super.c 42.1 KB
Newer Older
D
David Teigland 已提交
1 2
/*
 * Copyright (C) Sistina Software, Inc.  1997-2003 All rights reserved.
B
Bob Peterson 已提交
3
 * Copyright (C) 2004-2007 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

12
#include <linux/bio.h>
13
#include <linux/sched/signal.h>
D
David Teigland 已提交
14 15 16 17
#include <linux/slab.h>
#include <linux/spinlock.h>
#include <linux/completion.h>
#include <linux/buffer_head.h>
18 19 20 21 22
#include <linux/statfs.h>
#include <linux/seq_file.h>
#include <linux/mount.h>
#include <linux/kthread.h>
#include <linux/delay.h>
23
#include <linux/gfs2_ondisk.h>
24 25
#include <linux/crc32.h>
#include <linux/time.h>
26
#include <linux/wait.h>
27
#include <linux/writeback.h>
28
#include <linux/backing-dev.h>
29
#include <linux/kernel.h>
D
David Teigland 已提交
30 31

#include "gfs2.h"
32
#include "incore.h"
D
David Teigland 已提交
33 34 35 36 37 38 39 40 41 42 43 44
#include "bmap.h"
#include "dir.h"
#include "glock.h"
#include "glops.h"
#include "inode.h"
#include "log.h"
#include "meta_io.h"
#include "quota.h"
#include "recovery.h"
#include "rgrp.h"
#include "super.h"
#include "trans.h"
45
#include "util.h"
46
#include "sys.h"
47
#include "xattr.h"
48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76

#define args_neq(a1, a2, x) ((a1)->ar_##x != (a2)->ar_##x)

enum {
	Opt_lockproto,
	Opt_locktable,
	Opt_hostdata,
	Opt_spectator,
	Opt_ignore_local_fs,
	Opt_localflocks,
	Opt_localcaching,
	Opt_debug,
	Opt_nodebug,
	Opt_upgrade,
	Opt_acl,
	Opt_noacl,
	Opt_quota_off,
	Opt_quota_account,
	Opt_quota_on,
	Opt_quota,
	Opt_noquota,
	Opt_suiddir,
	Opt_nosuiddir,
	Opt_data_writeback,
	Opt_data_ordered,
	Opt_meta,
	Opt_discard,
	Opt_nodiscard,
	Opt_commit,
77 78
	Opt_err_withdraw,
	Opt_err_panic,
79 80 81
	Opt_statfs_quantum,
	Opt_statfs_percent,
	Opt_quota_quantum,
82 83
	Opt_barrier,
	Opt_nobarrier,
84 85
	Opt_rgrplvb,
	Opt_norgrplvb,
86 87
	Opt_loccookie,
	Opt_noloccookie,
88 89 90 91 92 93 94 95
	Opt_error,
};

static const match_table_t tokens = {
	{Opt_lockproto, "lockproto=%s"},
	{Opt_locktable, "locktable=%s"},
	{Opt_hostdata, "hostdata=%s"},
	{Opt_spectator, "spectator"},
96
	{Opt_spectator, "norecovery"},
97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117
	{Opt_ignore_local_fs, "ignore_local_fs"},
	{Opt_localflocks, "localflocks"},
	{Opt_localcaching, "localcaching"},
	{Opt_debug, "debug"},
	{Opt_nodebug, "nodebug"},
	{Opt_upgrade, "upgrade"},
	{Opt_acl, "acl"},
	{Opt_noacl, "noacl"},
	{Opt_quota_off, "quota=off"},
	{Opt_quota_account, "quota=account"},
	{Opt_quota_on, "quota=on"},
	{Opt_quota, "quota"},
	{Opt_noquota, "noquota"},
	{Opt_suiddir, "suiddir"},
	{Opt_nosuiddir, "nosuiddir"},
	{Opt_data_writeback, "data=writeback"},
	{Opt_data_ordered, "data=ordered"},
	{Opt_meta, "meta"},
	{Opt_discard, "discard"},
	{Opt_nodiscard, "nodiscard"},
	{Opt_commit, "commit=%d"},
118 119
	{Opt_err_withdraw, "errors=withdraw"},
	{Opt_err_panic, "errors=panic"},
120 121 122
	{Opt_statfs_quantum, "statfs_quantum=%d"},
	{Opt_statfs_percent, "statfs_percent=%d"},
	{Opt_quota_quantum, "quota_quantum=%d"},
123 124
	{Opt_barrier, "barrier"},
	{Opt_nobarrier, "nobarrier"},
125 126
	{Opt_rgrplvb, "rgrplvb"},
	{Opt_norgrplvb, "norgrplvb"},
127 128
	{Opt_loccookie, "loccookie"},
	{Opt_noloccookie, "noloccookie"},
129 130 131 132 133
	{Opt_error, NULL}
};

/**
 * gfs2_mount_args - Parse mount options
134 135
 * @args: The structure into which the parsed options will be written
 * @options: The options to parse
136 137 138 139
 *
 * Return: errno
 */

140
int gfs2_mount_args(struct gfs2_args *args, char *options)
141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174
{
	char *o;
	int token;
	substring_t tmp[MAX_OPT_ARGS];
	int rv;

	/* Split the options into tokens with the "," character and
	   process them */

	while (1) {
		o = strsep(&options, ",");
		if (o == NULL)
			break;
		if (*o == '\0')
			continue;

		token = match_token(o, tokens, tmp);
		switch (token) {
		case Opt_lockproto:
			match_strlcpy(args->ar_lockproto, &tmp[0],
				      GFS2_LOCKNAME_LEN);
			break;
		case Opt_locktable:
			match_strlcpy(args->ar_locktable, &tmp[0],
				      GFS2_LOCKNAME_LEN);
			break;
		case Opt_hostdata:
			match_strlcpy(args->ar_hostdata, &tmp[0],
				      GFS2_LOCKNAME_LEN);
			break;
		case Opt_spectator:
			args->ar_spectator = 1;
			break;
		case Opt_ignore_local_fs:
175
			/* Retained for backwards compat only */
176 177 178 179 180
			break;
		case Opt_localflocks:
			args->ar_localflocks = 1;
			break;
		case Opt_localcaching:
181
			/* Retained for backwards compat only */
182 183
			break;
		case Opt_debug:
184
			if (args->ar_errors == GFS2_ERRORS_PANIC) {
185
				pr_warn("-o debug and -o errors=panic are mutually exclusive\n");
186 187
				return -EINVAL;
			}
188 189 190 191 192 193
			args->ar_debug = 1;
			break;
		case Opt_nodebug:
			args->ar_debug = 0;
			break;
		case Opt_upgrade:
194
			/* Retained for backwards compat only */
195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236
			break;
		case Opt_acl:
			args->ar_posix_acl = 1;
			break;
		case Opt_noacl:
			args->ar_posix_acl = 0;
			break;
		case Opt_quota_off:
		case Opt_noquota:
			args->ar_quota = GFS2_QUOTA_OFF;
			break;
		case Opt_quota_account:
			args->ar_quota = GFS2_QUOTA_ACCOUNT;
			break;
		case Opt_quota_on:
		case Opt_quota:
			args->ar_quota = GFS2_QUOTA_ON;
			break;
		case Opt_suiddir:
			args->ar_suiddir = 1;
			break;
		case Opt_nosuiddir:
			args->ar_suiddir = 0;
			break;
		case Opt_data_writeback:
			args->ar_data = GFS2_DATA_WRITEBACK;
			break;
		case Opt_data_ordered:
			args->ar_data = GFS2_DATA_ORDERED;
			break;
		case Opt_meta:
			args->ar_meta = 1;
			break;
		case Opt_discard:
			args->ar_discard = 1;
			break;
		case Opt_nodiscard:
			args->ar_discard = 0;
			break;
		case Opt_commit:
			rv = match_int(&tmp[0], &args->ar_commit);
			if (rv || args->ar_commit <= 0) {
237
				pr_warn("commit mount option requires a positive numeric argument\n");
238 239 240
				return rv ? rv : -EINVAL;
			}
			break;
241 242 243
		case Opt_statfs_quantum:
			rv = match_int(&tmp[0], &args->ar_statfs_quantum);
			if (rv || args->ar_statfs_quantum < 0) {
244
				pr_warn("statfs_quantum mount option requires a non-negative numeric argument\n");
245 246 247 248 249 250
				return rv ? rv : -EINVAL;
			}
			break;
		case Opt_quota_quantum:
			rv = match_int(&tmp[0], &args->ar_quota_quantum);
			if (rv || args->ar_quota_quantum <= 0) {
251
				pr_warn("quota_quantum mount option requires a positive numeric argument\n");
252 253 254 255 256 257 258
				return rv ? rv : -EINVAL;
			}
			break;
		case Opt_statfs_percent:
			rv = match_int(&tmp[0], &args->ar_statfs_percent);
			if (rv || args->ar_statfs_percent < 0 ||
			    args->ar_statfs_percent > 100) {
259
				pr_warn("statfs_percent mount option requires a numeric argument between 0 and 100\n");
260 261 262
				return rv ? rv : -EINVAL;
			}
			break;
263 264 265 266 267
		case Opt_err_withdraw:
			args->ar_errors = GFS2_ERRORS_WITHDRAW;
			break;
		case Opt_err_panic:
			if (args->ar_debug) {
268
				pr_warn("-o debug and -o errors=panic are mutually exclusive\n");
269 270 271 272
				return -EINVAL;
			}
			args->ar_errors = GFS2_ERRORS_PANIC;
			break;
273 274 275 276 277 278
		case Opt_barrier:
			args->ar_nobarrier = 0;
			break;
		case Opt_nobarrier:
			args->ar_nobarrier = 1;
			break;
279 280 281 282 283 284
		case Opt_rgrplvb:
			args->ar_rgrplvb = 1;
			break;
		case Opt_norgrplvb:
			args->ar_rgrplvb = 0;
			break;
285 286 287 288 289 290
		case Opt_loccookie:
			args->ar_loccookie = 1;
			break;
		case Opt_noloccookie:
			args->ar_loccookie = 0;
			break;
291 292
		case Opt_error:
		default:
293
			pr_warn("invalid mount option: %s\n", o);
294 295 296 297 298 299
			return -EINVAL;
		}
	}

