xfs_iomap.c 33.2 KB
Newer Older
Linus Torvalds's avatar
Linus Torvalds committed
1
/*
2
 * Copyright (c) 2000-2006 Silicon Graphics, Inc.
3
 * Copyright (c) 2016 Christoph Hellwig.
4
 * All Rights Reserved.
Linus Torvalds's avatar
Linus Torvalds committed
5
 *
6 7
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License as
Linus Torvalds's avatar
Linus Torvalds committed
8 9
 * published by the Free Software Foundation.
 *
10 11 12 13
 * 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.
Linus Torvalds's avatar
Linus Torvalds committed
14
 *
15 16 17
 * 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
Linus Torvalds's avatar
Linus Torvalds committed
18
 */
19
#include <linux/iomap.h>
Linus Torvalds's avatar
Linus Torvalds committed
20 21
#include "xfs.h"
#include "xfs_fs.h"
22
#include "xfs_shared.h"
23 24 25
#include "xfs_format.h"
#include "xfs_log_format.h"
#include "xfs_trans_resv.h"
Linus Torvalds's avatar
Linus Torvalds committed
26
#include "xfs_mount.h"
27
#include "xfs_defer.h"
Linus Torvalds's avatar
Linus Torvalds committed
28
#include "xfs_inode.h"
29
#include "xfs_btree.h"
30
#include "xfs_bmap_btree.h"
Linus Torvalds's avatar
Linus Torvalds committed
31
#include "xfs_bmap.h"
32
#include "xfs_bmap_util.h"
Linus Torvalds's avatar
Linus Torvalds committed
33
#include "xfs_error.h"
34
#include "xfs_trans.h"
Linus Torvalds's avatar
Linus Torvalds committed
35
#include "xfs_trans_space.h"
36
#include "xfs_inode_item.h"
Linus Torvalds's avatar
Linus Torvalds committed
37
#include "xfs_iomap.h"
38
#include "xfs_trace.h"
39
#include "xfs_icache.h"
40
#include "xfs_quota.h"
41 42
#include "xfs_dquot_item.h"
#include "xfs_dquot.h"
43
#include "xfs_reflink.h"
Linus Torvalds's avatar
Linus Torvalds committed
44 45 46 47 48


#define XFS_WRITEIO_ALIGN(mp,off)	(((off) >> mp->m_writeio_log) \
						<< mp->m_writeio_log)

49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72
void
xfs_bmbt_to_iomap(
	struct xfs_inode	*ip,
	struct iomap		*iomap,
	struct xfs_bmbt_irec	*imap)
{
	struct xfs_mount	*mp = ip->i_mount;

	if (imap->br_startblock == HOLESTARTBLOCK) {
		iomap->blkno = IOMAP_NULL_BLOCK;
		iomap->type = IOMAP_HOLE;
	} else if (imap->br_startblock == DELAYSTARTBLOCK) {
		iomap->blkno = IOMAP_NULL_BLOCK;
		iomap->type = IOMAP_DELALLOC;
	} else {
		iomap->blkno = xfs_fsb_to_db(ip, imap->br_startblock);
		if (imap->br_state == XFS_EXT_UNWRITTEN)
			iomap->type = IOMAP_UNWRITTEN;
		else
			iomap->type = IOMAP_MAPPED;
	}
	iomap->offset = XFS_FSB_TO_B(mp, imap->br_startoff);
	iomap->length = XFS_FSB_TO_B(mp, imap->br_blockcount);
	iomap->bdev = xfs_find_bdev_for_inode(VFS_I(ip));
73
	iomap->dax_dev = xfs_find_daxdev_for_inode(VFS_I(ip));
74 75
}

76
xfs_extlen_t
77 78 79
xfs_eof_alignment(
	struct xfs_inode	*ip,
	xfs_extlen_t		extsize)
80
{
81 82
	struct xfs_mount	*mp = ip->i_mount;
	xfs_extlen_t		align = 0;
83

84 85 86 87 88 89 90 91 92 93 94 95 96 97
	if (!XFS_IS_REALTIME_INODE(ip)) {
		/*
		 * Round up the allocation request to a stripe unit
		 * (m_dalign) boundary if the file size is >= stripe unit
		 * size, and we are allocating past the allocation eof.
		 *
		 * If mounted with the "-o swalloc" option the alignment is
		 * increased from the strip unit size to the stripe width.
		 */
		if (mp->m_swidth && (mp->m_flags & XFS_MOUNT_SWALLOC))
			align = mp->m_swidth;
		else if (mp->m_dalign)
			align = mp->m_dalign;

98 99
		if (align && XFS_ISIZE(ip) < XFS_FSB_TO_B(mp, align))
			align = 0;
100
	}
101 102 103 104 105 106

	/*
	 * Always round up the allocation request to an extent boundary
	 * (when file on a real-time subvolume or has di_extsize hint).
	 */
	if (extsize) {
107 108
		if (align)
			align = roundup_64(align, extsize);
109 110 111 112
		else
			align = extsize;
	}

113 114 115 116 117 118 119 120 121 122 123
	return align;
}

STATIC int
xfs_iomap_eof_align_last_fsb(
	struct xfs_inode	*ip,
	xfs_extlen_t		extsize,
	xfs_fileoff_t		*last_fsb)
{
	xfs_extlen_t		align = xfs_eof_alignment(ip, extsize);

124 125
	if (align) {
		xfs_fileoff_t	new_last_fsb = roundup_64(*last_fsb, align);
126 127
		int		eof, error;

128
		error = xfs_bmap_eof(ip, new_last_fsb, XFS_DATA_FORK, &eof);
129 130 131 132 133 134 135 136
		if (error)
			return error;
		if (eof)
			*last_fsb = new_last_fsb;
	}
	return 0;
}

137
STATIC int
138
xfs_alert_fsblock_zero(
139 140 141
	xfs_inode_t	*ip,
	xfs_bmbt_irec_t	*imap)
{
142
	xfs_alert_tag(ip->i_mount, XFS_PTAG_FSBLOCK_ZERO,
143 144
			"Access to block zero in inode %llu "
			"start_block: %llx start_off: %llx "
145
			"blkcnt: %llx extent-state: %x",
146 147 148 149 150
		(unsigned long long)ip->i_ino,
		(unsigned long long)imap->br_startblock,
		(unsigned long long)imap->br_startoff,
		(unsigned long long)imap->br_blockcount,
		imap->br_state);
151
	return -EFSCORRUPTED;
152 153
}

