xfs_sync.c 14.1 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
/*
 * Copyright (c) 2000-2005 Silicon Graphics, Inc.
 * All Rights Reserved.
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License as
 * published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it would be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write the Free Software Foundation,
 * Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
 */
#include "xfs.h"
#include "xfs_fs.h"
#include "xfs_types.h"
#include "xfs_bit.h"
#include "xfs_log.h"
#include "xfs_inum.h"
#include "xfs_trans.h"
#include "xfs_sb.h"
#include "xfs_ag.h"
#include "xfs_dir2.h"
#include "xfs_dmapi.h"
#include "xfs_mount.h"
#include "xfs_bmap_btree.h"
#include "xfs_alloc_btree.h"
#include "xfs_ialloc_btree.h"
#include "xfs_btree.h"
#include "xfs_dir2_sf.h"
#include "xfs_attr_sf.h"
#include "xfs_inode.h"
#include "xfs_dinode.h"
#include "xfs_error.h"
#include "xfs_mru_cache.h"
#include "xfs_filestream.h"
#include "xfs_vnodeops.h"
#include "xfs_utils.h"
#include "xfs_buf_item.h"
#include "xfs_inode_item.h"
#include "xfs_rw.h"

47 48 49
#include <linux/kthread.h>
#include <linux/freezer.h>

50
/*
51 52
 * Sync all the inodes in the given AG according to the
 * direction given by the flags.
53
 */
54 55
STATIC int
xfs_sync_inodes_ag(
56
	xfs_mount_t	*mp,
57
	int		ag,
58
	int		flags)
59
{
60 61 62 63 64 65 66
	xfs_perag_t	*pag = &mp->m_perag[ag];
	int		nr_found;
	int		first_index = 0;
	int		error = 0;
	int		last_error = 0;
	int		fflag = XFS_B_ASYNC;
	int		lock_flags = XFS_ILOCK_SHARED;
67 68 69 70 71 72

	if (flags & SYNC_DELWRI)
		fflag = XFS_B_DELWRI;
	if (flags & SYNC_WAIT)
		fflag = 0;		/* synchronous overrides all */

D
David Chinner 已提交
73
	if (flags & SYNC_DELWRI) {
74 75 76 77
		/*
		 * We need the I/O lock if we're going to call any of
		 * the flush/inval routines.
		 */
78
		lock_flags |= XFS_IOLOCK_SHARED;
79 80 81
	}

	do {
82 83 84 85
		struct inode	*inode;
		boolean_t	inode_refed;
		xfs_inode_t	*ip = NULL;

86
		/*
87 88 89
		 * use a gang lookup to find the next inode in the tree
		 * as the tree is sparse and a gang lookup walks to find
		 * the number of objects requested.
90
		 */
91 92 93
		read_lock(&pag->pag_ici_lock);
		nr_found = radix_tree_gang_lookup(&pag->pag_ici_root,
				(void**)&ip, first_index, 1);
94

95 96 97
		if (!nr_found) {
			read_unlock(&pag->pag_ici_lock);
			break;
98 99
		}

100 101
		/* update the index for the next lookup */
		first_index = XFS_INO_TO_AGINO(mp, ip->i_ino + 1);
102 103

		/*
104 105
		 * skip inodes in reclaim. Let xfs_syncsub do that for
		 * us so we don't need to worry.
106
		 */
107
		if (xfs_iflags_test(ip, (XFS_IRECLAIM|XFS_IRECLAIMABLE))) {
108
			read_unlock(&pag->pag_ici_lock);
109 110 111
			continue;
		}

112
		/* bad inodes are dealt with elsewhere */
113 114
		inode = VFS_I(ip);
		if (is_bad_inode(inode)) {
115
			read_unlock(&pag->pag_ici_lock);
116 117 118
			continue;
		}

119
		/* nothing to sync during shutdown */
D
David Chinner 已提交
120
		if (XFS_FORCED_SHUTDOWN(mp)) {
121
			read_unlock(&pag->pag_ici_lock);
122 123 124 125
			return 0;
		}

