
	const struct gfs2_glock_operations *glops = gl->gl_ops;
	unsigned long long dtime;
	const struct gfs2_holder *gh;
	char gflags_buf[32];
D
David Teigland 已提交
1709

1710 1711 1712 1713
	dtime = jiffies - gl->gl_demote_time;
	dtime *= 1000000/HZ; /* demote time in uSec */
	if (!test_bit(GLF_DEMOTE, &gl->gl_flags))
		dtime = 0;
1714
	gfs2_print_dbg(seq, "G:  s:%s n:%u/%llx f:%s t:%s d:%s/%llu a:%d v:%d r:%d m:%ld\n",
1715 1716 1717
		  state2str(gl->gl_state),
		  gl->gl_name.ln_type,
		  (unsigned long long)gl->gl_name.ln_number,
1718
		  gflags2str(gflags_buf, gl),
1719 1720 1721
		  state2str(gl->gl_target),
		  state2str(gl->gl_demote_state), dtime,
		  atomic_read(&gl->gl_ail_count),
1722
		  atomic_read(&gl->gl_revokes),
S
Steven Whitehouse 已提交
1723
		  (int)gl->gl_lockref.count, gl->gl_hold_time);
D
David Teigland 已提交
1724

1725 1726 1727
	list_for_each_entry(gh, &gl->gl_holders, gh_list)
		dump_holder(seq, gh);

1728
	if (gl->gl_state != LM_ST_UNLOCKED && glops->go_dump)
1729
		glops->go_dump(seq, gl);
D
David Teigland 已提交
1730 1731
}

1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787
static int gfs2_glstats_seq_show(struct seq_file *seq, void *iter_ptr)
{
	struct gfs2_glock *gl = iter_ptr;

	seq_printf(seq, "G: n:%u/%llx rtt:%lld/%lld rttb:%lld/%lld irt:%lld/%lld dcnt: %lld qcnt: %lld\n",
		   gl->gl_name.ln_type,
		   (unsigned long long)gl->gl_name.ln_number,
		   (long long)gl->gl_stats.stats[GFS2_LKS_SRTT],
		   (long long)gl->gl_stats.stats[GFS2_LKS_SRTTVAR],
		   (long long)gl->gl_stats.stats[GFS2_LKS_SRTTB],
		   (long long)gl->gl_stats.stats[GFS2_LKS_SRTTVARB],
		   (long long)gl->gl_stats.stats[GFS2_LKS_SIRT],
		   (long long)gl->gl_stats.stats[GFS2_LKS_SIRTVAR],
		   (long long)gl->gl_stats.stats[GFS2_LKS_DCOUNT],
		   (long long)gl->gl_stats.stats[GFS2_LKS_QCOUNT]);
	return 0;
}

static const char *gfs2_gltype[] = {
	"type",
	"reserved",
	"nondisk",
	"inode",
	"rgrp",
	"meta",
	"iopen",
	"flock",
	"plock",
	"quota",
	"journal",
};

static const char *gfs2_stype[] = {
	[GFS2_LKS_SRTT]		= "srtt",
	[GFS2_LKS_SRTTVAR]	= "srttvar",
	[GFS2_LKS_SRTTB]	= "srttb",
	[GFS2_LKS_SRTTVARB]	= "srttvarb",
	[GFS2_LKS_SIRT]		= "sirt",
	[GFS2_LKS_SIRTVAR]	= "sirtvar",
	[GFS2_LKS_DCOUNT]	= "dlm",
	[GFS2_LKS_QCOUNT]	= "queue",
};

#define GFS2_NR_SBSTATS (ARRAY_SIZE(gfs2_gltype) * ARRAY_SIZE(gfs2_stype))

static int gfs2_sbstats_seq_show(struct seq_file *seq, void *iter_ptr)
{
	struct gfs2_glock_iter *gi = seq->private;
	struct gfs2_sbd *sdp = gi->sdp;
	unsigned index = gi->hash >> 3;
	unsigned subindex = gi->hash & 0x07;
	s64 value;
	int i;

	if (index == 0 && subindex != 0)
		return 0;
1788

1789 1790
	seq_printf(seq, "%-10s %8s:", gfs2_gltype[index],
		   (index == 0) ? "cpu": gfs2_stype[subindex]);
D
David Teigland 已提交
1791