Christoph Hellwig's avatar
Christoph Hellwig committed
154
int
Linus Torvalds's avatar
Linus Torvalds committed
155 156
xfs_iomap_write_direct(
	xfs_inode_t	*ip,
157
	xfs_off_t	offset,
Linus Torvalds's avatar
Linus Torvalds committed
158
	size_t		count,
159
	xfs_bmbt_irec_t *imap,
160
	int		nmaps)
Linus Torvalds's avatar
Linus Torvalds committed
161 162 163 164
{
	xfs_mount_t	*mp = ip->i_mount;
	xfs_fileoff_t	offset_fsb;
	xfs_fileoff_t	last_fsb;
165
	xfs_filblks_t	count_fsb, resaligned;
Linus Torvalds's avatar
Linus Torvalds committed
166
	xfs_fsblock_t	firstfsb;
167
	xfs_extlen_t	extsz;
168
	int		nimaps;
169
	int		quota_flag;
Linus Torvalds's avatar
Linus Torvalds committed
170 171
	int		rt;
	xfs_trans_t	*tp;
172
	struct xfs_defer_ops dfops;
173 174
	uint		qblocks, resblks, resrtextents;
	int		error;
175
	int		lockmode;
176
	int		bmapi_flags = XFS_BMAPI_PREALLOC;
177
	uint		tflags = 0;
Linus Torvalds's avatar
Linus Torvalds committed
178

179
	rt = XFS_IS_REALTIME_INODE(ip);
180
	extsz = xfs_get_extsz_hint(ip);
181 182 183
	lockmode = XFS_ILOCK_SHARED;	/* locked by caller */

	ASSERT(xfs_isilocked(ip, lockmode));
Linus Torvalds's avatar
Linus Torvalds committed
184

185 186
	offset_fsb = XFS_B_TO_FSBT(mp, offset);
	last_fsb = XFS_B_TO_FSB(mp, ((xfs_ufsize_t)(offset + count)));
187
	if ((offset + count) > XFS_ISIZE(ip)) {
188 189 190 191 192 193 194 195 196
		/*
		 * Assert that the in-core extent list is present since this can
		 * call xfs_iread_extents() and we only have the ilock shared.
		 * This should be safe because the lock was held around a bmapi
		 * call in the caller and we only need it to access the in-core
		 * list.
		 */
		ASSERT(XFS_IFORK_PTR(ip, XFS_DATA_FORK)->if_flags &
								XFS_IFEXTENTS);
197
		error = xfs_iomap_eof_align_last_fsb(ip, extsz, &last_fsb);
198
		if (error)
199
			goto out_unlock;
Linus Torvalds's avatar
Linus Torvalds committed
200
	} else {
201
		if (nmaps && (imap->br_startblock == HOLESTARTBLOCK))
202
			last_fsb = MIN(last_fsb, (xfs_fileoff_t)
203 204
					imap->br_blockcount +
					imap->br_startoff);
Linus Torvalds's avatar
Linus Torvalds committed
205
	}
206 207
	count_fsb = last_fsb - offset_fsb;
	ASSERT(count_fsb > 0);
208
	resaligned = xfs_aligned_fsb_count(offset_fsb, count_fsb, extsz);
209 210 211 212

	if (unlikely(rt)) {
		resrtextents = qblocks = resaligned;
		resrtextents /= mp->m_sb.sb_rextsize;
213 214 215 216
		resblks = XFS_DIOSTRAT_SPACE_RES(mp, 0);
		quota_flag = XFS_QMOPT_RES_RTBLKS;
	} else {
		resrtextents = 0;
217
		resblks = qblocks = XFS_DIOSTRAT_SPACE_RES(mp, resaligned);
218 219
		quota_flag = XFS_QMOPT_RES_REGBLKS;
	}
Linus Torvalds's avatar
Linus Torvalds committed
220

221 222 223 224 225 226 227 228 229
	/*
	 * Drop the shared lock acquired by the caller, attach the dquot if
	 * necessary and move on to transaction setup.
	 */
	xfs_iunlock(ip, lockmode);
	error = xfs_qm_dqattach(ip, 0);
	if (error)
		return error;

230 231 232 233 234 235
	/*
	 * For DAX, we do not allocate unwritten extents, but instead we zero
	 * the block before we commit the transaction.  Ideally we'd like to do
	 * this outside the transaction context, but if we commit and then crash
	 * we may not have zeroed the blocks and this will be exposed on
	 * recovery of the allocation. Hence we must zero before commit.
236
	 *
237 238 239
	 * Further, if we are mapping unwritten extents here, we need to zero
	 * and convert them to written so that we don't need an unwritten extent
	 * callback for DAX. This also means that we need to be able to dip into
240 241
	 * the reserve block pool for bmbt block allocation if there is no space
	 * left but we need to do unwritten extent conversion.
242 243 244
	 */
	if (IS_DAX(VFS_I(ip))) {
		bmapi_flags = XFS_BMAPI_CONVERT | XFS_BMAPI_ZERO;
245
		if (imap->br_state == XFS_EXT_UNWRITTEN) {
246
			tflags |= XFS_TRANS_RESERVE;
247 248
			resblks = XFS_DIOSTRAT_SPACE_RES(mp, 0) << 1;
		}
249
	}
250 251 252
	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_write, resblks, resrtextents,
			tflags, &tp);
	if (error)
Eric Sandeen's avatar
Eric Sandeen committed
253
		return error;
254

255 256
	lockmode = XFS_ILOCK_EXCL;
	xfs_ilock(ip, lockmode);
Linus Torvalds's avatar
Linus Torvalds committed
257

Christoph Hellwig's avatar
Christoph Hellwig committed
258
	error = xfs_trans_reserve_quota_nblks(tp, ip, qblocks, 0, quota_flag);
259
	if (error)
260
		goto out_trans_cancel;
Linus Torvalds's avatar
Linus Torvalds committed
261

262
	xfs_trans_ijoin(tp, ip, 0);
Linus Torvalds's avatar
Linus Torvalds committed
263 264

	/*
265 266
	 * From this point onwards we overwrite the imap pointer that the
	 * caller gave to us.
Linus Torvalds's avatar
Linus Torvalds committed
267
	 */
268
	xfs_defer_init(&dfops, &firstfsb);
269
	nimaps = 1;
270
	error = xfs_bmapi_write(tp, ip, offset_fsb, count_fsb,
271
				bmapi_flags, &firstfsb, resblks, imap,
272
				&nimaps, &dfops);
273
	if (error)
274
		goto out_bmap_cancel;
Linus Torvalds's avatar
Linus Torvalds committed
275 276

	/*
277
	 * Complete the transaction
Linus Torvalds's avatar
Linus Torvalds committed
278
	 */
279
	error = xfs_defer_finish(&tp, &dfops);
280
	if (error)
281
		goto out_bmap_cancel;
282

283
	error = xfs_trans_commit(tp);
284
	if (error)
285
		goto out_unlock;
Linus Torvalds's avatar
Linus Torvalds committed
286

287 288 289
	/*
	 * Copy any maps to caller's array and return any error.
	 */
Linus Torvalds's avatar
Linus Torvalds committed
290
	if (nimaps == 0) {
291
		error = -ENOSPC;
292
		goto out_unlock;
293 294
	}

295
	if (!(imap->br_startblock || XFS_IS_REALTIME_INODE(ip)))
296
		error = xfs_alert_fsblock_zero(ip, imap);
Linus Torvalds's avatar
Linus Torvalds committed
297