		/*
126 127 128 129 130 131 132 133
		 * If we can't get a reference on the VFS_I, the inode must be
		 * in reclaim. If we can get the inode lock without blocking,
		 * it is safe to flush the inode because we hold the tree lock
		 * and xfs_iextract will block right now. Hence if we lock the
		 * inode while holding the tree lock, xfs_ireclaim() is
		 * guaranteed to block on the inode lock we now hold and hence
		 * it is safe to reference the inode until we drop the inode
		 * locks completely.
134
		 */
135 136
		inode_refed = B_FALSE;
		if (igrab(inode)) {
137
			read_unlock(&pag->pag_ici_lock);
138
			xfs_ilock(ip, lock_flags);
139
			inode_refed = B_TRUE;
140
		} else {
141 142 143 144 145
			if (!xfs_ilock_nowait(ip, lock_flags)) {
				/* leave it to reclaim */
				read_unlock(&pag->pag_ici_lock);
				continue;
			}
146
			read_unlock(&pag->pag_ici_lock);
147
		}
148

149 150 151 152 153 154
		/*
		 * If we have to flush data or wait for I/O completion
		 * we need to drop the ilock that we currently hold.
		 * If we need to drop the lock, insert a marker if we
		 * have not already done so.
		 */
155
		if ((flags & SYNC_DELWRI) && VN_DIRTY(inode)) {
156 157 158 159 160 161
			xfs_iunlock(ip, XFS_ILOCK_SHARED);
			error = xfs_flush_pages(ip, 0, -1, fflag, FI_NONE);
			if (flags & SYNC_IOWAIT)
				vn_iowait(ip);
			xfs_ilock(ip, XFS_ILOCK_SHARED);
		}
162

163
		if ((flags & SYNC_ATTR) && !xfs_inode_clean(ip)) {
164 165
			if (flags & SYNC_WAIT) {
				xfs_iflock(ip);
166 167 168 169
				if (!xfs_inode_clean(ip))
					error = xfs_iflush(ip, XFS_IFLUSH_SYNC);
				else
					xfs_ifunlock(ip);
170
			} else if (xfs_iflock_nowait(ip)) {
171 172 173 174
				if (!xfs_inode_clean(ip))
					error = xfs_iflush(ip, XFS_IFLUSH_DELWRI);
				else
					xfs_ifunlock(ip);
175 176 177
			}
		}

178
		if (lock_flags)
179 180
			xfs_iunlock(ip, lock_flags);

181
		if (inode_refed) {
182 183 184
			IRELE(ip);
		}

185
		if (error)
186 187 188 189
			last_error = error;
		/*
		 * bail out if the filesystem is corrupted.
		 */
190
		if (error == EFSCORRUPTED)
191 192
			return XFS_ERROR(error);

193
	} while (nr_found);
194

195 196
	return last_error;
}
197

198 199 200
int
xfs_sync_inodes(
	xfs_mount_t	*mp,
201
	int		flags)
202 203 204 205
{
	int		error;
	int		last_error;
	int		i;
206
	int		lflags = XFS_LOG_FORCE;
207

208 209 210 211
	if (mp->m_flags & XFS_MOUNT_RDONLY)
		return 0;
	error = 0;
	last_error = 0;
212

213 214 215
	if (flags & SYNC_WAIT)
		lflags |= XFS_LOG_SYNC;

216 217 218
	for (i = 0; i < mp->m_sb.sb_agcount; i++) {
		if (!mp->m_perag[i].pag_ici_init)
			continue;
219
		error = xfs_sync_inodes_ag(mp, i, flags);
220 221 222 223 224
		if (error)
			last_error = error;
		if (error == EFSCORRUPTED)
			break;
	}
225 226 227
	if (flags & SYNC_DELWRI)
		xfs_log_force(mp, 0, lflags);

228 229 230
	return XFS_ERROR(last_error);
}

231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264
STATIC int
xfs_commit_dummy_trans(
	struct xfs_mount	*mp,
	uint			log_flags)
{
	struct xfs_inode	*ip = mp->m_rootip;
	struct xfs_trans	*tp;
	int			error;

