xfs_bmap.c 173.4 KB
Newer Older
L
Linus Torvalds 已提交
1
/*
2
 * Copyright (c) 2000-2006 Silicon Graphics, Inc.
3
 * All Rights Reserved.
L
Linus Torvalds 已提交
4
 *
5 6
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License as
L
Linus Torvalds 已提交
7 8
 * published by the Free Software Foundation.
 *
9 10 11 12
 * 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.
L
Linus Torvalds 已提交
13
 *
14 15 16
 * 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
L
Linus Torvalds 已提交
17 18
 */
#include "xfs.h"
19
#include "xfs_fs.h"
20
#include "xfs_shared.h"
21 22 23
#include "xfs_format.h"
#include "xfs_log_format.h"
#include "xfs_trans_resv.h"
24
#include "xfs_bit.h"
L
Linus Torvalds 已提交
25
#include "xfs_sb.h"
26
#include "xfs_mount.h"
27
#include "xfs_da_format.h"
28
#include "xfs_da_btree.h"
29
#include "xfs_dir2.h"
L
Linus Torvalds 已提交
30
#include "xfs_inode.h"
31
#include "xfs_btree.h"
32
#include "xfs_trans.h"
33
#include "xfs_inode_item.h"
L
Linus Torvalds 已提交
34 35 36
#include "xfs_extfree_item.h"
#include "xfs_alloc.h"
#include "xfs_bmap.h"
D
Dave Chinner 已提交
37
#include "xfs_bmap_util.h"
38
#include "xfs_bmap_btree.h"
L
Linus Torvalds 已提交
39 40 41 42 43
#include "xfs_rtalloc.h"
#include "xfs_error.h"
#include "xfs_quota.h"
#include "xfs_trans_space.h"
#include "xfs_buf_item.h"
C
Christoph Hellwig 已提交
44
#include "xfs_trace.h"
45
#include "xfs_symlink.h"
46 47
#include "xfs_attr_leaf.h"
#include "xfs_filestream.h"
L
Linus Torvalds 已提交
48 49 50 51 52


kmem_zone_t		*xfs_bmap_free_item_zone;

/*
53
 * Miscellaneous helper functions
L
Linus Torvalds 已提交
54 55 56
 */

/*
57 58
 * Compute and fill in the value of the maximum depth of a bmap btree
 * in this filesystem.  Done once, during mount.
L
Linus Torvalds 已提交
59
 */
60 61 62 63 64 65 66 67 68 69 70 71
void
xfs_bmap_compute_maxlevels(
	xfs_mount_t	*mp,		/* file system mount structure */
	int		whichfork)	/* data or attr fork */
{
	int		level;		/* btree level */
	uint		maxblocks;	/* max blocks at this level */
	uint		maxleafents;	/* max leaf entries possible */
	int		maxrootrecs;	/* max records in root block */
	int		minleafrecs;	/* min records in leaf block */
	int		minnoderecs;	/* min records in node block */
	int		sz;		/* root block size */
L
Linus Torvalds 已提交
72

73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93
	/*
	 * The maximum number of extents in a file, hence the maximum
	 * number of leaf entries, is controlled by the type of di_nextents
	 * (a signed 32-bit number, xfs_extnum_t), or by di_anextents
	 * (a signed 16-bit number, xfs_aextnum_t).
	 *
	 * Note that we can no longer assume that if we are in ATTR1 that
	 * the fork offset of all the inodes will be
	 * (xfs_default_attroffset(ip) >> 3) because we could have mounted
	 * with ATTR2 and then mounted back with ATTR1, keeping the
	 * di_forkoff's fixed but probably at various positions. Therefore,
	 * for both ATTR1 and ATTR2 we have to assume the worst case scenario
	 * of a minimum size available.
	 */
	if (whichfork == XFS_DATA_FORK) {
		maxleafents = MAXEXTNUM;
		sz = XFS_BMDR_SPACE_CALC(MINDBTPTRS);
	} else {
		maxleafents = MAXAEXTNUM;
		sz = XFS_BMDR_SPACE_CALC(MINABTPTRS);
	}
94
	maxrootrecs = xfs_bmdr_maxrecs(sz, 0);
95 96 97 98 99 100 101 102 103 104 105
	minleafrecs = mp->m_bmap_dmnr[0];
	minnoderecs = mp->m_bmap_dmnr[1];
	maxblocks = (maxleafents + minleafrecs - 1) / minleafrecs;
	for (level = 1; maxblocks > 1; level++) {
		if (maxblocks <= maxrootrecs)
			maxblocks = 1;
		else
			maxblocks = (maxblocks + minnoderecs - 1) / minnoderecs;
	}
	mp->m_bm_maxlevels[whichfork] = level;
}
106

107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134
STATIC int				/* error */
xfs_bmbt_lookup_eq(
	struct xfs_btree_cur	*cur,
	xfs_fileoff_t		off,
	xfs_fsblock_t		bno,
	xfs_filblks_t		len,
	int			*stat)	/* success/failure */
{
	cur->bc_rec.b.br_startoff = off;
	cur->bc_rec.b.br_startblock = bno;
	cur->bc_rec.b.br_blockcount = len;
	return xfs_btree_lookup(cur, XFS_LOOKUP_EQ, stat);
}

STATIC int				/* error */
xfs_bmbt_lookup_ge(
	struct xfs_btree_cur	*cur,
	xfs_fileoff_t		off,
	xfs_fsblock_t		bno,
	xfs_filblks_t		len,
	int			*stat)	/* success/failure */
{
	cur->bc_rec.b.br_startoff = off;
	cur->bc_rec.b.br_startblock = bno;
	cur->bc_rec.b.br_blockcount = len;
	return xfs_btree_lookup(cur, XFS_LOOKUP_GE, stat);
}

135
/*
136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156
 * Check if the inode needs to be converted to btree format.
 */
static inline bool xfs_bmap_needs_btree(struct xfs_inode *ip, int whichfork)
{
	return XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_EXTENTS &&
		XFS_IFORK_NEXTENTS(ip, whichfork) >
			XFS_IFORK_MAXEXT(ip, whichfork);
}

/*
 * Check if the inode should be converted to extent format.
 */
static inline bool xfs_bmap_wants_extents(struct xfs_inode *ip, int whichfork)
{
	return XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_BTREE &&
		XFS_IFORK_NEXTENTS(ip, whichfork) <=
			XFS_IFORK_MAXEXT(ip, whichfork);
}

/*
 * Update the record referred to by cur to the value given
157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172
 * by [off, bno, len, state].
 * This either works (return 0) or gets an EFSCORRUPTED error.
 */
STATIC int
xfs_bmbt_update(
	struct xfs_btree_cur	*cur,
	xfs_fileoff_t		off,
	xfs_fsblock_t		bno,
	xfs_filblks_t		len,
	xfs_exntst_t		state)
{
	union xfs_btree_rec	rec;

	xfs_bmbt_disk_set_allf(&rec.bmbt, off, bno, len, state);
	return xfs_btree_update(cur, &rec);
}
173

L
Linus Torvalds 已提交
174
/*
175 176
 * Compute the worst-case number of indirect blocks that will be used
 * for ip's delayed extent of length "len".
L
Linus Torvalds 已提交
177
 */
178 179 180 181
STATIC xfs_filblks_t
xfs_bmap_worst_indlen(
	xfs_inode_t	*ip,		/* incore inode pointer */
	xfs_filblks_t	len)		/* delayed extent length */
L
Linus Torvalds 已提交
182
{
183 184 185 186
	int		level;		/* btree level number */
	int		maxrecs;	/* maximum record count at this level */
	xfs_mount_t	*mp;		/* mount structure */
	xfs_filblks_t	rval;		/* return value */
L
Linus Torvalds 已提交
187 188

	mp = ip->i_mount;
189 190 191 192 193 194 195 196 197 198 199 200
	maxrecs = mp->m_bmap_dmxr[0];
	for (level = 0, rval = 0;
	     level < XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK);
	     level++) {
		len += maxrecs - 1;
		do_div(len, maxrecs);
		rval += len;
		if (len == 1)
			return rval + XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK) -
				level - 1;
		if (level == 0)
			maxrecs = mp->m_bmap_dmxr[1];
L
Linus Torvalds 已提交
201
	}
202
	return rval;
L
Linus Torvalds 已提交
203 204 205
}

/*
206
 * Calculate the default attribute fork offset for newly created inodes.
L
Linus Torvalds 已提交
207
 */
208 209 210
uint
xfs_default_attroffset(
	struct xfs_inode	*ip)
L
Linus Torvalds 已提交
211
{
212 213
	struct xfs_mount	*mp = ip->i_mount;
	uint			offset;
L
Linus Torvalds 已提交
214

215
	if (mp->m_sb.sb_inodesize == 256) {
216
		offset = XFS_LITINO(mp, ip->i_d.di_version) -
217 218 219
				XFS_BMDR_SPACE_CALC(MINABTPTRS);
	} else {
		offset = XFS_BMDR_SPACE_CALC(6 * MINABTPTRS);
L
Linus Torvalds 已提交
220
	}
221

222
	ASSERT(offset < XFS_LITINO(mp, ip->i_d.di_version));
223
	return offset;
L
Linus Torvalds 已提交
224 225 226
}

/*
227 228 229
 * Helper routine to reset inode di_forkoff field when switching
 * attribute fork from local to extent format - we reset it where
 * possible to make space available for inline data fork extents.
230 231
 */
STATIC void
232 233 234
xfs_bmap_forkoff_reset(
	xfs_inode_t	*ip,
	int		whichfork)
235
{
236 237 238 239 240 241 242 243 244
	if (whichfork == XFS_ATTR_FORK &&
	    ip->i_d.di_format != XFS_DINODE_FMT_DEV &&
	    ip->i_d.di_format != XFS_DINODE_FMT_UUID &&
	    ip->i_d.di_format != XFS_DINODE_FMT_BTREE) {
		uint	dfl_forkoff = xfs_default_attroffset(ip) >> 3;

		if (dfl_forkoff > ip->i_d.di_forkoff)
			ip->i_d.di_forkoff = dfl_forkoff;
	}
245 246
}

247 248 249 250 251 252 253 254
#ifdef DEBUG
STATIC struct xfs_buf *
xfs_bmap_get_bp(
	struct xfs_btree_cur	*cur,
	xfs_fsblock_t		bno)
{
	struct xfs_log_item_desc *lidp;
	int			i;
255

256 257 258 259 260 261 262 263
	if (!cur)
		return NULL;

	for (i = 0; i < XFS_BTREE_MAXLEVELS; i++) {
		if (!cur->bc_bufs[i])
			break;
		if (XFS_BUF_ADDR(cur->bc_bufs[i]) == bno)
			return cur->bc_bufs[i];
L
Linus Torvalds 已提交
264
	}
265

266 267 268 269 270 271 272 273
	/* Chase down all the log items to see if the bp is there */
	list_for_each_entry(lidp, &cur->bc_tp->t_items, lid_trans) {
		struct xfs_buf_log_item	*bip;
		bip = (struct xfs_buf_log_item *)lidp->lid_item;
		if (bip->bli_item.li_type == XFS_LI_BUF &&
		    XFS_BUF_ADDR(bip->bli_buf) == bno)
			return bip->bli_buf;
	}
274

275 276
	return NULL;
}
C
Christoph Hellwig 已提交
277

278 279 280 281 282 283 284 285 286 287
STATIC void
xfs_check_block(
	struct xfs_btree_block	*block,
	xfs_mount_t		*mp,
	int			root,
	short			sz)
{
	int			i, j, dmxr;
	__be64			*pp, *thispa;	/* pointer to block address */
	xfs_bmbt_key_t		*prevp, *keyp;
L
Linus Torvalds 已提交
288

289
	ASSERT(be16_to_cpu(block->bb_level) > 0);
C
Christoph Hellwig 已提交
290

291 292 293 294
	prevp = NULL;
	for( i = 1; i <= xfs_btree_get_numrecs(block); i++) {
		dmxr = mp->m_bmap_dmxr[0];
		keyp = XFS_BMBT_KEY_ADDR(mp, block, i);
C
Christoph Hellwig 已提交
295

296 297 298
		if (prevp) {
			ASSERT(be64_to_cpu(prevp->br_startoff) <
			       be64_to_cpu(keyp->br_startoff));
L
Linus Torvalds 已提交
299
		}
300
		prevp = keyp;
L
Linus Torvalds 已提交
301 302

		/*
303
		 * Compare the block numbers to see if there are dups.
L
Linus Torvalds 已提交
304
		 */
305 306 307 308
		if (root)
			pp = XFS_BMAP_BROOT_PTR_ADDR(mp, block, i, sz);
		else
			pp = XFS_BMBT_PTR_ADDR(mp, block, i, dmxr);
C
Christoph Hellwig 已提交
309

310 311 312 313 314 315 316 317 318 319 320 321
		for (j = i+1; j <= be16_to_cpu(block->bb_numrecs); j++) {
			if (root)
				thispa = XFS_BMAP_BROOT_PTR_ADDR(mp, block, j, sz);
			else
				thispa = XFS_BMBT_PTR_ADDR(mp, block, j, dmxr);
			if (*thispa == *pp) {
				xfs_warn(mp, "%s: thispa(%d) == pp(%d) %Ld",
					__func__, j, i,
					(unsigned long long)be64_to_cpu(*thispa));
				panic("%s: ptrs are equal in node\n",
					__func__);
			}
L
Linus Torvalds 已提交
322
		}
323 324
	}
}
L
Linus Torvalds 已提交
325

326 327
/*
 * Check that the extents for the inode ip are in the right order in all
328 329 330 331
 * btree leaves. THis becomes prohibitively expensive for large extent count
 * files, so don't bother with inodes that have more than 10,000 extents in
 * them. The btree record ordering checks will still be done, so for such large
 * bmapbt constructs that is going to catch most corruptions.
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
 */
STATIC void
xfs_bmap_check_leaf_extents(
	xfs_btree_cur_t		*cur,	/* btree cursor or null */
	xfs_inode_t		*ip,		/* incore inode pointer */
	int			whichfork)	/* data or attr fork */
{
	struct xfs_btree_block	*block;	/* current btree block */
	xfs_fsblock_t		bno;	/* block # of "block" */
	xfs_buf_t		*bp;	/* buffer for "block" */
	int			error;	/* error return value */
	xfs_extnum_t		i=0, j;	/* index into the extents list */
	xfs_ifork_t		*ifp;	/* fork structure */
	int			level;	/* btree level, for checking */
	xfs_mount_t		*mp;	/* file system mount structure */
	__be64			*pp;	/* pointer to block address */
	xfs_bmbt_rec_t		*ep;	/* pointer to current extent */
	xfs_bmbt_rec_t		last = {0, 0}; /* last extent in prev block */
	xfs_bmbt_rec_t		*nextp;	/* pointer to next extent */
	int			bp_release = 0;

	if (XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE) {
		return;
	}

357 358 359 360
	/* skip large extent count inodes */
	if (ip->i_d.di_nextents > 10000)
		return;

361 362 363 364 365 366 367 368 369 370 371 372 373
	bno = NULLFSBLOCK;
	mp = ip->i_mount;
	ifp = XFS_IFORK_PTR(ip, whichfork);
	block = ifp->if_broot;
	/*
	 * Root level must use BMAP_BROOT_PTR_ADDR macro to get ptr out.
	 */
	level = be16_to_cpu(block->bb_level);
	ASSERT(level > 0);
	xfs_check_block(block, mp, 1, ifp->if_broot_bytes);
	pp = XFS_BMAP_BROOT_PTR_ADDR(mp, block, 1, ifp->if_broot_bytes);
	bno = be64_to_cpu(*pp);

C
Christoph Hellwig 已提交
374
	ASSERT(bno != NULLFSBLOCK);
375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390
	ASSERT(XFS_FSB_TO_AGNO(mp, bno) < mp->m_sb.sb_agcount);
	ASSERT(XFS_FSB_TO_AGBNO(mp, bno) < mp->m_sb.sb_agblocks);

	/*
	 * Go down the tree until leaf level is reached, following the first
	 * pointer (leftmost) at each level.
	 */
	while (level-- > 0) {
		/* See if buf is in cur first */
		bp_release = 0;
		bp = xfs_bmap_get_bp(cur, XFS_FSB_TO_DADDR(mp, bno));
		if (!bp) {
			bp_release = 1;
			error = xfs_btree_read_bufl(mp, NULL, bno, 0, &bp,
						XFS_BMAP_BTREE_REF,
						&xfs_bmbt_buf_ops);
391
			if (error)
392
				goto error_norelse;
L
Linus Torvalds 已提交
393
		}
394 395 396
		block = XFS_BUF_TO_BLOCK(bp);
		if (level == 0)
			break;
L
Linus Torvalds 已提交
397 398

		/*
399 400
		 * Check this block for basic sanity (increasing keys and
		 * no duplicate blocks).
L
Linus Torvalds 已提交
401
		 */
C
Christoph Hellwig 已提交
402

403 404 405
		xfs_check_block(block, mp, 0, 0);
		pp = XFS_BMBT_PTR_ADDR(mp, block, 1, mp->m_bmap_dmxr[1]);
		bno = be64_to_cpu(*pp);
406 407
		XFS_WANT_CORRUPTED_GOTO(mp,
					XFS_FSB_SANITY_CHECK(mp, bno), error0);
408 409 410
		if (bp_release) {
			bp_release = 0;
			xfs_trans_brelse(NULL, bp);
L
Linus Torvalds 已提交
411
		}
412
	}
C
Christoph Hellwig 已提交
413

414 415 416 417 418 419 420 421 422 423 424 425 426 427
	/*
	 * Here with bp and block set to the leftmost leaf node in the tree.
	 */
	i = 0;

	/*
	 * Loop over all leaf nodes checking that all extents are in the right order.
	 */
	for (;;) {
		xfs_fsblock_t	nextbno;
		xfs_extnum_t	num_recs;


		num_recs = xfs_btree_get_numrecs(block);
L
Linus Torvalds 已提交
428 429

		/*
430
		 * Read-ahead the next leaf block, if any.
L
Linus Torvalds 已提交
431
		 */
432

433
		nextbno = be64_to_cpu(block->bb_u.l.bb_rightsib);
L
Linus Torvalds 已提交
434 435

		/*
436 437 438
		 * Check all the extents to make sure they are OK.
		 * If we had a previous block, the last entry should
		 * conform with the first entry in this one.
L
Linus Torvalds 已提交
439
		 */
C
Christoph Hellwig 已提交
440

441 442 443 444 445 446 447 448 449 450 451 452 453
		ep = XFS_BMBT_REC_ADDR(mp, block, 1);
		if (i) {
			ASSERT(xfs_bmbt_disk_get_startoff(&last) +
			       xfs_bmbt_disk_get_blockcount(&last) <=
			       xfs_bmbt_disk_get_startoff(ep));
		}
		for (j = 1; j < num_recs; j++) {
			nextp = XFS_BMBT_REC_ADDR(mp, block, j + 1);
			ASSERT(xfs_bmbt_disk_get_startoff(ep) +
			       xfs_bmbt_disk_get_blockcount(ep) <=
			       xfs_bmbt_disk_get_startoff(nextp));
			ep = nextp;
		}
L
Linus Torvalds 已提交
454

455 456 457 458 459 460 461
		last = *ep;
		i += num_recs;
		if (bp_release) {
			bp_release = 0;
			xfs_trans_brelse(NULL, bp);
		}
		bno = nextbno;
L
Linus Torvalds 已提交
462
		/*
463
		 * If we've reached the end, stop.
L
Linus Torvalds 已提交
464
		 */
465 466
		if (bno == NULLFSBLOCK)
			break;
467

468 469 470 471 472 473 474
		bp_release = 0;
		bp = xfs_bmap_get_bp(cur, XFS_FSB_TO_DADDR(mp, bno));
		if (!bp) {
			bp_release = 1;
			error = xfs_btree_read_bufl(mp, NULL, bno, 0, &bp,
						XFS_BMAP_BTREE_REF,
						&xfs_bmbt_buf_ops);
475
			if (error)
476
				goto error_norelse;
L
Linus Torvalds 已提交
477
		}
478
		block = XFS_BUF_TO_BLOCK(bp);
479
	}
480

481
	return;
482

483 484 485 486 487 488 489 490 491
error0:
	xfs_warn(mp, "%s: at error0", __func__);
	if (bp_release)
		xfs_trans_brelse(NULL, bp);
error_norelse:
	xfs_warn(mp, "%s: BAD after btree leaves for %d extents",
		__func__, i);
	panic("%s: CORRUPTED BTREE OR SOMETHING", __func__);
	return;
L
Linus Torvalds 已提交
492 493 494
}

/*
495
 * Add bmap trace insert entries for all the contents of the extent records.
L
Linus Torvalds 已提交
496
 */
497 498 499 500 501 502
void
xfs_bmap_trace_exlist(
	xfs_inode_t	*ip,		/* incore inode pointer */
	xfs_extnum_t	cnt,		/* count of entries in the list */
	int		whichfork,	/* data or attr fork */
	unsigned long	caller_ip)
L
Linus Torvalds 已提交
503
{
504 505 506
	xfs_extnum_t	idx;		/* extent record index */
	xfs_ifork_t	*ifp;		/* inode fork pointer */
	int		state = 0;
507

508 509
	if (whichfork == XFS_ATTR_FORK)
		state |= BMAP_ATTRFORK;
510

511 512 513 514 515
	ifp = XFS_IFORK_PTR(ip, whichfork);
	ASSERT(cnt == (ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t)));
	for (idx = 0; idx < cnt; idx++)
		trace_xfs_extlist(ip, idx, whichfork, caller_ip);
}
516

517 518
/*
 * Validate that the bmbt_irecs being returned from bmapi are valid
519 520
 * given the caller's original parameters.  Specifically check the
 * ranges of the returned irecs to ensure that they only extend beyond
521 522 523 524 525 526 527 528 529 530 531 532
 * the given parameters if the XFS_BMAPI_ENTIRE flag was set.
 */
STATIC void
xfs_bmap_validate_ret(
	xfs_fileoff_t		bno,
	xfs_filblks_t		len,
	int			flags,
	xfs_bmbt_irec_t		*mval,
	int			nmap,
	int			ret_nmap)
{
	int			i;		/* index to map values */
533

534
	ASSERT(ret_nmap <= nmap);
535

536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556
	for (i = 0; i < ret_nmap; i++) {
		ASSERT(mval[i].br_blockcount > 0);
		if (!(flags & XFS_BMAPI_ENTIRE)) {
			ASSERT(mval[i].br_startoff >= bno);
			ASSERT(mval[i].br_blockcount <= len);
			ASSERT(mval[i].br_startoff + mval[i].br_blockcount <=
			       bno + len);
		} else {
			ASSERT(mval[i].br_startoff < bno + len);
			ASSERT(mval[i].br_startoff + mval[i].br_blockcount >
			       bno);
		}
		ASSERT(i == 0 ||
		       mval[i - 1].br_startoff + mval[i - 1].br_blockcount ==
		       mval[i].br_startoff);
		ASSERT(mval[i].br_startblock != DELAYSTARTBLOCK &&
		       mval[i].br_startblock != HOLESTARTBLOCK);
		ASSERT(mval[i].br_state == XFS_EXT_NORM ||
		       mval[i].br_state == XFS_EXT_UNWRITTEN);
	}
}
557

558 559 560 561
#else
#define xfs_bmap_check_leaf_extents(cur, ip, whichfork)		do { } while (0)
#define	xfs_bmap_validate_ret(bno,len,flags,mval,onmap,nmap)
#endif /* DEBUG */
562

563 564 565
/*
 * bmap free list manipulation functions
 */
566

567 568 569 570 571 572
/*
 * Add the extent to the list of extents to be free at transaction end.
 * The list is maintained sorted (by block number).
 */
void
xfs_bmap_add_free(
573 574
	struct xfs_mount	*mp,		/* mount point structure */
	struct xfs_bmap_free	*flist,		/* list of extents */
575
	xfs_fsblock_t		bno,		/* fs block number of extent */
576
	xfs_filblks_t		len)		/* length of extent */
577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604
{
	xfs_bmap_free_item_t	*cur;		/* current (next) element */
	xfs_bmap_free_item_t	*new;		/* new element */
	xfs_bmap_free_item_t	*prev;		/* previous element */
#ifdef DEBUG
	xfs_agnumber_t		agno;
	xfs_agblock_t		agbno;

	ASSERT(bno != NULLFSBLOCK);
	ASSERT(len > 0);
	ASSERT(len <= MAXEXTLEN);
	ASSERT(!isnullstartblock(bno));
	agno = XFS_FSB_TO_AGNO(mp, bno);
	agbno = XFS_FSB_TO_AGBNO(mp, bno);
	ASSERT(agno < mp->m_sb.sb_agcount);
	ASSERT(agbno < mp->m_sb.sb_agblocks);
	ASSERT(len < mp->m_sb.sb_agblocks);
	ASSERT(agbno + len <= mp->m_sb.sb_agblocks);
#endif
	ASSERT(xfs_bmap_free_item_zone != NULL);
	new = kmem_zone_alloc(xfs_bmap_free_item_zone, KM_SLEEP);
	new->xbfi_startblock = bno;
	new->xbfi_blockcount = (xfs_extlen_t)len;
	for (prev = NULL, cur = flist->xbf_first;
	     cur != NULL;
	     prev = cur, cur = cur->xbfi_next) {
		if (cur->xbfi_startblock >= bno)
			break;
L
Linus Torvalds 已提交
605
	}
606 607 608 609 610 611 612
	if (prev)
		prev->xbfi_next = new;
	else
		flist->xbf_first = new;
	new->xbfi_next = cur;
	flist->xbf_count++;
}
613

614 615 616 617
/*
 * Remove the entry "free" from the free item list.  Prev points to the
 * previous entry, unless "free" is the head of the list.
 */
D
Dave Chinner 已提交
618
void
619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640
xfs_bmap_del_free(
	xfs_bmap_free_t		*flist,	/* free item list header */
	xfs_bmap_free_item_t	*prev,	/* previous item on list, if any */
	xfs_bmap_free_item_t	*free)	/* list item to be freed */
{
	if (prev)
		prev->xbfi_next = free->xbfi_next;
	else
		flist->xbf_first = free->xbfi_next;
	flist->xbf_count--;
	kmem_zone_free(xfs_bmap_free_item_zone, free);
}

/*
 * Free up any items left in the list.
 */
void
xfs_bmap_cancel(
	xfs_bmap_free_t		*flist)	/* list of bmap_free_items */
{
	xfs_bmap_free_item_t	*free;	/* free list item */
	xfs_bmap_free_item_t	*next;
C
Christoph Hellwig 已提交
641

642 643 644 645 646 647 648 649 650
	if (flist->xbf_count == 0)
		return;
	ASSERT(flist->xbf_first != NULL);
	for (free = flist->xbf_first; free; free = next) {
		next = free->xbfi_next;
		xfs_bmap_del_free(flist, NULL, free);
	}
	ASSERT(flist->xbf_count == 0);
}
C
Christoph Hellwig 已提交
651

652 653 654
/*
 * Inode fork format manipulation functions
 */
L
Linus Torvalds 已提交
655

656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678
/*
 * Transform a btree format file with only one leaf node, where the
 * extents list will fit in the inode, into an extents format file.
 * Since the file extents are already in-core, all we have to do is
 * give up the space for the btree root and pitch the leaf block.
 */
STATIC int				/* error */
xfs_bmap_btree_to_extents(
	xfs_trans_t		*tp,	/* transaction pointer */
	xfs_inode_t		*ip,	/* incore inode pointer */
	xfs_btree_cur_t		*cur,	/* btree cursor */
	int			*logflagsp, /* inode logging flags */
	int			whichfork)  /* data or attr fork */
{
	/* REFERENCED */
	struct xfs_btree_block	*cblock;/* child btree block */
	xfs_fsblock_t		cbno;	/* child block number */
	xfs_buf_t		*cbp;	/* child block's buffer */
	int			error;	/* error return value */
	xfs_ifork_t		*ifp;	/* inode fork data */
	xfs_mount_t		*mp;	/* mount point structure */
	__be64			*pp;	/* ptr to block address */
	struct xfs_btree_block	*rblock;/* root btree block */
L
Linus Torvalds 已提交
679