298
out_unlock:
299
	xfs_iunlock(ip, lockmode);
300
	return error;
Linus Torvalds's avatar
Linus Torvalds committed
301

302
out_bmap_cancel:
303
	xfs_defer_cancel(&dfops);
304
	xfs_trans_unreserve_quota_nblks(tp, ip, (long)qblocks, 0, quota_flag);
305
out_trans_cancel:
306
	xfs_trans_cancel(tp);
307
	goto out_unlock;
Linus Torvalds's avatar
Linus Torvalds committed
308 309
}

310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336
STATIC bool
xfs_quota_need_throttle(
	struct xfs_inode *ip,
	int type,
	xfs_fsblock_t alloc_blocks)
{
	struct xfs_dquot *dq = xfs_inode_dquot(ip, type);

	if (!dq || !xfs_this_quota_on(ip->i_mount, type))
		return false;

	/* no hi watermark, no throttle */
	if (!dq->q_prealloc_hi_wmark)
		return false;

	/* under the lo watermark, no throttle */
	if (dq->q_res_bcount + alloc_blocks < dq->q_prealloc_lo_wmark)
		return false;

	return true;
}

STATIC void
xfs_quota_calc_throttle(
	struct xfs_inode *ip,
	int type,
	xfs_fsblock_t *qblocks,
337 338
	int *qshift,
	int64_t	*qfreesp)
339 340 341 342 343
{
	int64_t freesp;
	int shift = 0;
	struct xfs_dquot *dq = xfs_inode_dquot(ip, type);

344 345
	/* no dq, or over hi wmark, squash the prealloc completely */
	if (!dq || dq->q_res_bcount >= dq->q_prealloc_hi_wmark) {
346
		*qblocks = 0;
347
		*qfreesp = 0;
348 349 350 351 352 353 354 355 356 357 358 359
		return;
	}

	freesp = dq->q_prealloc_hi_wmark - dq->q_res_bcount;
	if (freesp < dq->q_low_space[XFS_QLOWSP_5_PCNT]) {
		shift = 2;
		if (freesp < dq->q_low_space[XFS_QLOWSP_3_PCNT])
			shift += 2;
		if (freesp < dq->q_low_space[XFS_QLOWSP_1_PCNT])
			shift += 2;
	}

360 361 362
	if (freesp < *qfreesp)
		*qfreesp = freesp;

363 364 365 366 367 368 369
	/* only overwrite the throttle values if we are more aggressive */
	if ((freesp >> shift) < (*qblocks >> *qshift)) {
		*qblocks = freesp;
		*qshift = shift;
	}
}

370
/*
371 372 373 374
 * If we are doing a write at the end of the file and there are no allocations
 * past this one, then extend the allocation out to the file system's write
 * iosize.
 *
375
 * If we don't have a user specified preallocation size, dynamically increase
376
 * the preallocation size as the size of the file grows.  Cap the maximum size
377 378
 * at a single extent or less if the filesystem is near full. The closer the
 * filesystem is to full, the smaller the maximum prealocation.
379 380 381 382 383 384 385 386
 *
 * As an exception we don't do any preallocation at all if the file is smaller
 * than the minimum preallocation and we are using the default dynamic
 * preallocation scheme, as it is likely this is the only write to the file that
 * is going to be done.
 *
 * We clean up any extra space left over when the file is closed in
 * xfs_inactive().
387 388 389
 */
STATIC xfs_fsblock_t
xfs_iomap_prealloc_size(
390
	struct xfs_inode	*ip,
391 392
	loff_t			offset,
	loff_t			count,
393
	xfs_extnum_t		idx)
394
{
395
	struct xfs_mount	*mp = ip->i_mount;
396
	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
397
	xfs_fileoff_t		offset_fsb = XFS_B_TO_FSBT(mp, offset);
398
	struct xfs_bmbt_irec	prev;
399 400
	int			shift = 0;
	int64_t			freesp;
401 402
	xfs_fsblock_t		qblocks;
	int			qshift = 0;
403 404 405 406 407 408 409 410 411 412 413 414 415 416 417
	xfs_fsblock_t		alloc_blocks = 0;

	if (offset + count <= XFS_ISIZE(ip))
		return 0;

	if (!(mp->m_flags & XFS_MOUNT_DFLT_IOSIZE) &&
	    (XFS_ISIZE(ip) < XFS_FSB_TO_B(mp, mp->m_writeio_blocks)))
		return 0;

	/*
	 * If an explicit allocsize is set, the file is small, or we
	 * are writing behind a hole, then use the minimum prealloc:
	 */
	if ((mp->m_flags & XFS_MOUNT_DFLT_IOSIZE) ||
	    XFS_ISIZE(ip) < XFS_FSB_TO_B(mp, mp->m_dalign) ||
418 419
	    !xfs_iext_get_extent(ifp, idx - 1, &prev) ||
	    prev.br_startoff + prev.br_blockcount < offset_fsb)
420
		return mp->m_writeio_blocks;
421

422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437
	/*
	 * Determine the initial size of the preallocation. We are beyond the
	 * current EOF here, but we need to take into account whether this is
	 * a sparse write or an extending write when determining the
	 * preallocation size.  Hence we need to look up the extent that ends
	 * at the current write offset and use the result to determine the
	 * preallocation size.
	 *
	 * If the extent is a hole, then preallocation is essentially disabled.
	 * Otherwise we take the size of the preceding data extent as the basis
	 * for the preallocation size. If the size of the extent is greater than
	 * half the maximum extent length, then use the current offset as the
	 * basis. This ensures that for large files the preallocation size
	 * always extends to MAXEXTLEN rather than falling short due to things
	 * like stripe unit/width alignment of real extents.
	 */
438 439
	if (prev.br_blockcount <= (MAXEXTLEN >> 1))
		alloc_blocks = prev.br_blockcount << 1;
440 441
	else
		alloc_blocks = XFS_B_TO_FSB(mp, offset);
442 443
	if (!alloc_blocks)
		goto check_writeio;
444
	qblocks = alloc_blocks;
445

446 447 448 449 450 451 452 453 454
	/*
	 * MAXEXTLEN is not a power of two value but we round the prealloc down
	 * to the nearest power of two value after throttling. To prevent the
	 * round down from unconditionally reducing the maximum supported prealloc
	 * size, we round up first, apply appropriate throttling, round down and
	 * cap the value to MAXEXTLEN.
	 */
	alloc_blocks = XFS_FILEOFF_MIN(roundup_pow_of_two(MAXEXTLEN),
				       alloc_blocks);
455

456
	freesp = percpu_counter_read_positive(&mp->m_fdblocks);
457 458 459 460 461 462 463 464 465 466
	if (freesp < mp->m_low_space[XFS_LOWSP_5_PCNT]) {
		shift = 2;
		if (freesp < mp->m_low_space[XFS_LOWSP_4_PCNT])
			shift++;
		if (freesp < mp->m_low_space[XFS_LOWSP_3_PCNT])
			shift++;
		if (freesp < mp->m_low_space[XFS_LOWSP_2_PCNT])
			shift++;
		if (freesp < mp->m_low_space[XFS_LOWSP_1_PCNT])
			shift++;
467
	}
468 469