	/*
	 * Put a dummy transaction in the log to tell recovery
	 * that all others are OK.
	 */
	tp = xfs_trans_alloc(mp, XFS_TRANS_DUMMY1);
	error = xfs_trans_reserve(tp, 0, XFS_ICHANGE_LOG_RES(mp), 0, 0, 0);
	if (error) {
		xfs_trans_cancel(tp, 0);
		return error;
	}

	xfs_ilock(ip, XFS_ILOCK_EXCL);

	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
	xfs_trans_ihold(tp, ip);
	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
	/* XXX(hch): ignoring the error here.. */
	error = xfs_trans_commit(tp, 0);

	xfs_iunlock(ip, XFS_ILOCK_EXCL);

	xfs_log_force(mp, 0, log_flags);
	return 0;
}

265
int
266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315
xfs_sync_fsdata(
	struct xfs_mount	*mp,
	int			flags)
{
	struct xfs_buf		*bp;
	struct xfs_buf_log_item	*bip;
	int			error = 0;

	/*
	 * If this is xfssyncd() then only sync the superblock if we can
	 * lock it without sleeping and it is not pinned.
	 */
	if (flags & SYNC_BDFLUSH) {
		ASSERT(!(flags & SYNC_WAIT));

		bp = xfs_getsb(mp, XFS_BUF_TRYLOCK);
		if (!bp)
			goto out;

		bip = XFS_BUF_FSPRIVATE(bp, struct xfs_buf_log_item *);
		if (!bip || !xfs_buf_item_dirty(bip) || XFS_BUF_ISPINNED(bp))
			goto out_brelse;
	} else {
		bp = xfs_getsb(mp, 0);

		/*
		 * If the buffer is pinned then push on the log so we won't
		 * get stuck waiting in the write for someone, maybe
		 * ourselves, to flush the log.
		 *
		 * Even though we just pushed the log above, we did not have
		 * the superblock buffer locked at that point so it can
		 * become pinned in between there and here.
		 */
		if (XFS_BUF_ISPINNED(bp))
			xfs_log_force(mp, 0, XFS_LOG_FORCE);
	}


	if (flags & SYNC_WAIT)
		XFS_BUF_UNASYNC(bp);
	else
		XFS_BUF_ASYNC(bp);

	return xfs_bwrite(mp, bp);

 out_brelse:
	xfs_buf_relse(bp);
 out:
	return error;
316 317 318
}

/*
D
David Chinner 已提交
319 320 321 322 323 324 325 326 327 328 329
 * When remounting a filesystem read-only or freezing the filesystem, we have
 * two phases to execute. This first phase is syncing the data before we
 * quiesce the filesystem, and the second is flushing all the inodes out after
 * we've waited for all the transactions created by the first phase to
 * complete. The second phase ensures that the inodes are written to their
 * location on disk rather than just existing in transactions in the log. This
 * means after a quiesce there is no log replay required to write the inodes to
 * disk (this is the main difference between a sync and a quiesce).
 */
/*
 * First stage of freeze - no writers will make progress now we are here,
330 331
 * so we flush delwri and delalloc buffers here, then wait for all I/O to
 * complete.  Data is frozen at that point. Metadata is not frozen,
D
David Chinner 已提交
332 333
 * transactions can still occur here so don't bother flushing the buftarg
 * because it'll just get dirty again.
334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349
 */
int
xfs_quiesce_data(
	struct xfs_mount	*mp)
{
	int error;

	/* push non-blocking */
	xfs_sync_inodes(mp, SYNC_DELWRI|SYNC_BDFLUSH);
	XFS_QM_DQSYNC(mp, SYNC_BDFLUSH);
	xfs_filestream_flush(mp);

	/* push and block */
	xfs_sync_inodes(mp, SYNC_DELWRI|SYNC_WAIT|SYNC_IOWAIT);
	XFS_QM_DQSYNC(mp, SYNC_WAIT);