680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701
	mp = ip->i_mount;
	ifp = XFS_IFORK_PTR(ip, whichfork);
	ASSERT(ifp->if_flags & XFS_IFEXTENTS);
	ASSERT(XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_BTREE);
	rblock = ifp->if_broot;
	ASSERT(be16_to_cpu(rblock->bb_level) == 1);
	ASSERT(be16_to_cpu(rblock->bb_numrecs) == 1);
	ASSERT(xfs_bmbt_maxrecs(mp, ifp->if_broot_bytes, 0) == 1);
	pp = XFS_BMAP_BROOT_PTR_ADDR(mp, rblock, 1, ifp->if_broot_bytes);
	cbno = be64_to_cpu(*pp);
	*logflagsp = 0;
#ifdef DEBUG
	if ((error = xfs_btree_check_lptr(cur, cbno, 1)))
		return error;
#endif
	error = xfs_btree_read_bufl(mp, tp, cbno, 0, &cbp, XFS_BMAP_BTREE_REF,
				&xfs_bmbt_buf_ops);
	if (error)
		return error;
	cblock = XFS_BUF_TO_BLOCK(cbp);
	if ((error = xfs_btree_check_block(cur, cblock, 0, cbp)))
		return error;
702
	xfs_bmap_add_free(mp, cur->bc_private.b.flist, cbno, 1);
703 704 705 706 707 708 709 710 711 712 713 714
	ip->i_d.di_nblocks--;
	xfs_trans_mod_dquot_byino(tp, ip, XFS_TRANS_DQ_BCOUNT, -1L);
	xfs_trans_binval(tp, cbp);
	if (cur->bc_bufs[0] == cbp)
		cur->bc_bufs[0] = NULL;
	xfs_iroot_realloc(ip, -1, whichfork);
	ASSERT(ifp->if_broot == NULL);
	ASSERT((ifp->if_flags & XFS_IFBROOT) == 0);
	XFS_IFORK_FMT_SET(ip, whichfork, XFS_DINODE_FMT_EXTENTS);
	*logflagsp = XFS_ILOG_CORE | xfs_ilog_fext(whichfork);
	return 0;
}
C
Christoph Hellwig 已提交
715

716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744
/*
 * Convert an extents-format file into a btree-format file.
 * The new file will have a root block (in the inode) and a single child block.
 */
STATIC int					/* error */
xfs_bmap_extents_to_btree(
	xfs_trans_t		*tp,		/* transaction pointer */
	xfs_inode_t		*ip,		/* incore inode pointer */
	xfs_fsblock_t		*firstblock,	/* first-block-allocated */
	xfs_bmap_free_t		*flist,		/* blocks freed in xaction */
	xfs_btree_cur_t		**curp,		/* cursor returned to caller */
	int			wasdel,		/* converting a delayed alloc */
	int			*logflagsp,	/* inode logging flags */
	int			whichfork)	/* data or attr fork */
{
	struct xfs_btree_block	*ablock;	/* allocated (child) bt block */
	xfs_buf_t		*abp;		/* buffer for ablock */
	xfs_alloc_arg_t		args;		/* allocation arguments */
	xfs_bmbt_rec_t		*arp;		/* child record pointer */
	struct xfs_btree_block	*block;		/* btree root block */
	xfs_btree_cur_t		*cur;		/* bmap btree cursor */
	xfs_bmbt_rec_host_t	*ep;		/* extent record pointer */
	int			error;		/* error return value */
	xfs_extnum_t		i, cnt;		/* extent record index */
	xfs_ifork_t		*ifp;		/* inode fork pointer */
	xfs_bmbt_key_t		*kp;		/* root block key pointer */
	xfs_mount_t		*mp;		/* mount structure */
	xfs_extnum_t		nextents;	/* number of file extents */
	xfs_bmbt_ptr_t		*pp;		/* root block address pointer */
L
Linus Torvalds 已提交
745

746
	mp = ip->i_mount;
747 748
	ifp = XFS_IFORK_PTR(ip, whichfork);
	ASSERT(XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_EXTENTS);
C
Christoph Hellwig 已提交
749

750 751 752 753 754
	/*
	 * Make space in the inode incore.
	 */
	xfs_iroot_realloc(ip, 1, whichfork);
	ifp->if_flags |= XFS_IFBROOT;
C
Christoph Hellwig 已提交
755

756 757 758 759
	/*
	 * Fill in the root.
	 */
	block = ifp->if_broot;
760 761 762 763 764 765 766 767
	if (xfs_sb_version_hascrc(&mp->m_sb))
		xfs_btree_init_block_int(mp, block, XFS_BUF_DADDR_NULL,
				 XFS_BMAP_CRC_MAGIC, 1, 1, ip->i_ino,
				 XFS_BTREE_LONG_PTRS | XFS_BTREE_CRC_BLOCKS);
	else
		xfs_btree_init_block_int(mp, block, XFS_BUF_DADDR_NULL,
				 XFS_BMAP_MAGIC, 1, 1, ip->i_ino,
				 XFS_BTREE_LONG_PTRS);
C
Christoph Hellwig 已提交
768

769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819
	/*
	 * Need a cursor.  Can't allocate until bb_level is filled in.
	 */
	cur = xfs_bmbt_init_cursor(mp, tp, ip, whichfork);
	cur->bc_private.b.firstblock = *firstblock;
	cur->bc_private.b.flist = flist;
	cur->bc_private.b.flags = wasdel ? XFS_BTCUR_BPRV_WASDEL : 0;
	/*
	 * Convert to a btree with two levels, one record in root.
	 */
	XFS_IFORK_FMT_SET(ip, whichfork, XFS_DINODE_FMT_BTREE);
	memset(&args, 0, sizeof(args));
	args.tp = tp;
	args.mp = mp;
	args.firstblock = *firstblock;
	if (*firstblock == NULLFSBLOCK) {
		args.type = XFS_ALLOCTYPE_START_BNO;
		args.fsbno = XFS_INO_TO_FSB(mp, ip->i_ino);
	} else if (flist->xbf_low) {
		args.type = XFS_ALLOCTYPE_START_BNO;
		args.fsbno = *firstblock;
	} else {
		args.type = XFS_ALLOCTYPE_NEAR_BNO;
		args.fsbno = *firstblock;
	}
	args.minlen = args.maxlen = args.prod = 1;
	args.wasdel = wasdel;
	*logflagsp = 0;
	if ((error = xfs_alloc_vextent(&args))) {
		xfs_iroot_realloc(ip, -1, whichfork);
		xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
		return error;
	}
	/*
	 * Allocation can't fail, the space was reserved.
	 */
	ASSERT(args.fsbno != NULLFSBLOCK);
	ASSERT(*firstblock == NULLFSBLOCK ||
	       args.agno == XFS_FSB_TO_AGNO(mp, *firstblock) ||
	       (flist->xbf_low &&
		args.agno > XFS_FSB_TO_AGNO(mp, *firstblock)));
	*firstblock = cur->bc_private.b.firstblock = args.fsbno;
	cur->bc_private.b.allocated++;
	ip->i_d.di_nblocks++;
	xfs_trans_mod_dquot_byino(tp, ip, XFS_TRANS_DQ_BCOUNT, 1L);
	abp = xfs_btree_get_bufl(mp, tp, args.fsbno, 0);
	/*
	 * Fill in the child block.
	 */
	abp->b_ops = &xfs_bmbt_buf_ops;
	ablock = XFS_BUF_TO_BLOCK(abp);
820 821 822 823 824 825 826 827 828
	if (xfs_sb_version_hascrc(&mp->m_sb))
		xfs_btree_init_block_int(mp, ablock, abp->b_bn,
				XFS_BMAP_CRC_MAGIC, 0, 0, ip->i_ino,
				XFS_BTREE_LONG_PTRS | XFS_BTREE_CRC_BLOCKS);
	else
		xfs_btree_init_block_int(mp, ablock, abp->b_bn,
				XFS_BMAP_MAGIC, 0, 0, ip->i_ino,
				XFS_BTREE_LONG_PTRS);

829 830 831 832 833 834 835 836
	arp = XFS_BMBT_REC_ADDR(mp, ablock, 1);
	nextents = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t);
	for (cnt = i = 0; i < nextents; i++) {
		ep = xfs_iext_get_ext(ifp, i);
		if (!isnullstartblock(xfs_bmbt_get_startblock(ep))) {
			arp->l0 = cpu_to_be64(ep->l0);
			arp->l1 = cpu_to_be64(ep->l1);
			arp++; cnt++;
L
Linus Torvalds 已提交
837
		}
838 839 840
	}
	ASSERT(cnt == XFS_IFORK_NEXTENTS(ip, whichfork));
	xfs_btree_set_numrecs(ablock, cnt);
L
Linus Torvalds 已提交
841

842 843 844 845 846 847 848 849 850
	/*
	 * Fill in the root key and pointer.
	 */
	kp = XFS_BMBT_KEY_ADDR(mp, block, 1);
	arp = XFS_BMBT_REC_ADDR(mp, ablock, 1);
	kp->br_startoff = cpu_to_be64(xfs_bmbt_disk_get_startoff(arp));
	pp = XFS_BMBT_PTR_ADDR(mp, block, 1, xfs_bmbt_get_maxrecs(cur,
						be16_to_cpu(block->bb_level)));
	*pp = cpu_to_be64(args.fsbno);
C
Christoph Hellwig 已提交
851

852 853 854 855 856 857 858 859 860 861 862
	/*
	 * Do all this logging at the end so that
	 * the root is at the right level.
	 */
	xfs_btree_log_block(cur, abp, XFS_BB_ALL_BITS);
	xfs_btree_log_recs(cur, abp, 1, be16_to_cpu(ablock->bb_numrecs));
	ASSERT(*curp == NULL);
	*curp = cur;
	*logflagsp = XFS_ILOG_CORE | xfs_ilog_fbroot(whichfork);
	return 0;
}
C
Christoph Hellwig 已提交
863

864 865 866 867 868 869
/*
 * Convert a local file to an extents file.
 * This code is out of bounds for data forks of regular files,
 * since the file data needs to get logged so things will stay consistent.
 * (The bmap-level manipulations are ok, though).
 */
870 871 872 873 874 875 876 877 878 879 880
void
xfs_bmap_local_to_extents_empty(
	struct xfs_inode	*ip,
	int			whichfork)
{
	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);

	ASSERT(XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL);
	ASSERT(ifp->if_bytes == 0);
	ASSERT(XFS_IFORK_NEXTENTS(ip, whichfork) == 0);

881
	xfs_bmap_forkoff_reset(ip, whichfork);
882 883 884 885 886 887
	ifp->if_flags &= ~XFS_IFINLINE;
	ifp->if_flags |= XFS_IFEXTENTS;
	XFS_IFORK_FMT_SET(ip, whichfork, XFS_DINODE_FMT_EXTENTS);
}


888 889 890 891 892 893 894 895
STATIC int				/* error */
xfs_bmap_local_to_extents(
	xfs_trans_t	*tp,		/* transaction pointer */
	xfs_inode_t	*ip,		/* incore inode pointer */
	xfs_fsblock_t	*firstblock,	/* first block allocated in xaction */
	xfs_extlen_t	total,		/* total blocks needed by transaction */
	int		*logflagsp,	/* inode logging flags */
	int		whichfork,
896 897
	void		(*init_fn)(struct xfs_trans *tp,
				   struct xfs_buf *bp,
898 899 900
				   struct xfs_inode *ip,
				   struct xfs_ifork *ifp))
{
901
	int		error = 0;
902 903
	int		flags;		/* logging flags returned */
	xfs_ifork_t	*ifp;		/* inode fork pointer */
904 905 906
	xfs_alloc_arg_t	args;		/* allocation arguments */
	xfs_buf_t	*bp;		/* buffer for extent block */
	xfs_bmbt_rec_host_t *ep;	/* extent record pointer */
C
Christoph Hellwig 已提交
907

908 909 910 911
	/*
	 * We don't want to deal with the case of keeping inode data inline yet.
	 * So sending the data fork of a regular inode is invalid.
	 */
D
Dave Chinner 已提交
912
	ASSERT(!(S_ISREG(VFS_I(ip)->i_mode) && whichfork == XFS_DATA_FORK));
913 914
	ifp = XFS_IFORK_PTR(ip, whichfork);
	ASSERT(XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL);
915 916 917 918 919 920 921

	if (!ifp->if_bytes) {
		xfs_bmap_local_to_extents_empty(ip, whichfork);
		flags = XFS_ILOG_CORE;
		goto done;
	}

922 923
	flags = 0;
	error = 0;
924 925 926 927 928 929 930 931 932 933 934 935 936
	ASSERT((ifp->if_flags & (XFS_IFINLINE|XFS_IFEXTENTS|XFS_IFEXTIREC)) ==
								XFS_IFINLINE);
	memset(&args, 0, sizeof(args));
	args.tp = tp;
	args.mp = ip->i_mount;
	args.firstblock = *firstblock;
	/*
	 * Allocate a block.  We know we need only one, since the
	 * file currently fits in an inode.
	 */
	if (*firstblock == NULLFSBLOCK) {
		args.fsbno = XFS_INO_TO_FSB(args.mp, ip->i_ino);
		args.type = XFS_ALLOCTYPE_START_BNO;
937
	} else {
938 939
		args.fsbno = *firstblock;
		args.type = XFS_ALLOCTYPE_NEAR_BNO;
940
	}
941 942 943 944 945 946 947 948 949 950 951 952
	args.total = total;
	args.minlen = args.maxlen = args.prod = 1;
	error = xfs_alloc_vextent(&args);
	if (error)
		goto done;

	/* Can't fail, the space was reserved. */
	ASSERT(args.fsbno != NULLFSBLOCK);
	ASSERT(args.len == 1);
	*firstblock = args.fsbno;
	bp = xfs_btree_get_bufl(args.mp, tp, args.fsbno, 0);

953
	/*
954
	 * Initialize the block, copy the data and log the remote buffer.
955
	 *
956 957 958 959
	 * The callout is responsible for logging because the remote format
	 * might differ from the local format and thus we don't know how much to
	 * log here. Note that init_fn must also set the buffer log item type
	 * correctly.
960
	 */
961 962
	init_fn(tp, bp, ip, ifp);

963
	/* account for the change in fork size */
964 965
	xfs_idata_realloc(ip, -ifp->if_bytes, whichfork);
	xfs_bmap_local_to_extents_empty(ip, whichfork);
966
	flags |= XFS_ILOG_CORE;
967 968 969 970 971 972 973 974 975 976 977 978 979

	xfs_iext_add(ifp, 0, 1);
	ep = xfs_iext_get_ext(ifp, 0);
	xfs_bmbt_set_allf(ep, 0, args.fsbno, 1, XFS_EXT_NORM);
	trace_xfs_bmap_post_update(ip, 0,
			whichfork == XFS_ATTR_FORK ? BMAP_ATTRFORK : 0,
			_THIS_IP_);
	XFS_IFORK_NEXT_SET(ip, whichfork, 1);
	ip->i_d.di_nblocks = 1;
	xfs_trans_mod_dquot_byino(tp, ip,
		XFS_TRANS_DQ_BCOUNT, 1L);
	flags |= xfs_ilog_fext(whichfork);

980 981 982 983
done:
	*logflagsp = flags;
	return error;
}
C
Christoph Hellwig 已提交
984

985 986 987 988 989 990 991 992 993 994 995 996 997 998 999
/*
 * Called from xfs_bmap_add_attrfork to handle btree format files.
 */
STATIC int					/* error */
xfs_bmap_add_attrfork_btree(
	xfs_trans_t		*tp,		/* transaction pointer */
	xfs_inode_t		*ip,		/* incore inode pointer */
	xfs_fsblock_t		*firstblock,	/* first block allocated */
	xfs_bmap_free_t		*flist,		/* blocks to free at commit */
	int			*flags)		/* inode logging flags */
{
	xfs_btree_cur_t		*cur;		/* btree cursor */
	int			error;		/* error return value */
	xfs_mount_t		*mp;		/* file system mount struct */
	int			stat;		/* newroot status */
C
Christoph Hellwig 已提交
1000

1001 1002 1003 1004 1005 1006 1007 1008 1009 1010
	mp = ip->i_mount;
	if (ip->i_df.if_broot_bytes <= XFS_IFORK_DSIZE(ip))
		*flags |= XFS_ILOG_DBROOT;
	else {
		cur = xfs_bmbt_init_cursor(mp, tp, ip, XFS_DATA_FORK);
		cur->bc_private.b.flist = flist;
		cur->bc_private.b.firstblock = *firstblock;
		if ((error = xfs_bmbt_lookup_ge(cur, 0, 0, 0, &stat)))
			goto error0;
		/* must be at least one entry */
1011
		XFS_WANT_CORRUPTED_GOTO(mp, stat == 1, error0);
1012 1013 1014 1015
		if ((error = xfs_btree_new_iroot(cur, flags, &stat)))
			goto error0;
		if (stat == 0) {
			xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR);
D
Dave Chinner 已提交
1016
			return -ENOSPC;
L
Linus Torvalds 已提交
1017
		}
1018 1019 1020
		*firstblock = cur->bc_private.b.firstblock;
		cur->bc_private.b.allocated = 0;
		xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR);
L
Linus Torvalds 已提交
1021
	}
1022 1023 1024 1025 1026
	return 0;
error0:
	xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
	return error;
}
1027

1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040
/*
 * Called from xfs_bmap_add_attrfork to handle extents format files.
 */
STATIC int					/* error */
xfs_bmap_add_attrfork_extents(
	xfs_trans_t		*tp,		/* transaction pointer */
	xfs_inode_t		*ip,		/* incore inode pointer */
	xfs_fsblock_t		*firstblock,	/* first block allocated */
	xfs_bmap_free_t		*flist,		/* blocks to free at commit */
	int			*flags)		/* inode logging flags */
{
	xfs_btree_cur_t		*cur;		/* bmap btree cursor */
	int			error;		/* error return value */
1041

1042 1043 1044 1045 1046
	if (ip->i_d.di_nextents * sizeof(xfs_bmbt_rec_t) <= XFS_IFORK_DSIZE(ip))
		return 0;
	cur = NULL;
	error = xfs_bmap_extents_to_btree(tp, ip, firstblock, flist, &cur, 0,
		flags, XFS_DATA_FORK);
1047 1048
	if (cur) {
		cur->bc_private.b.allocated = 0;
1049 1050
		xfs_btree_del_cursor(cur,
			error ? XFS_BTREE_ERROR : XFS_BTREE_NOERROR);
1051
	}
L
Linus Torvalds 已提交
1052 1053 1054
	return error;
}

1055 1056 1057 1058 1059 1060 1061 1062 1063
/*
 * Called from xfs_bmap_add_attrfork to handle local format files. Each
 * different data fork content type needs a different callout to do the
 * conversion. Some are basic and only require special block initialisation
 * callouts for the data formating, others (directories) are so specialised they
 * handle everything themselves.
 *
 * XXX (dgc): investigate whether directory conversion can use the generic
 * formatting callout. It should be possible - it's just a very complex
1064
 * formatter.
1065 1066 1067 1068 1069 1070 1071 1072 1073 1074
 */
STATIC int					/* error */
xfs_bmap_add_attrfork_local(
	xfs_trans_t		*tp,		/* transaction pointer */
	xfs_inode_t		*ip,		/* incore inode pointer */
	xfs_fsblock_t		*firstblock,	/* first block allocated */
	xfs_bmap_free_t		*flist,		/* blocks to free at commit */
	int			*flags)		/* inode logging flags */
{
	xfs_da_args_t		dargs;		/* args for dir/attr code */
1075

1076 1077
	if (ip->i_df.if_bytes <= XFS_IFORK_DSIZE(ip))
		return 0;
1078

D
Dave Chinner 已提交
1079
	if (S_ISDIR(VFS_I(ip)->i_mode)) {
1080
		memset(&dargs, 0, sizeof(dargs));
1081
		dargs.geo = ip->i_mount->m_dir_geo;
1082 1083 1084
		dargs.dp = ip;
		dargs.firstblock = firstblock;
		dargs.flist = flist;
1085
		dargs.total = dargs.geo->fsbcount;
1086 1087 1088
		dargs.whichfork = XFS_DATA_FORK;
		dargs.trans = tp;
		return xfs_dir2_sf_to_block(&dargs);
L
Linus Torvalds 已提交
1089
	}
1090

D
Dave Chinner 已提交
1091
	if (S_ISLNK(VFS_I(ip)->i_mode))
1092 1093 1094
		return xfs_bmap_local_to_extents(tp, ip, firstblock, 1,
						 flags, XFS_DATA_FORK,
						 xfs_symlink_local_to_remote);
1095

1096 1097
	/* should only be called for types that support local format data */
	ASSERT(0);
D
Dave Chinner 已提交
1098
	return -EFSCORRUPTED;
1099
}
C
Christoph Hellwig 已提交
1100

1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118
/*
 * Convert inode from non-attributed to attributed.
 * Must not be in a transaction, ip must not be locked.
 */
int						/* error code */
xfs_bmap_add_attrfork(
	xfs_inode_t		*ip,		/* incore inode pointer */
	int			size,		/* space new attribute needs */
	int			rsvd)		/* xact may use reserved blks */
{
	xfs_fsblock_t		firstblock;	/* 1st block/ag allocated */
	xfs_bmap_free_t		flist;		/* freed extent records */
	xfs_mount_t		*mp;		/* mount structure */
	xfs_trans_t		*tp;		/* transaction pointer */
	int			blks;		/* space reservation */
	int			version = 1;	/* superblock attr version */
	int			logflags;	/* logging flags */
	int			error;		/* error return value */
C
Christoph Hellwig 已提交
1119

1120
	ASSERT(XFS_IFORK_Q(ip) == 0);
L
Linus Torvalds 已提交
1121

1122 1123
	mp = ip->i_mount;
	ASSERT(!XFS_NOT_DQATTACHED(mp, ip));
1124

1125
	blks = XFS_ADDAFORK_SPACE_RES(mp);
1126 1127 1128 1129

	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_addafork, blks, 0,
			rsvd ? XFS_TRANS_RESERVE : 0, &tp);
	if (error)
1130
		return error;
1131

1132 1133 1134 1135
	xfs_ilock(ip, XFS_ILOCK_EXCL);
	error = xfs_trans_reserve_quota_nblks(tp, ip, blks, 0, rsvd ?
			XFS_QMOPT_RES_REGBLKS | XFS_QMOPT_FORCE_RES :
			XFS_QMOPT_RES_REGBLKS);
1136 1137
	if (error)
		goto trans_cancel;
1138
	if (XFS_IFORK_Q(ip))
1139
		goto trans_cancel;
1140
	if (ip->i_d.di_aformat != XFS_DINODE_FMT_EXTENTS) {
L
Linus Torvalds 已提交
1141
		/*
1142
		 * For inodes coming from pre-6.2 filesystems.
L
Linus Torvalds 已提交
1143
		 */
1144 1145 1146 1147
		ASSERT(ip->i_d.di_aformat == 0);
		ip->i_d.di_aformat = XFS_DINODE_FMT_EXTENTS;
	}
	ASSERT(ip->i_d.di_anextents == 0);
C
Christoph Hellwig 已提交
1148

1149
	xfs_trans_ijoin(tp, ip, 0);
1150
	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
L
Linus Torvalds 已提交
1151

1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166
	switch (ip->i_d.di_format) {
	case XFS_DINODE_FMT_DEV:
		ip->i_d.di_forkoff = roundup(sizeof(xfs_dev_t), 8) >> 3;
		break;
	case XFS_DINODE_FMT_UUID:
		ip->i_d.di_forkoff = roundup(sizeof(uuid_t), 8) >> 3;
		break;
	case XFS_DINODE_FMT_LOCAL:
	case XFS_DINODE_FMT_EXTENTS:
	case XFS_DINODE_FMT_BTREE:
		ip->i_d.di_forkoff = xfs_attr_shortform_bytesfit(ip, size);
		if (!ip->i_d.di_forkoff)
			ip->i_d.di_forkoff = xfs_default_attroffset(ip) >> 3;
		else if (mp->m_flags & XFS_MOUNT_ATTR2)
			version = 2;
L
Linus Torvalds 已提交
1167
		break;
1168 1169
	default:
		ASSERT(0);
D
Dave Chinner 已提交
1170
		error = -EINVAL;
1171
		goto trans_cancel;
1172
	}
L
Linus Torvalds 已提交
1173

1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193
	ASSERT(ip->i_afp == NULL);
	ip->i_afp = kmem_zone_zalloc(xfs_ifork_zone, KM_SLEEP);
	ip->i_afp->if_flags = XFS_IFEXTENTS;
	logflags = 0;
	xfs_bmap_init(&flist, &firstblock);
	switch (ip->i_d.di_format) {
	case XFS_DINODE_FMT_LOCAL:
		error = xfs_bmap_add_attrfork_local(tp, ip, &firstblock, &flist,
			&logflags);
		break;
	case XFS_DINODE_FMT_EXTENTS:
		error = xfs_bmap_add_attrfork_extents(tp, ip, &firstblock,
			&flist, &logflags);
		break;
	case XFS_DINODE_FMT_BTREE:
		error = xfs_bmap_add_attrfork_btree(tp, ip, &firstblock, &flist,
			&logflags);
		break;
	default:
		error = 0;
L
Linus Torvalds 已提交
1194 1195
		break;
	}
1196 1197 1198
	if (logflags)
		xfs_trans_log_inode(tp, ip, logflags);
	if (error)
1199
		goto bmap_cancel;
1200 1201
	if (!xfs_sb_version_hasattr(&mp->m_sb) ||
	   (!xfs_sb_version_hasattr2(&mp->m_sb) && version == 2)) {
1202
		bool log_sb = false;
1203 1204 1205 1206

		spin_lock(&mp->m_sb_lock);
		if (!xfs_sb_version_hasattr(&mp->m_sb)) {
			xfs_sb_version_addattr(&mp->m_sb);
1207
			log_sb = true;
1208 1209 1210
		}
		if (!xfs_sb_version_hasattr2(&mp->m_sb) && version == 2) {
			xfs_sb_version_addattr2(&mp->m_sb);
1211
			log_sb = true;
1212
		}
1213
		spin_unlock(&mp->m_sb_lock);
1214 1215
		if (log_sb)
			xfs_log_sb(tp);
L
Linus Torvalds 已提交
1216
	}
1217

1218
	error = xfs_bmap_finish(&tp, &flist, NULL);
1219
	if (error)
1220
		goto bmap_cancel;
1221
	error = xfs_trans_commit(tp);
1222 1223 1224 1225
	xfs_iunlock(ip, XFS_ILOCK_EXCL);
	return error;

bmap_cancel:
1226
	xfs_bmap_cancel(&flist);
1227
trans_cancel:
1228
	xfs_trans_cancel(tp);
1229 1230
	xfs_iunlock(ip, XFS_ILOCK_EXCL);
	return error;
L
Linus Torvalds 已提交
1231 1232 1233
}

/*
1234
 * Internal and external extent tree search functions.
L
Linus Torvalds 已提交
1235
 */
1236

1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260
/*
 * Read in the extents to if_extents.
 * All inode fields are set up by caller, we just traverse the btree
 * and copy the records in. If the file system cannot contain unwritten
 * extents, the records are checked for no "state" flags.
 */
int					/* error */
xfs_bmap_read_extents(
	xfs_trans_t		*tp,	/* transaction pointer */
	xfs_inode_t		*ip,	/* incore inode */
	int			whichfork) /* data or attr fork */
{
	struct xfs_btree_block	*block;	/* current btree block */
	xfs_fsblock_t		bno;	/* block # of "block" */
	xfs_buf_t		*bp;	/* buffer for "block" */
	int			error;	/* error return value */
	xfs_exntfmt_t		exntf;	/* XFS_EXTFMT_NOSTATE, if checking */
	xfs_extnum_t		i, j;	/* index into the extents list */
	xfs_ifork_t		*ifp;	/* fork structure */
	int			level;	/* btree level, for checking */
	xfs_mount_t		*mp;	/* file system mount structure */
	__be64			*pp;	/* pointer to block address */
	/* REFERENCED */
	xfs_extnum_t		room;	/* number of entries there's room for */
1261