	/*
470 471
	 * Check each quota to cap the prealloc size, provide a shift value to
	 * throttle with and adjust amount of available space.
472 473
	 */
	if (xfs_quota_need_throttle(ip, XFS_DQ_USER, alloc_blocks))
474 475
		xfs_quota_calc_throttle(ip, XFS_DQ_USER, &qblocks, &qshift,
					&freesp);
476
	if (xfs_quota_need_throttle(ip, XFS_DQ_GROUP, alloc_blocks))
477 478
		xfs_quota_calc_throttle(ip, XFS_DQ_GROUP, &qblocks, &qshift,
					&freesp);
479
	if (xfs_quota_need_throttle(ip, XFS_DQ_PROJ, alloc_blocks))
480 481
		xfs_quota_calc_throttle(ip, XFS_DQ_PROJ, &qblocks, &qshift,
					&freesp);
482 483 484 485 486 487 488 489 490 491 492

	/*
	 * The final prealloc size is set to the minimum of free space available
	 * in each of the quotas and the overall filesystem.
	 *
	 * The shift throttle value is set to the maximum value as determined by
	 * the global low free space values and per-quota low free space values.
	 */
	alloc_blocks = MIN(alloc_blocks, qblocks);
	shift = MAX(shift, qshift);

493 494
	if (shift)
		alloc_blocks >>= shift;
495 496 497 498 499 500 501 502
	/*
	 * rounddown_pow_of_two() returns an undefined result if we pass in
	 * alloc_blocks = 0.
	 */
	if (alloc_blocks)
		alloc_blocks = rounddown_pow_of_two(alloc_blocks);
	if (alloc_blocks > MAXEXTLEN)
		alloc_blocks = MAXEXTLEN;
503 504 505 506 507 508 509 510 511 512

	/*
	 * If we are still trying to allocate more space than is
	 * available, squash the prealloc hard. This can happen if we
	 * have a large file on a small filesystem and the above
	 * lowspace thresholds are smaller than MAXEXTLEN.
	 */
	while (alloc_blocks && alloc_blocks >= freesp)
		alloc_blocks >>= 4;
check_writeio:
513 514
	if (alloc_blocks < mp->m_writeio_blocks)
		alloc_blocks = mp->m_writeio_blocks;
515 516
	trace_xfs_iomap_prealloc_size(ip, alloc_blocks, shift,
				      mp->m_writeio_blocks);
517 518 519
	return alloc_blocks;
}

520 521 522 523 524 525
static int
xfs_file_iomap_begin_delay(
	struct inode		*inode,
	loff_t			offset,
	loff_t			count,
	struct iomap		*iomap)
Linus Torvalds's avatar
Linus Torvalds committed
526
{
527 528 529 530 531 532
	struct xfs_inode	*ip = XFS_I(inode);
	struct xfs_mount	*mp = ip->i_mount;
	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
	xfs_fileoff_t		offset_fsb = XFS_B_TO_FSBT(mp, offset);
	xfs_fileoff_t		maxbytes_fsb =
		XFS_B_TO_FSB(mp, mp->m_super->s_maxbytes);
533
	xfs_fileoff_t		end_fsb;
534 535 536
	int			error = 0, eof = 0;
	struct xfs_bmbt_irec	got;
	xfs_extnum_t		idx;
537
	xfs_fsblock_t		prealloc_blocks = 0;
538 539 540

	ASSERT(!XFS_IS_REALTIME_INODE(ip));
	ASSERT(!xfs_get_extsz_hint(ip));
541

542
	xfs_ilock(ip, XFS_ILOCK_EXCL);
Linus Torvalds's avatar
Linus Torvalds committed
543

544 545 546
	if (unlikely(XFS_TEST_ERROR(
	    (XFS_IFORK_FORMAT(ip, XFS_DATA_FORK) != XFS_DINODE_FMT_EXTENTS &&
	     XFS_IFORK_FORMAT(ip, XFS_DATA_FORK) != XFS_DINODE_FMT_BTREE),
547
	     mp, XFS_ERRTAG_BMAPIFORMAT))) {
548 549 550 551
		XFS_ERROR_REPORT(__func__, XFS_ERRLEVEL_LOW, mp);
		error = -EFSCORRUPTED;
		goto out_unlock;
	}
552

553
	XFS_STATS_INC(mp, xs_blk_mapw);
554

555 556 557 558
	if (!(ifp->if_flags & XFS_IFEXTENTS)) {
		error = xfs_iread_extents(NULL, ip, XFS_DATA_FORK);
		if (error)
			goto out_unlock;
Linus Torvalds's avatar
Linus Torvalds committed
559 560
	}

561
	eof = !xfs_iext_lookup_extent(ip, ifp, offset_fsb, &idx, &got);
562
	if (!eof && got.br_startoff <= offset_fsb) {
563 564 565 566 567 568 569 570 571 572 573
		if (xfs_is_reflink_inode(ip)) {
			bool		shared;

			end_fsb = min(XFS_B_TO_FSB(mp, offset + count),
					maxbytes_fsb);
			xfs_trim_extent(&got, offset_fsb, end_fsb - offset_fsb);
			error = xfs_reflink_reserve_cow(ip, &got, &shared);
			if (error)
				goto out_unlock;
		}

574 575
		trace_xfs_iomap_found(ip, offset, count, 0, &got);
		goto done;
Linus Torvalds's avatar
Linus Torvalds committed
576
	}
577

578 579 580 581
	error = xfs_qm_dqattach_locked(ip, 0);
	if (error)
		goto out_unlock;

582
	/*
583 584 585 586 587 588 589
	 * We cap the maximum length we map here to MAX_WRITEBACK_PAGES pages
	 * to keep the chunks of work done where somewhat symmetric with the
	 * work writeback does. This is a completely arbitrary number pulled
	 * out of thin air as a best guess for initial testing.
	 *
	 * Note that the values needs to be less than 32-bits wide until
	 * the lower level functions are updated.
590
	 */
591
	count = min_t(loff_t, count, 1024 * PAGE_SIZE);
592
	end_fsb = min(XFS_B_TO_FSB(mp, offset + count), maxbytes_fsb);
593 594

	if (eof) {
595
		prealloc_blocks = xfs_iomap_prealloc_size(ip, offset, count, idx);
596 597 598
		if (prealloc_blocks) {
			xfs_extlen_t	align;
			xfs_off_t	end_offset;
599
			xfs_fileoff_t	p_end_fsb;
600

601
			end_offset = XFS_WRITEIO_ALIGN(mp, offset + count - 1);
602 603
			p_end_fsb = XFS_B_TO_FSBT(mp, end_offset) +
					prealloc_blocks;
604 605 606

			align = xfs_eof_alignment(ip, 0);
			if (align)
607
				p_end_fsb = roundup_64(p_end_fsb, align);
608