	return 0;
}
D
David Teigland 已提交
300

301 302 303 304 305 306 307 308
/**
 * gfs2_jindex_free - Clear all the journal index information
 * @sdp: The GFS2 superblock
 *
 */

void gfs2_jindex_free(struct gfs2_sbd *sdp)
{
309
	struct list_head list;
310 311 312 313 314 315 316 317 318 319
	struct gfs2_jdesc *jd;

	spin_lock(&sdp->sd_jindex_spin);
	list_add(&list, &sdp->sd_jindex_list);
	list_del_init(&sdp->sd_jindex_list);
	sdp->sd_journals = 0;
	spin_unlock(&sdp->sd_jindex_spin);

	while (!list_empty(&list)) {
		jd = list_entry(list.next, struct gfs2_jdesc, jd_list);
320
		gfs2_free_journal_extents(jd);
321 322 323 324 325 326
		list_del(&jd->jd_list);
		iput(jd->jd_inode);
		kfree(jd);
	}
}

D
David Teigland 已提交
327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357
static struct gfs2_jdesc *jdesc_find_i(struct list_head *head, unsigned int jid)
{
	struct gfs2_jdesc *jd;
	int found = 0;

	list_for_each_entry(jd, head, jd_list) {
		if (jd->jd_jid == jid) {
			found = 1;
			break;
		}
	}

	if (!found)
		jd = NULL;

	return jd;
}

struct gfs2_jdesc *gfs2_jdesc_find(struct gfs2_sbd *sdp, unsigned int jid)
{
	struct gfs2_jdesc *jd;

	spin_lock(&sdp->sd_jindex_spin);
	jd = jdesc_find_i(&sdp->sd_jindex_list, jid);
	spin_unlock(&sdp->sd_jindex_spin);

	return jd;
}

int gfs2_jdesc_check(struct gfs2_jdesc *jd)
{
358 359
	struct gfs2_inode *ip = GFS2_I(jd->jd_inode);
	struct gfs2_sbd *sdp = GFS2_SB(jd->jd_inode);
S
Steven Whitehouse 已提交
360
	u64 size = i_size_read(jd->jd_inode);
D
David Teigland 已提交
361

F
Fabian Frederick 已提交
362
	if (gfs2_check_internal_file_size(jd->jd_inode, 8 << 20, BIT(30)))
D
David Teigland 已提交
363 364
		return -EIO;

S
Steven Whitehouse 已提交
365 366 367
	jd->jd_blocks = size >> sdp->sd_sb.sb_bsize_shift;

	if (gfs2_write_alloc_required(ip, 0, size)) {
D
David Teigland 已提交
368
		gfs2_consist_inode(ip);
369
		return -EIO;
D
David Teigland 已提交
370 371
	}

372
	return 0;
D
David Teigland 已提交
373 374
}

375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401
static int init_threads(struct gfs2_sbd *sdp)
{
	struct task_struct *p;
	int error = 0;

	p = kthread_run(gfs2_logd, sdp, "gfs2_logd");
	if (IS_ERR(p)) {
		error = PTR_ERR(p);
		fs_err(sdp, "can't start logd thread: %d\n", error);
		return error;
	}
	sdp->sd_logd_process = p;

	p = kthread_run(gfs2_quotad, sdp, "gfs2_quotad");
	if (IS_ERR(p)) {
		error = PTR_ERR(p);
		fs_err(sdp, "can't start quotad thread: %d\n", error);
		goto fail;
	}
	sdp->sd_quotad_process = p;
	return 0;

fail:
	kthread_stop(sdp->sd_logd_process);
	return error;
}

D
David Teigland 已提交
402 403 404 405 406 407 408 409 410
/**
 * gfs2_make_fs_rw - Turn a Read-Only FS into a Read-Write one
 * @sdp: the filesystem
 *
 * Returns: errno
 */

int gfs2_make_fs_rw(struct gfs2_sbd *sdp)
{
411
	struct gfs2_inode *ip = GFS2_I(sdp->sd_jdesc->jd_inode);
412
	struct gfs2_glock *j_gl = ip->i_gl;
413
	struct gfs2_holder freeze_gh;
A
Al Viro 已提交
414
	struct gfs2_log_header_host head;
D
David Teigland 已提交
415 416
	int error;

417
	error = init_threads(sdp);
D
David Teigland 已提交
418 419 420
	if (error)
		return error;

421
	error = gfs2_glock_nq_init(sdp->sd_freeze_gl, LM_ST_SHARED, 0,
422
				   &freeze_gh);
423 424 425
	if (error)
		goto fail_threads;

426
	j_gl->gl_ops->go_inval(j_gl, DIO_METADATA);
D
David Teigland 已提交
427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443

	error = gfs2_find_jhead(sdp->sd_jdesc, &head);
	if (error)
		goto fail;

	if (!(head.lh_flags & GFS2_LOG_HEAD_UNMOUNT)) {
		gfs2_consist(sdp);
		error = -EIO;
		goto fail;
	}

	/*  Initialize some head of the log stuff  */
	sdp->sd_log_sequence = head.lh_sequence + 1;
	gfs2_log_pointers_init(sdp, head.lh_blkno);

	error = gfs2_quota_init(sdp);
	if (error)
444
		goto fail;
D
David Teigland 已提交
445 446 447

	set_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags);

448
	gfs2_glock_dq_uninit(&freeze_gh);
D
David Teigland 已提交
449 450 451

	return 0;

452
fail:
453 454
	freeze_gh.gh_flags |= GL_NOCACHE;
	gfs2_glock_dq_uninit(&freeze_gh);
455 456 457
fail_threads:
	kthread_stop(sdp->sd_quotad_process);
	kthread_stop(sdp->sd_logd_process);
D
David Teigland 已提交
458 459 460
	return error;
}

461
void gfs2_statfs_change_in(struct gfs2_statfs_change_host *sc, const void *buf)
462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478
{
	const struct gfs2_statfs_change *str = buf;

	sc->sc_total = be64_to_cpu(str->sc_total);
	sc->sc_free = be64_to_cpu(str->sc_free);
	sc->sc_dinodes = be64_to_cpu(str->sc_dinodes);
}

static void gfs2_statfs_change_out(const struct gfs2_statfs_change_host *sc, void *buf)
{
	struct gfs2_statfs_change *str = buf;

	str->sc_total = cpu_to_be64(sc->sc_total);
	str->sc_free = cpu_to_be64(sc->sc_free);
	str->sc_dinodes = cpu_to_be64(sc->sc_dinodes);
}