1262 1263 1264 1265 1266 1267
	bno = NULLFSBLOCK;
	mp = ip->i_mount;
	ifp = XFS_IFORK_PTR(ip, whichfork);
	exntf = (whichfork != XFS_DATA_FORK) ? XFS_EXTFMT_NOSTATE :
					XFS_EXTFMT_INODE(ip);
	block = ifp->if_broot;
L
Linus Torvalds 已提交
1268
	/*
1269
	 * Root level must use BMAP_BROOT_PTR_ADDR macro to get ptr out.
L
Linus Torvalds 已提交
1270
	 */
1271 1272 1273 1274
	level = be16_to_cpu(block->bb_level);
	ASSERT(level > 0);
	pp = XFS_BMAP_BROOT_PTR_ADDR(mp, block, 1, ifp->if_broot_bytes);
	bno = be64_to_cpu(*pp);
C
Christoph Hellwig 已提交
1275
	ASSERT(bno != NULLFSBLOCK);
1276 1277
	ASSERT(XFS_FSB_TO_AGNO(mp, bno) < mp->m_sb.sb_agcount);
	ASSERT(XFS_FSB_TO_AGBNO(mp, bno) < mp->m_sb.sb_agblocks);
L
Linus Torvalds 已提交
1278
	/*
1279 1280
	 * Go down the tree until leaf level is reached, following the first
	 * pointer (leftmost) at each level.
L
Linus Torvalds 已提交
1281
	 */
1282 1283 1284 1285 1286 1287 1288 1289 1290 1291
	while (level-- > 0) {
		error = xfs_btree_read_bufl(mp, tp, bno, 0, &bp,
				XFS_BMAP_BTREE_REF, &xfs_bmbt_buf_ops);
		if (error)
			return error;
		block = XFS_BUF_TO_BLOCK(bp);
		if (level == 0)
			break;
		pp = XFS_BMBT_PTR_ADDR(mp, block, 1, mp->m_bmap_dmxr[1]);
		bno = be64_to_cpu(*pp);
1292 1293
		XFS_WANT_CORRUPTED_GOTO(mp,
			XFS_FSB_SANITY_CHECK(mp, bno), error0);
1294
		xfs_trans_brelse(tp, bp);
L
Linus Torvalds 已提交
1295 1296
	}
	/*
1297
	 * Here with bp and block set to the leftmost leaf node in the tree.
L
Linus Torvalds 已提交
1298
	 */
1299 1300
	room = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t);
	i = 0;
L
Linus Torvalds 已提交
1301
	/*
1302
	 * Loop over all leaf nodes.  Copy information to the extent records.
L
Linus Torvalds 已提交
1303
	 */
1304 1305 1306 1307 1308
	for (;;) {
		xfs_bmbt_rec_t	*frp;
		xfs_fsblock_t	nextbno;
		xfs_extnum_t	num_recs;
		xfs_extnum_t	start;
C
Christoph Hellwig 已提交
1309

1310 1311 1312 1313 1314 1315 1316 1317 1318
		num_recs = xfs_btree_get_numrecs(block);
		if (unlikely(i + num_recs > room)) {
			ASSERT(i + num_recs <= room);
			xfs_warn(ip->i_mount,
				"corrupt dinode %Lu, (btree extents).",
				(unsigned long long) ip->i_ino);
			XFS_CORRUPTION_ERROR("xfs_bmap_read_extents(1)",
				XFS_ERRLEVEL_LOW, ip->i_mount, block);
			goto error0;
L
Linus Torvalds 已提交
1319 1320
		}
		/*
1321
		 * Read-ahead the next leaf block, if any.
L
Linus Torvalds 已提交
1322
		 */
1323 1324 1325 1326
		nextbno = be64_to_cpu(block->bb_u.l.bb_rightsib);
		if (nextbno != NULLFSBLOCK)
			xfs_btree_reada_bufl(mp, nextbno, 1,
					     &xfs_bmbt_buf_ops);
L
Linus Torvalds 已提交
1327
		/*
1328
		 * Copy records into the extent records.
L
Linus Torvalds 已提交
1329
		 */
1330 1331 1332 1333 1334 1335
		frp = XFS_BMBT_REC_ADDR(mp, block, 1);
		start = i;
		for (j = 0; j < num_recs; j++, i++, frp++) {
			xfs_bmbt_rec_host_t *trp = xfs_iext_get_ext(ifp, i);
			trp->l0 = be64_to_cpu(frp->l0);
			trp->l1 = be64_to_cpu(frp->l1);
L
Linus Torvalds 已提交
1336
		}
1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352
		if (exntf == XFS_EXTFMT_NOSTATE) {
			/*
			 * Check all attribute bmap btree records and
			 * any "older" data bmap btree records for a
			 * set bit in the "extent flag" position.
			 */
			if (unlikely(xfs_check_nostate_extents(ifp,
					start, num_recs))) {
				XFS_ERROR_REPORT("xfs_bmap_read_extents(2)",
						 XFS_ERRLEVEL_LOW,
						 ip->i_mount);
				goto error0;
			}
		}
		xfs_trans_brelse(tp, bp);
		bno = nextbno;
L
Linus Torvalds 已提交
1353
		/*
1354
		 * If we've reached the end, stop.
L
Linus Torvalds 已提交
1355
		 */
1356 1357 1358 1359 1360 1361 1362
		if (bno == NULLFSBLOCK)
			break;
		error = xfs_btree_read_bufl(mp, tp, bno, 0, &bp,
				XFS_BMAP_BTREE_REF, &xfs_bmbt_buf_ops);
		if (error)
			return error;
		block = XFS_BUF_TO_BLOCK(bp);
L
Linus Torvalds 已提交
1363
	}
1364 1365 1366 1367 1368 1369
	ASSERT(i == (ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t)));
	ASSERT(i == XFS_IFORK_NEXTENTS(ip, whichfork));
	XFS_BMAP_TRACE_EXLIST(ip, i, whichfork);
	return 0;
error0:
	xfs_trans_brelse(tp, bp);
D
Dave Chinner 已提交
1370
	return -EFSCORRUPTED;
1371
}
1372 1373


1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391
/*
 * Search the extent records for the entry containing block bno.
 * If bno lies in a hole, point to the next entry.  If bno lies
 * past eof, *eofp will be set, and *prevp will contain the last
 * entry (null if none).  Else, *lastxp will be set to the index
 * of the found entry; *gotp will contain the entry.
 */
STATIC xfs_bmbt_rec_host_t *		/* pointer to found extent entry */
xfs_bmap_search_multi_extents(
	xfs_ifork_t	*ifp,		/* inode fork pointer */
	xfs_fileoff_t	bno,		/* block number searched for */
	int		*eofp,		/* out: end of file found */
	xfs_extnum_t	*lastxp,	/* out: last extent index */
	xfs_bmbt_irec_t	*gotp,		/* out: extent entry found */
	xfs_bmbt_irec_t	*prevp)		/* out: previous extent entry found */
{
	xfs_bmbt_rec_host_t *ep;		/* extent record pointer */
	xfs_extnum_t	lastx;		/* last extent index */
1392

1393 1394 1395 1396 1397 1398 1399 1400 1401
	/*
	 * Initialize the extent entry structure to catch access to
	 * uninitialized br_startblock field.
	 */
	gotp->br_startoff = 0xffa5a5a5a5a5a5a5LL;
	gotp->br_blockcount = 0xa55a5a5a5a5a5a5aLL;
	gotp->br_state = XFS_EXT_INVALID;
	gotp->br_startblock = 0xffffa5a5a5a5a5a5LL;
	prevp->br_startoff = NULLFILEOFF;
1402

1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418
	ep = xfs_iext_bno_to_ext(ifp, bno, &lastx);
	if (lastx > 0) {
		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, lastx - 1), prevp);
	}
	if (lastx < (ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t))) {
		xfs_bmbt_get_all(ep, gotp);
		*eofp = 0;
	} else {
		if (lastx > 0) {
			*gotp = *prevp;
		}
		*eofp = 1;
		ep = NULL;
	}
	*lastxp = lastx;
	return ep;
L
Linus Torvalds 已提交
1419 1420
}

1421
/*
1422 1423 1424 1425 1426
 * Search the extents list for the inode, for the extent containing bno.
 * If bno lies in a hole, point to the next entry.  If bno lies past eof,
 * *eofp will be set, and *prevp will contain the last entry (null if none).
 * Else, *lastxp will be set to the index of the found
 * entry; *gotp will contain the entry.
1427
 */
1428 1429 1430 1431 1432 1433 1434 1435 1436
STATIC xfs_bmbt_rec_host_t *                 /* pointer to found extent entry */
xfs_bmap_search_extents(
	xfs_inode_t     *ip,            /* incore inode pointer */
	xfs_fileoff_t   bno,            /* block number searched for */
	int             fork,      	/* data or attr fork */
	int             *eofp,          /* out: end of file found */
	xfs_extnum_t    *lastxp,        /* out: last extent index */
	xfs_bmbt_irec_t *gotp,          /* out: extent entry found */
	xfs_bmbt_irec_t *prevp)         /* out: previous extent entry found */
1437
{
1438 1439
	xfs_ifork_t	*ifp;		/* inode fork pointer */
	xfs_bmbt_rec_host_t  *ep;            /* extent record pointer */
1440

1441
	XFS_STATS_INC(ip->i_mount, xs_look_exlist);
1442
	ifp = XFS_IFORK_PTR(ip, fork);
1443

1444
	ep = xfs_bmap_search_multi_extents(ifp, bno, eofp, lastxp, gotp, prevp);
1445

1446 1447 1448 1449 1450
	if (unlikely(!(gotp->br_startblock) && (*lastxp != NULLEXTNUM) &&
		     !(XFS_IS_REALTIME_INODE(ip) && fork == XFS_DATA_FORK))) {
		xfs_alert_tag(ip->i_mount, XFS_PTAG_FSBLOCK_ZERO,
				"Access to block zero in inode %llu "
				"start_block: %llx start_off: %llx "
1451
				"blkcnt: %llx extent-state: %x lastx: %x",
1452 1453 1454 1455 1456 1457 1458 1459
			(unsigned long long)ip->i_ino,
			(unsigned long long)gotp->br_startblock,
			(unsigned long long)gotp->br_startoff,
			(unsigned long long)gotp->br_blockcount,
			gotp->br_state, *lastxp);
		*lastxp = NULLEXTNUM;
		*eofp = 1;
		return NULL;
1460
	}
1461 1462
	return ep;
}
1463

1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493
/*
 * Returns the file-relative block number of the first unused block(s)
 * in the file with at least "len" logically contiguous blocks free.
 * This is the lowest-address hole if the file has holes, else the first block
 * past the end of file.
 * Return 0 if the file is currently local (in-inode).
 */
int						/* error */
xfs_bmap_first_unused(
	xfs_trans_t	*tp,			/* transaction pointer */
	xfs_inode_t	*ip,			/* incore inode */
	xfs_extlen_t	len,			/* size of hole to find */
	xfs_fileoff_t	*first_unused,		/* unused block */
	int		whichfork)		/* data or attr fork */
{
	int		error;			/* error return value */
	int		idx;			/* extent record index */
	xfs_ifork_t	*ifp;			/* inode fork pointer */
	xfs_fileoff_t	lastaddr;		/* last block number seen */
	xfs_fileoff_t	lowest;			/* lowest useful block */
	xfs_fileoff_t	max;			/* starting useful block */
	xfs_fileoff_t	off;			/* offset for this block */
	xfs_extnum_t	nextents;		/* number of extent entries */

	ASSERT(XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_BTREE ||
	       XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_EXTENTS ||
	       XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL);
	if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL) {
		*first_unused = 0;
		return 0;
1494
	}
1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512
	ifp = XFS_IFORK_PTR(ip, whichfork);
	if (!(ifp->if_flags & XFS_IFEXTENTS) &&
	    (error = xfs_iread_extents(tp, ip, whichfork)))
		return error;
	lowest = *first_unused;
	nextents = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t);
	for (idx = 0, lastaddr = 0, max = lowest; idx < nextents; idx++) {
		xfs_bmbt_rec_host_t *ep = xfs_iext_get_ext(ifp, idx);
		off = xfs_bmbt_get_startoff(ep);
		/*
		 * See if the hole before this extent will work.
		 */
		if (off >= lowest + len && off - max >= len) {
			*first_unused = max;
			return 0;
		}
		lastaddr = off + xfs_bmbt_get_blockcount(ep);
		max = XFS_FILEOFF_MAX(lastaddr, lowest);
1513
	}
1514 1515 1516 1517 1518
	*first_unused = max;
	return 0;
}

/*
1519
 * Returns the file-relative block number of the last block - 1 before
1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542
 * last_block (input value) in the file.
 * This is not based on i_size, it is based on the extent records.
 * Returns 0 for local files, as they do not have extent records.
 */
int						/* error */
xfs_bmap_last_before(
	xfs_trans_t	*tp,			/* transaction pointer */
	xfs_inode_t	*ip,			/* incore inode */
	xfs_fileoff_t	*last_block,		/* last block */
	int		whichfork)		/* data or attr fork */
{
	xfs_fileoff_t	bno;			/* input file offset */
	int		eof;			/* hit end of file */
	xfs_bmbt_rec_host_t *ep;		/* pointer to last extent */
	int		error;			/* error return value */
	xfs_bmbt_irec_t	got;			/* current extent value */
	xfs_ifork_t	*ifp;			/* inode fork pointer */
	xfs_extnum_t	lastx;			/* last extent used */
	xfs_bmbt_irec_t	prev;			/* previous extent value */

	if (XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE &&
	    XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS &&
	    XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_LOCAL)
D
Dave Chinner 已提交
1543
	       return -EIO;
1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557
	if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL) {
		*last_block = 0;
		return 0;
	}
	ifp = XFS_IFORK_PTR(ip, whichfork);
	if (!(ifp->if_flags & XFS_IFEXTENTS) &&
	    (error = xfs_iread_extents(tp, ip, whichfork)))
		return error;
	bno = *last_block - 1;
	ep = xfs_bmap_search_extents(ip, bno, whichfork, &eof, &lastx, &got,
		&prev);
	if (eof || xfs_bmbt_get_startoff(ep) > bno) {
		if (prev.br_startoff == NULLFILEOFF)
			*last_block = 0;
1558
		else
1559
			*last_block = prev.br_startoff + prev.br_blockcount;
1560 1561
	}
	/*
1562
	 * Otherwise *last_block is already the right answer.
1563
	 */
1564 1565 1566
	return 0;
}

D
Dave Chinner 已提交
1567
int
1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582
xfs_bmap_last_extent(
	struct xfs_trans	*tp,
	struct xfs_inode	*ip,
	int			whichfork,
	struct xfs_bmbt_irec	*rec,
	int			*is_empty)
{
	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
	int			error;
	int			nextents;

	if (!(ifp->if_flags & XFS_IFEXTENTS)) {
		error = xfs_iread_extents(tp, ip, whichfork);
		if (error)
			return error;
1583 1584
	}

1585 1586 1587 1588 1589
	nextents = ifp->if_bytes / sizeof(xfs_bmbt_rec_t);
	if (nextents == 0) {
		*is_empty = 1;
		return 0;
	}
1590

1591 1592
	xfs_bmbt_get_all(xfs_iext_get_ext(ifp, nextents - 1), rec);
	*is_empty = 0;
1593 1594 1595
	return 0;
}

1596 1597 1598 1599 1600 1601
/*
 * Check the last inode extent to determine whether this allocation will result
 * in blocks being allocated at the end of the file. When we allocate new data
 * blocks at the end of the file which do not start at the previous data block,
 * we will try to align the new blocks at stripe unit boundaries.
 *
1602
 * Returns 1 in bma->aeof if the file (fork) is empty as any new write will be
1603 1604 1605 1606 1607 1608
 * at, or past the EOF.
 */
STATIC int
xfs_bmap_isaeof(
	struct xfs_bmalloca	*bma,
	int			whichfork)
L
Linus Torvalds 已提交
1609
{
1610 1611 1612
	struct xfs_bmbt_irec	rec;
	int			is_empty;
	int			error;
L
Linus Torvalds 已提交
1613

1614 1615 1616
	bma->aeof = 0;
	error = xfs_bmap_last_extent(NULL, bma->ip, whichfork, &rec,
				     &is_empty);
1617
	if (error)
1618
		return error;
L
Linus Torvalds 已提交
1619

1620 1621 1622 1623 1624
	if (is_empty) {
		bma->aeof = 1;
		return 0;
	}

L
Linus Torvalds 已提交
1625
	/*
1626 1627
	 * Check if we are allocation or past the last extent, or at least into
	 * the last delayed allocated extent.
L
Linus Torvalds 已提交
1628
	 */
1629 1630 1631 1632 1633
	bma->aeof = bma->offset >= rec.br_startoff + rec.br_blockcount ||
		(bma->offset >= rec.br_startoff &&
		 isnullstartblock(rec.br_startblock));
	return 0;
}
L
Linus Torvalds 已提交
1634

1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648
/*
 * Returns the file-relative block number of the first block past eof in
 * the file.  This is not based on i_size, it is based on the extent records.
 * Returns 0 for local files, as they do not have extent records.
 */
int
xfs_bmap_last_offset(
	struct xfs_inode	*ip,
	xfs_fileoff_t		*last_block,
	int			whichfork)
{
	struct xfs_bmbt_irec	rec;
	int			is_empty;
	int			error;
1649

1650
	*last_block = 0;
1651

1652 1653
	if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL)
		return 0;
1654

1655 1656
	if (XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE &&
	    XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS)
D
Dave Chinner 已提交
1657
	       return -EIO;
1658

1659 1660
	error = xfs_bmap_last_extent(NULL, ip, whichfork, &rec, &is_empty);
	if (error || is_empty)
1661
		return error;
1662 1663

	*last_block = rec.br_startoff + rec.br_blockcount;
1664 1665 1666
	return 0;
}

1667 1668 1669 1670 1671 1672 1673 1674 1675
/*
 * Returns whether the selected fork of the inode has exactly one
 * block or not.  For the data fork we check this matches di_size,
 * implying the file's range is 0..bsize-1.
 */
int					/* 1=>1 block, 0=>otherwise */
xfs_bmap_one_block(
	xfs_inode_t	*ip,		/* incore inode */
	int		whichfork)	/* data or attr fork */
C
Christoph Hellwig 已提交
1676
{
1677 1678 1679 1680
	xfs_bmbt_rec_host_t *ep;	/* ptr to fork's extent */
	xfs_ifork_t	*ifp;		/* inode fork pointer */
	int		rval;		/* return value */
	xfs_bmbt_irec_t	s;		/* internal version of extent */
C
Christoph Hellwig 已提交
1681

1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698
#ifndef DEBUG
	if (whichfork == XFS_DATA_FORK)
		return XFS_ISIZE(ip) == ip->i_mount->m_sb.sb_blocksize;
#endif	/* !DEBUG */
	if (XFS_IFORK_NEXTENTS(ip, whichfork) != 1)
		return 0;
	if (XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS)
		return 0;
	ifp = XFS_IFORK_PTR(ip, whichfork);
	ASSERT(ifp->if_flags & XFS_IFEXTENTS);
	ep = xfs_iext_get_ext(ifp, 0);
	xfs_bmbt_get_all(ep, &s);
	rval = s.br_startoff == 0 && s.br_blockcount == 1;
	if (rval && whichfork == XFS_DATA_FORK)
		ASSERT(XFS_ISIZE(ip) == ip->i_mount->m_sb.sb_blocksize);
	return rval;
}
C
Christoph Hellwig 已提交
1699

1700 1701 1702
/*
 * Extent tree manipulation functions used during allocation.
 */
C
Christoph Hellwig 已提交
1703

1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726
/*
 * Convert a delayed allocation to a real allocation.
 */
STATIC int				/* error */
xfs_bmap_add_extent_delay_real(
	struct xfs_bmalloca	*bma)
{
	struct xfs_bmbt_irec	*new = &bma->got;
	int			diff;	/* temp value */
	xfs_bmbt_rec_host_t	*ep;	/* extent entry for idx */
	int			error;	/* error return value */
	int			i;	/* temp state */
	xfs_ifork_t		*ifp;	/* inode fork pointer */
	xfs_fileoff_t		new_endoff;	/* end offset of new entry */
	xfs_bmbt_irec_t		r[3];	/* neighbor extent entries */
					/* left is 0, right is 1, prev is 2 */
	int			rval=0;	/* return value (logging flags) */
	int			state = 0;/* state bits, accessed thru macros */
	xfs_filblks_t		da_new; /* new count del alloc blocks used */
	xfs_filblks_t		da_old; /* old count del alloc blocks used */
	xfs_filblks_t		temp=0;	/* value for da_new calculations */
	xfs_filblks_t		temp2=0;/* value for da_new calculations */
	int			tmp_rval;	/* partial logging flags */
1727
	int			whichfork = XFS_DATA_FORK;
1728
	struct xfs_mount	*mp;
C
Christoph Hellwig 已提交
1729

1730
	mp = bma->ip->i_mount;
1731
	ifp = XFS_IFORK_PTR(bma->ip, whichfork);
C
Christoph Hellwig 已提交
1732

1733 1734 1735 1736 1737
	ASSERT(bma->idx >= 0);
	ASSERT(bma->idx <= ifp->if_bytes / sizeof(struct xfs_bmbt_rec));
	ASSERT(!isnullstartblock(new->br_startblock));
	ASSERT(!bma->cur ||
	       (bma->cur->bc_private.b.flags & XFS_BTCUR_BPRV_WASDEL));
C
Christoph Hellwig 已提交
1738

1739
	XFS_STATS_INC(mp, xs_add_exlist);
C
Christoph Hellwig 已提交
1740

1741 1742 1743
#define	LEFT		r[0]
#define	RIGHT		r[1]
#define	PREV		r[2]
C
Christoph Hellwig 已提交
1744 1745

	/*
1746
	 * Set up a bunch of variables to make the tests simpler.
C
Christoph Hellwig 已提交
1747
	 */
1748 1749 1750 1751 1752 1753 1754 1755 1756
	ep = xfs_iext_get_ext(ifp, bma->idx);
	xfs_bmbt_get_all(ep, &PREV);
	new_endoff = new->br_startoff + new->br_blockcount;
	ASSERT(PREV.br_startoff <= new->br_startoff);
	ASSERT(PREV.br_startoff + PREV.br_blockcount >= new_endoff);

	da_old = startblockval(PREV.br_startblock);
	da_new = 0;

C
Christoph Hellwig 已提交
1757
	/*
1758 1759
	 * Set flags determining what part of the previous delayed allocation
	 * extent is being replaced by a real allocation.
C
Christoph Hellwig 已提交
1760
	 */
1761 1762 1763 1764
	if (PREV.br_startoff == new->br_startoff)
		state |= BMAP_LEFT_FILLING;
	if (PREV.br_startoff + PREV.br_blockcount == new_endoff)
		state |= BMAP_RIGHT_FILLING;
C
Christoph Hellwig 已提交
1765 1766

	/*
1767 1768
	 * Check and set flags if this segment has a left neighbor.
	 * Don't set contiguous if the combined extent would be too large.
C
Christoph Hellwig 已提交
1769
	 */
1770 1771 1772
	if (bma->idx > 0) {
		state |= BMAP_LEFT_VALID;
		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx - 1), &LEFT);
1773

1774 1775
		if (isnullstartblock(LEFT.br_startblock))
			state |= BMAP_LEFT_DELAY;
1776 1777
	}

1778 1779 1780 1781 1782 1783
	if ((state & BMAP_LEFT_VALID) && !(state & BMAP_LEFT_DELAY) &&
	    LEFT.br_startoff + LEFT.br_blockcount == new->br_startoff &&
	    LEFT.br_startblock + LEFT.br_blockcount == new->br_startblock &&
	    LEFT.br_state == new->br_state &&
	    LEFT.br_blockcount + new->br_blockcount <= MAXEXTLEN)
		state |= BMAP_LEFT_CONTIG;
1784

L
Linus Torvalds 已提交
1785
	/*
1786 1787 1788
	 * Check and set flags if this segment has a right neighbor.
	 * Don't set contiguous if the combined extent would be too large.
	 * Also check for all-three-contiguous being too large.
L
Linus Torvalds 已提交
1789
	 */
1790
	if (bma->idx < ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t) - 1) {
1791 1792
		state |= BMAP_RIGHT_VALID;
		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx + 1), &RIGHT);
1793

1794 1795
		if (isnullstartblock(RIGHT.br_startblock))
			state |= BMAP_RIGHT_DELAY;
L
Linus Torvalds 已提交
1796
	}
1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811

	if ((state & BMAP_RIGHT_VALID) && !(state & BMAP_RIGHT_DELAY) &&
	    new_endoff == RIGHT.br_startoff &&
	    new->br_startblock + new->br_blockcount == RIGHT.br_startblock &&
	    new->br_state == RIGHT.br_state &&
	    new->br_blockcount + RIGHT.br_blockcount <= MAXEXTLEN &&
	    ((state & (BMAP_LEFT_CONTIG | BMAP_LEFT_FILLING |
		       BMAP_RIGHT_FILLING)) !=
		      (BMAP_LEFT_CONTIG | BMAP_LEFT_FILLING |
		       BMAP_RIGHT_FILLING) ||
	     LEFT.br_blockcount + new->br_blockcount + RIGHT.br_blockcount
			<= MAXEXTLEN))
		state |= BMAP_RIGHT_CONTIG;

	error = 0;
L
Linus Torvalds 已提交
1812
	/*
1813
	 * Switch out based on the FILLING and CONTIG state bits.
L
Linus Torvalds 已提交
1814
	 */
1815 1816 1817 1818
	switch (state & (BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG |
			 BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG)) {
	case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG |
	     BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG:
L
Linus Torvalds 已提交
1819
		/*
1820 1821
		 * Filling in all of a previously delayed allocation extent.
		 * The left and right neighbors are both contiguous with new.
L
Linus Torvalds 已提交
1822
		 */
1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840
		bma->idx--;
		trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
		xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, bma->idx),
			LEFT.br_blockcount + PREV.br_blockcount +
			RIGHT.br_blockcount);
		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);

		xfs_iext_remove(bma->ip, bma->idx + 1, 2, state);
		bma->ip->i_d.di_nextents--;
		if (bma->cur == NULL)
			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
		else {
			rval = XFS_ILOG_CORE;
			error = xfs_bmbt_lookup_eq(bma->cur, RIGHT.br_startoff,
					RIGHT.br_startblock,
					RIGHT.br_blockcount, &i);
			if (error)
				goto done;
1841
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
1842 1843 1844
			error = xfs_btree_delete(bma->cur, &i);
			if (error)
				goto done;
1845
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
1846 1847 1848
			error = xfs_btree_decrement(bma->cur, 0, &i);
			if (error)
				goto done;
1849
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860
			error = xfs_bmbt_update(bma->cur, LEFT.br_startoff,
					LEFT.br_startblock,
					LEFT.br_blockcount +
					PREV.br_blockcount +
					RIGHT.br_blockcount, LEFT.br_state);
			if (error)
				goto done;
		}
		break;

	case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING | BMAP_LEFT_CONTIG:
1861
		/*
1862 1863
		 * Filling in all of a previously delayed allocation extent.
		 * The left neighbor is contiguous, the right is not.
1864
		 */
1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881
		bma->idx--;

		trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
		xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, bma->idx),
			LEFT.br_blockcount + PREV.br_blockcount);
		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);

		xfs_iext_remove(bma->ip, bma->idx + 1, 1, state);
		if (bma->cur == NULL)
			rval = XFS_ILOG_DEXT;
		else {
			rval = 0;
			error = xfs_bmbt_lookup_eq(bma->cur, LEFT.br_startoff,
					LEFT.br_startblock, LEFT.br_blockcount,
					&i);
			if (error)
				goto done;
1882
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
1883 1884 1885 1886 1887 1888 1889 1890 1891 1892
			error = xfs_bmbt_update(bma->cur, LEFT.br_startoff,
					LEFT.br_startblock,
					LEFT.br_blockcount +
					PREV.br_blockcount, LEFT.br_state);
			if (error)
				goto done;
		}
		break;

	case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG:
1893
		/*
1894 1895
		 * Filling in all of a previously delayed allocation extent.
		 * The right neighbor is contiguous, the left is not.
1896
		 */
1897 1898 1899 1900 1901
		trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
		xfs_bmbt_set_startblock(ep, new->br_startblock);
		xfs_bmbt_set_blockcount(ep,
			PREV.br_blockcount + RIGHT.br_blockcount);
		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
1902