609 610 611
			p_end_fsb = min(p_end_fsb, maxbytes_fsb);
			ASSERT(p_end_fsb > offset_fsb);
			prealloc_blocks = p_end_fsb - end_fsb;
612 613 614 615
		}
	}

retry:
616
	error = xfs_bmapi_reserve_delalloc(ip, XFS_DATA_FORK, offset_fsb,
617
			end_fsb - offset_fsb, prealloc_blocks, &got, &idx, eof);
618 619
	switch (error) {
	case 0:
620
		break;
621 622
	case -ENOSPC:
	case -EDQUOT:
623
		/* retry without any preallocation */
624
		trace_xfs_delalloc_enospc(ip, offset, count);
625 626
		if (prealloc_blocks) {
			prealloc_blocks = 0;
627
			goto retry;
628
		}
629 630 631
		/*FALLTHRU*/
	default:
		goto out_unlock;
Linus Torvalds's avatar
Linus Torvalds committed
632 633
	}

634 635 636 637 638
	/*
	 * Flag newly allocated delalloc blocks with IOMAP_F_NEW so we punch
	 * them out if the write happens to fail.
	 */
	iomap->flags = IOMAP_F_NEW;
639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654
	trace_xfs_iomap_alloc(ip, offset, count, 0, &got);
done:
	if (isnullstartblock(got.br_startblock))
		got.br_startblock = DELAYSTARTBLOCK;

	if (!got.br_startblock) {
		error = xfs_alert_fsblock_zero(ip, &got);
		if (error)
			goto out_unlock;
	}

	xfs_bmbt_to_iomap(ip, iomap, &got);

out_unlock:
	xfs_iunlock(ip, XFS_ILOCK_EXCL);
	return error;
Linus Torvalds's avatar
Linus Torvalds committed
655 656 657 658 659 660 661 662
}

/*
 * Pass in a delayed allocate extent, convert it to real extents;
 * return to the caller the extent we create which maps on top of
 * the originating callers request.
 *
 * Called without a lock on the inode.
663 664 665
 *
 * We no longer bother to look at the incoming map - all we have to
 * guarantee is that whatever we allocate fills the required range.
Linus Torvalds's avatar
Linus Torvalds committed
666
 */
Christoph Hellwig's avatar
Christoph Hellwig committed
667
int
Linus Torvalds's avatar
Linus Torvalds committed
668 669
xfs_iomap_write_allocate(
	xfs_inode_t	*ip,
670
	int		whichfork,
671
	xfs_off_t	offset,
672
	xfs_bmbt_irec_t *imap)
Linus Torvalds's avatar
Linus Torvalds committed
673 674 675 676 677
{
	xfs_mount_t	*mp = ip->i_mount;
	xfs_fileoff_t	offset_fsb, last_block;
	xfs_fileoff_t	end_fsb, map_start_fsb;
	xfs_fsblock_t	first_block;
678
	struct xfs_defer_ops	dfops;
Linus Torvalds's avatar
Linus Torvalds committed
679 680
	xfs_filblks_t	count_fsb;
	xfs_trans_t	*tp;
681
	int		nimaps;
Linus Torvalds's avatar
Linus Torvalds committed
682
	int		error = 0;
683
	int		flags = XFS_BMAPI_DELALLOC;
Linus Torvalds's avatar
Linus Torvalds committed
684 685
	int		nres;

686
	if (whichfork == XFS_COW_FORK)
687
		flags |= XFS_BMAPI_COWFORK | XFS_BMAPI_PREALLOC;
688

Linus Torvalds's avatar
Linus Torvalds committed
689 690 691
	/*
	 * Make sure that the dquots are there.
	 */
Christoph Hellwig's avatar
Christoph Hellwig committed
692 693
	error = xfs_qm_dqattach(ip, 0);
	if (error)
Eric Sandeen's avatar
Eric Sandeen committed
694
		return error;
Linus Torvalds's avatar
Linus Torvalds committed
695

696
	offset_fsb = XFS_B_TO_FSBT(mp, offset);
697 698
	count_fsb = imap->br_blockcount;
	map_start_fsb = imap->br_startoff;
Linus Torvalds's avatar
Linus Torvalds committed
699

700
	XFS_STATS_ADD(mp, xs_xstrat_bytes, XFS_FSB_TO_B(mp, count_fsb));
Linus Torvalds's avatar
Linus Torvalds committed
701 702 703 704 705 706 707 708 709 710 711 712 713

	while (count_fsb != 0) {
		/*
		 * Set up a transaction with which to allocate the
		 * backing store for the file.  Do allocations in a
		 * loop until we get some space in the range we are
		 * interested in.  The other space that might be allocated
		 * is in the delayed allocation extent on which we sit
		 * but before our buffer starts.
		 */
		nimaps = 0;
		while (nimaps == 0) {
			nres = XFS_EXTENTADD_SPACE_RES(mp, XFS_DATA_FORK);
714 715 716 717 718 719 720
			/*
			 * We have already reserved space for the extent and any
			 * indirect blocks when creating the delalloc extent,
			 * there is no need to reserve space in this transaction
			 * again.
			 */
			error = xfs_trans_alloc(mp, &M_RES(mp)->tr_write, 0,
721 722
					0, XFS_TRANS_RESERVE, &tp);
			if (error)
Eric Sandeen's avatar
Eric Sandeen committed
723
				return error;
724

Linus Torvalds's avatar
Linus Torvalds committed
725
			xfs_ilock(ip, XFS_ILOCK_EXCL);
726
			xfs_trans_ijoin(tp, ip, 0);
Linus Torvalds's avatar
Linus Torvalds committed
727

728
			xfs_defer_init(&dfops, &first_block);
Linus Torvalds's avatar
Linus Torvalds committed
729 730

			/*
731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759
			 * it is possible that the extents have changed since
			 * we did the read call as we dropped the ilock for a
			 * while. We have to be careful about truncates or hole
			 * punchs here - we are not allowed to allocate
			 * non-delalloc blocks here.
			 *
			 * The only protection against truncation is the pages
			 * for the range we are being asked to convert are
			 * locked and hence a truncate will block on them
			 * first.
			 *
			 * As a result, if we go beyond the range we really
			 * need and hit an delalloc extent boundary followed by
			 * a hole while we have excess blocks in the map, we
			 * will fill the hole incorrectly and overrun the
			 * transaction reservation.
			 *
			 * Using a single map prevents this as we are forced to
			 * check each map we look for overlap with the desired
			 * range and abort as soon as we find it. Also, given
			 * that we only return a single map, having one beyond
			 * what we can return is probably a bit silly.
			 *
			 * We also need to check that we don't go beyond EOF;
			 * this is a truncate optimisation as a truncate sets
			 * the new file size before block on the pages we
			 * currently have locked under writeback. Because they
			 * are about to be tossed, we don't need to write them
			 * back....
Linus Torvalds's avatar
Linus Torvalds committed
760
			 */
761
			nimaps = 1;
762
			end_fsb = XFS_B_TO_FSB(mp, XFS_ISIZE(ip));
763
			error = xfs_bmap_last_offset(ip, &last_block,
764 765 766 767
							XFS_DATA_FORK);
			if (error)
				goto trans_cancel;