D
David Teigland 已提交
479 480
int gfs2_statfs_init(struct gfs2_sbd *sdp)
{
481
	struct gfs2_inode *m_ip = GFS2_I(sdp->sd_statfs_inode);
482
	struct gfs2_statfs_change_host *m_sc = &sdp->sd_statfs_master;
483
	struct gfs2_inode *l_ip = GFS2_I(sdp->sd_sc_inode);
484
	struct gfs2_statfs_change_host *l_sc = &sdp->sd_statfs_local;
D
David Teigland 已提交
485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517
	struct buffer_head *m_bh, *l_bh;
	struct gfs2_holder gh;
	int error;

	error = gfs2_glock_nq_init(m_ip->i_gl, LM_ST_EXCLUSIVE, GL_NOCACHE,
				   &gh);
	if (error)
		return error;

	error = gfs2_meta_inode_buffer(m_ip, &m_bh);
	if (error)
		goto out;

	if (sdp->sd_args.ar_spectator) {
		spin_lock(&sdp->sd_statfs_spin);
		gfs2_statfs_change_in(m_sc, m_bh->b_data +
				      sizeof(struct gfs2_dinode));
		spin_unlock(&sdp->sd_statfs_spin);
	} else {
		error = gfs2_meta_inode_buffer(l_ip, &l_bh);
		if (error)
			goto out_m_bh;

		spin_lock(&sdp->sd_statfs_spin);
		gfs2_statfs_change_in(m_sc, m_bh->b_data +
				      sizeof(struct gfs2_dinode));
		gfs2_statfs_change_in(l_sc, l_bh->b_data +
				      sizeof(struct gfs2_dinode));
		spin_unlock(&sdp->sd_statfs_spin);

		brelse(l_bh);
	}

518
out_m_bh:
D
David Teigland 已提交
519
	brelse(m_bh);
520
out:
D
David Teigland 已提交
521 522 523 524
	gfs2_glock_dq_uninit(&gh);
	return 0;
}

525 526
void gfs2_statfs_change(struct gfs2_sbd *sdp, s64 total, s64 free,
			s64 dinodes)
D
David Teigland 已提交
527
{
528
	struct gfs2_inode *l_ip = GFS2_I(sdp->sd_sc_inode);
529
	struct gfs2_statfs_change_host *l_sc = &sdp->sd_statfs_local;
530
	struct gfs2_statfs_change_host *m_sc = &sdp->sd_statfs_master;
D
David Teigland 已提交
531
	struct buffer_head *l_bh;
532 533
	s64 x, y;
	int need_sync = 0;
D
David Teigland 已提交
534 535 536 537 538 539
	int error;

	error = gfs2_meta_inode_buffer(l_ip, &l_bh);
	if (error)
		return;

540
	gfs2_trans_add_meta(l_ip->i_gl, l_bh);
D
David Teigland 已提交
541 542 543 544 545

	spin_lock(&sdp->sd_statfs_spin);
	l_sc->sc_total += total;
	l_sc->sc_free += free;
	l_sc->sc_dinodes += dinodes;
546
	gfs2_statfs_change_out(l_sc, l_bh->b_data + sizeof(struct gfs2_dinode));
547 548 549 550 551 552
	if (sdp->sd_args.ar_statfs_percent) {
		x = 100 * l_sc->sc_free;
		y = m_sc->sc_free * sdp->sd_args.ar_statfs_percent;
		if (x >= y || x <= -y)
			need_sync = 1;
	}
D
David Teigland 已提交
553 554 555
	spin_unlock(&sdp->sd_statfs_spin);

	brelse(l_bh);
556
	if (need_sync)
557
		gfs2_wake_up_statfs(sdp);
D
David Teigland 已提交
558 559
}

560 561 562 563 564 565 566 567
void update_statfs(struct gfs2_sbd *sdp, struct buffer_head *m_bh,
		   struct buffer_head *l_bh)
{
	struct gfs2_inode *m_ip = GFS2_I(sdp->sd_statfs_inode);
	struct gfs2_inode *l_ip = GFS2_I(sdp->sd_sc_inode);
	struct gfs2_statfs_change_host *m_sc = &sdp->sd_statfs_master;
	struct gfs2_statfs_change_host *l_sc = &sdp->sd_statfs_local;

568
	gfs2_trans_add_meta(l_ip->i_gl, l_bh);
569
	gfs2_trans_add_meta(m_ip->i_gl, m_bh);
570 571 572 573 574 575 576 577 578

	spin_lock(&sdp->sd_statfs_spin);
	m_sc->sc_total += l_sc->sc_total;
	m_sc->sc_free += l_sc->sc_free;
	m_sc->sc_dinodes += l_sc->sc_dinodes;
	memset(l_sc, 0, sizeof(struct gfs2_statfs_change));
	memset(l_bh->b_data + sizeof(struct gfs2_dinode),
	       0, sizeof(struct gfs2_statfs_change));
	gfs2_statfs_change_out(m_sc, m_bh->b_data + sizeof(struct gfs2_dinode));
579
	spin_unlock(&sdp->sd_statfs_spin);
580 581
}

582
int gfs2_statfs_sync(struct super_block *sb, int type)
D
David Teigland 已提交
583
{
584
	struct gfs2_sbd *sdp = sb->s_fs_info;
585 586
	struct gfs2_inode *m_ip = GFS2_I(sdp->sd_statfs_inode);
	struct gfs2_inode *l_ip = GFS2_I(sdp->sd_sc_inode);
587 588
	struct gfs2_statfs_change_host *m_sc = &sdp->sd_statfs_master;
	struct gfs2_statfs_change_host *l_sc = &sdp->sd_statfs_local;
D
David Teigland 已提交
589 590 591 592
	struct gfs2_holder gh;
	struct buffer_head *m_bh, *l_bh;
	int error;

593
	sb_start_write(sb);
D
David Teigland 已提交
594 595 596
	error = gfs2_glock_nq_init(m_ip->i_gl, LM_ST_EXCLUSIVE, GL_NOCACHE,
				   &gh);
	if (error)
597
		goto out;
D
David Teigland 已提交
598 599 600

	error = gfs2_meta_inode_buffer(m_ip, &m_bh);
	if (error)
601
		goto out_unlock;
D
David Teigland 已提交
602 603 604

	spin_lock(&sdp->sd_statfs_spin);
	gfs2_statfs_change_in(m_sc, m_bh->b_data +
605
			      sizeof(struct gfs2_dinode));
D
David Teigland 已提交
606 607 608 609 610 611 612 613 614 615 616 617 618 619
	if (!l_sc->sc_total && !l_sc->sc_free && !l_sc->sc_dinodes) {
		spin_unlock(&sdp->sd_statfs_spin);
		goto out_bh;
	}
	spin_unlock(&sdp->sd_statfs_spin);

	error = gfs2_meta_inode_buffer(l_ip, &l_bh);
	if (error)
		goto out_bh;

	error = gfs2_trans_begin(sdp, 2 * RES_DINODE, 0);
	if (error)
		goto out_bh2;

620
	update_statfs(sdp, m_bh, l_bh);
621
	sdp->sd_statfs_force_sync = 0;
D
David Teigland 已提交
622 623 624

	gfs2_trans_end(sdp);

625
out_bh2:
D
David Teigland 已提交
626
	brelse(l_bh);
627
out_bh:
D
David Teigland 已提交
628
	brelse(m_bh);
629
out_unlock:
D
David Teigland 已提交
630
	gfs2_glock_dq_uninit(&gh);
631 632
out:
	sb_end_write(sb);
D
David Teigland 已提交
633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650
	return error;
}

struct lfcc {
	struct list_head list;
	struct gfs2_holder gh;
};

/**
 * gfs2_lock_fs_check_clean - Stop all writes to the FS and check that all
 *                            journals are clean
 * @sdp: the file system
 * @state: the state to put the transaction lock into
 * @t_gh: the hold on the transaction lock
 *
 * Returns: errno
 */

651
static int gfs2_lock_fs_check_clean(struct gfs2_sbd *sdp,
652
				    struct gfs2_holder *freeze_gh)
D
David Teigland 已提交
653
{
654
	struct gfs2_inode *ip;
D
David Teigland 已提交
655 656 657
	struct gfs2_jdesc *jd;
	struct lfcc *lfcc;
	LIST_HEAD(list);
A
Al Viro 已提交
658
	struct gfs2_log_header_host lh;
D
David Teigland 已提交
659 660 661 662 663 664 665 666
	int error;

	list_for_each_entry(jd, &sdp->sd_jindex_list, jd_list) {
		lfcc = kmalloc(sizeof(struct lfcc), GFP_KERNEL);
		if (!lfcc) {
			error = -ENOMEM;
			goto out;
		}
667 668
		ip = GFS2_I(jd->jd_inode);
		error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, 0, &lfcc->gh);
D
David Teigland 已提交
669 670 671 672 673 674 675
		if (error) {
			kfree(lfcc);
			goto out;
		}
		list_add(&lfcc->list, &list);
	}