1903 1904 1905 1906 1907 1908 1909 1910 1911 1912
		xfs_iext_remove(bma->ip, bma->idx + 1, 1, state);
		if (bma->cur == NULL)
			rval = XFS_ILOG_DEXT;
		else {
			rval = 0;
			error = xfs_bmbt_lookup_eq(bma->cur, RIGHT.br_startoff,
					RIGHT.br_startblock,
					RIGHT.br_blockcount, &i);
			if (error)
				goto done;
1913
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
1914 1915 1916 1917 1918 1919 1920 1921 1922 1923
			error = xfs_bmbt_update(bma->cur, PREV.br_startoff,
					new->br_startblock,
					PREV.br_blockcount +
					RIGHT.br_blockcount, PREV.br_state);
			if (error)
				goto done;
		}
		break;

	case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING:
1924
		/*
1925 1926 1927
		 * Filling in all of a previously delayed allocation extent.
		 * Neither the left nor right neighbors are contiguous with
		 * the new one.
1928
		 */
1929 1930 1931
		trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
		xfs_bmbt_set_startblock(ep, new->br_startblock);
		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
1932

1933 1934 1935 1936 1937 1938 1939 1940 1941 1942
		bma->ip->i_d.di_nextents++;
		if (bma->cur == NULL)
			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
		else {
			rval = XFS_ILOG_CORE;
			error = xfs_bmbt_lookup_eq(bma->cur, new->br_startoff,
					new->br_startblock, new->br_blockcount,
					&i);
			if (error)
				goto done;
1943
			XFS_WANT_CORRUPTED_GOTO(mp, i == 0, done);
1944 1945 1946 1947
			bma->cur->bc_rec.b.br_state = XFS_EXT_NORM;
			error = xfs_btree_insert(bma->cur, &i);
			if (error)
				goto done;
1948
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
1949 1950
		}
		break;
L
Linus Torvalds 已提交
1951

1952
	case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG:
L
Linus Torvalds 已提交
1953
		/*
1954 1955
		 * Filling in the first part of a previous delayed allocation.
		 * The left neighbor is contiguous.
L
Linus Torvalds 已提交
1956
		 */
1957 1958 1959 1960 1961 1962
		trace_xfs_bmap_pre_update(bma->ip, bma->idx - 1, state, _THIS_IP_);
		xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, bma->idx - 1),
			LEFT.br_blockcount + new->br_blockcount);
		xfs_bmbt_set_startoff(ep,
			PREV.br_startoff + new->br_blockcount);
		trace_xfs_bmap_post_update(bma->ip, bma->idx - 1, state, _THIS_IP_);
L
Linus Torvalds 已提交
1963

1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975
		temp = PREV.br_blockcount - new->br_blockcount;
		trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
		xfs_bmbt_set_blockcount(ep, temp);
		if (bma->cur == NULL)
			rval = XFS_ILOG_DEXT;
		else {
			rval = 0;
			error = xfs_bmbt_lookup_eq(bma->cur, LEFT.br_startoff,
					LEFT.br_startblock, LEFT.br_blockcount,
					&i);
			if (error)
				goto done;
1976
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
1977 1978 1979 1980 1981
			error = xfs_bmbt_update(bma->cur, LEFT.br_startoff,
					LEFT.br_startblock,
					LEFT.br_blockcount +
					new->br_blockcount,
					LEFT.br_state);
C
Christoph Hellwig 已提交
1982
			if (error)
L
Linus Torvalds 已提交
1983 1984
				goto done;
		}
1985 1986 1987 1988 1989 1990 1991 1992 1993
		da_new = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(bma->ip, temp),
			startblockval(PREV.br_startblock));
		xfs_bmbt_set_startblock(ep, nullstartblock(da_new));
		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);

		bma->idx--;
		break;

	case BMAP_LEFT_FILLING:
L
Linus Torvalds 已提交
1994
		/*
1995 1996
		 * Filling in the first part of a previous delayed allocation.
		 * The left neighbor is not contiguous.
L
Linus Torvalds 已提交
1997
		 */
1998 1999 2000 2001 2002 2003 2004 2005
		trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
		xfs_bmbt_set_startoff(ep, new_endoff);
		temp = PREV.br_blockcount - new->br_blockcount;
		xfs_bmbt_set_blockcount(ep, temp);
		xfs_iext_insert(bma->ip, bma->idx, 1, new, state);
		bma->ip->i_d.di_nextents++;
		if (bma->cur == NULL)
			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
L
Linus Torvalds 已提交
2006
		else {
2007 2008 2009 2010 2011 2012
			rval = XFS_ILOG_CORE;
			error = xfs_bmbt_lookup_eq(bma->cur, new->br_startoff,
					new->br_startblock, new->br_blockcount,
					&i);
			if (error)
				goto done;
2013
			XFS_WANT_CORRUPTED_GOTO(mp, i == 0, done);
2014 2015 2016
			bma->cur->bc_rec.b.br_state = XFS_EXT_NORM;
			error = xfs_btree_insert(bma->cur, &i);
			if (error)
L
Linus Torvalds 已提交
2017
				goto done;
2018
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
L
Linus Torvalds 已提交
2019
		}
2020

2021
		if (xfs_bmap_needs_btree(bma->ip, whichfork)) {
2022 2023
			error = xfs_bmap_extents_to_btree(bma->tp, bma->ip,
					bma->firstblock, bma->flist,
2024
					&bma->cur, 1, &tmp_rval, whichfork);
2025 2026 2027
			rval |= tmp_rval;
			if (error)
				goto done;
L
Linus Torvalds 已提交
2028
		}
2029 2030 2031 2032 2033 2034
		da_new = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(bma->ip, temp),
			startblockval(PREV.br_startblock) -
			(bma->cur ? bma->cur->bc_private.b.allocated : 0));
		ep = xfs_iext_get_ext(ifp, bma->idx + 1);
		xfs_bmbt_set_startblock(ep, nullstartblock(da_new));
		trace_xfs_bmap_post_update(bma->ip, bma->idx + 1, state, _THIS_IP_);
L
Linus Torvalds 已提交
2035 2036
		break;

2037
	case BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG:
L
Linus Torvalds 已提交
2038
		/*
2039 2040
		 * Filling in the last part of a previous delayed allocation.
		 * The right neighbor is contiguous with the new allocation.
L
Linus Torvalds 已提交
2041
		 */
2042 2043
		temp = PREV.br_blockcount - new->br_blockcount;
		trace_xfs_bmap_pre_update(bma->ip, bma->idx + 1, state, _THIS_IP_);
L
Linus Torvalds 已提交
2044
		xfs_bmbt_set_blockcount(ep, temp);
2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058
		xfs_bmbt_set_allf(xfs_iext_get_ext(ifp, bma->idx + 1),
			new->br_startoff, new->br_startblock,
			new->br_blockcount + RIGHT.br_blockcount,
			RIGHT.br_state);
		trace_xfs_bmap_post_update(bma->ip, bma->idx + 1, state, _THIS_IP_);
		if (bma->cur == NULL)
			rval = XFS_ILOG_DEXT;
		else {
			rval = 0;
			error = xfs_bmbt_lookup_eq(bma->cur, RIGHT.br_startoff,
					RIGHT.br_startblock,
					RIGHT.br_blockcount, &i);
			if (error)
				goto done;
2059
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
2060 2061 2062 2063 2064 2065 2066
			error = xfs_bmbt_update(bma->cur, new->br_startoff,
					new->br_startblock,
					new->br_blockcount +
					RIGHT.br_blockcount,
					RIGHT.br_state);
			if (error)
				goto done;
L
Linus Torvalds 已提交
2067
		}
2068 2069 2070 2071 2072 2073 2074 2075

		da_new = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(bma->ip, temp),
			startblockval(PREV.br_startblock));
		trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
		xfs_bmbt_set_startblock(ep, nullstartblock(da_new));
		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);

		bma->idx++;
L
Linus Torvalds 已提交
2076 2077
		break;

2078
	case BMAP_RIGHT_FILLING:
L
Linus Torvalds 已提交
2079
		/*
2080 2081
		 * Filling in the last part of a previous delayed allocation.
		 * The right neighbor is not contiguous.
L
Linus Torvalds 已提交
2082
		 */
2083 2084
		temp = PREV.br_blockcount - new->br_blockcount;
		trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
L
Linus Torvalds 已提交
2085
		xfs_bmbt_set_blockcount(ep, temp);
2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096
		xfs_iext_insert(bma->ip, bma->idx + 1, 1, new, state);
		bma->ip->i_d.di_nextents++;
		if (bma->cur == NULL)
			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
		else {
			rval = XFS_ILOG_CORE;
			error = xfs_bmbt_lookup_eq(bma->cur, new->br_startoff,
					new->br_startblock, new->br_blockcount,
					&i);
			if (error)
				goto done;
2097
			XFS_WANT_CORRUPTED_GOTO(mp, i == 0, done);
2098 2099 2100 2101
			bma->cur->bc_rec.b.br_state = XFS_EXT_NORM;
			error = xfs_btree_insert(bma->cur, &i);
			if (error)
				goto done;
2102
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
L
Linus Torvalds 已提交
2103
		}
2104

2105
		if (xfs_bmap_needs_btree(bma->ip, whichfork)) {
2106 2107
			error = xfs_bmap_extents_to_btree(bma->tp, bma->ip,
				bma->firstblock, bma->flist, &bma->cur, 1,
2108
				&tmp_rval, whichfork);
2109 2110 2111
			rval |= tmp_rval;
			if (error)
				goto done;
L
Linus Torvalds 已提交
2112
		}
2113 2114 2115 2116 2117 2118 2119 2120
		da_new = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(bma->ip, temp),
			startblockval(PREV.br_startblock) -
			(bma->cur ? bma->cur->bc_private.b.allocated : 0));
		ep = xfs_iext_get_ext(ifp, bma->idx);
		xfs_bmbt_set_startblock(ep, nullstartblock(da_new));
		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);

		bma->idx++;
L
Linus Torvalds 已提交
2121 2122 2123 2124
		break;

	case 0:
		/*
2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142
		 * Filling in the middle part of a previous delayed allocation.
		 * Contiguity is impossible here.
		 * This case is avoided almost all the time.
		 *
		 * We start with a delayed allocation:
		 *
		 * +ddddddddddddddddddddddddddddddddddddddddddddddddddddddd+
		 *  PREV @ idx
		 *
	         * and we are allocating:
		 *                     +rrrrrrrrrrrrrrrrr+
		 *			      new
		 *
		 * and we set it up for insertion as:
		 * +ddddddddddddddddddd+rrrrrrrrrrrrrrrrr+ddddddddddddddddd+
		 *                            new
		 *  PREV @ idx          LEFT              RIGHT
		 *                      inserted at idx + 1
L
Linus Torvalds 已提交
2143
		 */
2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165
		temp = new->br_startoff - PREV.br_startoff;
		temp2 = PREV.br_startoff + PREV.br_blockcount - new_endoff;
		trace_xfs_bmap_pre_update(bma->ip, bma->idx, 0, _THIS_IP_);
		xfs_bmbt_set_blockcount(ep, temp);	/* truncate PREV */
		LEFT = *new;
		RIGHT.br_state = PREV.br_state;
		RIGHT.br_startblock = nullstartblock(
				(int)xfs_bmap_worst_indlen(bma->ip, temp2));
		RIGHT.br_startoff = new_endoff;
		RIGHT.br_blockcount = temp2;
		/* insert LEFT (r[0]) and RIGHT (r[1]) at the same time */
		xfs_iext_insert(bma->ip, bma->idx + 1, 2, &LEFT, state);
		bma->ip->i_d.di_nextents++;
		if (bma->cur == NULL)
			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
		else {
			rval = XFS_ILOG_CORE;
			error = xfs_bmbt_lookup_eq(bma->cur, new->br_startoff,
					new->br_startblock, new->br_blockcount,
					&i);
			if (error)
				goto done;
2166
			XFS_WANT_CORRUPTED_GOTO(mp, i == 0, done);
2167 2168 2169 2170
			bma->cur->bc_rec.b.br_state = XFS_EXT_NORM;
			error = xfs_btree_insert(bma->cur, &i);
			if (error)
				goto done;
2171
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
L
Linus Torvalds 已提交
2172 2173
		}

2174
		if (xfs_bmap_needs_btree(bma->ip, whichfork)) {
2175 2176
			error = xfs_bmap_extents_to_btree(bma->tp, bma->ip,
					bma->firstblock, bma->flist, &bma->cur,
2177
					1, &tmp_rval, whichfork);
2178 2179 2180 2181 2182 2183 2184 2185 2186
			rval |= tmp_rval;
			if (error)
				goto done;
		}
		temp = xfs_bmap_worst_indlen(bma->ip, temp);
		temp2 = xfs_bmap_worst_indlen(bma->ip, temp2);
		diff = (int)(temp + temp2 - startblockval(PREV.br_startblock) -
			(bma->cur ? bma->cur->bc_private.b.allocated : 0));
		if (diff > 0) {
2187 2188
			error = xfs_mod_fdblocks(bma->ip->i_mount,
						 -((int64_t)diff), false);
2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219
			ASSERT(!error);
			if (error)
				goto done;
		}

		ep = xfs_iext_get_ext(ifp, bma->idx);
		xfs_bmbt_set_startblock(ep, nullstartblock((int)temp));
		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
		trace_xfs_bmap_pre_update(bma->ip, bma->idx + 2, state, _THIS_IP_);
		xfs_bmbt_set_startblock(xfs_iext_get_ext(ifp, bma->idx + 2),
			nullstartblock((int)temp2));
		trace_xfs_bmap_post_update(bma->ip, bma->idx + 2, state, _THIS_IP_);

		bma->idx++;
		da_new = temp + temp2;
		break;

	case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
	case BMAP_RIGHT_FILLING | BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
	case BMAP_LEFT_FILLING | BMAP_RIGHT_CONTIG:
	case BMAP_RIGHT_FILLING | BMAP_LEFT_CONTIG:
	case BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
	case BMAP_LEFT_CONTIG:
	case BMAP_RIGHT_CONTIG:
		/*
		 * These cases are all impossible.
		 */
		ASSERT(0);
	}

	/* convert to a btree if necessary */
2220
	if (xfs_bmap_needs_btree(bma->ip, whichfork)) {
2221 2222 2223 2224 2225
		int	tmp_logflags;	/* partial log flag return val */

		ASSERT(bma->cur == NULL);
		error = xfs_bmap_extents_to_btree(bma->tp, bma->ip,
				bma->firstblock, bma->flist, &bma->cur,
2226
				da_old > 0, &tmp_logflags, whichfork);
2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238
		bma->logflags |= tmp_logflags;
		if (error)
			goto done;
	}

	/* adjust for changes in reserved delayed indirect blocks */
	if (da_old || da_new) {
		temp = da_new;
		if (bma->cur)
			temp += bma->cur->bc_private.b.allocated;
		ASSERT(temp <= da_old);
		if (temp < da_old)
2239 2240
			xfs_mod_fdblocks(bma->ip->i_mount,
					(int64_t)(da_old - temp), false);
2241
	}
2242 2243 2244 2245 2246

	/* clear out the allocated field, done with it now in any case. */
	if (bma->cur)
		bma->cur->bc_private.b.allocated = 0;

2247
	xfs_bmap_check_leaf_extents(bma->cur, bma->ip, whichfork);
L
Linus Torvalds 已提交
2248
done:
2249
	bma->logflags |= rval;
L
Linus Torvalds 已提交
2250
	return error;
2251 2252 2253
#undef	LEFT
#undef	RIGHT
#undef	PREV
L
Linus Torvalds 已提交
2254 2255 2256
}

/*
2257
 * Convert an unwritten allocation to a real allocation or vice versa.
L
Linus Torvalds 已提交
2258
 */
2259 2260 2261 2262 2263 2264 2265 2266 2267 2268
STATIC int				/* error */
xfs_bmap_add_extent_unwritten_real(
	struct xfs_trans	*tp,
	xfs_inode_t		*ip,	/* incore inode pointer */
	xfs_extnum_t		*idx,	/* extent number to update/insert */
	xfs_btree_cur_t		**curp,	/* if *curp is null, not a btree */
	xfs_bmbt_irec_t		*new,	/* new data to add to file extents */
	xfs_fsblock_t		*first,	/* pointer to firstblock variable */
	xfs_bmap_free_t		*flist,	/* list of extents to be freed */
	int			*logflagsp) /* inode logging flags */
L
Linus Torvalds 已提交
2269
{
2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281
	xfs_btree_cur_t		*cur;	/* btree cursor */
	xfs_bmbt_rec_host_t	*ep;	/* extent entry for idx */
	int			error;	/* error return value */
	int			i;	/* temp state */
	xfs_ifork_t		*ifp;	/* inode fork pointer */
	xfs_fileoff_t		new_endoff;	/* end offset of new entry */
	xfs_exntst_t		newext;	/* new extent state */
	xfs_exntst_t		oldext;	/* old extent state */
	xfs_bmbt_irec_t		r[3];	/* neighbor extent entries */
					/* left is 0, right is 1, prev is 2 */
	int			rval=0;	/* return value (logging flags) */
	int			state = 0;/* state bits, accessed thru macros */
2282
	struct xfs_mount	*mp = tp->t_mountp;
L
Linus Torvalds 已提交
2283

2284
	*logflagsp = 0;
L
Linus Torvalds 已提交
2285

2286 2287
	cur = *curp;
	ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
2288

2289 2290 2291 2292
	ASSERT(*idx >= 0);
	ASSERT(*idx <= ifp->if_bytes / sizeof(struct xfs_bmbt_rec));
	ASSERT(!isnullstartblock(new->br_startblock));

2293
	XFS_STATS_INC(mp, xs_add_exlist);
2294 2295 2296 2297

#define	LEFT		r[0]
#define	RIGHT		r[1]
#define	PREV		r[2]
2298

L
Linus Torvalds 已提交
2299
	/*
2300
	 * Set up a bunch of variables to make the tests simpler.
L
Linus Torvalds 已提交
2301
	 */
2302 2303 2304 2305 2306 2307 2308 2309 2310 2311
	error = 0;
	ep = xfs_iext_get_ext(ifp, *idx);
	xfs_bmbt_get_all(ep, &PREV);
	newext = new->br_state;
	oldext = (newext == XFS_EXT_UNWRITTEN) ?
		XFS_EXT_NORM : XFS_EXT_UNWRITTEN;
	ASSERT(PREV.br_state == oldext);
	new_endoff = new->br_startoff + new->br_blockcount;
	ASSERT(PREV.br_startoff <= new->br_startoff);
	ASSERT(PREV.br_startoff + PREV.br_blockcount >= new_endoff);
2312

L
Linus Torvalds 已提交
2313
	/*
2314 2315
	 * Set flags determining what part of the previous oldext allocation
	 * extent is being replaced by a newext allocation.
L
Linus Torvalds 已提交
2316
	 */
2317 2318 2319 2320 2321
	if (PREV.br_startoff == new->br_startoff)
		state |= BMAP_LEFT_FILLING;
	if (PREV.br_startoff + PREV.br_blockcount == new_endoff)
		state |= BMAP_RIGHT_FILLING;

L
Linus Torvalds 已提交
2322
	/*
2323 2324
	 * Check and set flags if this segment has a left neighbor.
	 * Don't set contiguous if the combined extent would be too large.
L
Linus Torvalds 已提交
2325
	 */
2326 2327 2328 2329 2330 2331
	if (*idx > 0) {
		state |= BMAP_LEFT_VALID;
		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx - 1), &LEFT);

		if (isnullstartblock(LEFT.br_startblock))
			state |= BMAP_LEFT_DELAY;
L
Linus Torvalds 已提交
2332
	}
2333 2334 2335 2336 2337 2338 2339 2340

	if ((state & BMAP_LEFT_VALID) && !(state & BMAP_LEFT_DELAY) &&
	    LEFT.br_startoff + LEFT.br_blockcount == new->br_startoff &&
	    LEFT.br_startblock + LEFT.br_blockcount == new->br_startblock &&
	    LEFT.br_state == newext &&
	    LEFT.br_blockcount + new->br_blockcount <= MAXEXTLEN)
		state |= BMAP_LEFT_CONTIG;

L
Linus Torvalds 已提交
2341
	/*
2342 2343 2344
	 * Check and set flags if this segment has a right neighbor.
	 * Don't set contiguous if the combined extent would be too large.
	 * Also check for all-three-contiguous being too large.
L
Linus Torvalds 已提交
2345
	 */
2346 2347 2348 2349 2350
	if (*idx < ip->i_df.if_bytes / (uint)sizeof(xfs_bmbt_rec_t) - 1) {
		state |= BMAP_RIGHT_VALID;
		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx + 1), &RIGHT);
		if (isnullstartblock(RIGHT.br_startblock))
			state |= BMAP_RIGHT_DELAY;
L
Linus Torvalds 已提交
2351
	}
2352

2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364
	if ((state & BMAP_RIGHT_VALID) && !(state & BMAP_RIGHT_DELAY) &&
	    new_endoff == RIGHT.br_startoff &&
	    new->br_startblock + new->br_blockcount == RIGHT.br_startblock &&
	    newext == RIGHT.br_state &&
	    new->br_blockcount + RIGHT.br_blockcount <= MAXEXTLEN &&
	    ((state & (BMAP_LEFT_CONTIG | BMAP_LEFT_FILLING |
		       BMAP_RIGHT_FILLING)) !=
		      (BMAP_LEFT_CONTIG | BMAP_LEFT_FILLING |
		       BMAP_RIGHT_FILLING) ||
	     LEFT.br_blockcount + new->br_blockcount + RIGHT.br_blockcount
			<= MAXEXTLEN))
		state |= BMAP_RIGHT_CONTIG;
2365

L
Linus Torvalds 已提交
2366
	/*
2367
	 * Switch out based on the FILLING and CONTIG state bits.
L
Linus Torvalds 已提交
2368
	 */
2369 2370 2371 2372 2373 2374 2375 2376 2377
	switch (state & (BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG |
			 BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG)) {
	case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG |
	     BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG:
		/*
		 * Setting all of a previous oldext extent to newext.
		 * The left and right neighbors are both contiguous with new.
		 */
		--*idx;
C
Christoph Hellwig 已提交
2378

2379 2380 2381 2382 2383
		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
		xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx),
			LEFT.br_blockcount + PREV.br_blockcount +
			RIGHT.br_blockcount);
		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
C
Christoph Hellwig 已提交
2384

2385 2386 2387 2388 2389 2390 2391 2392 2393 2394
		xfs_iext_remove(ip, *idx + 1, 2, state);
		ip->i_d.di_nextents -= 2;
		if (cur == NULL)
			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
		else {
			rval = XFS_ILOG_CORE;
			if ((error = xfs_bmbt_lookup_eq(cur, RIGHT.br_startoff,
					RIGHT.br_startblock,
					RIGHT.br_blockcount, &i)))
				goto done;
2395
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
2396 2397
			if ((error = xfs_btree_delete(cur, &i)))
				goto done;
2398
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
2399 2400
			if ((error = xfs_btree_decrement(cur, 0, &i)))
				goto done;
2401
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
2402 2403
			if ((error = xfs_btree_delete(cur, &i)))
				goto done;
2404
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
2405 2406
			if ((error = xfs_btree_decrement(cur, 0, &i)))
				goto done;
2407
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
2408 2409 2410 2411 2412 2413 2414
			if ((error = xfs_bmbt_update(cur, LEFT.br_startoff,
				LEFT.br_startblock,
				LEFT.br_blockcount + PREV.br_blockcount +
				RIGHT.br_blockcount, LEFT.br_state)))
				goto done;
		}
		break;
C
Christoph Hellwig 已提交
2415

2416 2417 2418 2419 2420 2421
	case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING | BMAP_LEFT_CONTIG:
		/*
		 * Setting all of a previous oldext extent to newext.
		 * The left neighbor is contiguous, the right is not.
		 */
		--*idx;
2422

2423 2424 2425 2426
		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
		xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx),
			LEFT.br_blockcount + PREV.br_blockcount);
		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
L
Linus Torvalds 已提交
2427

2428 2429 2430 2431 2432 2433 2434 2435 2436 2437
		xfs_iext_remove(ip, *idx + 1, 1, state);
		ip->i_d.di_nextents--;
		if (cur == NULL)
			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
		else {
			rval = XFS_ILOG_CORE;
			if ((error = xfs_bmbt_lookup_eq(cur, PREV.br_startoff,
					PREV.br_startblock, PREV.br_blockcount,
					&i)))
				goto done;
2438
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
2439 2440
			if ((error = xfs_btree_delete(cur, &i)))
				goto done;
2441
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
2442 2443
			if ((error = xfs_btree_decrement(cur, 0, &i)))
				goto done;
2444
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
2445 2446 2447 2448 2449 2450 2451
			if ((error = xfs_bmbt_update(cur, LEFT.br_startoff,
				LEFT.br_startblock,
				LEFT.br_blockcount + PREV.br_blockcount,
				LEFT.br_state)))
				goto done;
		}
		break;
L
Linus Torvalds 已提交
2452

2453
	case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG:
L
Linus Torvalds 已提交
2454
		/*
2455 2456
		 * Setting all of a previous oldext extent to newext.
		 * The right neighbor is contiguous, the left is not.
L
Linus Torvalds 已提交
2457
		 */
2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472
		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
		xfs_bmbt_set_blockcount(ep,
			PREV.br_blockcount + RIGHT.br_blockcount);
		xfs_bmbt_set_state(ep, newext);
		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
		xfs_iext_remove(ip, *idx + 1, 1, state);
		ip->i_d.di_nextents--;
		if (cur == NULL)
			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
		else {
			rval = XFS_ILOG_CORE;
			if ((error = xfs_bmbt_lookup_eq(cur, RIGHT.br_startoff,
					RIGHT.br_startblock,
					RIGHT.br_blockcount, &i)))
				goto done;
2473
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
2474 2475
			if ((error = xfs_btree_delete(cur, &i)))
				goto done;
2476
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
2477 2478
			if ((error = xfs_btree_decrement(cur, 0, &i)))
				goto done;
2479
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
2480 2481 2482 2483 2484
			if ((error = xfs_bmbt_update(cur, new->br_startoff,
				new->br_startblock,
				new->br_blockcount + RIGHT.br_blockcount,
				newext)))
				goto done;
L
Linus Torvalds 已提交
2485
		}
2486
		break;
2487

2488 2489 2490 2491 2492 2493 2494 2495 2496
	case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING:
		/*
		 * Setting all of a previous oldext extent to newext.
		 * Neither the left nor right neighbors are contiguous with
		 * the new one.
		 */
		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
		xfs_bmbt_set_state(ep, newext);
		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
2497

2498 2499 2500 2501 2502 2503 2504 2505
		if (cur == NULL)
			rval = XFS_ILOG_DEXT;
		else {
			rval = 0;
			if ((error = xfs_bmbt_lookup_eq(cur, new->br_startoff,
					new->br_startblock, new->br_blockcount,
					&i)))
				goto done;
2506
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
2507 2508 2509 2510 2511 2512
			if ((error = xfs_bmbt_update(cur, new->br_startoff,
				new->br_startblock, new->br_blockcount,
				newext)))
				goto done;
		}
		break;
2513

2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524
	case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG:
		/*
		 * Setting the first part of a previous oldext extent to newext.
		 * The left neighbor is contiguous.
		 */
		trace_xfs_bmap_pre_update(ip, *idx - 1, state, _THIS_IP_);
		xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx - 1),
			LEFT.br_blockcount + new->br_blockcount);
		xfs_bmbt_set_startoff(ep,
			PREV.br_startoff + new->br_blockcount);
		trace_xfs_bmap_post_update(ip, *idx - 1, state, _THIS_IP_);
L
Linus Torvalds 已提交
2525