Linus Torvalds's avatar
Linus Torvalds committed
768 769 770 771
			last_block = XFS_FILEOFF_MAX(last_block, end_fsb);
			if ((map_start_fsb + count_fsb) > last_block) {
				count_fsb = last_block - map_start_fsb;
				if (count_fsb == 0) {
772
					error = -EAGAIN;
Linus Torvalds's avatar
Linus Torvalds committed
773 774 775 776
					goto trans_cancel;
				}
			}

777 778 779 780
			/*
			 * From this point onwards we overwrite the imap
			 * pointer that the caller gave to us.
			 */
781
			error = xfs_bmapi_write(tp, ip, map_start_fsb,
782
						count_fsb, flags, &first_block,
783
						nres, imap, &nimaps,
784
						&dfops);
Linus Torvalds's avatar
Linus Torvalds committed
785 786 787
			if (error)
				goto trans_cancel;

788
			error = xfs_defer_finish(&tp, &dfops);
Linus Torvalds's avatar
Linus Torvalds committed
789 790 791
			if (error)
				goto trans_cancel;

792
			error = xfs_trans_commit(tp);
Linus Torvalds's avatar
Linus Torvalds committed
793 794 795 796 797 798 799 800 801 802
			if (error)
				goto error0;

			xfs_iunlock(ip, XFS_ILOCK_EXCL);
		}

		/*
		 * See if we were able to allocate an extent that
		 * covers at least part of the callers request
		 */
803
		if (!(imap->br_startblock || XFS_IS_REALTIME_INODE(ip)))
804
			return xfs_alert_fsblock_zero(ip, imap);
805

806 807 808
		if ((offset_fsb >= imap->br_startoff) &&
		    (offset_fsb < (imap->br_startoff +
				   imap->br_blockcount))) {
809
			XFS_STATS_INC(mp, xs_xstrat_quick);
810
			return 0;
Linus Torvalds's avatar
Linus Torvalds committed
811 812
		}

813 814
		/*
		 * So far we have not mapped the requested part of the
Linus Torvalds's avatar
Linus Torvalds committed
815 816
		 * file, just surrounding data, try again.
		 */
817 818
		count_fsb -= imap->br_blockcount;
		map_start_fsb = imap->br_startoff + imap->br_blockcount;
Linus Torvalds's avatar
Linus Torvalds committed
819 820 821
	}

trans_cancel:
822
	xfs_defer_cancel(&dfops);
823
	xfs_trans_cancel(tp);
Linus Torvalds's avatar
Linus Torvalds committed
824 825
error0:
	xfs_iunlock(ip, XFS_ILOCK_EXCL);
Eric Sandeen's avatar
Eric Sandeen committed
826
	return error;
Linus Torvalds's avatar
Linus Torvalds committed
827 828 829 830 831
}

int
xfs_iomap_write_unwritten(
	xfs_inode_t	*ip,
832
	xfs_off_t	offset,
833 834
	xfs_off_t	count,
	bool		update_isize)
Linus Torvalds's avatar
Linus Torvalds committed
835 836 837 838 839
{
	xfs_mount_t	*mp = ip->i_mount;
	xfs_fileoff_t	offset_fsb;
	xfs_filblks_t	count_fsb;
	xfs_filblks_t	numblks_fsb;
840 841 842 843
	xfs_fsblock_t	firstfsb;
	int		nimaps;
	xfs_trans_t	*tp;
	xfs_bmbt_irec_t imap;
844
	struct xfs_defer_ops dfops;
845
	struct inode	*inode = VFS_I(ip);
846
	xfs_fsize_t	i_size;
847
	uint		resblks;
Linus Torvalds's avatar
Linus Torvalds committed
848 849
	int		error;

850
	trace_xfs_unwritten_convert(ip, offset, count);
Linus Torvalds's avatar
Linus Torvalds committed
851 852 853 854 855

	offset_fsb = XFS_B_TO_FSBT(mp, offset);
	count_fsb = XFS_B_TO_FSB(mp, (xfs_ufsize_t)offset + count);
	count_fsb = (xfs_filblks_t)(count_fsb - offset_fsb);

856 857 858 859 860 861 862 863 864 865
	/*
	 * Reserve enough blocks in this transaction for two complete extent
	 * btree splits.  We may be converting the middle part of an unwritten
	 * extent and in this case we will insert two new extents in the btree
	 * each of which could cause a full split.
	 *
	 * This reservation amount will be used in the first call to
	 * xfs_bmbt_split() to select an AG with enough space to satisfy the
	 * rest of the operation.
	 */
866
	resblks = XFS_DIOSTRAT_SPACE_RES(mp, 0) << 1;
Linus Torvalds's avatar
Linus Torvalds committed
867

868
	do {
Linus Torvalds's avatar
Linus Torvalds committed
869
		/*
870
		 * Set up a transaction to convert the range of extents
Linus Torvalds's avatar
Linus Torvalds committed
871 872
		 * from unwritten to real. Do allocations in a loop until
		 * we have covered the range passed in.
873
		 *
874 875 876
		 * Note that we can't risk to recursing back into the filesystem
		 * here as we might be asked to write out the same inode that we
		 * complete here and might deadlock on the iolock.
Linus Torvalds's avatar
Linus Torvalds committed
877
		 */
878 879 880
		error = xfs_trans_alloc(mp, &M_RES(mp)->tr_write, resblks, 0,
				XFS_TRANS_RESERVE | XFS_TRANS_NOFS, &tp);
		if (error)
Eric Sandeen's avatar
Eric Sandeen committed
881
			return error;
Linus Torvalds's avatar
Linus Torvalds committed
882 883

		xfs_ilock(ip, XFS_ILOCK_EXCL);
884
		xfs_trans_ijoin(tp, ip, 0);
Linus Torvalds's avatar
Linus Torvalds committed
885 886 887 888

		/*
		 * Modify the unwritten extent state of the buffer.
		 */
889
		xfs_defer_init(&dfops, &firstfsb);
Linus Torvalds's avatar
Linus Torvalds committed
890
		nimaps = 1;
891
		error = xfs_bmapi_write(tp, ip, offset_fsb, count_fsb,
892
					XFS_BMAPI_CONVERT, &firstfsb, resblks,
893
					&imap, &nimaps, &dfops);
Linus Torvalds's avatar
Linus Torvalds committed
894 895 896
		if (error)
			goto error_on_bmapi_transaction;