676 677
	error = gfs2_glock_nq_init(sdp->sd_freeze_gl, LM_ST_EXCLUSIVE,
				   GL_NOCACHE, freeze_gh);
D
David Teigland 已提交
678 679 680 681 682 683 684 685 686 687 688 689 690 691 692

	list_for_each_entry(jd, &sdp->sd_jindex_list, jd_list) {
		error = gfs2_jdesc_check(jd);
		if (error)
			break;
		error = gfs2_find_jhead(jd, &lh);
		if (error)
			break;
		if (!(lh.lh_flags & GFS2_LOG_HEAD_UNMOUNT)) {
			error = -EBUSY;
			break;
		}
	}

	if (error)
693
		gfs2_glock_dq_uninit(freeze_gh);
D
David Teigland 已提交
694

695
out:
D
David Teigland 已提交
696 697 698 699 700 701 702 703 704
	while (!list_empty(&list)) {
		lfcc = list_entry(list.next, struct lfcc, list);
		list_del(&lfcc->list);
		gfs2_glock_dq_uninit(&lfcc->gh);
		kfree(lfcc);
	}
	return error;
}

705 706 707 708 709 710 711 712 713 714
void gfs2_dinode_out(const struct gfs2_inode *ip, void *buf)
{
	struct gfs2_dinode *str = buf;

	str->di_header.mh_magic = cpu_to_be32(GFS2_MAGIC);
	str->di_header.mh_type = cpu_to_be32(GFS2_METATYPE_DI);
	str->di_header.mh_format = cpu_to_be32(GFS2_FORMAT_DI);
	str->di_num.no_addr = cpu_to_be64(ip->i_no_addr);
	str->di_num.no_formal_ino = cpu_to_be64(ip->i_no_formal_ino);
	str->di_mode = cpu_to_be32(ip->i_inode.i_mode);
715 716
	str->di_uid = cpu_to_be32(i_uid_read(&ip->i_inode));
	str->di_gid = cpu_to_be32(i_gid_read(&ip->i_inode));
717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740
	str->di_nlink = cpu_to_be32(ip->i_inode.i_nlink);
	str->di_size = cpu_to_be64(i_size_read(&ip->i_inode));
	str->di_blocks = cpu_to_be64(gfs2_get_inode_blocks(&ip->i_inode));
	str->di_atime = cpu_to_be64(ip->i_inode.i_atime.tv_sec);
	str->di_mtime = cpu_to_be64(ip->i_inode.i_mtime.tv_sec);
	str->di_ctime = cpu_to_be64(ip->i_inode.i_ctime.tv_sec);

	str->di_goal_meta = cpu_to_be64(ip->i_goal);
	str->di_goal_data = cpu_to_be64(ip->i_goal);
	str->di_generation = cpu_to_be64(ip->i_generation);

	str->di_flags = cpu_to_be32(ip->i_diskflags);
	str->di_height = cpu_to_be16(ip->i_height);
	str->di_payload_format = cpu_to_be32(S_ISDIR(ip->i_inode.i_mode) &&
					     !(ip->i_diskflags & GFS2_DIF_EXHASH) ?
					     GFS2_FORMAT_DE : 0);
	str->di_depth = cpu_to_be16(ip->i_depth);
	str->di_entries = cpu_to_be32(ip->i_entries);

	str->di_eattr = cpu_to_be64(ip->i_eattr);
	str->di_atime_nsec = cpu_to_be32(ip->i_inode.i_atime.tv_nsec);
	str->di_mtime_nsec = cpu_to_be32(ip->i_inode.i_mtime.tv_nsec);
	str->di_ctime_nsec = cpu_to_be32(ip->i_inode.i_ctime.tv_nsec);
}
741 742 743 744

/**
 * gfs2_write_inode - Make sure the inode is stable on the disk
 * @inode: The inode
745
 * @wbc: The writeback control structure
746 747 748 749
 *
 * Returns: errno
 */

750
static int gfs2_write_inode(struct inode *inode, struct writeback_control *wbc)
751 752 753
{
	struct gfs2_inode *ip = GFS2_I(inode);
	struct gfs2_sbd *sdp = GFS2_SB(inode);
754
	struct address_space *metamapping = gfs2_glock2aspace(ip->i_gl);
755
	struct backing_dev_info *bdi = inode_to_bdi(metamapping->host);
S
Steven Whitehouse 已提交
756 757
	int ret = 0;

758
	if (wbc->sync_mode == WB_SYNC_ALL)
759
		gfs2_log_flush(GFS2_SB(inode), ip->i_gl, NORMAL_FLUSH);
760
	if (bdi->wb.dirty_exceeded)
761
		gfs2_ail1_flush(sdp, wbc);
762 763
	else
		filemap_fdatawrite(metamapping);
S
Steven Whitehouse 已提交
764
	if (wbc->sync_mode == WB_SYNC_ALL)
765 766 767
		ret = filemap_fdatawait(metamapping);
	if (ret)
		mark_inode_dirty_sync(inode);
768 769 770
	return ret;
}

S
Steven Whitehouse 已提交
771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795
/**
 * gfs2_dirty_inode - check for atime updates
 * @inode: The inode in question
 * @flags: The type of dirty
 *
 * Unfortunately it can be called under any combination of inode
 * glock and transaction lock, so we have to check carefully.
 *
 * At the moment this deals only with atime - it should be possible
 * to expand that role in future, once a review of the locking has
 * been carried out.
 */

static void gfs2_dirty_inode(struct inode *inode, int flags)
{
	struct gfs2_inode *ip = GFS2_I(inode);
	struct gfs2_sbd *sdp = GFS2_SB(inode);
	struct buffer_head *bh;
	struct gfs2_holder gh;
	int need_unlock = 0;
	int need_endtrans = 0;
	int ret;

	if (!(flags & (I_DIRTY_DATASYNC|I_DIRTY_SYNC)))
		return;
796 797
	if (unlikely(test_bit(SDF_SHUTDOWN, &sdp->sd_flags)))
		return;
S
Steven Whitehouse 已提交
798 799 800 801 802 803 804
	if (!gfs2_glock_is_locked_by_me(ip->i_gl)) {
		ret = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh);
		if (ret) {
			fs_err(sdp, "dirty_inode: glock %d\n", ret);
			return;
		}
		need_unlock = 1;
805 806
	} else if (WARN_ON_ONCE(ip->i_gl->gl_state != LM_ST_EXCLUSIVE))
		return;
S
Steven Whitehouse 已提交
807 808 809 810 811 812 813 814 815 816 817 818

	if (current->journal_info == NULL) {
		ret = gfs2_trans_begin(sdp, RES_DINODE, 0);
		if (ret) {
			fs_err(sdp, "dirty_inode: gfs2_trans_begin %d\n", ret);
			goto out;
		}
		need_endtrans = 1;
	}

	ret = gfs2_meta_inode_buffer(ip, &bh);
	if (ret == 0) {
819
		gfs2_trans_add_meta(ip->i_gl, bh);
S
Steven Whitehouse 已提交
820 821 822 823 824 825 826 827 828 829 830
		gfs2_dinode_out(ip, bh->b_data);
		brelse(bh);
	}

	if (need_endtrans)
		gfs2_trans_end(sdp);
out:
	if (need_unlock)
		gfs2_glock_dq_uninit(&gh);
}

831 832 833 834 835 836 837 838 839
/**
 * gfs2_make_fs_ro - Turn a Read-Write FS into a Read-Only one
 * @sdp: the filesystem
 *
 * Returns: errno
 */

static int gfs2_make_fs_ro(struct gfs2_sbd *sdp)
{
840
	struct gfs2_holder freeze_gh;
841 842
	int error;

843
	error = gfs2_glock_nq_init(sdp->sd_freeze_gl, LM_ST_SHARED, GL_NOCACHE,
844
				   &freeze_gh);
845 846 847
	if (error && !test_bit(SDF_SHUTDOWN, &sdp->sd_flags))
		return error;

848 849 850
	kthread_stop(sdp->sd_quotad_process);
	kthread_stop(sdp->sd_logd_process);

851
	flush_workqueue(gfs2_delete_workqueue);
852
	gfs2_quota_sync(sdp->sd_vfs, 0);
853
	gfs2_statfs_sync(sdp->sd_vfs, 0);
854

855
	gfs2_log_flush(sdp, NULL, SHUTDOWN_FLUSH);
856
	wait_event(sdp->sd_reserving_log_wait, atomic_read(&sdp->sd_reserving_log) == 0);
857
	gfs2_assert_warn(sdp, atomic_read(&sdp->sd_log_blks_free) == sdp->sd_jdesc->jd_blocks);
858

A
Andreas Gruenbacher 已提交
859
	if (gfs2_holder_initialized(&freeze_gh))
860
		gfs2_glock_dq_uninit(&freeze_gh);
861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890

	gfs2_quota_cleanup(sdp);

	return error;
}

/**
 * gfs2_put_super - Unmount the filesystem
 * @sb: The VFS superblock
 *
 */

static void gfs2_put_super(struct super_block *sb)
{
	struct gfs2_sbd *sdp = sb->s_fs_info;
	int error;
	struct gfs2_jdesc *jd;