2526 2527 2528 2529 2530 2531
		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
		xfs_bmbt_set_startblock(ep,
			new->br_startblock + new->br_blockcount);
		xfs_bmbt_set_blockcount(ep,
			PREV.br_blockcount - new->br_blockcount);
		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
2532

2533
		--*idx;
2534

2535 2536 2537 2538 2539 2540 2541 2542
		if (cur == NULL)
			rval = XFS_ILOG_DEXT;
		else {
			rval = 0;
			if ((error = xfs_bmbt_lookup_eq(cur, PREV.br_startoff,
					PREV.br_startblock, PREV.br_blockcount,
					&i)))
				goto done;
2543
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557
			if ((error = xfs_bmbt_update(cur,
				PREV.br_startoff + new->br_blockcount,
				PREV.br_startblock + new->br_blockcount,
				PREV.br_blockcount - new->br_blockcount,
				oldext)))
				goto done;
			if ((error = xfs_btree_decrement(cur, 0, &i)))
				goto done;
			error = xfs_bmbt_update(cur, LEFT.br_startoff,
				LEFT.br_startblock,
				LEFT.br_blockcount + new->br_blockcount,
				LEFT.br_state);
			if (error)
				goto done;
2558
		}
2559
		break;
2560

2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573
	case BMAP_LEFT_FILLING:
		/*
		 * Setting the first part of a previous oldext extent to newext.
		 * The left neighbor is not contiguous.
		 */
		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
		ASSERT(ep && xfs_bmbt_get_state(ep) == oldext);
		xfs_bmbt_set_startoff(ep, new_endoff);
		xfs_bmbt_set_blockcount(ep,
			PREV.br_blockcount - new->br_blockcount);
		xfs_bmbt_set_startblock(ep,
			new->br_startblock + new->br_blockcount);
		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
L
Linus Torvalds 已提交
2574

2575 2576 2577 2578 2579 2580 2581 2582 2583 2584
		xfs_iext_insert(ip, *idx, 1, new, state);
		ip->i_d.di_nextents++;
		if (cur == NULL)
			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
		else {
			rval = XFS_ILOG_CORE;
			if ((error = xfs_bmbt_lookup_eq(cur, PREV.br_startoff,
					PREV.br_startblock, PREV.br_blockcount,
					&i)))
				goto done;
2585
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
2586 2587 2588 2589 2590 2591 2592 2593 2594
			if ((error = xfs_bmbt_update(cur,
				PREV.br_startoff + new->br_blockcount,
				PREV.br_startblock + new->br_blockcount,
				PREV.br_blockcount - new->br_blockcount,
				oldext)))
				goto done;
			cur->bc_rec.b = *new;
			if ((error = xfs_btree_insert(cur, &i)))
				goto done;
2595
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
2596 2597
		}
		break;
L
Linus Torvalds 已提交
2598

2599 2600 2601 2602 2603 2604 2605 2606 2607
	case BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG:
		/*
		 * Setting the last part of a previous oldext extent to newext.
		 * The right neighbor is contiguous with the new allocation.
		 */
		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
		xfs_bmbt_set_blockcount(ep,
			PREV.br_blockcount - new->br_blockcount);
		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
2608

2609
		++*idx;
L
Linus Torvalds 已提交
2610

2611 2612 2613 2614 2615
		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
		xfs_bmbt_set_allf(xfs_iext_get_ext(ifp, *idx),
			new->br_startoff, new->br_startblock,
			new->br_blockcount + RIGHT.br_blockcount, newext);
		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
L
Linus Torvalds 已提交
2616

2617 2618 2619 2620 2621 2622 2623 2624
		if (cur == NULL)
			rval = XFS_ILOG_DEXT;
		else {
			rval = 0;
			if ((error = xfs_bmbt_lookup_eq(cur, PREV.br_startoff,
					PREV.br_startblock,
					PREV.br_blockcount, &i)))
				goto done;
2625
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639
			if ((error = xfs_bmbt_update(cur, PREV.br_startoff,
				PREV.br_startblock,
				PREV.br_blockcount - new->br_blockcount,
				oldext)))
				goto done;
			if ((error = xfs_btree_increment(cur, 0, &i)))
				goto done;
			if ((error = xfs_bmbt_update(cur, new->br_startoff,
				new->br_startblock,
				new->br_blockcount + RIGHT.br_blockcount,
				newext)))
				goto done;
		}
		break;
L
Linus Torvalds 已提交
2640

2641 2642 2643 2644 2645 2646 2647 2648 2649
	case BMAP_RIGHT_FILLING:
		/*
		 * Setting the last part of a previous oldext extent to newext.
		 * The right neighbor is not contiguous.
		 */
		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
		xfs_bmbt_set_blockcount(ep,
			PREV.br_blockcount - new->br_blockcount);
		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
L
Linus Torvalds 已提交
2650

2651 2652
		++*idx;
		xfs_iext_insert(ip, *idx, 1, new, state);
2653

2654 2655 2656 2657 2658 2659 2660 2661 2662
		ip->i_d.di_nextents++;
		if (cur == NULL)
			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
		else {
			rval = XFS_ILOG_CORE;
			if ((error = xfs_bmbt_lookup_eq(cur, PREV.br_startoff,
					PREV.br_startblock, PREV.br_blockcount,
					&i)))
				goto done;
2663
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
2664 2665 2666 2667 2668 2669 2670 2671 2672
			if ((error = xfs_bmbt_update(cur, PREV.br_startoff,
				PREV.br_startblock,
				PREV.br_blockcount - new->br_blockcount,
				oldext)))
				goto done;
			if ((error = xfs_bmbt_lookup_eq(cur, new->br_startoff,
					new->br_startblock, new->br_blockcount,
					&i)))
				goto done;
2673
			XFS_WANT_CORRUPTED_GOTO(mp, i == 0, done);
2674 2675 2676
			cur->bc_rec.b.br_state = XFS_EXT_NORM;
			if ((error = xfs_btree_insert(cur, &i)))
				goto done;
2677
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
2678 2679 2680 2681
		}
		break;

	case 0:
L
Linus Torvalds 已提交
2682
		/*
2683 2684 2685
		 * Setting the middle part of a previous oldext extent to
		 * newext.  Contiguity is impossible here.
		 * One extent becomes three extents.
L
Linus Torvalds 已提交
2686
		 */
2687 2688 2689 2690
		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
		xfs_bmbt_set_blockcount(ep,
			new->br_startoff - PREV.br_startoff);
		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
2691

2692 2693 2694 2695 2696 2697
		r[0] = *new;
		r[1].br_startoff = new_endoff;
		r[1].br_blockcount =
			PREV.br_startoff + PREV.br_blockcount - new_endoff;
		r[1].br_startblock = new->br_startblock + new->br_blockcount;
		r[1].br_state = oldext;
2698

2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710
		++*idx;
		xfs_iext_insert(ip, *idx, 2, &r[0], state);

		ip->i_d.di_nextents += 2;
		if (cur == NULL)
			rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
		else {
			rval = XFS_ILOG_CORE;
			if ((error = xfs_bmbt_lookup_eq(cur, PREV.br_startoff,
					PREV.br_startblock, PREV.br_blockcount,
					&i)))
				goto done;
2711
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722
			/* new right extent - oldext */
			if ((error = xfs_bmbt_update(cur, r[1].br_startoff,
				r[1].br_startblock, r[1].br_blockcount,
				r[1].br_state)))
				goto done;
			/* new left extent - oldext */
			cur->bc_rec.b = PREV;
			cur->bc_rec.b.br_blockcount =
				new->br_startoff - PREV.br_startoff;
			if ((error = xfs_btree_insert(cur, &i)))
				goto done;
2723
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
2724 2725 2726 2727 2728 2729 2730 2731 2732
			/*
			 * Reset the cursor to the position of the new extent
			 * we are about to insert as we can't trust it after
			 * the previous insert.
			 */
			if ((error = xfs_bmbt_lookup_eq(cur, new->br_startoff,
					new->br_startblock, new->br_blockcount,
					&i)))
				goto done;
2733
			XFS_WANT_CORRUPTED_GOTO(mp, i == 0, done);
2734 2735 2736 2737
			/* new middle extent - newext */
			cur->bc_rec.b.br_state = new->br_state;
			if ((error = xfs_btree_insert(cur, &i)))
				goto done;
2738
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
2739
		}
L
Linus Torvalds 已提交
2740
		break;
2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751

	case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
	case BMAP_RIGHT_FILLING | BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
	case BMAP_LEFT_FILLING | BMAP_RIGHT_CONTIG:
	case BMAP_RIGHT_FILLING | BMAP_LEFT_CONTIG:
	case BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
	case BMAP_LEFT_CONTIG:
	case BMAP_RIGHT_CONTIG:
		/*
		 * These cases are all impossible.
		 */
L
Linus Torvalds 已提交
2752 2753
		ASSERT(0);
	}
2754

2755 2756 2757 2758 2759 2760 2761 2762 2763 2764
	/* convert to a btree if necessary */
	if (xfs_bmap_needs_btree(ip, XFS_DATA_FORK)) {
		int	tmp_logflags;	/* partial log flag return val */

		ASSERT(cur == NULL);
		error = xfs_bmap_extents_to_btree(tp, ip, first, flist, &cur,
				0, &tmp_logflags, XFS_DATA_FORK);
		*logflagsp |= tmp_logflags;
		if (error)
			goto done;
L
Linus Torvalds 已提交
2765
	}
2766

2767 2768 2769 2770
	/* clear out the allocated field, done with it now in any case. */
	if (cur) {
		cur->bc_private.b.allocated = 0;
		*curp = cur;
L
Linus Torvalds 已提交
2771
	}
2772

2773 2774 2775
	xfs_bmap_check_leaf_extents(*curp, ip, XFS_DATA_FORK);
done:
	*logflagsp |= rval;
L
Linus Torvalds 已提交
2776
	return error;
2777 2778 2779
#undef	LEFT
#undef	RIGHT
#undef	PREV
L
Linus Torvalds 已提交
2780 2781 2782
}

/*
2783
 * Convert a hole to a delayed allocation.
L
Linus Torvalds 已提交
2784
 */
2785 2786 2787 2788 2789
STATIC void
xfs_bmap_add_extent_hole_delay(
	xfs_inode_t		*ip,	/* incore inode pointer */
	xfs_extnum_t		*idx,	/* extent number to update/insert */
	xfs_bmbt_irec_t		*new)	/* new data to add to file extents */
L
Linus Torvalds 已提交
2790
{
2791 2792 2793 2794 2795 2796 2797
	xfs_ifork_t		*ifp;	/* inode fork pointer */
	xfs_bmbt_irec_t		left;	/* left neighbor extent entry */
	xfs_filblks_t		newlen=0;	/* new indirect size */
	xfs_filblks_t		oldlen=0;	/* old indirect size */
	xfs_bmbt_irec_t		right;	/* right neighbor extent entry */
	int			state;  /* state bits, accessed thru macros */
	xfs_filblks_t		temp=0;	/* temp for indirect calculations */
L
Linus Torvalds 已提交
2798

2799 2800 2801
	ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
	state = 0;
	ASSERT(isnullstartblock(new->br_startblock));
L
Linus Torvalds 已提交
2802 2803

	/*
2804
	 * Check and set flags if this segment has a left neighbor
L
Linus Torvalds 已提交
2805
	 */
2806 2807 2808 2809 2810 2811
	if (*idx > 0) {
		state |= BMAP_LEFT_VALID;
		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx - 1), &left);

		if (isnullstartblock(left.br_startblock))
			state |= BMAP_LEFT_DELAY;
L
Linus Torvalds 已提交
2812 2813
	}

2814 2815 2816 2817 2818 2819 2820
	/*
	 * Check and set flags if the current (right) segment exists.
	 * If it doesn't exist, we're converting the hole at end-of-file.
	 */
	if (*idx < ip->i_df.if_bytes / (uint)sizeof(xfs_bmbt_rec_t)) {
		state |= BMAP_RIGHT_VALID;
		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx), &right);
L
Linus Torvalds 已提交
2821

2822 2823
		if (isnullstartblock(right.br_startblock))
			state |= BMAP_RIGHT_DELAY;
L
Linus Torvalds 已提交
2824
	}
2825

L
Linus Torvalds 已提交
2826
	/*
2827 2828
	 * Set contiguity flags on the left and right neighbors.
	 * Don't let extents get too large, even if the pieces are contiguous.
L
Linus Torvalds 已提交
2829
	 */
2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841
	if ((state & BMAP_LEFT_VALID) && (state & BMAP_LEFT_DELAY) &&
	    left.br_startoff + left.br_blockcount == new->br_startoff &&
	    left.br_blockcount + new->br_blockcount <= MAXEXTLEN)
		state |= BMAP_LEFT_CONTIG;

	if ((state & BMAP_RIGHT_VALID) && (state & BMAP_RIGHT_DELAY) &&
	    new->br_startoff + new->br_blockcount == right.br_startoff &&
	    new->br_blockcount + right.br_blockcount <= MAXEXTLEN &&
	    (!(state & BMAP_LEFT_CONTIG) ||
	     (left.br_blockcount + new->br_blockcount +
	      right.br_blockcount <= MAXEXTLEN)))
		state |= BMAP_RIGHT_CONTIG;
2842 2843

	/*
2844
	 * Switch out based on the contiguity flags.
2845
	 */
2846 2847 2848 2849 2850 2851 2852 2853 2854 2855
	switch (state & (BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG)) {
	case BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
		/*
		 * New allocation is contiguous with delayed allocations
		 * on the left and on the right.
		 * Merge all three into a single extent record.
		 */
		--*idx;
		temp = left.br_blockcount + new->br_blockcount +
			right.br_blockcount;
2856

2857 2858 2859 2860 2861 2862 2863 2864 2865
		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
		xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx), temp);
		oldlen = startblockval(left.br_startblock) +
			startblockval(new->br_startblock) +
			startblockval(right.br_startblock);
		newlen = xfs_bmap_worst_indlen(ip, temp);
		xfs_bmbt_set_startblock(xfs_iext_get_ext(ifp, *idx),
			nullstartblock((int)newlen));
		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
L
Linus Torvalds 已提交
2866

2867 2868
		xfs_iext_remove(ip, *idx + 1, 1, state);
		break;
L
Linus Torvalds 已提交
2869

2870 2871 2872 2873 2874 2875 2876 2877
	case BMAP_LEFT_CONTIG:
		/*
		 * New allocation is contiguous with a delayed allocation
		 * on the left.
		 * Merge the new allocation with the left neighbor.
		 */
		--*idx;
		temp = left.br_blockcount + new->br_blockcount;
L
Linus Torvalds 已提交
2878

2879 2880 2881 2882 2883 2884 2885 2886 2887
		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
		xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx), temp);
		oldlen = startblockval(left.br_startblock) +
			startblockval(new->br_startblock);
		newlen = xfs_bmap_worst_indlen(ip, temp);
		xfs_bmbt_set_startblock(xfs_iext_get_ext(ifp, *idx),
			nullstartblock((int)newlen));
		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
		break;
L
Linus Torvalds 已提交
2888

2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914
	case BMAP_RIGHT_CONTIG:
		/*
		 * New allocation is contiguous with a delayed allocation
		 * on the right.
		 * Merge the new allocation with the right neighbor.
		 */
		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
		temp = new->br_blockcount + right.br_blockcount;
		oldlen = startblockval(new->br_startblock) +
			startblockval(right.br_startblock);
		newlen = xfs_bmap_worst_indlen(ip, temp);
		xfs_bmbt_set_allf(xfs_iext_get_ext(ifp, *idx),
			new->br_startoff,
			nullstartblock((int)newlen), temp, right.br_state);
		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
		break;

	case 0:
		/*
		 * New allocation is not contiguous with another
		 * delayed allocation.
		 * Insert a new entry.
		 */
		oldlen = newlen = 0;
		xfs_iext_insert(ip, *idx, 1, new, state);
		break;
L
Linus Torvalds 已提交
2915
	}
2916 2917
	if (oldlen != newlen) {
		ASSERT(oldlen > newlen);
2918 2919
		xfs_mod_fdblocks(ip->i_mount, (int64_t)(oldlen - newlen),
				 false);
L
Linus Torvalds 已提交
2920
		/*
2921
		 * Nothing to do for disk quota accounting here.
L
Linus Torvalds 已提交
2922 2923 2924 2925 2926
		 */
	}
}

/*
2927
 * Convert a hole to a real allocation.
L
Linus Torvalds 已提交
2928
 */
2929 2930 2931 2932
STATIC int				/* error */
xfs_bmap_add_extent_hole_real(
	struct xfs_bmalloca	*bma,
	int			whichfork)
2933
{
2934 2935 2936 2937 2938 2939 2940 2941
	struct xfs_bmbt_irec	*new = &bma->got;
	int			error;	/* error return value */
	int			i;	/* temp state */
	xfs_ifork_t		*ifp;	/* inode fork pointer */
	xfs_bmbt_irec_t		left;	/* left neighbor extent entry */
	xfs_bmbt_irec_t		right;	/* right neighbor extent entry */
	int			rval=0;	/* return value (logging flags) */
	int			state;	/* state bits, accessed thru macros */
2942
	struct xfs_mount	*mp;
2943

2944
	mp = bma->ip->i_mount;
2945
	ifp = XFS_IFORK_PTR(bma->ip, whichfork);
2946

2947 2948 2949 2950 2951
	ASSERT(bma->idx >= 0);
	ASSERT(bma->idx <= ifp->if_bytes / sizeof(struct xfs_bmbt_rec));
	ASSERT(!isnullstartblock(new->br_startblock));
	ASSERT(!bma->cur ||
	       !(bma->cur->bc_private.b.flags & XFS_BTCUR_BPRV_WASDEL));
2952

2953
	XFS_STATS_INC(mp, xs_add_exlist);
2954

2955 2956 2957
	state = 0;
	if (whichfork == XFS_ATTR_FORK)
		state |= BMAP_ATTRFORK;
2958

2959 2960 2961 2962 2963 2964 2965 2966 2967
	/*
	 * Check and set flags if this segment has a left neighbor.
	 */
	if (bma->idx > 0) {
		state |= BMAP_LEFT_VALID;
		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx - 1), &left);
		if (isnullstartblock(left.br_startblock))
			state |= BMAP_LEFT_DELAY;
	}
2968 2969

	/*
2970 2971
	 * Check and set flags if this segment has a current value.
	 * Not true if we're inserting into the "hole" at eof.
2972
	 */
2973 2974 2975 2976 2977 2978
	if (bma->idx < ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t)) {
		state |= BMAP_RIGHT_VALID;
		xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx), &right);
		if (isnullstartblock(right.br_startblock))
			state |= BMAP_RIGHT_DELAY;
	}
2979

2980 2981 2982 2983 2984 2985 2986 2987 2988 2989
	/*
	 * We're inserting a real allocation between "left" and "right".
	 * Set the contiguity flags.  Don't let extents get too large.
	 */
	if ((state & BMAP_LEFT_VALID) && !(state & BMAP_LEFT_DELAY) &&
	    left.br_startoff + left.br_blockcount == new->br_startoff &&
	    left.br_startblock + left.br_blockcount == new->br_startblock &&
	    left.br_state == new->br_state &&
	    left.br_blockcount + new->br_blockcount <= MAXEXTLEN)
		state |= BMAP_LEFT_CONTIG;
2990

2991 2992 2993 2994 2995 2996 2997 2998 2999
	if ((state & BMAP_RIGHT_VALID) && !(state & BMAP_RIGHT_DELAY) &&
	    new->br_startoff + new->br_blockcount == right.br_startoff &&
	    new->br_startblock + new->br_blockcount == right.br_startblock &&
	    new->br_state == right.br_state &&
	    new->br_blockcount + right.br_blockcount <= MAXEXTLEN &&
	    (!(state & BMAP_LEFT_CONTIG) ||
	     left.br_blockcount + new->br_blockcount +
	     right.br_blockcount <= MAXEXTLEN))
		state |= BMAP_RIGHT_CONTIG;
3000

3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017
	error = 0;
	/*
	 * Select which case we're in here, and implement it.
	 */
	switch (state & (BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG)) {
	case BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
		/*
		 * New allocation is contiguous with real allocations on the
		 * left and on the right.
		 * Merge all three into a single extent record.
		 */
		--bma->idx;
		trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
		xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, bma->idx),
			left.br_blockcount + new->br_blockcount +
			right.br_blockcount);
		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
3018

3019
		xfs_iext_remove(bma->ip, bma->idx + 1, 1, state);
3020

3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031
		XFS_IFORK_NEXT_SET(bma->ip, whichfork,
			XFS_IFORK_NEXTENTS(bma->ip, whichfork) - 1);
		if (bma->cur == NULL) {
			rval = XFS_ILOG_CORE | xfs_ilog_fext(whichfork);
		} else {
			rval = XFS_ILOG_CORE;
			error = xfs_bmbt_lookup_eq(bma->cur, right.br_startoff,
					right.br_startblock, right.br_blockcount,
					&i);
			if (error)
				goto done;
3032
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
3033 3034 3035
			error = xfs_btree_delete(bma->cur, &i);
			if (error)
				goto done;
3036
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
3037 3038 3039
			error = xfs_btree_decrement(bma->cur, 0, &i);
			if (error)
				goto done;
3040
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
3041 3042 3043 3044 3045 3046 3047 3048 3049 3050
			error = xfs_bmbt_update(bma->cur, left.br_startoff,
					left.br_startblock,
					left.br_blockcount +
						new->br_blockcount +
						right.br_blockcount,
					left.br_state);
			if (error)
				goto done;
		}
		break;
3051

3052 3053 3054 3055 3056 3057 3058 3059 3060 3061 3062
	case BMAP_LEFT_CONTIG:
		/*
		 * New allocation is contiguous with a real allocation
		 * on the left.
		 * Merge the new allocation with the left neighbor.
		 */
		--bma->idx;
		trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
		xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, bma->idx),
			left.br_blockcount + new->br_blockcount);
		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
L
Linus Torvalds 已提交
3063

3064 3065 3066 3067 3068 3069 3070 3071 3072
		if (bma->cur == NULL) {
			rval = xfs_ilog_fext(whichfork);
		} else {
			rval = 0;
			error = xfs_bmbt_lookup_eq(bma->cur, left.br_startoff,
					left.br_startblock, left.br_blockcount,
					&i);
			if (error)
				goto done;
3073
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
3074 3075 3076 3077 3078 3079 3080 3081 3082
			error = xfs_bmbt_update(bma->cur, left.br_startoff,
					left.br_startblock,
					left.br_blockcount +
						new->br_blockcount,
					left.br_state);
			if (error)
				goto done;
		}
		break;
3083

3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095
	case BMAP_RIGHT_CONTIG:
		/*
		 * New allocation is contiguous with a real allocation
		 * on the right.
		 * Merge the new allocation with the right neighbor.
		 */
		trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
		xfs_bmbt_set_allf(xfs_iext_get_ext(ifp, bma->idx),
			new->br_startoff, new->br_startblock,
			new->br_blockcount + right.br_blockcount,
			right.br_state);
		trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
3096

3097 3098 3099 3100 3101 3102 3103 3104 3105 3106
		if (bma->cur == NULL) {
			rval = xfs_ilog_fext(whichfork);
		} else {
			rval = 0;
			error = xfs_bmbt_lookup_eq(bma->cur,
					right.br_startoff,
					right.br_startblock,
					right.br_blockcount, &i);
			if (error)
				goto done;
3107
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
3108 3109 3110 3111 3112 3113 3114 3115 3116 3117 3118 3119 3120 3121 3122 3123 3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134 3135 3136
			error = xfs_bmbt_update(bma->cur, new->br_startoff,
					new->br_startblock,
					new->br_blockcount +
						right.br_blockcount,
					right.br_state);
			if (error)
				goto done;
		}
		break;

	case 0:
		/*
		 * New allocation is not contiguous with another
		 * real allocation.
		 * Insert a new entry.
		 */
		xfs_iext_insert(bma->ip, bma->idx, 1, new, state);
		XFS_IFORK_NEXT_SET(bma->ip, whichfork,
			XFS_IFORK_NEXTENTS(bma->ip, whichfork) + 1);
		if (bma->cur == NULL) {
			rval = XFS_ILOG_CORE | xfs_ilog_fext(whichfork);
		} else {
			rval = XFS_ILOG_CORE;
			error = xfs_bmbt_lookup_eq(bma->cur,
					new->br_startoff,
					new->br_startblock,
					new->br_blockcount, &i);
			if (error)
				goto done;
3137
			XFS_WANT_CORRUPTED_GOTO(mp, i == 0, done);
3138 3139 3140 3141
			bma->cur->bc_rec.b.br_state = new->br_state;
			error = xfs_btree_insert(bma->cur, &i);
			if (error)
				goto done;
3142
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167
		}
		break;
	}

	/* convert to a btree if necessary */
	if (xfs_bmap_needs_btree(bma->ip, whichfork)) {
		int	tmp_logflags;	/* partial log flag return val */

		ASSERT(bma->cur == NULL);
		error = xfs_bmap_extents_to_btree(bma->tp, bma->ip,
				bma->firstblock, bma->flist, &bma->cur,
				0, &tmp_logflags, whichfork);
		bma->logflags |= tmp_logflags;
		if (error)
			goto done;
	}

	/* clear out the allocated field, done with it now in any case. */
	if (bma->cur)
		bma->cur->bc_private.b.allocated = 0;

	xfs_bmap_check_leaf_extents(bma->cur, bma->ip, whichfork);
done:
	bma->logflags |= rval;
	return error;
L
Linus Torvalds 已提交
3168 3169 3170
}

/*
3171
 * Functions used in the extent read, allocate and remove paths
L
Linus Torvalds 已提交
3172 3173
 */

3174 3175 3176
/*
 * Adjust the size of the new extent based on di_extsize and rt extsize.
 */
D
Dave Chinner 已提交
3177
int
3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188
xfs_bmap_extsize_align(
	xfs_mount_t	*mp,
	xfs_bmbt_irec_t	*gotp,		/* next extent pointer */
	xfs_bmbt_irec_t	*prevp,		/* previous extent pointer */
	xfs_extlen_t	extsz,		/* align to this extent size */
	int		rt,		/* is this a realtime inode? */
	int		eof,		/* is extent at end-of-file? */
	int		delay,		/* creating delalloc extent? */
	int		convert,	/* overwriting unwritten extent? */
	xfs_fileoff_t	*offp,		/* in/out: aligned offset */
	xfs_extlen_t	*lenp)		/* in/out: aligned length */
3189
{
3190 3191 3192 3193 3194 3195 3196 3197
	xfs_fileoff_t	orig_off;	/* original offset */
	xfs_extlen_t	orig_alen;	/* original length */
	xfs_fileoff_t	orig_end;	/* original off+len */
	xfs_fileoff_t	nexto;		/* next file offset */
	xfs_fileoff_t	prevo;		/* previous file offset */
	xfs_fileoff_t	align_off;	/* temp for offset */
	xfs_extlen_t	align_alen;	/* temp for length */
	xfs_extlen_t	temp;		/* temp for calculations */
3198

3199
	if (convert)
3200 3201
		return 0;

3202 3203 3204
	orig_off = align_off = *offp;
	orig_alen = align_alen = *lenp;
	orig_end = orig_off + orig_alen;
L
Linus Torvalds 已提交
3205 3206

	/*
3207 3208
	 * If this request overlaps an existing extent, then don't
	 * attempt to perform any additional alignment.
L
Linus Torvalds 已提交
3209
	 */
3210 3211 3212 3213 3214 3215
	if (!delay && !eof &&
	    (orig_off >= gotp->br_startoff) &&
	    (orig_end <= gotp->br_startoff + gotp->br_blockcount)) {
		return 0;
	}

L
Linus Torvalds 已提交
3216
	/*
3217 3218 3219 3220 3221
	 * If the file offset is unaligned vs. the extent size
	 * we need to align it.  This will be possible unless
	 * the file was previously written with a kernel that didn't
	 * perform this alignment, or if a truncate shot us in the
	 * foot.
L
Linus Torvalds 已提交
3222
	 */
3223 3224 3225 3226
	temp = do_mod(orig_off, extsz);
	if (temp) {
		align_alen += temp;
		align_off -= temp;
L
Linus Torvalds 已提交
3227
	}
3228 3229 3230 3231 3232 3233