897 898 899 900 901 902 903 904
		/*
		 * Log the updated inode size as we go.  We have to be careful
		 * to only log it up to the actual write offset if it is
		 * halfway into a block.
		 */
		i_size = XFS_FSB_TO_B(mp, offset_fsb + count_fsb);
		if (i_size > offset + count)
			i_size = offset + count;
905 906
		if (update_isize && i_size > i_size_read(inode))
			i_size_write(inode, i_size);
907 908 909 910 911 912
		i_size = xfs_new_eof(ip, i_size);
		if (i_size) {
			ip->i_d.di_size = i_size;
			xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
		}

913
		error = xfs_defer_finish(&tp, &dfops);
Linus Torvalds's avatar
Linus Torvalds committed
914 915 916
		if (error)
			goto error_on_bmapi_transaction;

917
		error = xfs_trans_commit(tp);
Linus Torvalds's avatar
Linus Torvalds committed
918 919
		xfs_iunlock(ip, XFS_ILOCK_EXCL);
		if (error)
Eric Sandeen's avatar
Eric Sandeen committed
920
			return error;
921

922
		if (!(imap.br_startblock || XFS_IS_REALTIME_INODE(ip)))
923
			return xfs_alert_fsblock_zero(ip, &imap);
Linus Torvalds's avatar
Linus Torvalds committed
924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939

		if ((numblks_fsb = imap.br_blockcount) == 0) {
			/*
			 * The numblks_fsb value should always get
			 * smaller, otherwise the loop is stuck.
			 */
			ASSERT(imap.br_blockcount);
			break;
		}
		offset_fsb += numblks_fsb;
		count_fsb -= numblks_fsb;
	} while (count_fsb > 0);

	return 0;

error_on_bmapi_transaction:
940
	xfs_defer_cancel(&dfops);
941
	xfs_trans_cancel(tp);
Linus Torvalds's avatar
Linus Torvalds committed
942
	xfs_iunlock(ip, XFS_ILOCK_EXCL);
Eric Sandeen's avatar
Eric Sandeen committed
943
	return error;
Linus Torvalds's avatar
Linus Torvalds committed
944
}
945

946 947
static inline bool imap_needs_alloc(struct inode *inode,
		struct xfs_bmbt_irec *imap, int nimaps)
948 949 950
{
	return !nimaps ||
		imap->br_startblock == HOLESTARTBLOCK ||
951
		imap->br_startblock == DELAYSTARTBLOCK ||
952
		(IS_DAX(inode) && imap->br_state == XFS_EXT_UNWRITTEN);
953 954
}

Christoph Hellwig's avatar
Christoph Hellwig committed
955 956 957 958 959 960 961 962 963 964 965 966 967
static inline bool need_excl_ilock(struct xfs_inode *ip, unsigned flags)
{
	/*
	 * COW writes will allocate delalloc space, so we need to make sure
	 * to take the lock exclusively here.
	 */
	if (xfs_is_reflink_inode(ip) && (flags & (IOMAP_WRITE | IOMAP_ZERO)))
		return true;
	if ((flags & IOMAP_DIRECT) && (flags & IOMAP_WRITE))
		return true;
	return false;
}

968 969 970 971 972 973 974 975 976 977 978 979 980
static int
xfs_file_iomap_begin(
	struct inode		*inode,
	loff_t			offset,
	loff_t			length,
	unsigned		flags,
	struct iomap		*iomap)
{
	struct xfs_inode	*ip = XFS_I(inode);
	struct xfs_mount	*mp = ip->i_mount;
	struct xfs_bmbt_irec	imap;
	xfs_fileoff_t		offset_fsb, end_fsb;
	int			nimaps = 1, error = 0;
981
	bool			shared = false, trimmed = false;
982
	unsigned		lockmode;
983 984 985 986

	if (XFS_FORCED_SHUTDOWN(mp))
		return -EIO;

Christoph Hellwig's avatar
Christoph Hellwig committed
987 988
	if (((flags & (IOMAP_WRITE | IOMAP_DIRECT)) == IOMAP_WRITE) &&
			!IS_DAX(inode) && !xfs_get_extsz_hint(ip)) {
989
		/* Reserve delalloc blocks for regular writeback. */
990
		return xfs_file_iomap_begin_delay(inode, offset, length, iomap);
991 992
	}

Christoph Hellwig's avatar
Christoph Hellwig committed
993
	if (need_excl_ilock(ip, flags)) {
994 995 996 997 998
		lockmode = XFS_ILOCK_EXCL;
		xfs_ilock(ip, XFS_ILOCK_EXCL);
	} else {
		lockmode = xfs_ilock_data_map_shared(ip);
	}
999

1000 1001 1002 1003 1004
	if ((flags & IOMAP_NOWAIT) && !(ip->i_df.if_flags & XFS_IFEXTENTS)) {
		error = -EAGAIN;
		goto out_unlock;
	}

1005 1006 1007 1008 1009 1010 1011
	ASSERT(offset <= mp->m_super->s_maxbytes);
	if ((xfs_fsize_t)offset + length > mp->m_super->s_maxbytes)
		length = mp->m_super->s_maxbytes - offset;
	offset_fsb = XFS_B_TO_FSBT(mp, offset);
	end_fsb = XFS_B_TO_FSB(mp, offset + length);

	error = xfs_bmapi_read(ip, offset_fsb, end_fsb - offset_fsb, &imap,
1012
			       &nimaps, 0);
1013 1014
	if (error)
		goto out_unlock;
1015

1016
	if (flags & IOMAP_REPORT) {
1017 1018 1019
		/* Trim the mapping to the nearest shared extent boundary. */
		error = xfs_reflink_trim_around_shared(ip, &imap, &shared,
				&trimmed);
1020 1021 1022 1023 1024
		if (error)
			goto out_unlock;
	}

	if ((flags & (IOMAP_WRITE | IOMAP_ZERO)) && xfs_is_reflink_inode(ip)) {
1025
		if (flags & IOMAP_DIRECT) {
1026 1027 1028 1029 1030 1031 1032 1033 1034
			/*
			 * A reflinked inode will result in CoW alloc.
			 * FIXME: It could still overwrite on unshared extents
			 * and not need allocation.
			 */
			if (flags & IOMAP_NOWAIT) {
				error = -EAGAIN;
				goto out_unlock;
			}
1035 1036 1037 1038 1039 1040 1041 1042 1043 1044
			/* may drop and re-acquire the ilock */
			error = xfs_reflink_allocate_cow(ip, &imap, &shared,
					&lockmode);
			if (error)
				goto out_unlock;
		} else {
			error = xfs_reflink_reserve_cow(ip, &imap, &shared);
			if (error)
				goto out_unlock;
		}
1045 1046 1047

		end_fsb = imap.br_startoff + imap.br_blockcount;
		length = XFS_FSB_TO_B(mp, end_fsb) - offset;
1048 1049
	}