	/* No more recovery requests */
	set_bit(SDF_NORECOVERY, &sdp->sd_flags);
	smp_mb();

	/* Wait on outstanding recovery */
restart:
	spin_lock(&sdp->sd_jindex_spin);
	list_for_each_entry(jd, &sdp->sd_jindex_list, jd_list) {
		if (!test_bit(JDF_RECOVERY, &jd->jd_flags))
			continue;
		spin_unlock(&sdp->sd_jindex_spin);
		wait_on_bit(&jd->jd_flags, JDF_RECOVERY,
891
			    TASK_UNINTERRUPTIBLE);
892 893 894 895
		goto restart;
	}
	spin_unlock(&sdp->sd_jindex_spin);

896
	if (!sb_rdonly(sb)) {
897 898 899 900 901 902 903 904 905 906 907 908 909 910
		error = gfs2_make_fs_ro(sdp);
		if (error)
			gfs2_io_error(sdp);
	}
	/*  At this point, we're through modifying the disk  */

	/*  Release stuff  */

	iput(sdp->sd_jindex);
	iput(sdp->sd_statfs_inode);
	iput(sdp->sd_rindex);
	iput(sdp->sd_quota_inode);

	gfs2_glock_put(sdp->sd_rename_gl);
911
	gfs2_glock_put(sdp->sd_freeze_gl);
912 913 914 915 916 917 918 919 920 921 922 923 924 925 926

	if (!sdp->sd_args.ar_spectator) {
		gfs2_glock_dq_uninit(&sdp->sd_journal_gh);
		gfs2_glock_dq_uninit(&sdp->sd_jinode_gh);
		gfs2_glock_dq_uninit(&sdp->sd_sc_gh);
		gfs2_glock_dq_uninit(&sdp->sd_qc_gh);
		iput(sdp->sd_sc_inode);
		iput(sdp->sd_qc_inode);
	}

	gfs2_glock_dq_uninit(&sdp->sd_live_gh);
	gfs2_clear_rgrpd(sdp);
	gfs2_jindex_free(sdp);
	/*  Take apart glock structures and buffer lists  */
	gfs2_gl_hash_clear(sdp);
927
	gfs2_delete_debugfs_file(sdp);
928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943
	/*  Unmount the locking protocol  */
	gfs2_lm_unmount(sdp);

	/*  At this point, we're through participating in the lockspace  */
	gfs2_sys_fs_del(sdp);
}

/**
 * gfs2_sync_fs - sync the filesystem
 * @sb: the superblock
 *
 * Flushes the log to disk.
 */

static int gfs2_sync_fs(struct super_block *sb, int wait)
{
944
	struct gfs2_sbd *sdp = sb->s_fs_info;
945 946

	gfs2_quota_sync(sb, -1);
947
	if (wait)
948
		gfs2_log_flush(sdp, NULL, NORMAL_FLUSH);
949
	return sdp->sd_log_error;
950 951
}

952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981
void gfs2_freeze_func(struct work_struct *work)
{
	int error;
	struct gfs2_holder freeze_gh;
	struct gfs2_sbd *sdp = container_of(work, struct gfs2_sbd, sd_freeze_work);
	struct super_block *sb = sdp->sd_vfs;

	atomic_inc(&sb->s_active);
	error = gfs2_glock_nq_init(sdp->sd_freeze_gl, LM_ST_SHARED, 0,
				   &freeze_gh);
	if (error) {
		printk(KERN_INFO "GFS2: couln't get freeze lock : %d\n", error);
		gfs2_assert_withdraw(sdp, 0);
	}
	else {
		atomic_set(&sdp->sd_freeze_state, SFS_UNFROZEN);
		error = thaw_super(sb);
		if (error) {
			printk(KERN_INFO "GFS2: couldn't thaw filesystem: %d\n",
			       error);
			gfs2_assert_withdraw(sdp, 0);
		}
		if (!test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags))
			freeze_gh.gh_flags |= GL_NOCACHE;
		gfs2_glock_dq_uninit(&freeze_gh);
	}
	deactivate_super(sb);
	return;
}

982 983 984 985 986 987 988 989 990
/**
 * gfs2_freeze - prevent further writes to the filesystem
 * @sb: the VFS structure for the filesystem
 *
 */

static int gfs2_freeze(struct super_block *sb)
{
	struct gfs2_sbd *sdp = sb->s_fs_info;
991
	int error = 0;
992

993 994 995 996 997 998 999 1000
	mutex_lock(&sdp->sd_freeze_mutex);
	if (atomic_read(&sdp->sd_freeze_state) != SFS_UNFROZEN)
		goto out;

	if (test_bit(SDF_SHUTDOWN, &sdp->sd_flags)) {
		error = -EINVAL;
		goto out;
	}
1001 1002

	for (;;) {
S
Steven Whitehouse 已提交
1003
		error = gfs2_lock_fs_check_clean(sdp, &sdp->sd_freeze_gh);
1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019
		if (!error)
			break;

		switch (error) {
		case -EBUSY:
			fs_err(sdp, "waiting for recovery before freeze\n");
			break;

		default:
			fs_err(sdp, "error freezing FS: %d\n", error);
			break;
		}

		fs_err(sdp, "retrying...\n");
		msleep(1000);
	}
1020 1021 1022 1023
	error = 0;
out:
	mutex_unlock(&sdp->sd_freeze_mutex);
	return error;
1024 1025 1026 1027 1028 1029 1030 1031 1032 1033
}

/**
 * gfs2_unfreeze - reallow writes to the filesystem
 * @sb: the VFS structure for the filesystem
 *
 */

static int gfs2_unfreeze(struct super_block *sb)
{
S
Steven Whitehouse 已提交
1034 1035
	struct gfs2_sbd *sdp = sb->s_fs_info;

1036 1037
	mutex_lock(&sdp->sd_freeze_mutex);
        if (atomic_read(&sdp->sd_freeze_state) != SFS_FROZEN ||
A
Andreas Gruenbacher 已提交
1038
	    !gfs2_holder_initialized(&sdp->sd_freeze_gh)) {
1039 1040 1041 1042
		mutex_unlock(&sdp->sd_freeze_mutex);
                return 0;
	}

S
Steven Whitehouse 已提交
1043
	gfs2_glock_dq_uninit(&sdp->sd_freeze_gh);
1044
	mutex_unlock(&sdp->sd_freeze_mutex);
1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088
	return 0;
}

/**
 * statfs_fill - fill in the sg for a given RG
 * @rgd: the RG
 * @sc: the sc structure
 *
 * Returns: 0 on success, -ESTALE if the LVB is invalid
 */

static int statfs_slow_fill(struct gfs2_rgrpd *rgd,
			    struct gfs2_statfs_change_host *sc)
{
	gfs2_rgrp_verify(rgd);
	sc->sc_total += rgd->rd_data;
	sc->sc_free += rgd->rd_free;
	sc->sc_dinodes += rgd->rd_dinodes;
	return 0;
}

/**
 * gfs2_statfs_slow - Stat a filesystem using asynchronous locking
 * @sdp: the filesystem
 * @sc: the sc info that will be returned
 *
 * Any error (other than a signal) will cause this routine to fall back
 * to the synchronous version.
 *
 * FIXME: This really shouldn't busy wait like this.
 *
 * Returns: errno
 */

static int gfs2_statfs_slow(struct gfs2_sbd *sdp, struct gfs2_statfs_change_host *sc)
{
	struct gfs2_rgrpd *rgd_next;
	struct gfs2_holder *gha, *gh;
	unsigned int slots = 64;
	unsigned int x;
	int done;
	int error = 0, err;

	memset(sc, 0, sizeof(struct gfs2_statfs_change_host));
A
Andreas Gruenbacher 已提交
1089
	gha = kmalloc(slots * sizeof(struct gfs2_holder), GFP_KERNEL);
1090 1091
	if (!gha)
		return -ENOMEM;
A
Andreas Gruenbacher 已提交
1092 1093
	for (x = 0; x < slots; x++)
		gfs2_holder_mark_uninitialized(gha + x);
1094 1095 1096 1097 1098 1099 1100 1101 1102

	rgd_next = gfs2_rgrpd_get_first(sdp);

	for (;;) {
		done = 1;

		for (x = 0; x < slots; x++) {
			gh = gha + x;

A
Andreas Gruenbacher 已提交
1103
			if (gfs2_holder_initialized(gh) && gfs2_glock_poll(gh)) {
1104 1105 1106 1107 1108
				err = gfs2_glock_wait(gh);
				if (err) {
					gfs2_holder_uninit(gh);
					error = err;
				} else {
1109 1110 1111 1112 1113 1114
					if (!error) {
						struct gfs2_rgrpd *rgd =
							gfs2_glock2rgrp(gh->gh_gl);

						error = statfs_slow_fill(rgd, sc);
					}
1115 1116 1117 1118
					gfs2_glock_dq_uninit(gh);
				}
			}