	/* Same adjustment for the end of the requested area. */
	temp = (align_alen % extsz);
	if (temp)
		align_alen += extsz - temp;

L
Linus Torvalds 已提交
3234
	/*
3235 3236 3237 3238 3239 3240
	 * For large extent hint sizes, the aligned extent might be larger than
	 * MAXEXTLEN. In that case, reduce the size by an extsz so that it pulls
	 * the length back under MAXEXTLEN. The outer allocation loops handle
	 * short allocation just fine, so it is safe to do this. We only want to
	 * do it when we are forced to, though, because it means more allocation
	 * operations are required.
L
Linus Torvalds 已提交
3241
	 */
3242 3243 3244 3245
	while (align_alen > MAXEXTLEN)
		align_alen -= extsz;
	ASSERT(align_alen <= MAXEXTLEN);

L
Linus Torvalds 已提交
3246
	/*
3247 3248
	 * If the previous block overlaps with this proposed allocation
	 * then move the start forward without adjusting the length.
L
Linus Torvalds 已提交
3249
	 */
3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292
	if (prevp->br_startoff != NULLFILEOFF) {
		if (prevp->br_startblock == HOLESTARTBLOCK)
			prevo = prevp->br_startoff;
		else
			prevo = prevp->br_startoff + prevp->br_blockcount;
	} else
		prevo = 0;
	if (align_off != orig_off && align_off < prevo)
		align_off = prevo;
	/*
	 * If the next block overlaps with this proposed allocation
	 * then move the start back without adjusting the length,
	 * but not before offset 0.
	 * This may of course make the start overlap previous block,
	 * and if we hit the offset 0 limit then the next block
	 * can still overlap too.
	 */
	if (!eof && gotp->br_startoff != NULLFILEOFF) {
		if ((delay && gotp->br_startblock == HOLESTARTBLOCK) ||
		    (!delay && gotp->br_startblock == DELAYSTARTBLOCK))
			nexto = gotp->br_startoff + gotp->br_blockcount;
		else
			nexto = gotp->br_startoff;
	} else
		nexto = NULLFILEOFF;
	if (!eof &&
	    align_off + align_alen != orig_end &&
	    align_off + align_alen > nexto)
		align_off = nexto > align_alen ? nexto - align_alen : 0;
	/*
	 * If we're now overlapping the next or previous extent that
	 * means we can't fit an extsz piece in this hole.  Just move
	 * the start forward to the first valid spot and set
	 * the length so we hit the end.
	 */
	if (align_off != orig_off && align_off < prevo)
		align_off = prevo;
	if (align_off + align_alen != orig_end &&
	    align_off + align_alen > nexto &&
	    nexto != NULLFILEOFF) {
		ASSERT(nexto > prevo);
		align_alen = nexto - align_off;
	}
L
Linus Torvalds 已提交
3293

3294 3295 3296 3297 3298
	/*
	 * If realtime, and the result isn't a multiple of the realtime
	 * extent size we need to remove blocks until it is.
	 */
	if (rt && (temp = (align_alen % mp->m_sb.sb_rextsize))) {
L
Linus Torvalds 已提交
3299
		/*
3300 3301
		 * We're not covering the original request, or
		 * we won't be able to once we fix the length.
L
Linus Torvalds 已提交
3302
		 */
3303 3304 3305
		if (orig_off < align_off ||
		    orig_end > align_off + align_alen ||
		    align_alen - temp < orig_alen)
D
Dave Chinner 已提交
3306
			return -EINVAL;
L
Linus Torvalds 已提交
3307
		/*
3308
		 * Try to fix it by moving the start up.
L
Linus Torvalds 已提交
3309
		 */
3310 3311 3312
		if (align_off + temp <= orig_off) {
			align_alen -= temp;
			align_off += temp;
L
Linus Torvalds 已提交
3313
		}
3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325
		/*
		 * Try to fix it by moving the end in.
		 */
		else if (align_off + align_alen - temp >= orig_end)
			align_alen -= temp;
		/*
		 * Set the start to the minimum then trim the length.
		 */
		else {
			align_alen -= orig_off - align_off;
			align_off = orig_off;
			align_alen -= align_alen % mp->m_sb.sb_rextsize;
L
Linus Torvalds 已提交
3326 3327
		}
		/*
3328
		 * Result doesn't cover the request, fail it.
L
Linus Torvalds 已提交
3329
		 */
3330
		if (orig_off < align_off || orig_end > align_off + align_alen)
D
Dave Chinner 已提交
3331
			return -EINVAL;
3332 3333
	} else {
		ASSERT(orig_off >= align_off);
3334 3335 3336
		/* see MAXEXTLEN handling above */
		ASSERT(orig_end <= align_off + align_alen ||
		       align_alen + extsz > MAXEXTLEN);
L
Linus Torvalds 已提交
3337 3338
	}

C
Christoph Hellwig 已提交
3339
#ifdef DEBUG
3340 3341 3342 3343 3344
	if (!eof && gotp->br_startoff != NULLFILEOFF)
		ASSERT(align_off + align_alen <= gotp->br_startoff);
	if (prevp->br_startoff != NULLFILEOFF)
		ASSERT(align_off >= prevp->br_startoff + prevp->br_blockcount);
#endif
L
Linus Torvalds 已提交
3345

3346 3347 3348
	*lenp = align_alen;
	*offp = align_off;
	return 0;
L
Linus Torvalds 已提交
3349 3350
}

3351 3352
#define XFS_ALLOC_GAP_UNITS	4

D
Dave Chinner 已提交
3353
void
3354
xfs_bmap_adjacent(
D
Dave Chinner 已提交
3355
	struct xfs_bmalloca	*ap)	/* bmap alloc argument struct */
L
Linus Torvalds 已提交
3356
{
3357 3358 3359 3360 3361
	xfs_fsblock_t	adjust;		/* adjustment to block numbers */
	xfs_agnumber_t	fb_agno;	/* ag number of ap->firstblock */
	xfs_mount_t	*mp;		/* mount point structure */
	int		nullfb;		/* true if ap->firstblock isn't set */
	int		rt;		/* true if inode is realtime */
L
Linus Torvalds 已提交
3362

3363 3364 3365 3366 3367 3368
#define	ISVALID(x,y)	\
	(rt ? \
		(x) < mp->m_sb.sb_rblocks : \
		XFS_FSB_TO_AGNO(mp, x) == XFS_FSB_TO_AGNO(mp, y) && \
		XFS_FSB_TO_AGNO(mp, x) < mp->m_sb.sb_agcount && \
		XFS_FSB_TO_AGBNO(mp, x) < mp->m_sb.sb_agblocks)
L
Linus Torvalds 已提交
3369

3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437 3438
	mp = ap->ip->i_mount;
	nullfb = *ap->firstblock == NULLFSBLOCK;
	rt = XFS_IS_REALTIME_INODE(ap->ip) && ap->userdata;
	fb_agno = nullfb ? NULLAGNUMBER : XFS_FSB_TO_AGNO(mp, *ap->firstblock);
	/*
	 * If allocating at eof, and there's a previous real block,
	 * try to use its last block as our starting point.
	 */
	if (ap->eof && ap->prev.br_startoff != NULLFILEOFF &&
	    !isnullstartblock(ap->prev.br_startblock) &&
	    ISVALID(ap->prev.br_startblock + ap->prev.br_blockcount,
		    ap->prev.br_startblock)) {
		ap->blkno = ap->prev.br_startblock + ap->prev.br_blockcount;
		/*
		 * Adjust for the gap between prevp and us.
		 */
		adjust = ap->offset -
			(ap->prev.br_startoff + ap->prev.br_blockcount);
		if (adjust &&
		    ISVALID(ap->blkno + adjust, ap->prev.br_startblock))
			ap->blkno += adjust;
	}
	/*
	 * If not at eof, then compare the two neighbor blocks.
	 * Figure out whether either one gives us a good starting point,
	 * and pick the better one.
	 */
	else if (!ap->eof) {
		xfs_fsblock_t	gotbno;		/* right side block number */
		xfs_fsblock_t	gotdiff=0;	/* right side difference */
		xfs_fsblock_t	prevbno;	/* left side block number */
		xfs_fsblock_t	prevdiff=0;	/* left side difference */

		/*
		 * If there's a previous (left) block, select a requested
		 * start block based on it.
		 */
		if (ap->prev.br_startoff != NULLFILEOFF &&
		    !isnullstartblock(ap->prev.br_startblock) &&
		    (prevbno = ap->prev.br_startblock +
			       ap->prev.br_blockcount) &&
		    ISVALID(prevbno, ap->prev.br_startblock)) {
			/*
			 * Calculate gap to end of previous block.
			 */
			adjust = prevdiff = ap->offset -
				(ap->prev.br_startoff +
				 ap->prev.br_blockcount);
			/*
			 * Figure the startblock based on the previous block's
			 * end and the gap size.
			 * Heuristic!
			 * If the gap is large relative to the piece we're
			 * allocating, or using it gives us an invalid block
			 * number, then just use the end of the previous block.
			 */
			if (prevdiff <= XFS_ALLOC_GAP_UNITS * ap->length &&
			    ISVALID(prevbno + prevdiff,
				    ap->prev.br_startblock))
				prevbno += adjust;
			else
				prevdiff += adjust;
			/*
			 * If the firstblock forbids it, can't use it,
			 * must use default.
			 */
			if (!rt && !nullfb &&
			    XFS_FSB_TO_AGNO(mp, prevbno) != fb_agno)
				prevbno = NULLFSBLOCK;
L
Linus Torvalds 已提交
3439
		}
3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477 3478 3479 3480 3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496
		/*
		 * No previous block or can't follow it, just default.
		 */
		else
			prevbno = NULLFSBLOCK;
		/*
		 * If there's a following (right) block, select a requested
		 * start block based on it.
		 */
		if (!isnullstartblock(ap->got.br_startblock)) {
			/*
			 * Calculate gap to start of next block.
			 */
			adjust = gotdiff = ap->got.br_startoff - ap->offset;
			/*
			 * Figure the startblock based on the next block's
			 * start and the gap size.
			 */
			gotbno = ap->got.br_startblock;
			/*
			 * Heuristic!
			 * If the gap is large relative to the piece we're
			 * allocating, or using it gives us an invalid block
			 * number, then just use the start of the next block
			 * offset by our length.
			 */
			if (gotdiff <= XFS_ALLOC_GAP_UNITS * ap->length &&
			    ISVALID(gotbno - gotdiff, gotbno))
				gotbno -= adjust;
			else if (ISVALID(gotbno - ap->length, gotbno)) {
				gotbno -= ap->length;
				gotdiff += adjust - ap->length;
			} else
				gotdiff += adjust;
			/*
			 * If the firstblock forbids it, can't use it,
			 * must use default.
			 */
			if (!rt && !nullfb &&
			    XFS_FSB_TO_AGNO(mp, gotbno) != fb_agno)
				gotbno = NULLFSBLOCK;
		}
		/*
		 * No next block, just default.
		 */
		else
			gotbno = NULLFSBLOCK;
		/*
		 * If both valid, pick the better one, else the only good
		 * one, else ap->blkno is already set (to 0 or the inode block).
		 */
		if (prevbno != NULLFSBLOCK && gotbno != NULLFSBLOCK)
			ap->blkno = prevdiff <= gotdiff ? prevbno : gotbno;
		else if (prevbno != NULLFSBLOCK)
			ap->blkno = prevbno;
		else if (gotbno != NULLFSBLOCK)
			ap->blkno = gotbno;
L
Linus Torvalds 已提交
3497
	}
3498
#undef ISVALID
L
Linus Torvalds 已提交
3499 3500
}

3501 3502 3503 3504 3505 3506 3507 3508 3509 3510 3511 3512 3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523 3524
static int
xfs_bmap_longest_free_extent(
	struct xfs_trans	*tp,
	xfs_agnumber_t		ag,
	xfs_extlen_t		*blen,
	int			*notinit)
{
	struct xfs_mount	*mp = tp->t_mountp;
	struct xfs_perag	*pag;
	xfs_extlen_t		longest;
	int			error = 0;

	pag = xfs_perag_get(mp, ag);
	if (!pag->pagf_init) {
		error = xfs_alloc_pagf_init(mp, tp, ag, XFS_ALLOC_FLAG_TRYLOCK);
		if (error)
			goto out;

		if (!pag->pagf_init) {
			*notinit = 1;
			goto out;
		}
	}

3525
	longest = xfs_alloc_longest_free_extent(mp, pag,
3526
					xfs_alloc_min_freelist(mp, pag));
3527 3528 3529 3530 3531 3532 3533 3534 3535 3536 3537 3538 3539 3540 3541 3542 3543 3544 3545 3546 3547 3548 3549 3550 3551 3552 3553 3554 3555 3556 3557 3558 3559 3560 3561 3562
	if (*blen < longest)
		*blen = longest;

out:
	xfs_perag_put(pag);
	return error;
}

static void
xfs_bmap_select_minlen(
	struct xfs_bmalloca	*ap,
	struct xfs_alloc_arg	*args,
	xfs_extlen_t		*blen,
	int			notinit)
{
	if (notinit || *blen < ap->minlen) {
		/*
		 * Since we did a BUF_TRYLOCK above, it is possible that
		 * there is space for this request.
		 */
		args->minlen = ap->minlen;
	} else if (*blen < args->maxlen) {
		/*
		 * If the best seen length is less than the request length,
		 * use the best as the minimum.
		 */
		args->minlen = *blen;
	} else {
		/*
		 * Otherwise we've seen an extent as big as maxlen, use that
		 * as the minimum.
		 */
		args->minlen = args->maxlen;
	}
}

3563
STATIC int
3564 3565 3566 3567
xfs_bmap_btalloc_nullfb(
	struct xfs_bmalloca	*ap,
	struct xfs_alloc_arg	*args,
	xfs_extlen_t		*blen)
3568
{
3569 3570 3571
	struct xfs_mount	*mp = ap->ip->i_mount;
	xfs_agnumber_t		ag, startag;
	int			notinit = 0;
3572 3573
	int			error;

3574
	args->type = XFS_ALLOCTYPE_START_BNO;
3575
	args->total = ap->total;
3576

3577 3578 3579
	startag = ag = XFS_FSB_TO_AGNO(mp, args->fsbno);
	if (startag == NULLAGNUMBER)
		startag = ag = 0;
3580

3581
	while (*blen < args->maxlen) {
3582 3583 3584 3585
		error = xfs_bmap_longest_free_extent(args->tp, ag, blen,
						     &notinit);
		if (error)
			return error;
3586

3587 3588 3589 3590 3591
		if (++ag == mp->m_sb.sb_agcount)
			ag = 0;
		if (ag == startag)
			break;
	}
3592

3593 3594 3595 3596 3597 3598 3599 3600 3601 3602 3603 3604 3605 3606 3607 3608 3609 3610 3611 3612 3613 3614 3615 3616 3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627 3628 3629 3630 3631
	xfs_bmap_select_minlen(ap, args, blen, notinit);
	return 0;
}

STATIC int
xfs_bmap_btalloc_filestreams(
	struct xfs_bmalloca	*ap,
	struct xfs_alloc_arg	*args,
	xfs_extlen_t		*blen)
{
	struct xfs_mount	*mp = ap->ip->i_mount;
	xfs_agnumber_t		ag;
	int			notinit = 0;
	int			error;

	args->type = XFS_ALLOCTYPE_NEAR_BNO;
	args->total = ap->total;

	ag = XFS_FSB_TO_AGNO(mp, args->fsbno);
	if (ag == NULLAGNUMBER)
		ag = 0;

	error = xfs_bmap_longest_free_extent(args->tp, ag, blen, &notinit);
	if (error)
		return error;

	if (*blen < args->maxlen) {
		error = xfs_filestream_new_ag(ap, &ag);
		if (error)
			return error;

		error = xfs_bmap_longest_free_extent(args->tp, ag, blen,
						     &notinit);
		if (error)
			return error;

	}

	xfs_bmap_select_minlen(ap, args, blen, notinit);
3632 3633

	/*
3634 3635
	 * Set the failure fallback case to look in the selected AG as stream
	 * may have moved.
3636
	 */
3637
	ap->blkno = args->fsbno = XFS_AGB_TO_FSB(mp, ag, 0);
3638 3639 3640
	return 0;
}

3641 3642
STATIC int
xfs_bmap_btalloc(
D
Dave Chinner 已提交
3643
	struct xfs_bmalloca	*ap)	/* bmap alloc argument struct */
3644
{
3645 3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 3656
	xfs_mount_t	*mp;		/* mount point structure */
	xfs_alloctype_t	atype = 0;	/* type for allocation routines */
	xfs_extlen_t	align;		/* minimum allocation alignment */
	xfs_agnumber_t	fb_agno;	/* ag number of ap->firstblock */
	xfs_agnumber_t	ag;
	xfs_alloc_arg_t	args;
	xfs_extlen_t	blen;
	xfs_extlen_t	nextminlen = 0;
	int		nullfb;		/* true if ap->firstblock isn't set */
	int		isaligned;
	int		tryagain;
	int		error;
3657
	int		stripe_align;
3658

3659
	ASSERT(ap->length);
3660

3661
	mp = ap->ip->i_mount;
3662 3663 3664 3665 3666 3667 3668 3669

	/* stripe alignment for allocation is determined by mount parameters */
	stripe_align = 0;
	if (mp->m_swidth && (mp->m_flags & XFS_MOUNT_SWALLOC))
		stripe_align = mp->m_swidth;
	else if (mp->m_dalign)
		stripe_align = mp->m_dalign;

3670 3671 3672 3673 3674 3675 3676
	align = ap->userdata ? xfs_get_extsz_hint(ap->ip) : 0;
	if (unlikely(align)) {
		error = xfs_bmap_extsize_align(mp, &ap->got, &ap->prev,
						align, 0, ap->eof, 0, ap->conv,
						&ap->offset, &ap->length);
		ASSERT(!error);
		ASSERT(ap->length);
3677
	}
3678 3679


3680 3681 3682 3683 3684 3685 3686 3687 3688 3689 3690 3691
	nullfb = *ap->firstblock == NULLFSBLOCK;
	fb_agno = nullfb ? NULLAGNUMBER : XFS_FSB_TO_AGNO(mp, *ap->firstblock);
	if (nullfb) {
		if (ap->userdata && xfs_inode_is_filestream(ap->ip)) {
			ag = xfs_filestream_lookup_ag(ap->ip);
			ag = (ag != NULLAGNUMBER) ? ag : 0;
			ap->blkno = XFS_AGB_TO_FSB(mp, ag, 0);
		} else {
			ap->blkno = XFS_INO_TO_FSB(mp, ap->ip->i_ino);
		}
	} else
		ap->blkno = *ap->firstblock;
3692

3693
	xfs_bmap_adjacent(ap);
3694

3695 3696 3697 3698 3699 3700 3701 3702 3703 3704 3705 3706 3707 3708 3709 3710
	/*
	 * If allowed, use ap->blkno; otherwise must use firstblock since
	 * it's in the right allocation group.
	 */
	if (nullfb || XFS_FSB_TO_AGNO(mp, ap->blkno) == fb_agno)
		;
	else
		ap->blkno = *ap->firstblock;
	/*
	 * Normal allocation, done through xfs_alloc_vextent.
	 */
	tryagain = isaligned = 0;
	memset(&args, 0, sizeof(args));
	args.tp = ap->tp;
	args.mp = mp;
	args.fsbno = ap->blkno;
3711

3712 3713 3714 3715 3716
	/* Trim the allocation back to the maximum an AG can fit. */
	args.maxlen = MIN(ap->length, XFS_ALLOC_AG_MAX_USABLE(mp));
	args.firstblock = *ap->firstblock;
	blen = 0;
	if (nullfb) {
3717 3718 3719 3720 3721 3722 3723 3724 3725
		/*
		 * Search for an allocation group with a single extent large
		 * enough for the request.  If one isn't found, then adjust
		 * the minimum allocation size to the largest space found.
		 */
		if (ap->userdata && xfs_inode_is_filestream(ap->ip))
			error = xfs_bmap_btalloc_filestreams(ap, &args, &blen);
		else
			error = xfs_bmap_btalloc_nullfb(ap, &args, &blen);
3726 3727
		if (error)
			return error;
3728 3729 3730 3731 3732 3733 3734 3735 3736 3737
	} else if (ap->flist->xbf_low) {
		if (xfs_inode_is_filestream(ap->ip))
			args.type = XFS_ALLOCTYPE_FIRST_AG;
		else
			args.type = XFS_ALLOCTYPE_START_BNO;
		args.total = args.minlen = ap->minlen;
	} else {
		args.type = XFS_ALLOCTYPE_NEAR_BNO;
		args.total = ap->total;
		args.minlen = ap->minlen;
3738
	}
3739 3740 3741 3742 3743
	/* apply extent size hints if obtained earlier */
	if (unlikely(align)) {
		args.prod = align;
		if ((args.mod = (xfs_extlen_t)do_mod(ap->offset, args.prod)))
			args.mod = (xfs_extlen_t)(args.prod - args.mod);
3744
	} else if (mp->m_sb.sb_blocksize >= PAGE_SIZE) {
3745 3746 3747
		args.prod = 1;
		args.mod = 0;
	} else {
3748
		args.prod = PAGE_SIZE >> mp->m_sb.sb_blocklog;
3749 3750
		if ((args.mod = (xfs_extlen_t)(do_mod(ap->offset, args.prod))))
			args.mod = (xfs_extlen_t)(args.prod - args.mod);
3751
	}
3752
	/*
3753 3754 3755 3756 3757 3758 3759
	 * If we are not low on available data blocks, and the
	 * underlying logical volume manager is a stripe, and
	 * the file offset is zero then try to allocate data
	 * blocks on stripe unit boundary.
	 * NOTE: ap->aeof is only set if the allocation length
	 * is >= the stripe unit and the allocation offset is
	 * at the end of file.
3760
	 */
3761 3762
	if (!ap->flist->xbf_low && ap->aeof) {
		if (!ap->offset) {
3763
			args.alignment = stripe_align;
3764 3765 3766 3767 3768 3769 3770 3771 3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785 3786 3787
			atype = args.type;
			isaligned = 1;
			/*
			 * Adjust for alignment
			 */
			if (blen > args.alignment && blen <= args.maxlen)
				args.minlen = blen - args.alignment;
			args.minalignslop = 0;
		} else {
			/*
			 * First try an exact bno allocation.
			 * If it fails then do a near or start bno
			 * allocation with alignment turned on.
			 */
			atype = args.type;
			tryagain = 1;
			args.type = XFS_ALLOCTYPE_THIS_BNO;
			args.alignment = 1;
			/*
			 * Compute the minlen+alignment for the
			 * next case.  Set slop so that the value
			 * of minlen+alignment+slop doesn't go up
			 * between the calls.
			 */
3788 3789
			if (blen > stripe_align && blen <= args.maxlen)
				nextminlen = blen - stripe_align;
3790 3791
			else
				nextminlen = args.minlen;
3792
			if (nextminlen + stripe_align > args.minlen + 1)
3793
				args.minalignslop =
3794
					nextminlen + stripe_align -
3795 3796 3797
					args.minlen - 1;
			else
				args.minalignslop = 0;
3798 3799
		}
	} else {
3800 3801
		args.alignment = 1;
		args.minalignslop = 0;
3802
	}
3803 3804 3805 3806
	args.minleft = ap->minleft;
	args.wasdel = ap->wasdel;
	args.isfl = 0;
	args.userdata = ap->userdata;
3807 3808 3809 3810 3811
	if (ap->userdata & XFS_ALLOC_USERDATA_ZERO)
		args.ip = ap->ip;

	error = xfs_alloc_vextent(&args);
	if (error)
3812
		return error;
3813

3814 3815 3816 3817 3818 3819 3820
	if (tryagain && args.fsbno == NULLFSBLOCK) {
		/*
		 * Exact allocation failed. Now try with alignment
		 * turned on.
		 */
		args.type = atype;
		args.fsbno = ap->blkno;
3821
		args.alignment = stripe_align;
3822 3823 3824 3825 3826
		args.minlen = nextminlen;
		args.minalignslop = 0;
		isaligned = 1;
		if ((error = xfs_alloc_vextent(&args)))
			return error;
3827
	}
3828 3829 3830 3831 3832 3833 3834 3835 3836
	if (isaligned && args.fsbno == NULLFSBLOCK) {
		/*
		 * allocation failed, so turn off alignment and
		 * try again.
		 */
		args.type = atype;
		args.fsbno = ap->blkno;
		args.alignment = 0;
		if ((error = xfs_alloc_vextent(&args)))
3837 3838
			return error;
	}
3839 3840 3841 3842 3843 3844 3845
	if (args.fsbno == NULLFSBLOCK && nullfb &&
	    args.minlen > ap->minlen) {
		args.minlen = ap->minlen;
		args.type = XFS_ALLOCTYPE_START_BNO;
		args.fsbno = ap->blkno;
		if ((error = xfs_alloc_vextent(&args)))
			return error;
3846
	}
3847 3848 3849 3850 3851 3852 3853 3854 3855 3856 3857 3858 3859 3860 3861 3862 3863 3864 3865 3866
	if (args.fsbno == NULLFSBLOCK && nullfb) {
		args.fsbno = 0;
		args.type = XFS_ALLOCTYPE_FIRST_AG;
		args.total = ap->minlen;
		args.minleft = 0;
		if ((error = xfs_alloc_vextent(&args)))
			return error;
		ap->flist->xbf_low = 1;
	}
	if (args.fsbno != NULLFSBLOCK) {
		/*
		 * check the allocation happened at the same or higher AG than
		 * the first block that was allocated.
		 */
		ASSERT(*ap->firstblock == NULLFSBLOCK ||
		       XFS_FSB_TO_AGNO(mp, *ap->firstblock) ==
		       XFS_FSB_TO_AGNO(mp, args.fsbno) ||
		       (ap->flist->xbf_low &&
			XFS_FSB_TO_AGNO(mp, *ap->firstblock) <
			XFS_FSB_TO_AGNO(mp, args.fsbno)));
3867

3868 3869 3870 3871 3872 3873 3874 3875 3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891
		ap->blkno = args.fsbno;
		if (*ap->firstblock == NULLFSBLOCK)
			*ap->firstblock = args.fsbno;
		ASSERT(nullfb || fb_agno == args.agno ||
		       (ap->flist->xbf_low && fb_agno < args.agno));
		ap->length = args.len;
		ap->ip->i_d.di_nblocks += args.len;
		xfs_trans_log_inode(ap->tp, ap->ip, XFS_ILOG_CORE);
		if (ap->wasdel)
			ap->ip->i_delayed_blks -= args.len;
		/*
		 * Adjust the disk quota also. This was reserved
		 * earlier.
		 */
		xfs_trans_mod_dquot_byino(ap->tp, ap->ip,
			ap->wasdel ? XFS_TRANS_DQ_DELBCOUNT :
					XFS_TRANS_DQ_BCOUNT,
			(long) args.len);
	} else {
		ap->blkno = NULLFSBLOCK;
		ap->length = 0;
	}
	return 0;
}
3892

3893 3894 3895 3896 3897 3898
/*
 * xfs_bmap_alloc is called by xfs_bmapi to allocate an extent for a file.
 * It figures out where to ask the underlying allocator to put the new extent.
 */
STATIC int
xfs_bmap_alloc(
D
Dave Chinner 已提交
3899
	struct xfs_bmalloca	*ap)	/* bmap alloc argument struct */
3900 3901 3902 3903 3904
{
	if (XFS_IS_REALTIME_INODE(ap->ip) && ap->userdata)
		return xfs_bmap_rtalloc(ap);
	return xfs_bmap_btalloc(ap);
}
3905

3906 3907 3908 3909 3910 3911 3912 3913 3914 3915 3916 3917 3918 3919 3920 3921 3922 3923 3924 3925 3926
/*
 * Trim the returned map to the required bounds
 */
STATIC void
xfs_bmapi_trim_map(
	struct xfs_bmbt_irec	*mval,
	struct xfs_bmbt_irec	*got,
	xfs_fileoff_t		*bno,
	xfs_filblks_t		len,
	xfs_fileoff_t		obno,
	xfs_fileoff_t		end,
	int			n,
	int			flags)
{
	if ((flags & XFS_BMAPI_ENTIRE) ||
	    got->br_startoff + got->br_blockcount <= obno) {
		*mval = *got;
		if (isnullstartblock(got->br_startblock))
			mval->br_startblock = DELAYSTARTBLOCK;
		return;
	}
3927