1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803
	for_each_possible_cpu(i) {
                const struct gfs2_pcpu_lkstats *lkstats = per_cpu_ptr(sdp->sd_lkstats, i);
		if (index == 0) {
			value = i;
		} else {
			value = lkstats->lkstats[index - 1].stats[subindex];
		}
		seq_printf(seq, " %15lld", (long long)value);
	}
	seq_putc(seq, '\n');
	return 0;
}
1804

1805 1806 1807 1808
int __init gfs2_glock_init(void)
{
	unsigned i;
	for(i = 0; i < GFS2_GL_HASH_SIZE; i++) {
1809
		INIT_HLIST_BL_HEAD(&gl_hash_table[i]);
1810
	}
1811

1812
	glock_workqueue = alloc_workqueue("glock_workqueue", WQ_MEM_RECLAIM |
1813
					  WQ_HIGHPRI | WQ_FREEZABLE, 0);
1814 1815
	if (!glock_workqueue)
		return -ENOMEM;
1816
	gfs2_delete_workqueue = alloc_workqueue("delete_workqueue",
1817
						WQ_MEM_RECLAIM | WQ_FREEZABLE,
1818
						0);
1819
	if (!gfs2_delete_workqueue) {
1820
		destroy_workqueue(glock_workqueue);
1821
		return -ENOMEM;
1822
	}
1823 1824

	register_shrinker(&glock_shrinker);
1825

1826 1827 1828
	return 0;
}

1829 1830
void gfs2_glock_exit(void)
{
1831
	unregister_shrinker(&glock_shrinker);
1832
	destroy_workqueue(glock_workqueue);
1833
	destroy_workqueue(gfs2_delete_workqueue);
1834 1835
}

1836 1837 1838 1839 1840 1841 1842 1843
static inline struct gfs2_glock *glock_hash_chain(unsigned hash)
{
	return hlist_bl_entry(hlist_bl_first_rcu(&gl_hash_table[hash]),
			      struct gfs2_glock, gl_list);
}

static inline struct gfs2_glock *glock_hash_next(struct gfs2_glock *gl)
{
1844
	return hlist_bl_entry(rcu_dereference(gl->gl_list.next),
1845 1846 1847
			      struct gfs2_glock, gl_list);
}

1848
static int gfs2_glock_iter_next(struct gfs2_glock_iter *gi)
1849
{
1850 1851
	struct gfs2_glock *gl;

1852 1853 1854 1855
	do {
		gl = gi->gl;
		if (gl) {
			gi->gl = glock_hash_next(gl);
1856
			gi->nhash++;
1857
		} else {
1858 1859 1860 1861
			if (gi->hash >= GFS2_GL_HASH_SIZE) {
				rcu_read_unlock();
				return 1;
			}
1862
			gi->gl = glock_hash_chain(gi->hash);
1863
			gi->nhash = 0;
1864 1865 1866 1867 1868 1869 1870 1871
		}
		while (gi->gl == NULL) {
			gi->hash++;
			if (gi->hash >= GFS2_GL_HASH_SIZE) {
				rcu_read_unlock();
				return 1;
			}
			gi->gl = glock_hash_chain(gi->hash);
1872
			gi->nhash = 0;
1873 1874
		}
	/* Skip entries for other sb and dead entries */
1875 1876
	} while (gi->sdp != gi->gl->gl_sbd ||
		 __lockref_is_dead(&gi->gl->gl_lockref));
1877

1878 1879 1880
	return 0;
}

1881
static void *gfs2_glock_seq_start(struct seq_file *seq, loff_t *pos)
1882
{
1883
	struct gfs2_glock_iter *gi = seq->private;
1884 1885
	loff_t n = *pos;

1886 1887 1888 1889 1890 1891
	if (gi->last_pos <= *pos)
		n = gi->nhash + (*pos - gi->last_pos);
	else
		gi->hash = 0;

	gi->nhash = 0;
1892
	rcu_read_lock();
1893

1894
	do {
1895
		if (gfs2_glock_iter_next(gi))
1896
			return NULL;
1897
	} while (n--);
1898