A
Andreas Gruenbacher 已提交
1119
			if (gfs2_holder_initialized(gh))
1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185
				done = 0;
			else if (rgd_next && !error) {
				error = gfs2_glock_nq_init(rgd_next->rd_gl,
							   LM_ST_SHARED,
							   GL_ASYNC,
							   gh);
				rgd_next = gfs2_rgrpd_get_next(rgd_next);
				done = 0;
			}

			if (signal_pending(current))
				error = -ERESTARTSYS;
		}

		if (done)
			break;

		yield();
	}

	kfree(gha);
	return error;
}

/**
 * gfs2_statfs_i - Do a statfs
 * @sdp: the filesystem
 * @sg: the sg structure
 *
 * Returns: errno
 */

static int gfs2_statfs_i(struct gfs2_sbd *sdp, struct gfs2_statfs_change_host *sc)
{
	struct gfs2_statfs_change_host *m_sc = &sdp->sd_statfs_master;
	struct gfs2_statfs_change_host *l_sc = &sdp->sd_statfs_local;

	spin_lock(&sdp->sd_statfs_spin);

	*sc = *m_sc;
	sc->sc_total += l_sc->sc_total;
	sc->sc_free += l_sc->sc_free;
	sc->sc_dinodes += l_sc->sc_dinodes;

	spin_unlock(&sdp->sd_statfs_spin);

	if (sc->sc_free < 0)
		sc->sc_free = 0;
	if (sc->sc_free > sc->sc_total)
		sc->sc_free = sc->sc_total;
	if (sc->sc_dinodes < 0)
		sc->sc_dinodes = 0;

	return 0;
}

/**
 * gfs2_statfs - Gather and return stats about the filesystem
 * @sb: The superblock
 * @statfsbuf: The buffer
 *
 * Returns: 0 on success or error code
 */

static int gfs2_statfs(struct dentry *dentry, struct kstatfs *buf)
{
1186
	struct super_block *sb = dentry->d_sb;
1187 1188 1189 1190
	struct gfs2_sbd *sdp = sb->s_fs_info;
	struct gfs2_statfs_change_host sc;
	int error;

1191 1192 1193 1194
	error = gfs2_rindex_update(sdp);
	if (error)
		return error;

1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230
	if (gfs2_tune_get(sdp, gt_statfs_slow))
		error = gfs2_statfs_slow(sdp, &sc);
	else
		error = gfs2_statfs_i(sdp, &sc);

	if (error)
		return error;

	buf->f_type = GFS2_MAGIC;
	buf->f_bsize = sdp->sd_sb.sb_bsize;
	buf->f_blocks = sc.sc_total;
	buf->f_bfree = sc.sc_free;
	buf->f_bavail = sc.sc_free;
	buf->f_files = sc.sc_dinodes + sc.sc_free;
	buf->f_ffree = sc.sc_free;
	buf->f_namelen = GFS2_FNAMESIZE;

	return 0;
}

/**
 * gfs2_remount_fs - called when the FS is remounted
 * @sb:  the filesystem
 * @flags:  the remount flags
 * @data:  extra data passed in (not used right now)
 *
 * Returns: errno
 */

static int gfs2_remount_fs(struct super_block *sb, int *flags, char *data)
{
	struct gfs2_sbd *sdp = sb->s_fs_info;
	struct gfs2_args args = sdp->sd_args; /* Default to current settings */
	struct gfs2_tune *gt = &sdp->sd_tune;
	int error;

1231 1232
	sync_filesystem(sb);

1233
	spin_lock(&gt->gt_spin);
1234
	args.ar_commit = gt->gt_logd_secs;
1235 1236 1237 1238 1239
	args.ar_quota_quantum = gt->gt_quota_quantum;
	if (gt->gt_statfs_slow)
		args.ar_statfs_quantum = 0;
	else
		args.ar_statfs_quantum = gt->gt_statfs_quantum;
1240
	spin_unlock(&gt->gt_spin);
1241
	error = gfs2_mount_args(&args, data);
1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273
	if (error)
		return error;

	/* Not allowed to change locking details */
	if (strcmp(args.ar_lockproto, sdp->sd_args.ar_lockproto) ||
	    strcmp(args.ar_locktable, sdp->sd_args.ar_locktable) ||
	    strcmp(args.ar_hostdata, sdp->sd_args.ar_hostdata))
		return -EINVAL;

	/* Some flags must not be changed */
	if (args_neq(&args, &sdp->sd_args, spectator) ||
	    args_neq(&args, &sdp->sd_args, localflocks) ||
	    args_neq(&args, &sdp->sd_args, meta))
		return -EINVAL;

	if (sdp->sd_args.ar_spectator)
		*flags |= MS_RDONLY;

	if ((sb->s_flags ^ *flags) & MS_RDONLY) {
		if (*flags & MS_RDONLY)
			error = gfs2_make_fs_ro(sdp);
		else
			error = gfs2_make_fs_rw(sdp);
		if (error)
			return error;
	}

	sdp->sd_args = args;
	if (sdp->sd_args.ar_posix_acl)
		sb->s_flags |= MS_POSIXACL;
	else
		sb->s_flags &= ~MS_POSIXACL;
1274 1275 1276 1277
	if (sdp->sd_args.ar_nobarrier)
		set_bit(SDF_NOBARRIERS, &sdp->sd_flags);
	else
		clear_bit(SDF_NOBARRIERS, &sdp->sd_flags);
1278
	spin_lock(&gt->gt_spin);
1279
	gt->gt_logd_secs = args.ar_commit;
1280 1281 1282 1283 1284 1285 1286 1287 1288
	gt->gt_quota_quantum = args.ar_quota_quantum;
	if (args.ar_statfs_quantum) {
		gt->gt_statfs_slow = 0;
		gt->gt_statfs_quantum = args.ar_statfs_quantum;
	}
	else {
		gt->gt_statfs_slow = 1;
		gt->gt_statfs_quantum = 30;
	}
1289 1290
	spin_unlock(&gt->gt_spin);

1291
	gfs2_online_uevent(sdp);
1292 1293 1294 1295 1296 1297 1298
	return 0;
}

/**
 * gfs2_drop_inode - Drop an inode (test for remote unlink)
 * @inode: The inode to drop
 *
A
Andreas Gruenbacher 已提交
1299
 * If we've received a callback on an iopen lock then it's because a
1300 1301 1302 1303 1304 1305 1306 1307 1308 1309
 * remote node tried to deallocate the inode but failed due to this node
 * still having the inode open. Here we mark the link count zero
 * since we know that it must have reached zero if the GLF_DEMOTE flag
 * is set on the iopen glock. If we didn't do a disk read since the
 * remote node removed the final link then we might otherwise miss
 * this event. This check ensures that this node will deallocate the
 * inode's blocks, or alternatively pass the baton on to another
 * node for later deallocation.
 */

1310
static int gfs2_drop_inode(struct inode *inode)
1311 1312 1313
{
	struct gfs2_inode *ip = GFS2_I(inode);

A
Andreas Gruenbacher 已提交
1314 1315 1316
	if (!test_bit(GIF_FREE_VFS_INODE, &ip->i_flags) &&
	    inode->i_nlink &&
	    gfs2_holder_initialized(&ip->i_iopen_gh)) {
1317
		struct gfs2_glock *gl = ip->i_iopen_gh.gh_gl;
A
Andreas Gruenbacher 已提交
1318
		if (test_bit(GLF_DEMOTE, &gl->gl_flags))
1319 1320
			clear_nlink(inode);
	}
1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337

	/*
	 * When under memory pressure when an inode's link count has dropped to
	 * zero, defer deleting the inode to the delete workqueue.  This avoids
	 * calling into DLM under memory pressure, which can deadlock.
	 */
	if (!inode->i_nlink &&
	    unlikely(current->flags & PF_MEMALLOC) &&
	    gfs2_holder_initialized(&ip->i_iopen_gh)) {
		struct gfs2_glock *gl = ip->i_iopen_gh.gh_gl;

		gfs2_glock_hold(gl);
		if (queue_work(gfs2_delete_workqueue, &gl->gl_delete) == 0)
			gfs2_glock_queue_put(gl);
		return false;
	}

1338
	return generic_drop_inode(inode);
1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353
}

static int is_ancestor(const struct dentry *d1, const struct dentry *d2)
{
	do {
		if (d1 == d2)
			return 1;
		d1 = d1->d_parent;
	} while (!IS_ROOT(d1));
	return 0;
}