3928 3929 3930 3931 3932 3933 3934 3935 3936 3937
	if (obno > *bno)
		*bno = obno;
	ASSERT((*bno >= obno) || (n == 0));
	ASSERT(*bno < end);
	mval->br_startoff = *bno;
	if (isnullstartblock(got->br_startblock))
		mval->br_startblock = DELAYSTARTBLOCK;
	else
		mval->br_startblock = got->br_startblock +
					(*bno - got->br_startoff);
3938
	/*
3939 3940 3941 3942 3943
	 * Return the minimum of what we got and what we asked for for
	 * the length.  We can use the len variable here because it is
	 * modified below and we could have been there before coming
	 * here if the first part of the allocation didn't overlap what
	 * was asked for.
3944
	 */
3945 3946 3947 3948 3949
	mval->br_blockcount = XFS_FILBLKS_MIN(end - *bno,
			got->br_blockcount - (*bno - got->br_startoff));
	mval->br_state = got->br_state;
	ASSERT(mval->br_blockcount <= len);
	return;
3950 3951
}

3952 3953 3954 3955 3956 3957 3958 3959 3960 3961 3962 3963
/*
 * Update and validate the extent map to return
 */
STATIC void
xfs_bmapi_update_map(
	struct xfs_bmbt_irec	**map,
	xfs_fileoff_t		*bno,
	xfs_filblks_t		*len,
	xfs_fileoff_t		obno,
	xfs_fileoff_t		end,
	int			*n,
	int			flags)
3964
{
3965
	xfs_bmbt_irec_t	*mval = *map;
3966

3967 3968 3969 3970
	ASSERT((flags & XFS_BMAPI_ENTIRE) ||
	       ((mval->br_startoff + mval->br_blockcount) <= end));
	ASSERT((flags & XFS_BMAPI_ENTIRE) || (mval->br_blockcount <= *len) ||
	       (mval->br_startoff < obno));
3971

3972 3973 3974 3975 3976 3977 3978 3979 3980 3981 3982 3983 3984 3985 3986 3987 3988 3989 3990 3991 3992 3993 3994 3995 3996 3997 3998 3999 4000 4001 4002 4003 4004
	*bno = mval->br_startoff + mval->br_blockcount;
	*len = end - *bno;
	if (*n > 0 && mval->br_startoff == mval[-1].br_startoff) {
		/* update previous map with new information */
		ASSERT(mval->br_startblock == mval[-1].br_startblock);
		ASSERT(mval->br_blockcount > mval[-1].br_blockcount);
		ASSERT(mval->br_state == mval[-1].br_state);
		mval[-1].br_blockcount = mval->br_blockcount;
		mval[-1].br_state = mval->br_state;
	} else if (*n > 0 && mval->br_startblock != DELAYSTARTBLOCK &&
		   mval[-1].br_startblock != DELAYSTARTBLOCK &&
		   mval[-1].br_startblock != HOLESTARTBLOCK &&
		   mval->br_startblock == mval[-1].br_startblock +
					  mval[-1].br_blockcount &&
		   ((flags & XFS_BMAPI_IGSTATE) ||
			mval[-1].br_state == mval->br_state)) {
		ASSERT(mval->br_startoff ==
		       mval[-1].br_startoff + mval[-1].br_blockcount);
		mval[-1].br_blockcount += mval->br_blockcount;
	} else if (*n > 0 &&
		   mval->br_startblock == DELAYSTARTBLOCK &&
		   mval[-1].br_startblock == DELAYSTARTBLOCK &&
		   mval->br_startoff ==
		   mval[-1].br_startoff + mval[-1].br_blockcount) {
		mval[-1].br_blockcount += mval->br_blockcount;
		mval[-1].br_state = mval->br_state;
	} else if (!((*n == 0) &&
		     ((mval->br_startoff + mval->br_blockcount) <=
		      obno))) {
		mval++;
		(*n)++;
	}
	*map = mval;
4005 4006 4007
}

/*
4008
 * Map file blocks to filesystem blocks without allocation.
4009 4010
 */
int
4011 4012 4013
xfs_bmapi_read(
	struct xfs_inode	*ip,
	xfs_fileoff_t		bno,
4014
	xfs_filblks_t		len,
4015 4016
	struct xfs_bmbt_irec	*mval,
	int			*nmap,
4017
	int			flags)
4018
{
4019 4020 4021 4022 4023 4024 4025 4026 4027 4028
	struct xfs_mount	*mp = ip->i_mount;
	struct xfs_ifork	*ifp;
	struct xfs_bmbt_irec	got;
	struct xfs_bmbt_irec	prev;
	xfs_fileoff_t		obno;
	xfs_fileoff_t		end;
	xfs_extnum_t		lastx;
	int			error;
	int			eof;
	int			n = 0;
4029 4030 4031
	int			whichfork = (flags & XFS_BMAPI_ATTRFORK) ?
						XFS_ATTR_FORK : XFS_DATA_FORK;

4032 4033 4034
	ASSERT(*nmap >= 1);
	ASSERT(!(flags & ~(XFS_BMAPI_ATTRFORK|XFS_BMAPI_ENTIRE|
			   XFS_BMAPI_IGSTATE)));
4035
	ASSERT(xfs_isilocked(ip, XFS_ILOCK_SHARED|XFS_ILOCK_EXCL));
4036

4037 4038 4039 4040 4041
	if (unlikely(XFS_TEST_ERROR(
	    (XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS &&
	     XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE),
	     mp, XFS_ERRTAG_BMAPIFORMAT, XFS_RANDOM_BMAPIFORMAT))) {
		XFS_ERROR_REPORT("xfs_bmapi_read", XFS_ERRLEVEL_LOW, mp);
D
Dave Chinner 已提交
4042
		return -EFSCORRUPTED;
4043
	}
4044

4045
	if (XFS_FORCED_SHUTDOWN(mp))
D
Dave Chinner 已提交
4046
		return -EIO;
4047

4048
	XFS_STATS_INC(mp, xs_blk_mapr);
4049 4050 4051 4052 4053 4054 4055

	ifp = XFS_IFORK_PTR(ip, whichfork);

	if (!(ifp->if_flags & XFS_IFEXTENTS)) {
		error = xfs_iread_extents(NULL, ip, whichfork);
		if (error)
			return error;
4056 4057
	}

4058 4059 4060
	xfs_bmap_search_extents(ip, bno, whichfork, &eof, &lastx, &got, &prev);
	end = bno + len;
	obno = bno;
4061

4062 4063 4064 4065 4066 4067 4068 4069 4070 4071 4072 4073 4074 4075 4076 4077 4078
	while (bno < end && n < *nmap) {
		/* Reading past eof, act as though there's a hole up to end. */
		if (eof)
			got.br_startoff = end;
		if (got.br_startoff > bno) {
			/* Reading in a hole.  */
			mval->br_startoff = bno;
			mval->br_startblock = HOLESTARTBLOCK;
			mval->br_blockcount =
				XFS_FILBLKS_MIN(len, got.br_startoff - bno);
			mval->br_state = XFS_EXT_NORM;
			bno += mval->br_blockcount;
			len -= mval->br_blockcount;
			mval++;
			n++;
			continue;
		}
4079

4080 4081 4082 4083 4084 4085 4086 4087 4088 4089 4090 4091 4092 4093 4094
		/* set up the extent map to return. */
		xfs_bmapi_trim_map(mval, &got, &bno, len, obno, end, n, flags);
		xfs_bmapi_update_map(&mval, &bno, &len, obno, end, &n, flags);

		/* If we're done, stop now. */
		if (bno >= end || n >= *nmap)
			break;

		/* Else go on to the next record. */
		if (++lastx < ifp->if_bytes / sizeof(xfs_bmbt_rec_t))
			xfs_bmbt_get_all(xfs_iext_get_ext(ifp, lastx), &got);
		else
			eof = 1;
	}
	*nmap = n;
4095 4096 4097
	return 0;
}

4098 4099 4100 4101 4102 4103 4104 4105 4106
STATIC int
xfs_bmapi_reserve_delalloc(
	struct xfs_inode	*ip,
	xfs_fileoff_t		aoff,
	xfs_filblks_t		len,
	struct xfs_bmbt_irec	*got,
	struct xfs_bmbt_irec	*prev,
	xfs_extnum_t		*lastx,
	int			eof)
L
Linus Torvalds 已提交
4107
{
4108
	struct xfs_mount	*mp = ip->i_mount;
4109 4110 4111 4112 4113 4114
	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
	xfs_extlen_t		alen;
	xfs_extlen_t		indlen;
	char			rt = XFS_IS_REALTIME_INODE(ip);
	xfs_extlen_t		extsz;
	int			error;
4115

4116 4117 4118
	alen = XFS_FILBLKS_MIN(len, MAXEXTLEN);
	if (!eof)
		alen = XFS_FILBLKS_MIN(alen, got->br_startoff - aoff);
L
Linus Torvalds 已提交
4119

4120 4121 4122 4123 4124 4125 4126 4127 4128 4129 4130 4131 4132 4133 4134 4135 4136 4137 4138 4139 4140 4141 4142 4143 4144 4145 4146 4147 4148
	/* Figure out the extent size, adjust alen */
	extsz = xfs_get_extsz_hint(ip);
	if (extsz) {
		error = xfs_bmap_extsize_align(mp, got, prev, extsz, rt, eof,
					       1, 0, &aoff, &alen);
		ASSERT(!error);
	}

	if (rt)
		extsz = alen / mp->m_sb.sb_rextsize;

	/*
	 * Make a transaction-less quota reservation for delayed allocation
	 * blocks.  This number gets adjusted later.  We return if we haven't
	 * allocated blocks already inside this loop.
	 */
	error = xfs_trans_reserve_quota_nblks(NULL, ip, (long)alen, 0,
			rt ? XFS_QMOPT_RES_RTBLKS : XFS_QMOPT_RES_REGBLKS);
	if (error)
		return error;

	/*
	 * Split changing sb for alen and indlen since they could be coming
	 * from different places.
	 */
	indlen = (xfs_extlen_t)xfs_bmap_worst_indlen(ip, alen);
	ASSERT(indlen > 0);

	if (rt) {
D
Dave Chinner 已提交
4149
		error = xfs_mod_frextents(mp, -((int64_t)extsz));
4150
	} else {
4151
		error = xfs_mod_fdblocks(mp, -((int64_t)alen), false);
4152 4153 4154 4155 4156
	}

	if (error)
		goto out_unreserve_quota;

4157
	error = xfs_mod_fdblocks(mp, -((int64_t)indlen), false);
4158 4159 4160 4161 4162 4163 4164 4165 4166 4167 4168 4169 4170 4171 4172 4173 4174 4175 4176 4177 4178 4179 4180 4181 4182 4183
	if (error)
		goto out_unreserve_blocks;


	ip->i_delayed_blks += alen;

	got->br_startoff = aoff;
	got->br_startblock = nullstartblock(indlen);
	got->br_blockcount = alen;
	got->br_state = XFS_EXT_NORM;
	xfs_bmap_add_extent_hole_delay(ip, lastx, got);

	/*
	 * Update our extent pointer, given that xfs_bmap_add_extent_hole_delay
	 * might have merged it into one of the neighbouring ones.
	 */
	xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *lastx), got);

	ASSERT(got->br_startoff <= aoff);
	ASSERT(got->br_startoff + got->br_blockcount >= aoff + alen);
	ASSERT(isnullstartblock(got->br_startblock));
	ASSERT(got->br_state == XFS_EXT_NORM);
	return 0;

out_unreserve_blocks:
	if (rt)
D
Dave Chinner 已提交
4184
		xfs_mod_frextents(mp, extsz);
4185
	else
4186
		xfs_mod_fdblocks(mp, alen, false);
4187 4188 4189 4190 4191 4192 4193 4194 4195 4196 4197 4198 4199 4200 4201 4202 4203 4204 4205 4206 4207 4208 4209 4210 4211 4212 4213 4214 4215
out_unreserve_quota:
	if (XFS_IS_QUOTA_ON(mp))
		xfs_trans_unreserve_quota_nblks(NULL, ip, (long)alen, 0, rt ?
				XFS_QMOPT_RES_RTBLKS : XFS_QMOPT_RES_REGBLKS);
	return error;
}

/*
 * Map file blocks to filesystem blocks, adding delayed allocations as needed.
 */
int
xfs_bmapi_delay(
	struct xfs_inode	*ip,	/* incore inode */
	xfs_fileoff_t		bno,	/* starting file offs. mapped */
	xfs_filblks_t		len,	/* length to map in file */
	struct xfs_bmbt_irec	*mval,	/* output: map values */
	int			*nmap,	/* i/o: mval size/count */
	int			flags)	/* XFS_BMAPI_... */
{
	struct xfs_mount	*mp = ip->i_mount;
	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
	struct xfs_bmbt_irec	got;	/* current file extent record */
	struct xfs_bmbt_irec	prev;	/* previous file extent record */
	xfs_fileoff_t		obno;	/* old block number (offset) */
	xfs_fileoff_t		end;	/* end of mapped file region */
	xfs_extnum_t		lastx;	/* last useful extent number */
	int			eof;	/* we've hit the end of extents */
	int			n = 0;	/* current extent index */
	int			error = 0;
4216

L
Linus Torvalds 已提交
4217
	ASSERT(*nmap >= 1);
4218
	ASSERT(*nmap <= XFS_BMAP_MAX_NMAP);
4219
	ASSERT(!(flags & ~XFS_BMAPI_ENTIRE));
4220
	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
4221

L
Linus Torvalds 已提交
4222
	if (unlikely(XFS_TEST_ERROR(
4223 4224
	    (XFS_IFORK_FORMAT(ip, XFS_DATA_FORK) != XFS_DINODE_FMT_EXTENTS &&
	     XFS_IFORK_FORMAT(ip, XFS_DATA_FORK) != XFS_DINODE_FMT_BTREE),
L
Linus Torvalds 已提交
4225
	     mp, XFS_ERRTAG_BMAPIFORMAT, XFS_RANDOM_BMAPIFORMAT))) {
4226
		XFS_ERROR_REPORT("xfs_bmapi_delay", XFS_ERRLEVEL_LOW, mp);
D
Dave Chinner 已提交
4227
		return -EFSCORRUPTED;
L
Linus Torvalds 已提交
4228
	}
4229

L
Linus Torvalds 已提交
4230
	if (XFS_FORCED_SHUTDOWN(mp))
D
Dave Chinner 已提交
4231
		return -EIO;
4232

4233
	XFS_STATS_INC(mp, xs_blk_mapw);
4234 4235

	if (!(ifp->if_flags & XFS_IFEXTENTS)) {
4236
		error = xfs_iread_extents(NULL, ip, XFS_DATA_FORK);
4237
		if (error)
4238
			return error;
4239 4240
	}

4241
	xfs_bmap_search_extents(ip, bno, XFS_DATA_FORK, &eof, &lastx, &got, &prev);
L
Linus Torvalds 已提交
4242 4243
	end = bno + len;
	obno = bno;
4244

L
Linus Torvalds 已提交
4245
	while (bno < end && n < *nmap) {
4246 4247 4248 4249 4250 4251 4252 4253 4254 4255 4256
		if (eof || got.br_startoff > bno) {
			error = xfs_bmapi_reserve_delalloc(ip, bno, len, &got,
							   &prev, &lastx, eof);
			if (error) {
				if (n == 0) {
					*nmap = 0;
					return error;
				}
				break;
			}
		}
4257

4258 4259
		/* set up the extent map to return. */
		xfs_bmapi_trim_map(mval, &got, &bno, len, obno, end, n, flags);
4260 4261
		xfs_bmapi_update_map(&mval, &bno, &len, obno, end, &n, flags);

4262 4263
		/* If we're done, stop now. */
		if (bno >= end || n >= *nmap)
L
Linus Torvalds 已提交
4264
			break;
4265 4266

		/* Else go on to the next record. */
4267 4268 4269 4270
		prev = got;
		if (++lastx < ifp->if_bytes / sizeof(xfs_bmbt_rec_t))
			xfs_bmbt_get_all(xfs_iext_get_ext(ifp, lastx), &got);
		else
4271
			eof = 1;
L
Linus Torvalds 已提交
4272
	}
4273

L
Linus Torvalds 已提交
4274
	*nmap = n;
4275 4276
	return 0;
}
4277

4278

4279 4280
static int
xfs_bmapi_allocate(
4281 4282 4283 4284 4285 4286 4287 4288 4289 4290 4291
	struct xfs_bmalloca	*bma)
{
	struct xfs_mount	*mp = bma->ip->i_mount;
	int			whichfork = (bma->flags & XFS_BMAPI_ATTRFORK) ?
						XFS_ATTR_FORK : XFS_DATA_FORK;
	struct xfs_ifork	*ifp = XFS_IFORK_PTR(bma->ip, whichfork);
	int			tmp_logflags = 0;
	int			error;

	ASSERT(bma->length > 0);

L
Linus Torvalds 已提交
4292
	/*
4293 4294
	 * For the wasdelay case, we could also just allocate the stuff asked
	 * for in this bmap call but that wouldn't be as good.
L
Linus Torvalds 已提交
4295
	 */
4296 4297 4298 4299 4300 4301
	if (bma->wasdel) {
		bma->length = (xfs_extlen_t)bma->got.br_blockcount;
		bma->offset = bma->got.br_startoff;
		if (bma->idx != NULLEXTNUM && bma->idx) {
			xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx - 1),
					 &bma->prev);
L
Linus Torvalds 已提交
4302
		}
4303 4304 4305 4306 4307
	} else {
		bma->length = XFS_FILBLKS_MIN(bma->length, MAXEXTLEN);
		if (!bma->eof)
			bma->length = XFS_FILBLKS_MIN(bma->length,
					bma->got.br_startoff - bma->offset);
L
Linus Torvalds 已提交
4308 4309
	}

4310 4311
	/*
	 * Indicate if this is the first user data in the file, or just any
4312 4313
	 * user data. And if it is userdata, indicate whether it needs to
	 * be initialised to zero during allocation.
4314 4315 4316 4317
	 */
	if (!(bma->flags & XFS_BMAPI_METADATA)) {
		bma->userdata = (bma->offset == 0) ?
			XFS_ALLOC_INITIAL_USER_DATA : XFS_ALLOC_USERDATA;
4318 4319
		if (bma->flags & XFS_BMAPI_ZERO)
			bma->userdata |= XFS_ALLOC_USERDATA_ZERO;
4320
	}
L
Linus Torvalds 已提交
4321

4322
	bma->minlen = (bma->flags & XFS_BMAPI_CONTIG) ? bma->length : 1;
C
Christoph Hellwig 已提交
4323

4324 4325 4326 4327 4328 4329 4330 4331 4332
	/*
	 * Only want to do the alignment at the eof if it is userdata and
	 * allocation length is larger than a stripe unit.
	 */
	if (mp->m_dalign && bma->length >= mp->m_dalign &&
	    !(bma->flags & XFS_BMAPI_METADATA) && whichfork == XFS_DATA_FORK) {
		error = xfs_bmap_isaeof(bma, whichfork);
		if (error)
			return error;
L
Linus Torvalds 已提交
4333
	}
4334

4335 4336
	error = xfs_bmap_alloc(bma);
	if (error)
L
Linus Torvalds 已提交
4337
		return error;
4338 4339 4340 4341 4342 4343

	if (bma->flist->xbf_low)
		bma->minleft = 0;
	if (bma->cur)
		bma->cur->bc_private.b.firstblock = *bma->firstblock;
	if (bma->blkno == NULLFSBLOCK)
L
Linus Torvalds 已提交
4344
		return 0;
4345 4346 4347 4348
	if ((ifp->if_flags & XFS_IFBROOT) && !bma->cur) {
		bma->cur = xfs_bmbt_init_cursor(mp, bma->tp, bma->ip, whichfork);
		bma->cur->bc_private.b.firstblock = *bma->firstblock;
		bma->cur->bc_private.b.flist = bma->flist;
L
Linus Torvalds 已提交
4349
	}
4350 4351 4352 4353 4354 4355 4356 4357 4358 4359 4360 4361 4362 4363
	/*
	 * Bump the number of extents we've allocated
	 * in this call.
	 */
	bma->nallocs++;

	if (bma->cur)
		bma->cur->bc_private.b.flags =
			bma->wasdel ? XFS_BTCUR_BPRV_WASDEL : 0;

	bma->got.br_startoff = bma->offset;
	bma->got.br_startblock = bma->blkno;
	bma->got.br_blockcount = bma->length;
	bma->got.br_state = XFS_EXT_NORM;
4364

L
Linus Torvalds 已提交
4365
	/*
4366 4367
	 * A wasdelay extent has been initialized, so shouldn't be flagged
	 * as unwritten.
L
Linus Torvalds 已提交
4368
	 */
4369 4370 4371 4372 4373 4374 4375 4376 4377 4378 4379 4380 4381 4382 4383 4384 4385 4386 4387 4388 4389 4390 4391 4392 4393 4394 4395 4396 4397 4398 4399 4400 4401 4402 4403 4404 4405 4406 4407 4408 4409 4410 4411 4412 4413 4414 4415 4416 4417 4418 4419 4420 4421 4422 4423 4424 4425 4426 4427 4428 4429
	if (!bma->wasdel && (bma->flags & XFS_BMAPI_PREALLOC) &&
	    xfs_sb_version_hasextflgbit(&mp->m_sb))
		bma->got.br_state = XFS_EXT_UNWRITTEN;

	if (bma->wasdel)
		error = xfs_bmap_add_extent_delay_real(bma);
	else
		error = xfs_bmap_add_extent_hole_real(bma, whichfork);

	bma->logflags |= tmp_logflags;
	if (error)
		return error;

	/*
	 * Update our extent pointer, given that xfs_bmap_add_extent_delay_real
	 * or xfs_bmap_add_extent_hole_real might have merged it into one of
	 * the neighbouring ones.
	 */
	xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx), &bma->got);

	ASSERT(bma->got.br_startoff <= bma->offset);
	ASSERT(bma->got.br_startoff + bma->got.br_blockcount >=
	       bma->offset + bma->length);
	ASSERT(bma->got.br_state == XFS_EXT_NORM ||
	       bma->got.br_state == XFS_EXT_UNWRITTEN);
	return 0;
}

STATIC int
xfs_bmapi_convert_unwritten(
	struct xfs_bmalloca	*bma,
	struct xfs_bmbt_irec	*mval,
	xfs_filblks_t		len,
	int			flags)
{
	int			whichfork = (flags & XFS_BMAPI_ATTRFORK) ?
						XFS_ATTR_FORK : XFS_DATA_FORK;
	struct xfs_ifork	*ifp = XFS_IFORK_PTR(bma->ip, whichfork);
	int			tmp_logflags = 0;
	int			error;

	/* check if we need to do unwritten->real conversion */
	if (mval->br_state == XFS_EXT_UNWRITTEN &&
	    (flags & XFS_BMAPI_PREALLOC))
		return 0;

	/* check if we need to do real->unwritten conversion */
	if (mval->br_state == XFS_EXT_NORM &&
	    (flags & (XFS_BMAPI_PREALLOC | XFS_BMAPI_CONVERT)) !=
			(XFS_BMAPI_PREALLOC | XFS_BMAPI_CONVERT))
		return 0;

	/*
	 * Modify (by adding) the state flag, if writing.
	 */
	ASSERT(mval->br_blockcount <= len);
	if ((ifp->if_flags & XFS_IFBROOT) && !bma->cur) {
		bma->cur = xfs_bmbt_init_cursor(bma->ip->i_mount, bma->tp,
					bma->ip, whichfork);
		bma->cur->bc_private.b.firstblock = *bma->firstblock;
		bma->cur->bc_private.b.flist = bma->flist;
L
Linus Torvalds 已提交
4430
	}
4431 4432
	mval->br_state = (mval->br_state == XFS_EXT_UNWRITTEN)
				? XFS_EXT_NORM : XFS_EXT_UNWRITTEN;
4433

4434 4435 4436 4437 4438 4439 4440 4441 4442 4443 4444
	/*
	 * Before insertion into the bmbt, zero the range being converted
	 * if required.
	 */
	if (flags & XFS_BMAPI_ZERO) {
		error = xfs_zero_extent(bma->ip, mval->br_startblock,
					mval->br_blockcount);
		if (error)
			return error;
	}

4445 4446 4447
	error = xfs_bmap_add_extent_unwritten_real(bma->tp, bma->ip, &bma->idx,
			&bma->cur, mval, bma->firstblock, bma->flist,
			&tmp_logflags);
4448 4449 4450 4451 4452 4453 4454 4455 4456
	/*
	 * Log the inode core unconditionally in the unwritten extent conversion
	 * path because the conversion might not have done so (e.g., if the
	 * extent count hasn't changed). We need to make sure the inode is dirty
	 * in the transaction for the sake of fsync(), even if nothing has
	 * changed, because fsync() will not force the log for this transaction
	 * unless it sees the inode pinned.
	 */
	bma->logflags |= tmp_logflags | XFS_ILOG_CORE;
4457 4458 4459 4460 4461 4462 4463 4464 4465 4466 4467 4468 4469 4470 4471
	if (error)
		return error;

	/*
	 * Update our extent pointer, given that
	 * xfs_bmap_add_extent_unwritten_real might have merged it into one
	 * of the neighbouring ones.
	 */
	xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx), &bma->got);

	/*
	 * We may have combined previously unwritten space with written space,
	 * so generate another request.
	 */
	if (mval->br_blockcount < len)
D
Dave Chinner 已提交
4472
		return -EAGAIN;
4473 4474 4475 4476 4477 4478 4479 4480 4481 4482 4483 4484 4485 4486 4487 4488 4489 4490 4491 4492 4493 4494 4495 4496 4497 4498 4499 4500 4501 4502 4503
	return 0;
}

/*
 * Map file blocks to filesystem blocks, and allocate blocks or convert the
 * extent state if necessary.  Details behaviour is controlled by the flags
 * parameter.  Only allocates blocks from a single allocation group, to avoid
 * locking problems.
 *
 * The returned value in "firstblock" from the first call in a transaction
 * must be remembered and presented to subsequent calls in "firstblock".
 * An upper bound for the number of blocks to be allocated is supplied to
 * the first call in "total"; if no allocation group has that many free
 * blocks then the call will fail (return NULLFSBLOCK in "firstblock").
 */
int
xfs_bmapi_write(
	struct xfs_trans	*tp,		/* transaction pointer */
	struct xfs_inode	*ip,		/* incore inode */
	xfs_fileoff_t		bno,		/* starting file offs. mapped */
	xfs_filblks_t		len,		/* length to map in file */
	int			flags,		/* XFS_BMAPI_... */
	xfs_fsblock_t		*firstblock,	/* first allocated block
						   controls a.g. for allocs */
	xfs_extlen_t		total,		/* total blocks needed */
	struct xfs_bmbt_irec	*mval,		/* output: map values */
	int			*nmap,		/* i/o: mval size/count */
	struct xfs_bmap_free	*flist)		/* i/o: list extents to free */
{
	struct xfs_mount	*mp = ip->i_mount;
	struct xfs_ifork	*ifp;
4504
	struct xfs_bmalloca	bma = { NULL };	/* args for xfs_bmap_alloc */
4505 4506 4507 4508 4509 4510 4511 4512 4513 4514 4515 4516 4517 4518 4519 4520 4521 4522 4523 4524 4525 4526
	xfs_fileoff_t		end;		/* end of mapped file region */
	int			eof;		/* after the end of extents */
	int			error;		/* error return */
	int			n;		/* current extent index */
	xfs_fileoff_t		obno;		/* old block number (offset) */
	int			whichfork;	/* data or attr fork */
	char			inhole;		/* current location is hole in file */
	char			wasdelay;	/* old extent was delayed */