1899
	gi->last_pos = *pos;
1900
	return gi->gl;
1901 1902
}

1903
static void *gfs2_glock_seq_next(struct seq_file *seq, void *iter_ptr,
1904 1905
				 loff_t *pos)
{
1906
	struct gfs2_glock_iter *gi = seq->private;
1907 1908

	(*pos)++;
1909
	gi->last_pos = *pos;
1910
	if (gfs2_glock_iter_next(gi))
1911 1912
		return NULL;

1913
	return gi->gl;
1914 1915
}

1916
static void gfs2_glock_seq_stop(struct seq_file *seq, void *iter_ptr)
1917
{
1918
	struct gfs2_glock_iter *gi = seq->private;
1919 1920 1921 1922

	if (gi->gl)
		rcu_read_unlock();
	gi->gl = NULL;
1923 1924
}

1925
static int gfs2_glock_seq_show(struct seq_file *seq, void *iter_ptr)
1926
{
1927 1928
	dump_glock(seq, iter_ptr);
	return 0;
1929 1930
}

1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959
static void *gfs2_sbstats_seq_start(struct seq_file *seq, loff_t *pos)
{
	struct gfs2_glock_iter *gi = seq->private;

	gi->hash = *pos;
	if (*pos >= GFS2_NR_SBSTATS)
		return NULL;
	preempt_disable();
	return SEQ_START_TOKEN;
}

static void *gfs2_sbstats_seq_next(struct seq_file *seq, void *iter_ptr,
				   loff_t *pos)
{
	struct gfs2_glock_iter *gi = seq->private;
	(*pos)++;
	gi->hash++;
	if (gi->hash >= GFS2_NR_SBSTATS) {
		preempt_enable();
		return NULL;
	}
	return SEQ_START_TOKEN;
}

static void gfs2_sbstats_seq_stop(struct seq_file *seq, void *iter_ptr)
{
	preempt_enable();
}

1960
static const struct seq_operations gfs2_glock_seq_ops = {
1961 1962 1963 1964 1965 1966
	.start = gfs2_glock_seq_start,
	.next  = gfs2_glock_seq_next,
	.stop  = gfs2_glock_seq_stop,
	.show  = gfs2_glock_seq_show,
};

1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980
static const struct seq_operations gfs2_glstats_seq_ops = {
	.start = gfs2_glock_seq_start,
	.next  = gfs2_glock_seq_next,
	.stop  = gfs2_glock_seq_stop,
	.show  = gfs2_glstats_seq_show,
};

static const struct seq_operations gfs2_sbstats_seq_ops = {
	.start = gfs2_sbstats_seq_start,
	.next  = gfs2_sbstats_seq_next,
	.stop  = gfs2_sbstats_seq_stop,
	.show  = gfs2_sbstats_seq_show,
};

1981 1982
#define GFS2_SEQ_GOODSIZE min(PAGE_SIZE << PAGE_ALLOC_COSTLY_ORDER, 65536UL)

1983
static int gfs2_glocks_open(struct inode *inode, struct file *file)
1984
{
1985 1986 1987 1988 1989 1990
	int ret = seq_open_private(file, &gfs2_glock_seq_ops,
				   sizeof(struct gfs2_glock_iter));
	if (ret == 0) {
		struct seq_file *seq = file->private_data;
		struct gfs2_glock_iter *gi = seq->private;
		gi->sdp = inode->i_private;
1991
		seq->buf = kmalloc(GFS2_SEQ_GOODSIZE, GFP_KERNEL | __GFP_NOWARN);
1992
		if (seq->buf)
1993
			seq->size = GFS2_SEQ_GOODSIZE;
1994 1995
	}
	return ret;
1996 1997
}