/**
 * gfs2_show_options - Show mount options for /proc/mounts
 * @s: seq_file structure
1354
 * @root: root of this (sub)tree
1355 1356 1357 1358
 *
 * Returns: 0 on success or error code
 */

1359
static int gfs2_show_options(struct seq_file *s, struct dentry *root)
1360
{
1361
	struct gfs2_sbd *sdp = root->d_sb->s_fs_info;
1362
	struct gfs2_args *args = &sdp->sd_args;
1363
	int val;
1364

1365
	if (is_ancestor(root, sdp->sd_master_dir))
1366
		seq_puts(s, ",meta");
1367
	if (args->ar_lockproto[0])
1368
		seq_show_option(s, "lockproto", args->ar_lockproto);
1369
	if (args->ar_locktable[0])
1370
		seq_show_option(s, "locktable", args->ar_locktable);
1371
	if (args->ar_hostdata[0])
1372
		seq_show_option(s, "hostdata", args->ar_hostdata);
1373
	if (args->ar_spectator)
1374
		seq_puts(s, ",spectator");
1375
	if (args->ar_localflocks)
1376
		seq_puts(s, ",localflocks");
1377
	if (args->ar_debug)
1378
		seq_puts(s, ",debug");
1379
	if (args->ar_posix_acl)
1380
		seq_puts(s, ",acl");
1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399
	if (args->ar_quota != GFS2_QUOTA_DEFAULT) {
		char *state;
		switch (args->ar_quota) {
		case GFS2_QUOTA_OFF:
			state = "off";
			break;
		case GFS2_QUOTA_ACCOUNT:
			state = "account";
			break;
		case GFS2_QUOTA_ON:
			state = "on";
			break;
		default:
			state = "unknown";
			break;
		}
		seq_printf(s, ",quota=%s", state);
	}
	if (args->ar_suiddir)
1400
		seq_puts(s, ",suiddir");
1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416
	if (args->ar_data != GFS2_DATA_DEFAULT) {
		char *state;
		switch (args->ar_data) {
		case GFS2_DATA_WRITEBACK:
			state = "writeback";
			break;
		case GFS2_DATA_ORDERED:
			state = "ordered";
			break;
		default:
			state = "unknown";
			break;
		}
		seq_printf(s, ",data=%s", state);
	}
	if (args->ar_discard)
1417
		seq_puts(s, ",discard");
1418 1419
	val = sdp->sd_tune.gt_logd_secs;
	if (val != 30)
1420 1421 1422 1423
		seq_printf(s, ",commit=%d", val);
	val = sdp->sd_tune.gt_statfs_quantum;
	if (val != 30)
		seq_printf(s, ",statfs_quantum=%d", val);
1424 1425
	else if (sdp->sd_tune.gt_statfs_slow)
		seq_puts(s, ",statfs_quantum=0");
1426 1427 1428 1429 1430
	val = sdp->sd_tune.gt_quota_quantum;
	if (val != 60)
		seq_printf(s, ",quota_quantum=%d", val);
	if (args->ar_statfs_percent)
		seq_printf(s, ",statfs_percent=%d", args->ar_statfs_percent);
1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446
	if (args->ar_errors != GFS2_ERRORS_DEFAULT) {
		const char *state;

		switch (args->ar_errors) {
		case GFS2_ERRORS_WITHDRAW:
			state = "withdraw";
			break;
		case GFS2_ERRORS_PANIC:
			state = "panic";
			break;
		default:
			state = "unknown";
			break;
		}
		seq_printf(s, ",errors=%s", state);
	}
1447
	if (test_bit(SDF_NOBARRIERS, &sdp->sd_flags))
1448
		seq_puts(s, ",nobarrier");
1449
	if (test_bit(SDF_DEMOTE, &sdp->sd_flags))
1450
		seq_puts(s, ",demote_interface_used");
1451
	if (args->ar_rgrplvb)
1452
		seq_puts(s, ",rgrplvb");
1453 1454
	if (args->ar_loccookie)
		seq_puts(s, ",loccookie");
1455 1456 1457
	return 0;
}

1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475
static void gfs2_final_release_pages(struct gfs2_inode *ip)
{
	struct inode *inode = &ip->i_inode;
	struct gfs2_glock *gl = ip->i_gl;

	truncate_inode_pages(gfs2_glock2aspace(ip->i_gl), 0);
	truncate_inode_pages(&inode->i_data, 0);

	if (atomic_read(&gl->gl_revokes) == 0) {
		clear_bit(GLF_LFLUSH, &gl->gl_flags);
		clear_bit(GLF_DIRTY, &gl->gl_flags);
	}
}

static int gfs2_dinode_dealloc(struct gfs2_inode *ip)
{
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
	struct gfs2_rgrpd *rgd;
1476
	struct gfs2_holder gh;
1477 1478 1479
	int error;

	if (gfs2_get_inode_blocks(&ip->i_inode) != 1) {
1480
		gfs2_consist_inode(ip);
1481 1482 1483
		return -EIO;
	}

B
Bob Peterson 已提交
1484 1485 1486
	error = gfs2_rindex_update(sdp);
	if (error)
		return error;
1487

1488
	error = gfs2_quota_hold(ip, NO_UID_QUOTA_CHANGE, NO_GID_QUOTA_CHANGE);
1489
	if (error)
1490
		return error;
1491

S
Steven Whitehouse 已提交
1492
	rgd = gfs2_blk2rgrpd(sdp, ip->i_no_addr, 1);
1493 1494 1495
	if (!rgd) {
		gfs2_consist_inode(ip);
		error = -EIO;
1496
		goto out_qs;
1497 1498
	}

1499
	error = gfs2_glock_nq_init(rgd->rd_gl, LM_ST_EXCLUSIVE, 0, &gh);
1500
	if (error)
1501
		goto out_qs;
1502

1503 1504
	error = gfs2_trans_begin(sdp, RES_RG_BIT + RES_STATFS + RES_QUOTA,
				 sdp->sd_jdesc->jd_blocks);
1505 1506 1507 1508 1509 1510 1511 1512 1513 1514
	if (error)
		goto out_rg_gunlock;

	gfs2_free_di(rgd, ip);

	gfs2_final_release_pages(ip);

	gfs2_trans_end(sdp);

out_rg_gunlock:
1515
	gfs2_glock_dq_uninit(&gh);
1516 1517 1518 1519 1520
out_qs:
	gfs2_quota_unhold(ip);
	return error;
}

1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536
/**
 * gfs2_glock_put_eventually
 * @gl:	The glock to put
 *
 * When under memory pressure, trigger a deferred glock put to make sure we
 * won't call into DLM and deadlock.  Otherwise, put the glock directly.
 */

static void gfs2_glock_put_eventually(struct gfs2_glock *gl)
{
	if (current->flags & PF_MEMALLOC)
		gfs2_glock_queue_put(gl);
	else
		gfs2_glock_put(gl);
}

1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550
/**
 * gfs2_evict_inode - Remove an inode from cache
 * @inode: The inode to evict
 *
 * There are three cases to consider:
 * 1. i_nlink == 0, we are final opener (and must deallocate)
 * 2. i_nlink == 0, we are not the final opener (and cannot deallocate)
 * 3. i_nlink > 0
 *
 * If the fs is read only, then we have to treat all cases as per #3
 * since we are unable to do any deallocation. The inode will be
 * deallocated by the next read/write node to attempt an allocation
 * in the same resource group
 *
1551 1552 1553 1554 1555 1556 1557
 * We have to (at the moment) hold the inodes main lock to cover
 * the gap between unlocking the shared lock on the iopen lock and
 * taking the exclusive lock. I'd rather do a shared -> exclusive
 * conversion on the iopen lock, but we can change that later. This
 * is safe, just less efficient.
 */

A
Al Viro 已提交
1558
static void gfs2_evict_inode(struct inode *inode)
1559
{
1560 1561
	struct super_block *sb = inode->i_sb;
	struct gfs2_sbd *sdp = sb->s_fs_info;
1562 1563
	struct gfs2_inode *ip = GFS2_I(inode);
	struct gfs2_holder gh;
1564
	struct address_space *metamapping;
1565 1566
	int error;

1567 1568 1569 1570 1571
	if (test_bit(GIF_FREE_VFS_INODE, &ip->i_flags)) {
		clear_inode(inode);
		return;
	}

1572
	if (inode->i_nlink || sb_rdonly(sb))
A
Al Viro 已提交
1573 1574
		goto out;

1575 1576 1577 1578 1579 1580
	if (test_bit(GIF_ALLOC_FAILED, &ip->i_flags)) {
		BUG_ON(!gfs2_glock_is_locked_by_me(ip->i_gl));
		gfs2_holder_mark_uninitialized(&gh);
		goto alloc_failed;
	}