D
David Chinner 已提交
350
	/* write superblock and hoover up shutdown errors */
351 352
	error = xfs_sync_fsdata(mp, 0);

D
David Chinner 已提交
353
	/* flush data-only devices */
354 355 356 357
	if (mp->m_rtdev_targp)
		XFS_bflush(mp->m_rtdev_targp);

	return error;
358 359
}

D
David Chinner 已提交
360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414
STATIC void
xfs_quiesce_fs(
	struct xfs_mount	*mp)
{
	int	count = 0, pincount;

	xfs_flush_buftarg(mp->m_ddev_targp, 0);
	xfs_finish_reclaim_all(mp, 0, XFS_IFLUSH_DELWRI_ELSE_ASYNC);

	/*
	 * This loop must run at least twice.  The first instance of the loop
	 * will flush most meta data but that will generate more meta data
	 * (typically directory updates).  Which then must be flushed and
	 * logged before we can write the unmount record.
	 */
	do {
		xfs_sync_inodes(mp, SYNC_ATTR|SYNC_WAIT);
		pincount = xfs_flush_buftarg(mp->m_ddev_targp, 1);
		if (!pincount) {
			delay(50);
			count++;
		}
	} while (count < 2);
}

/*
 * Second stage of a quiesce. The data is already synced, now we have to take
 * care of the metadata. New transactions are already blocked, so we need to
 * wait for any remaining transactions to drain out before proceding.
 */
void
xfs_quiesce_attr(
	struct xfs_mount	*mp)
{
	int	error = 0;

	/* wait for all modifications to complete */
	while (atomic_read(&mp->m_active_trans) > 0)
		delay(100);

	/* flush inodes and push all remaining buffers out to disk */
	xfs_quiesce_fs(mp);

	ASSERT_ALWAYS(atomic_read(&mp->m_active_trans) == 0);

	/* Push the superblock and write an unmount record */
	error = xfs_log_sbcount(mp, 1);
	if (error)
		xfs_fs_cmn_err(CE_WARN, mp,
				"xfs_attr_quiesce: failed to log sb changes. "
				"Frozen image may not be consistent.");
	xfs_log_unmount_write(mp);
	xfs_unmountfs_writesb(mp);
}

415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493
/*
 * Enqueue a work item to be picked up by the vfs xfssyncd thread.
 * Doing this has two advantages:
 * - It saves on stack space, which is tight in certain situations
 * - It can be used (with care) as a mechanism to avoid deadlocks.
 * Flushing while allocating in a full filesystem requires both.
 */
STATIC void
xfs_syncd_queue_work(
	struct xfs_mount *mp,
	void		*data,
	void		(*syncer)(struct xfs_mount *, void *))
{
	struct bhv_vfs_sync_work *work;

	work = kmem_alloc(sizeof(struct bhv_vfs_sync_work), KM_SLEEP);
	INIT_LIST_HEAD(&work->w_list);
	work->w_syncer = syncer;
	work->w_data = data;
	work->w_mount = mp;
	spin_lock(&mp->m_sync_lock);
	list_add_tail(&work->w_list, &mp->m_sync_list);
	spin_unlock(&mp->m_sync_lock);
	wake_up_process(mp->m_sync_task);
}

/*
 * Flush delayed allocate data, attempting to free up reserved space
 * from existing allocations.  At this point a new allocation attempt
 * has failed with ENOSPC and we are in the process of scratching our
 * heads, looking about for more room...
 */
STATIC void
xfs_flush_inode_work(
	struct xfs_mount *mp,
	void		*arg)
{
	struct inode	*inode = arg;
	filemap_flush(inode->i_mapping);
	iput(inode);
}

void
xfs_flush_inode(
	xfs_inode_t	*ip)
{
	struct inode	*inode = VFS_I(ip);

	igrab(inode);
	xfs_syncd_queue_work(ip->i_mount, inode, xfs_flush_inode_work);
	delay(msecs_to_jiffies(500));
}