1050
	if ((flags & IOMAP_WRITE) && imap_needs_alloc(inode, &imap, nimaps)) {
1051 1052 1053 1054 1055 1056 1057 1058
		/*
		 * If nowait is set bail since we are going to make
		 * allocations.
		 */
		if (flags & IOMAP_NOWAIT) {
			error = -EAGAIN;
			goto out_unlock;
		}
1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069
		/*
		 * We cap the maximum length we map here to MAX_WRITEBACK_PAGES
		 * pages to keep the chunks of work done where somewhat symmetric
		 * with the work writeback does. This is a completely arbitrary
		 * number pulled out of thin air as a best guess for initial
		 * testing.
		 *
		 * Note that the values needs to be less than 32-bits wide until
		 * the lower level functions are updated.
		 */
		length = min_t(loff_t, length, 1024 * PAGE_SIZE);
1070 1071 1072 1073
		/*
		 * xfs_iomap_write_direct() expects the shared lock. It
		 * is unlocked on return.
		 */
1074 1075
		if (lockmode == XFS_ILOCK_EXCL)
			xfs_ilock_demote(ip, lockmode);
1076 1077
		error = xfs_iomap_write_direct(ip, offset, length, &imap,
				nimaps);
1078 1079 1080
		if (error)
			return error;

1081
		iomap->flags = IOMAP_F_NEW;
1082 1083
		trace_xfs_iomap_alloc(ip, offset, length, 0, &imap);
	} else {
1084 1085
		ASSERT(nimaps);

1086
		xfs_iunlock(ip, lockmode);
1087
		trace_xfs_iomap_found(ip, offset, length, 0, &imap);
1088 1089
	}

1090 1091 1092 1093
	if ((flags & IOMAP_WRITE) && xfs_ipincount(ip) &&
	    (ip->i_itemp->ili_fsync_fields & ~XFS_ILOG_TIMESTAMP))
		iomap->flags |= IOMAP_F_DIRTY;

1094
	xfs_bmbt_to_iomap(ip, iomap, &imap);
1095

1096 1097
	if (shared)
		iomap->flags |= IOMAP_F_SHARED;
1098
	return 0;
1099 1100 1101
out_unlock:
	xfs_iunlock(ip, lockmode);
	return error;
1102 1103 1104 1105 1106 1107 1108
}

static int
xfs_file_iomap_end_delalloc(
	struct xfs_inode	*ip,
	loff_t			offset,
	loff_t			length,
1109 1110
	ssize_t			written,
	struct iomap		*iomap)
1111 1112 1113 1114 1115 1116
{
	struct xfs_mount	*mp = ip->i_mount;
	xfs_fileoff_t		start_fsb;
	xfs_fileoff_t		end_fsb;
	int			error = 0;

1117 1118 1119 1120
	/*
	 * Behave as if the write failed if drop writes is enabled. Set the NEW
	 * flag to force delalloc cleanup.
	 */
1121
	if (XFS_TEST_ERROR(false, mp, XFS_ERRTAG_DROP_WRITES)) {
1122
		iomap->flags |= IOMAP_F_NEW;
1123
		written = 0;
1124
	}
1125

1126 1127 1128 1129 1130 1131 1132 1133 1134
	/*
	 * start_fsb refers to the first unused block after a short write. If
	 * nothing was written, round offset down to point at the first block in
	 * the range.
	 */
	if (unlikely(!written))
		start_fsb = XFS_B_TO_FSBT(mp, offset);
	else
		start_fsb = XFS_B_TO_FSB(mp, offset + written);
1135 1136 1137
	end_fsb = XFS_B_TO_FSB(mp, offset + length);

	/*
1138 1139
	 * Trim delalloc blocks if they were allocated by this write and we
	 * didn't manage to write the whole range.
1140 1141 1142 1143 1144
	 *
	 * We don't need to care about racing delalloc as we hold i_mutex
	 * across the reserve/allocate/unreserve calls. If there are delalloc
	 * blocks in the range, they are ours.
	 */
1145
	if ((iomap->flags & IOMAP_F_NEW) && start_fsb < end_fsb) {
1146 1147 1148
		truncate_pagecache_range(VFS_I(ip), XFS_FSB_TO_B(mp, start_fsb),
					 XFS_FSB_TO_B(mp, end_fsb) - 1);

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
		xfs_ilock(ip, XFS_ILOCK_EXCL);
		error = xfs_bmap_punch_delalloc_range(ip, start_fsb,
					       end_fsb - start_fsb);
		xfs_iunlock(ip, XFS_ILOCK_EXCL);

		if (error && !XFS_FORCED_SHUTDOWN(mp)) {
			xfs_alert(mp, "%s: unable to clean up ino %lld",
				__func__, ip->i_ino);
			return error;
		}
	}

	return 0;
}

static int
xfs_file_iomap_end(
	struct inode		*inode,
	loff_t			offset,
	loff_t			length,
	ssize_t			written,
	unsigned		flags,
	struct iomap		*iomap)
{
	if ((flags & IOMAP_WRITE) && iomap->type == IOMAP_DELALLOC)
		return xfs_file_iomap_end_delalloc(XFS_I(inode), offset,
1175
				length, written, iomap);
1176 1177 1178
	return 0;
}

1179
const struct iomap_ops xfs_iomap_ops = {
1180 1181 1182
	.iomap_begin		= xfs_file_iomap_begin,
	.iomap_end		= xfs_file_iomap_end,
};
1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202

static int
xfs_xattr_iomap_begin(
	struct inode		*inode,
	loff_t			offset,
	loff_t			length,
	unsigned		flags,
	struct iomap		*iomap)
{
	struct xfs_inode	*ip = XFS_I(inode);
	struct xfs_mount	*mp = ip->i_mount;
	xfs_fileoff_t		offset_fsb = XFS_B_TO_FSBT(mp, offset);
	xfs_fileoff_t		end_fsb = XFS_B_TO_FSB(mp, offset + length);
	struct xfs_bmbt_irec	imap;
	int			nimaps = 1, error = 0;
	unsigned		lockmode;

	if (XFS_FORCED_SHUTDOWN(mp))
		return -EIO;

1203
	lockmode = xfs_ilock_attr_map_shared(ip);
1204 1205

	/* if there are no attribute fork or extents, return ENOENT */
1206
	if (!XFS_IFORK_Q(ip) || !ip->i_d.di_anextents) {
1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224
		error = -ENOENT;
		goto out_unlock;
	}

	ASSERT(ip->i_d.di_aformat != XFS_DINODE_FMT_LOCAL);
	error = xfs_bmapi_read(ip, offset_fsb, end_fsb - offset_fsb, &imap,
			       &nimaps, XFS_BMAPI_ENTIRE | XFS_BMAPI_ATTRFORK);
out_unlock:
	xfs_iunlock(ip, lockmode);

	if (!error) {
		ASSERT(nimaps);
		xfs_bmbt_to_iomap(ip, iomap, &imap);
	}

	return error;
}

1225
const struct iomap_ops xfs_xattr_iomap_ops = {
1226 1227
	.iomap_begin		= xfs_xattr_iomap_begin,
};