1581 1582 1583 1584
	/* Deletes should never happen under memory pressure anymore.  */
	if (WARN_ON_ONCE(current->flags & PF_MEMALLOC))
		goto out;

1585 1586
	/* Must not read inode block until block type has been verified */
	error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, GL_SKIP, &gh);
1587
	if (unlikely(error)) {
1588
		glock_clear_object(ip->i_iopen_gh.gh_gl, ip);
B
Bob Peterson 已提交
1589
		ip->i_iopen_gh.gh_flags |= GL_NOCACHE;
1590
		gfs2_glock_dq_uninit(&ip->i_iopen_gh);
1591 1592 1593
		goto out;
	}

1594 1595 1596
	error = gfs2_check_blk_type(sdp, ip->i_no_addr, GFS2_BLKST_UNLINKED);
	if (error)
		goto out_truncate;
1597

1598 1599 1600 1601 1602 1603
	if (test_bit(GIF_INVALID, &ip->i_flags)) {
		error = gfs2_inode_refresh(ip);
		if (error)
			goto out_truncate;
	}

1604 1605 1606 1607 1608 1609
	/*
	 * The inode may have been recreated in the meantime.
	 */
	if (inode->i_nlink)
		goto out_truncate;

1610
alloc_failed:
A
Andreas Gruenbacher 已提交
1611
	if (gfs2_holder_initialized(&ip->i_iopen_gh) &&
1612 1613 1614 1615 1616 1617 1618 1619 1620
	    test_bit(HIF_HOLDER, &ip->i_iopen_gh.gh_iflags)) {
		ip->i_iopen_gh.gh_flags |= GL_NOCACHE;
		gfs2_glock_dq_wait(&ip->i_iopen_gh);
		gfs2_holder_reinit(LM_ST_EXCLUSIVE, LM_FLAG_TRY_1CB | GL_NOCACHE,
				   &ip->i_iopen_gh);
		error = gfs2_glock_nq(&ip->i_iopen_gh);
		if (error)
			goto out_truncate;
	}
1621

1622 1623
	/* Case 1 starts here */

1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642
	if (S_ISDIR(inode->i_mode) &&
	    (ip->i_diskflags & GFS2_DIF_EXHASH)) {
		error = gfs2_dir_exhash_dealloc(ip);
		if (error)
			goto out_unlock;
	}

	if (ip->i_eattr) {
		error = gfs2_ea_dealloc(ip);
		if (error)
			goto out_unlock;
	}

	if (!gfs2_is_stuffed(ip)) {
		error = gfs2_file_dealloc(ip);
		if (error)
			goto out_unlock;
	}

1643 1644 1645 1646 1647
	/* We're about to clear the bitmap for the dinode, but as soon as we
	   do, gfs2_create_inode can create another inode at the same block
	   location and try to set gl_object again. We clear gl_object here so
	   that subsequent inode creates don't see an old gl_object. */
	glock_clear_object(ip->i_gl, ip);
1648
	error = gfs2_dinode_dealloc(ip);
1649
	goto out_unlock;
1650 1651

out_truncate:
1652
	gfs2_log_flush(sdp, ip->i_gl, NORMAL_FLUSH);
1653
	metamapping = gfs2_glock2aspace(ip->i_gl);
1654 1655 1656 1657
	if (test_bit(GLF_DIRTY, &ip->i_gl->gl_flags)) {
		filemap_fdatawrite(metamapping);
		filemap_fdatawait(metamapping);
	}
1658
	write_inode_now(inode, 1);
1659
	gfs2_ail_flush(ip->i_gl, 0);
1660

1661
	/* Case 2 starts here */
1662 1663 1664
	error = gfs2_trans_begin(sdp, 0, sdp->sd_jdesc->jd_blocks);
	if (error)
		goto out_unlock;
1665 1666
	/* Needs to be done before glock release & also in a transaction */
	truncate_inode_pages(&inode->i_data, 0);
1667
	truncate_inode_pages(metamapping, 0);
1668 1669 1670
	gfs2_trans_end(sdp);

out_unlock:
1671
	/* Error path for case 1 */
1672 1673
	if (gfs2_rs_active(&ip->i_res))
		gfs2_rs_deltree(&ip->i_res);
B
Bob Peterson 已提交
1674

A
Andreas Gruenbacher 已提交
1675
	if (gfs2_holder_initialized(&ip->i_iopen_gh)) {
1676
		glock_clear_object(ip->i_iopen_gh.gh_gl, ip);
1677 1678
		if (test_bit(HIF_HOLDER, &ip->i_iopen_gh.gh_iflags)) {
			ip->i_iopen_gh.gh_flags |= GL_NOCACHE;
1679
			gfs2_glock_dq(&ip->i_iopen_gh);
1680 1681
		}
		gfs2_holder_uninit(&ip->i_iopen_gh);
B
Bob Peterson 已提交
1682
	}
1683 1684
	if (gfs2_holder_initialized(&gh)) {
		glock_clear_object(ip->i_gl, ip);
1685
		gfs2_glock_dq_uninit(&gh);
1686
	}
1687
	if (error && error != GLR_TRYFAILED && error != -EROFS)
A
Al Viro 已提交
1688
		fs_warn(sdp, "gfs2_evict_inode: %d\n", error);
1689
out:
1690
	/* Case 3 starts here */
1691
	truncate_inode_pages_final(&inode->i_data);
1692
	gfs2_rsqa_delete(ip, NULL);
1693
	gfs2_ordered_del_inode(ip);
1694
	clear_inode(inode);
1695
	gfs2_dir_hash_inval(ip);
1696
	glock_clear_object(ip->i_gl, ip);
1697
	wait_on_bit_io(&ip->i_flags, GIF_GLOP_PENDING, TASK_UNINTERRUPTIBLE);
1698
	gfs2_glock_add_to_lru(ip->i_gl);
1699
	gfs2_glock_put_eventually(ip->i_gl);
A
Al Viro 已提交
1700
	ip->i_gl = NULL;
A
Andreas Gruenbacher 已提交
1701
	if (gfs2_holder_initialized(&ip->i_iopen_gh)) {
1702 1703 1704
		struct gfs2_glock *gl = ip->i_iopen_gh.gh_gl;

		glock_clear_object(gl, ip);
B
Bob Peterson 已提交
1705
		ip->i_iopen_gh.gh_flags |= GL_NOCACHE;
1706
		gfs2_glock_hold(gl);
1707
		gfs2_glock_dq_uninit(&ip->i_iopen_gh);
1708
		gfs2_glock_put_eventually(gl);
A
Al Viro 已提交
1709
	}
1710 1711 1712 1713 1714 1715 1716 1717 1718 1719
}

static struct inode *gfs2_alloc_inode(struct super_block *sb)
{
	struct gfs2_inode *ip;

	ip = kmem_cache_alloc(gfs2_inode_cachep, GFP_KERNEL);
	if (ip) {
		ip->i_flags = 0;
		ip->i_gl = NULL;
1720
		ip->i_rgd = NULL;
1721 1722
		memset(&ip->i_res, 0, sizeof(ip->i_res));
		RB_CLEAR_NODE(&ip->i_res.rs_node);
1723
		ip->i_rahead = 0;
1724 1725 1726 1727
	}
	return &ip->i_inode;
}

N
Nick Piggin 已提交
1728
static void gfs2_i_callback(struct rcu_head *head)
1729
{
N
Nick Piggin 已提交
1730
	struct inode *inode = container_of(head, struct inode, i_rcu);
1731 1732 1733
	kmem_cache_free(gfs2_inode_cachep, inode);
}

N
Nick Piggin 已提交
1734 1735 1736 1737 1738
static void gfs2_destroy_inode(struct inode *inode)
{
	call_rcu(&inode->i_rcu, gfs2_i_callback);
}

1739 1740 1741 1742
const struct super_operations gfs2_super_ops = {
	.alloc_inode		= gfs2_alloc_inode,
	.destroy_inode		= gfs2_destroy_inode,
	.write_inode		= gfs2_write_inode,
S
Steven Whitehouse 已提交
1743
	.dirty_inode		= gfs2_dirty_inode,
A
Al Viro 已提交
1744
	.evict_inode		= gfs2_evict_inode,
1745 1746
	.put_super		= gfs2_put_super,
	.sync_fs		= gfs2_sync_fs,
1747 1748
	.freeze_super		= gfs2_freeze,
	.thaw_super		= gfs2_unfreeze,
1749 1750 1751 1752 1753 1754
	.statfs			= gfs2_statfs,
	.remount_fs		= gfs2_remount_fs,
	.drop_inode		= gfs2_drop_inode,
	.show_options		= gfs2_show_options,
};