/*
 * This is the "bigger hammer" version of xfs_flush_inode_work...
 * (IOW, "If at first you don't succeed, use a Bigger Hammer").
 */
STATIC void
xfs_flush_device_work(
	struct xfs_mount *mp,
	void		*arg)
{
	struct inode	*inode = arg;
	sync_blockdev(mp->m_super->s_bdev);
	iput(inode);
}

void
xfs_flush_device(
	xfs_inode_t	*ip)
{
	struct inode	*inode = VFS_I(ip);

	igrab(inode);
	xfs_syncd_queue_work(ip->i_mount, inode, xfs_flush_device_work);
	delay(msecs_to_jiffies(500));
	xfs_log_force(ip->i_mount, (xfs_lsn_t)0, XFS_LOG_FORCE|XFS_LOG_SYNC);
}

494 495 496 497 498
/*
 * Every sync period we need to unpin all items, reclaim inodes, sync
 * quota and write out the superblock. We might need to cover the log
 * to indicate it is idle.
 */
499 500 501 502 503 504 505
STATIC void
xfs_sync_worker(
	struct xfs_mount *mp,
	void		*unused)
{
	int		error;

506 507 508 509 510 511 512 513 514
	if (!(mp->m_flags & XFS_MOUNT_RDONLY)) {
		xfs_log_force(mp, (xfs_lsn_t)0, XFS_LOG_FORCE);
		xfs_finish_reclaim_all(mp, 1, XFS_IFLUSH_DELWRI_ELSE_ASYNC);
		/* dgc: errors ignored here */
		error = XFS_QM_DQSYNC(mp, SYNC_BDFLUSH);
		error = xfs_sync_fsdata(mp, SYNC_BDFLUSH);
		if (xfs_log_need_covered(mp))
			error = xfs_commit_dummy_trans(mp, XFS_LOG_FORCE);
	}
515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585
	mp->m_sync_seq++;
	wake_up(&mp->m_wait_single_sync_task);
}

STATIC int
xfssyncd(
	void			*arg)
{
	struct xfs_mount	*mp = arg;
	long			timeleft;
	bhv_vfs_sync_work_t	*work, *n;
	LIST_HEAD		(tmp);

	set_freezable();
	timeleft = xfs_syncd_centisecs * msecs_to_jiffies(10);
	for (;;) {
		timeleft = schedule_timeout_interruptible(timeleft);
		/* swsusp */
		try_to_freeze();
		if (kthread_should_stop() && list_empty(&mp->m_sync_list))
			break;

		spin_lock(&mp->m_sync_lock);
		/*
		 * We can get woken by laptop mode, to do a sync -
		 * that's the (only!) case where the list would be
		 * empty with time remaining.
		 */
		if (!timeleft || list_empty(&mp->m_sync_list)) {
			if (!timeleft)
				timeleft = xfs_syncd_centisecs *
							msecs_to_jiffies(10);
			INIT_LIST_HEAD(&mp->m_sync_work.w_list);
			list_add_tail(&mp->m_sync_work.w_list,
					&mp->m_sync_list);
		}
		list_for_each_entry_safe(work, n, &mp->m_sync_list, w_list)
			list_move(&work->w_list, &tmp);
		spin_unlock(&mp->m_sync_lock);

		list_for_each_entry_safe(work, n, &tmp, w_list) {
			(*work->w_syncer)(mp, work->w_data);
			list_del(&work->w_list);
			if (work == &mp->m_sync_work)
				continue;
			kmem_free(work);
		}
	}

	return 0;
}

int
xfs_syncd_init(
	struct xfs_mount	*mp)
{
	mp->m_sync_work.w_syncer = xfs_sync_worker;
	mp->m_sync_work.w_mount = mp;
	mp->m_sync_task = kthread_run(xfssyncd, mp, "xfssyncd");
	if (IS_ERR(mp->m_sync_task))
		return -PTR_ERR(mp->m_sync_task);
	return 0;
}

void
xfs_syncd_stop(
	struct xfs_mount	*mp)
{
	kthread_stop(mp->m_sync_task);
}