1998 1999 2000 2001 2002 2003 2004 2005
static int gfs2_glstats_open(struct inode *inode, struct file *file)
{
	int ret = seq_open_private(file, &gfs2_glstats_seq_ops,
				   sizeof(struct gfs2_glock_iter));
	if (ret == 0) {
		struct seq_file *seq = file->private_data;
		struct gfs2_glock_iter *gi = seq->private;
		gi->sdp = inode->i_private;
2006
		seq->buf = kmalloc(GFS2_SEQ_GOODSIZE, GFP_KERNEL | __GFP_NOWARN);
2007
		if (seq->buf)
2008
			seq->size = GFS2_SEQ_GOODSIZE;
2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033
	}
	return ret;
}

static int gfs2_sbstats_open(struct inode *inode, struct file *file)
{
	int ret = seq_open_private(file, &gfs2_sbstats_seq_ops,
				   sizeof(struct gfs2_glock_iter));
	if (ret == 0) {
		struct seq_file *seq = file->private_data;
		struct gfs2_glock_iter *gi = seq->private;
		gi->sdp = inode->i_private;
	}
	return ret;
}

static const struct file_operations gfs2_glocks_fops = {
	.owner   = THIS_MODULE,
	.open    = gfs2_glocks_open,
	.read    = seq_read,
	.llseek  = seq_lseek,
	.release = seq_release_private,
};

static const struct file_operations gfs2_glstats_fops = {
2034
	.owner   = THIS_MODULE,
2035 2036 2037 2038 2039 2040 2041 2042 2043
	.open    = gfs2_glstats_open,
	.read    = seq_read,
	.llseek  = seq_lseek,
	.release = seq_release_private,
};

static const struct file_operations gfs2_sbstats_fops = {
	.owner   = THIS_MODULE,
	.open	 = gfs2_sbstats_open,
2044 2045
	.read    = seq_read,
	.llseek  = seq_lseek,
2046
	.release = seq_release_private,
2047 2048 2049 2050
};

int gfs2_create_debugfs_file(struct gfs2_sbd *sdp)
{
R
Robert Peterson 已提交
2051 2052 2053 2054 2055 2056
	sdp->debugfs_dir = debugfs_create_dir(sdp->sd_table_name, gfs2_root);
	if (!sdp->debugfs_dir)
		return -ENOMEM;
	sdp->debugfs_dentry_glocks = debugfs_create_file("glocks",
							 S_IFREG | S_IRUGO,
							 sdp->debugfs_dir, sdp,
2057
							 &gfs2_glocks_fops);
R
Robert Peterson 已提交
2058
	if (!sdp->debugfs_dentry_glocks)
2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073
		goto fail;

	sdp->debugfs_dentry_glstats = debugfs_create_file("glstats",
							S_IFREG | S_IRUGO,
							sdp->debugfs_dir, sdp,
							&gfs2_glstats_fops);
	if (!sdp->debugfs_dentry_glstats)
		goto fail;

	sdp->debugfs_dentry_sbstats = debugfs_create_file("sbstats",
							S_IFREG | S_IRUGO,
							sdp->debugfs_dir, sdp,
							&gfs2_sbstats_fops);
	if (!sdp->debugfs_dentry_sbstats)
		goto fail;
2074 2075

	return 0;
2076 2077 2078
fail:
	gfs2_delete_debugfs_file(sdp);
	return -ENOMEM;
2079 2080 2081 2082
}

void gfs2_delete_debugfs_file(struct gfs2_sbd *sdp)
{
2083
	if (sdp->debugfs_dir) {
R
Robert Peterson 已提交
2084 2085 2086 2087
		if (sdp->debugfs_dentry_glocks) {
			debugfs_remove(sdp->debugfs_dentry_glocks);
			sdp->debugfs_dentry_glocks = NULL;
		}
2088 2089 2090 2091 2092 2093 2094 2095
		if (sdp->debugfs_dentry_glstats) {
			debugfs_remove(sdp->debugfs_dentry_glstats);
			sdp->debugfs_dentry_glstats = NULL;
		}
		if (sdp->debugfs_dentry_sbstats) {
			debugfs_remove(sdp->debugfs_dentry_sbstats);
			sdp->debugfs_dentry_sbstats = NULL;
		}
R
Robert Peterson 已提交
2096 2097 2098
		debugfs_remove(sdp->debugfs_dir);
		sdp->debugfs_dir = NULL;
	}
2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109
}

int gfs2_register_debugfs(void)
{
	gfs2_root = debugfs_create_dir("gfs2", NULL);
	return gfs2_root ? 0 : -ENOMEM;
}

void gfs2_unregister_debugfs(void)
{
	debugfs_remove(gfs2_root);
R
Robert Peterson 已提交
2110
	gfs2_root = NULL;
2111
}