#ifdef DEBUG
	xfs_fileoff_t		orig_bno;	/* original block number value */
	int			orig_flags;	/* original flags arg value */
	xfs_filblks_t		orig_len;	/* original value of len arg */
	struct xfs_bmbt_irec	*orig_mval;	/* original value of mval */
	int			orig_nmap;	/* original value of *nmap */

	orig_bno = bno;
	orig_len = len;
	orig_flags = flags;
	orig_mval = mval;
	orig_nmap = *nmap;
#endif
4527 4528
	whichfork = (flags & XFS_BMAPI_ATTRFORK) ?
		XFS_ATTR_FORK : XFS_DATA_FORK;
4529 4530 4531 4532 4533 4534

	ASSERT(*nmap >= 1);
	ASSERT(*nmap <= XFS_BMAP_MAX_NMAP);
	ASSERT(!(flags & XFS_BMAPI_IGSTATE));
	ASSERT(tp != NULL);
	ASSERT(len > 0);
4535
	ASSERT(XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_LOCAL);
4536
	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
4537

4538 4539 4540 4541 4542 4543 4544 4545 4546 4547 4548 4549
	/* zeroing is for currently only for data extents, not metadata */
	ASSERT((flags & (XFS_BMAPI_METADATA | XFS_BMAPI_ZERO)) !=
			(XFS_BMAPI_METADATA | XFS_BMAPI_ZERO));
	/*
	 * we can allocate unwritten extents or pre-zero allocated blocks,
	 * but it makes no sense to do both at once. This would result in
	 * zeroing the unwritten extent twice, but it still being an
	 * unwritten extent....
	 */
	ASSERT((flags & (XFS_BMAPI_PREALLOC | XFS_BMAPI_ZERO)) !=
			(XFS_BMAPI_PREALLOC | XFS_BMAPI_ZERO));

4550 4551
	if (unlikely(XFS_TEST_ERROR(
	    (XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS &&
4552
	     XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE),
4553 4554
	     mp, XFS_ERRTAG_BMAPIFORMAT, XFS_RANDOM_BMAPIFORMAT))) {
		XFS_ERROR_REPORT("xfs_bmapi_write", XFS_ERRLEVEL_LOW, mp);
D
Dave Chinner 已提交
4555
		return -EFSCORRUPTED;
4556 4557
	}

4558
	if (XFS_FORCED_SHUTDOWN(mp))
D
Dave Chinner 已提交
4559
		return -EIO;
4560 4561 4562

	ifp = XFS_IFORK_PTR(ip, whichfork);

4563
	XFS_STATS_INC(mp, xs_blk_mapw);
4564 4565 4566 4567 4568 4569 4570 4571 4572 4573 4574 4575 4576 4577 4578 4579 4580 4581 4582 4583 4584 4585 4586 4587 4588 4589 4590 4591 4592 4593 4594 4595 4596

	if (*firstblock == NULLFSBLOCK) {
		if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_BTREE)
			bma.minleft = be16_to_cpu(ifp->if_broot->bb_level) + 1;
		else
			bma.minleft = 1;
	} else {
		bma.minleft = 0;
	}

	if (!(ifp->if_flags & XFS_IFEXTENTS)) {
		error = xfs_iread_extents(tp, ip, whichfork);
		if (error)
			goto error0;
	}

	xfs_bmap_search_extents(ip, bno, whichfork, &eof, &bma.idx, &bma.got,
				&bma.prev);
	n = 0;
	end = bno + len;
	obno = bno;

	bma.tp = tp;
	bma.ip = ip;
	bma.total = total;
	bma.userdata = 0;
	bma.flist = flist;
	bma.firstblock = firstblock;

	while (bno < end && n < *nmap) {
		inhole = eof || bma.got.br_startoff > bno;
		wasdelay = !inhole && isnullstartblock(bma.got.br_startblock);

L
Linus Torvalds 已提交
4597
		/*
4598 4599
		 * First, deal with the hole before the allocated space
		 * that we found, if any.
L
Linus Torvalds 已提交
4600
		 */
4601 4602 4603 4604 4605 4606 4607
		if (inhole || wasdelay) {
			bma.eof = eof;
			bma.conv = !!(flags & XFS_BMAPI_CONVERT);
			bma.wasdel = wasdelay;
			bma.offset = bno;
			bma.flags = flags;

L
Linus Torvalds 已提交
4608
			/*
4609 4610 4611 4612 4613
			 * There's a 32/64 bit type mismatch between the
			 * allocation length request (which can be 64 bits in
			 * length) and the bma length request, which is
			 * xfs_extlen_t and therefore 32 bits. Hence we have to
			 * check for 32-bit overflows and handle them here.
L
Linus Torvalds 已提交
4614
			 */
4615 4616 4617 4618 4619 4620 4621 4622
			if (len > (xfs_filblks_t)MAXEXTLEN)
				bma.length = MAXEXTLEN;
			else
				bma.length = len;

			ASSERT(len > 0);
			ASSERT(bma.length > 0);
			error = xfs_bmapi_allocate(&bma);
L
Linus Torvalds 已提交
4623 4624
			if (error)
				goto error0;
4625 4626
			if (bma.blkno == NULLFSBLOCK)
				break;
L
Linus Torvalds 已提交
4627
		}
4628 4629 4630 4631 4632 4633 4634

		/* Deal with the allocated space we found.  */
		xfs_bmapi_trim_map(mval, &bma.got, &bno, len, obno,
							end, n, flags);

		/* Execute unwritten extent conversion if necessary */
		error = xfs_bmapi_convert_unwritten(&bma, mval, len, flags);
D
Dave Chinner 已提交
4635
		if (error == -EAGAIN)
4636 4637 4638 4639 4640 4641 4642 4643 4644 4645 4646 4647 4648 4649 4650 4651 4652 4653 4654 4655 4656 4657 4658 4659 4660 4661 4662 4663 4664 4665 4666 4667 4668 4669 4670 4671 4672 4673 4674 4675 4676 4677 4678 4679 4680 4681 4682 4683 4684 4685 4686 4687 4688 4689 4690 4691 4692 4693 4694 4695 4696 4697 4698 4699 4700 4701 4702 4703 4704 4705 4706 4707 4708
			continue;
		if (error)
			goto error0;

		/* update the extent map to return */
		xfs_bmapi_update_map(&mval, &bno, &len, obno, end, &n, flags);

		/*
		 * If we're done, stop now.  Stop when we've allocated
		 * XFS_BMAP_MAX_NMAP extents no matter what.  Otherwise
		 * the transaction may get too big.
		 */
		if (bno >= end || n >= *nmap || bma.nallocs >= *nmap)
			break;

		/* Else go on to the next record. */
		bma.prev = bma.got;
		if (++bma.idx < ifp->if_bytes / sizeof(xfs_bmbt_rec_t)) {
			xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma.idx),
					 &bma.got);
		} else
			eof = 1;
	}
	*nmap = n;

	/*
	 * Transform from btree to extents, give it cur.
	 */
	if (xfs_bmap_wants_extents(ip, whichfork)) {
		int		tmp_logflags = 0;

		ASSERT(bma.cur);
		error = xfs_bmap_btree_to_extents(tp, ip, bma.cur,
			&tmp_logflags, whichfork);
		bma.logflags |= tmp_logflags;
		if (error)
			goto error0;
	}

	ASSERT(XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE ||
	       XFS_IFORK_NEXTENTS(ip, whichfork) >
		XFS_IFORK_MAXEXT(ip, whichfork));
	error = 0;
error0:
	/*
	 * Log everything.  Do this after conversion, there's no point in
	 * logging the extent records if we've converted to btree format.
	 */
	if ((bma.logflags & xfs_ilog_fext(whichfork)) &&
	    XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS)
		bma.logflags &= ~xfs_ilog_fext(whichfork);
	else if ((bma.logflags & xfs_ilog_fbroot(whichfork)) &&
		 XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE)
		bma.logflags &= ~xfs_ilog_fbroot(whichfork);
	/*
	 * Log whatever the flags say, even if error.  Otherwise we might miss
	 * detecting a case where the data is changed, there's an error,
	 * and it's not logged so we don't shutdown when we should.
	 */
	if (bma.logflags)
		xfs_trans_log_inode(tp, ip, bma.logflags);

	if (bma.cur) {
		if (!error) {
			ASSERT(*firstblock == NULLFSBLOCK ||
			       XFS_FSB_TO_AGNO(mp, *firstblock) ==
			       XFS_FSB_TO_AGNO(mp,
				       bma.cur->bc_private.b.firstblock) ||
			       (flist->xbf_low &&
				XFS_FSB_TO_AGNO(mp, *firstblock) <
				XFS_FSB_TO_AGNO(mp,
					bma.cur->bc_private.b.firstblock)));
			*firstblock = bma.cur->bc_private.b.firstblock;
L
Linus Torvalds 已提交
4709
		}
4710 4711 4712 4713 4714 4715 4716 4717
		xfs_btree_del_cursor(bma.cur,
			error ? XFS_BTREE_ERROR : XFS_BTREE_NOERROR);
	}
	if (!error)
		xfs_bmap_validate_ret(orig_bno, orig_len, orig_flags, orig_mval,
			orig_nmap, *nmap);
	return error;
}
4718

4719 4720 4721 4722 4723 4724 4725
/*
 * When a delalloc extent is split (e.g., due to a hole punch), the original
 * indlen reservation must be shared across the two new extents that are left
 * behind.
 *
 * Given the original reservation and the worst case indlen for the two new
 * extents (as calculated by xfs_bmap_worst_indlen()), split the original
4726 4727 4728 4729
 * reservation fairly across the two new extents. If necessary, steal available
 * blocks from a deleted extent to make up a reservation deficiency (e.g., if
 * ores == 1). The number of stolen blocks is returned. The availability and
 * subsequent accounting of stolen blocks is the responsibility of the caller.
4730
 */
4731
static xfs_filblks_t
4732 4733 4734
xfs_bmap_split_indlen(
	xfs_filblks_t			ores,		/* original res. */
	xfs_filblks_t			*indlen1,	/* ext1 worst indlen */
4735 4736
	xfs_filblks_t			*indlen2,	/* ext2 worst indlen */
	xfs_filblks_t			avail)		/* stealable blocks */
4737 4738 4739 4740
{
	xfs_filblks_t			len1 = *indlen1;
	xfs_filblks_t			len2 = *indlen2;
	xfs_filblks_t			nres = len1 + len2; /* new total res. */
4741 4742 4743 4744 4745 4746 4747 4748 4749 4750 4751
	xfs_filblks_t			stolen = 0;

	/*
	 * Steal as many blocks as we can to try and satisfy the worst case
	 * indlen for both new extents.
	 */
	while (nres > ores && avail) {
		nres--;
		avail--;
		stolen++;
	}
4752 4753

	/*
4754 4755 4756 4757 4758
	 * The only blocks available are those reserved for the original
	 * extent and what we can steal from the extent being removed.
	 * If this still isn't enough to satisfy the combined
	 * requirements for the two new extents, skim blocks off of each
	 * of the new reservations until they match what is available.
4759 4760 4761 4762 4763 4764 4765 4766 4767 4768 4769 4770 4771 4772 4773 4774
	 */
	while (nres > ores) {
		if (len1) {
			len1--;
			nres--;
		}
		if (nres == ores)
			break;
		if (len2) {
			len2--;
			nres--;
		}
	}

	*indlen1 = len1;
	*indlen2 = len2;
4775 4776

	return stolen;
4777 4778
}

4779 4780 4781 4782 4783 4784 4785 4786 4787 4788 4789 4790 4791 4792 4793 4794 4795 4796 4797 4798 4799 4800 4801 4802 4803 4804 4805 4806 4807 4808 4809 4810 4811 4812 4813 4814 4815
/*
 * Called by xfs_bmapi to update file extent records and the btree
 * after removing space (or undoing a delayed allocation).
 */
STATIC int				/* error */
xfs_bmap_del_extent(
	xfs_inode_t		*ip,	/* incore inode pointer */
	xfs_trans_t		*tp,	/* current transaction pointer */
	xfs_extnum_t		*idx,	/* extent number to update/delete */
	xfs_bmap_free_t		*flist,	/* list of extents to be freed */
	xfs_btree_cur_t		*cur,	/* if null, not a btree */
	xfs_bmbt_irec_t		*del,	/* data to remove from extents */
	int			*logflagsp, /* inode logging flags */
	int			whichfork) /* data or attr fork */
{
	xfs_filblks_t		da_new;	/* new delay-alloc indirect blocks */
	xfs_filblks_t		da_old;	/* old delay-alloc indirect blocks */
	xfs_fsblock_t		del_endblock=0;	/* first block past del */
	xfs_fileoff_t		del_endoff;	/* first offset past del */
	int			delay;	/* current block is delayed allocated */
	int			do_fx;	/* free extent at end of routine */
	xfs_bmbt_rec_host_t	*ep;	/* current extent entry pointer */
	int			error;	/* error return value */
	int			flags;	/* inode logging flags */
	xfs_bmbt_irec_t		got;	/* current extent entry */
	xfs_fileoff_t		got_endoff;	/* first offset past got */
	int			i;	/* temp state */
	xfs_ifork_t		*ifp;	/* inode fork pointer */
	xfs_mount_t		*mp;	/* mount structure */
	xfs_filblks_t		nblks;	/* quota/sb block count */
	xfs_bmbt_irec_t		new;	/* new record to be inserted */
	/* REFERENCED */
	uint			qfield;	/* quota field to update */
	xfs_filblks_t		temp;	/* for indirect length calculations */
	xfs_filblks_t		temp2;	/* for indirect length calculations */
	int			state = 0;

4816 4817
	mp = ip->i_mount;
	XFS_STATS_INC(mp, xs_del_exlist);
4818 4819 4820 4821 4822 4823 4824 4825 4826 4827 4828 4829 4830 4831 4832 4833 4834 4835 4836 4837 4838 4839 4840 4841 4842 4843 4844 4845 4846 4847 4848 4849 4850 4851 4852 4853 4854 4855 4856 4857 4858 4859 4860 4861 4862 4863

	if (whichfork == XFS_ATTR_FORK)
		state |= BMAP_ATTRFORK;

	ifp = XFS_IFORK_PTR(ip, whichfork);
	ASSERT((*idx >= 0) && (*idx < ifp->if_bytes /
		(uint)sizeof(xfs_bmbt_rec_t)));
	ASSERT(del->br_blockcount > 0);
	ep = xfs_iext_get_ext(ifp, *idx);
	xfs_bmbt_get_all(ep, &got);
	ASSERT(got.br_startoff <= del->br_startoff);
	del_endoff = del->br_startoff + del->br_blockcount;
	got_endoff = got.br_startoff + got.br_blockcount;
	ASSERT(got_endoff >= del_endoff);
	delay = isnullstartblock(got.br_startblock);
	ASSERT(isnullstartblock(del->br_startblock) == delay);
	flags = 0;
	qfield = 0;
	error = 0;
	/*
	 * If deleting a real allocation, must free up the disk space.
	 */
	if (!delay) {
		flags = XFS_ILOG_CORE;
		/*
		 * Realtime allocation.  Free it and record di_nblocks update.
		 */
		if (whichfork == XFS_DATA_FORK && XFS_IS_REALTIME_INODE(ip)) {
			xfs_fsblock_t	bno;
			xfs_filblks_t	len;

			ASSERT(do_mod(del->br_blockcount,
				      mp->m_sb.sb_rextsize) == 0);
			ASSERT(do_mod(del->br_startblock,
				      mp->m_sb.sb_rextsize) == 0);
			bno = del->br_startblock;
			len = del->br_blockcount;
			do_div(bno, mp->m_sb.sb_rextsize);
			do_div(len, mp->m_sb.sb_rextsize);
			error = xfs_rtfree_extent(tp, bno, (xfs_extlen_t)len);
			if (error)
				goto done;
			do_fx = 0;
			nblks = len * mp->m_sb.sb_rextsize;
			qfield = XFS_TRANS_DQ_RTBCOUNT;
		}
L
Linus Torvalds 已提交
4864
		/*
4865
		 * Ordinary allocation.
L
Linus Torvalds 已提交
4866
		 */
4867 4868 4869 4870
		else {
			do_fx = 1;
			nblks = del->br_blockcount;
			qfield = XFS_TRANS_DQ_BCOUNT;
L
Linus Torvalds 已提交
4871 4872
		}
		/*
4873
		 * Set up del_endblock and cur for later.
L
Linus Torvalds 已提交
4874
		 */
4875 4876 4877 4878 4879 4880
		del_endblock = del->br_startblock + del->br_blockcount;
		if (cur) {
			if ((error = xfs_bmbt_lookup_eq(cur, got.br_startoff,
					got.br_startblock, got.br_blockcount,
					&i)))
				goto done;
4881
			XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
L
Linus Torvalds 已提交
4882
		}
4883 4884 4885 4886 4887 4888
		da_old = da_new = 0;
	} else {
		da_old = startblockval(got.br_startblock);
		da_new = 0;
		nblks = 0;
		do_fx = 0;
L
Linus Torvalds 已提交
4889 4890
	}
	/*
4891 4892
	 * Set flag value to use in switch statement.
	 * Left-contig is 2, right-contig is 1.
L
Linus Torvalds 已提交
4893
	 */
4894 4895 4896 4897 4898 4899 4900 4901 4902 4903 4904 4905 4906 4907 4908 4909 4910 4911 4912 4913 4914
	switch (((got.br_startoff == del->br_startoff) << 1) |
		(got_endoff == del_endoff)) {
	case 3:
		/*
		 * Matches the whole extent.  Delete the entry.
		 */
		xfs_iext_remove(ip, *idx, 1,
				whichfork == XFS_ATTR_FORK ? BMAP_ATTRFORK : 0);
		--*idx;
		if (delay)
			break;

		XFS_IFORK_NEXT_SET(ip, whichfork,
			XFS_IFORK_NEXTENTS(ip, whichfork) - 1);
		flags |= XFS_ILOG_CORE;
		if (!cur) {
			flags |= xfs_ilog_fext(whichfork);
			break;
		}
		if ((error = xfs_btree_delete(cur, &i)))
			goto done;
4915
		XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
4916 4917 4918 4919 4920 4921 4922 4923 4924 4925 4926 4927 4928 4929 4930 4931 4932 4933 4934 4935 4936 4937 4938 4939 4940 4941 4942 4943 4944 4945 4946 4947 4948 4949 4950 4951 4952 4953 4954 4955 4956 4957 4958 4959 4960 4961 4962 4963 4964 4965 4966 4967 4968 4969 4970 4971 4972 4973 4974 4975 4976 4977 4978 4979 4980 4981 4982 4983 4984 4985 4986 4987 4988 4989 4990 4991 4992 4993 4994 4995 4996
		break;

	case 2:
		/*
		 * Deleting the first part of the extent.
		 */
		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
		xfs_bmbt_set_startoff(ep, del_endoff);
		temp = got.br_blockcount - del->br_blockcount;
		xfs_bmbt_set_blockcount(ep, temp);
		if (delay) {
			temp = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip, temp),
				da_old);
			xfs_bmbt_set_startblock(ep, nullstartblock((int)temp));
			trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
			da_new = temp;
			break;
		}
		xfs_bmbt_set_startblock(ep, del_endblock);
		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
		if (!cur) {
			flags |= xfs_ilog_fext(whichfork);
			break;
		}
		if ((error = xfs_bmbt_update(cur, del_endoff, del_endblock,
				got.br_blockcount - del->br_blockcount,
				got.br_state)))
			goto done;
		break;

	case 1:
		/*
		 * Deleting the last part of the extent.
		 */
		temp = got.br_blockcount - del->br_blockcount;
		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
		xfs_bmbt_set_blockcount(ep, temp);
		if (delay) {
			temp = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip, temp),
				da_old);
			xfs_bmbt_set_startblock(ep, nullstartblock((int)temp));
			trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
			da_new = temp;
			break;
		}
		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
		if (!cur) {
			flags |= xfs_ilog_fext(whichfork);
			break;
		}
		if ((error = xfs_bmbt_update(cur, got.br_startoff,
				got.br_startblock,
				got.br_blockcount - del->br_blockcount,
				got.br_state)))
			goto done;
		break;

	case 0:
		/*
		 * Deleting the middle of the extent.
		 */
		temp = del->br_startoff - got.br_startoff;
		trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
		xfs_bmbt_set_blockcount(ep, temp);
		new.br_startoff = del_endoff;
		temp2 = got_endoff - del_endoff;
		new.br_blockcount = temp2;
		new.br_state = got.br_state;
		if (!delay) {
			new.br_startblock = del_endblock;
			flags |= XFS_ILOG_CORE;
			if (cur) {
				if ((error = xfs_bmbt_update(cur,
						got.br_startoff,
						got.br_startblock, temp,
						got.br_state)))
					goto done;
				if ((error = xfs_btree_increment(cur, 0, &i)))
					goto done;
				cur->bc_rec.b = new;
				error = xfs_btree_insert(cur, &i);
D
Dave Chinner 已提交
4997
				if (error && error != -ENOSPC)
4998 4999 5000 5001 5002 5003 5004
					goto done;
				/*
				 * If get no-space back from btree insert,
				 * it tried a split, and we have a zero
				 * block reservation.
				 * Fix up our state and return the error.
				 */
D
Dave Chinner 已提交
5005
				if (error == -ENOSPC) {
5006 5007 5008 5009 5010 5011 5012 5013 5014
					/*
					 * Reset the cursor, don't trust
					 * it after any insert operation.
					 */
					if ((error = xfs_bmbt_lookup_eq(cur,
							got.br_startoff,
							got.br_startblock,
							temp, &i)))
						goto done;
5015 5016
					XFS_WANT_CORRUPTED_GOTO(mp,
								i == 1, done);
5017 5018 5019 5020 5021 5022 5023 5024 5025 5026 5027 5028 5029 5030 5031 5032 5033
					/*
					 * Update the btree record back
					 * to the original value.
					 */
					if ((error = xfs_bmbt_update(cur,
							got.br_startoff,
							got.br_startblock,
							got.br_blockcount,
							got.br_state)))
						goto done;
					/*
					 * Reset the extent record back
					 * to the original value.
					 */
					xfs_bmbt_set_blockcount(ep,
						got.br_blockcount);
					flags = 0;
D
Dave Chinner 已提交
5034
					error = -ENOSPC;
5035 5036
					goto done;
				}
5037
				XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
5038 5039 5040 5041 5042
			} else
				flags |= xfs_ilog_fext(whichfork);
			XFS_IFORK_NEXT_SET(ip, whichfork,
				XFS_IFORK_NEXTENTS(ip, whichfork) + 1);
		} else {
5043
			xfs_filblks_t	stolen;
5044
			ASSERT(whichfork == XFS_DATA_FORK);
5045 5046 5047

			/*
			 * Distribute the original indlen reservation across the
5048 5049 5050
			 * two new extents. Steal blocks from the deleted extent
			 * if necessary. Stealing blocks simply fudges the
			 * fdblocks accounting in xfs_bunmapi().
5051 5052 5053
			 */
			temp = xfs_bmap_worst_indlen(ip, got.br_blockcount);
			temp2 = xfs_bmap_worst_indlen(ip, new.br_blockcount);
5054 5055 5056 5057
			stolen = xfs_bmap_split_indlen(da_old, &temp, &temp2,
						       del->br_blockcount);
			da_new = temp + temp2 - stolen;
			del->br_blockcount -= stolen;
5058 5059

			/*
5060 5061
			 * Set the reservation for each extent. Warn if either
			 * is zero as this can lead to delalloc problems.
5062
			 */
5063
			WARN_ON_ONCE(!temp || !temp2);
5064 5065 5066 5067 5068 5069 5070
			xfs_bmbt_set_startblock(ep, nullstartblock((int)temp));
			new.br_startblock = nullstartblock((int)temp2);
		}
		trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
		xfs_iext_insert(ip, *idx + 1, 1, &new, state);
		++*idx;
		break;
L
Linus Torvalds 已提交
5071 5072
	}
	/*
5073
	 * If we need to, add to list of extents to delete.
L
Linus Torvalds 已提交
5074
	 */
5075
	if (do_fx)
5076 5077
		xfs_bmap_add_free(mp, flist, del->br_startblock,
			del->br_blockcount);
L
Linus Torvalds 已提交
5078
	/*
5079
	 * Adjust inode # blocks in the file.
L
Linus Torvalds 已提交
5080
	 */
5081 5082
	if (nblks)
		ip->i_d.di_nblocks -= nblks;
L
Linus Torvalds 已提交
5083
	/*
5084
	 * Adjust quota data.
L
Linus Torvalds 已提交
5085
	 */
5086 5087 5088 5089 5090 5091 5092 5093
	if (qfield)
		xfs_trans_mod_dquot_byino(tp, ip, qfield, (long)-nblks);

	/*
	 * Account for change in delayed indirect blocks.
	 * Nothing to do for disk quota accounting here.
	 */
	ASSERT(da_old >= da_new);
5094 5095
	if (da_old > da_new)
		xfs_mod_fdblocks(mp, (int64_t)(da_old - da_new), false);
5096 5097
done:
	*logflagsp = flags;
L
Linus Torvalds 已提交
5098 5099 5100
	return error;
}

5101
/*
5102 5103 5104 5105
 * Unmap (remove) blocks from a file.
 * If nexts is nonzero then the number of extents to remove is limited to
 * that value.  If not all extents in the block range can be removed then
 * *done is set.
5106
 */
5107 5108 5109 5110 5111 5112 5113 5114 5115 5116 5117 5118
int						/* error */
xfs_bunmapi(
	xfs_trans_t		*tp,		/* transaction pointer */
	struct xfs_inode	*ip,		/* incore inode */
	xfs_fileoff_t		bno,		/* starting offset to unmap */
	xfs_filblks_t		len,		/* length to unmap in file */
	int			flags,		/* misc flags */
	xfs_extnum_t		nexts,		/* number of extents max */
	xfs_fsblock_t		*firstblock,	/* first allocated block
						   controls a.g. for allocs */
	xfs_bmap_free_t		*flist,		/* i/o: list extents to free */
	int			*done)		/* set if not done yet */
5119
{
5120 5121 5122 5123 5124 5125 5126
	xfs_btree_cur_t		*cur;		/* bmap btree cursor */
	xfs_bmbt_irec_t		del;		/* extent being deleted */
	int			eof;		/* is deleting at eof */
	xfs_bmbt_rec_host_t	*ep;		/* extent record pointer */
	int			error;		/* error return value */
	xfs_extnum_t		extno;		/* extent number in list */
	xfs_bmbt_irec_t		got;		/* current extent record */
E
Eric Sandeen 已提交
5127
	xfs_ifork_t		*ifp;		/* inode fork pointer */
5128 5129 5130 5131 5132 5133 5134 5135 5136 5137 5138 5139
	int			isrt;		/* freeing in rt area */
	xfs_extnum_t		lastx;		/* last extent index used */
	int			logflags;	/* transaction logging flags */
	xfs_extlen_t		mod;		/* rt extent offset */
	xfs_mount_t		*mp;		/* mount structure */
	xfs_extnum_t		nextents;	/* number of file extents */
	xfs_bmbt_irec_t		prev;		/* previous extent record */
	xfs_fileoff_t		start;		/* first file offset deleted */
	int			tmp_logflags;	/* partial logging flags */
	int			wasdel;		/* was a delayed alloc extent */
	int			whichfork;	/* data or attribute fork */
	xfs_fsblock_t		sum;
L
Linus Torvalds 已提交
5140

5141
	trace_xfs_bunmap(ip, bno, len, flags, _RET_IP_);
L
Linus Torvalds 已提交
5142

5143 5144 5145 5146 5147 5148 5149 5150
	whichfork = (flags & XFS_BMAPI_ATTRFORK) ?
		XFS_ATTR_FORK : XFS_DATA_FORK;
	ifp = XFS_IFORK_PTR(ip, whichfork);
	if (unlikely(
	    XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS &&
	    XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE)) {
		XFS_ERROR_REPORT("xfs_bunmapi", XFS_ERRLEVEL_LOW,
				 ip->i_mount);
D
Dave Chinner 已提交
5151
		return -EFSCORRUPTED;
L
Linus Torvalds 已提交
5152
	}
5153 5154
	mp = ip->i_mount;
	if (XFS_FORCED_SHUTDOWN(mp))
D
Dave Chinner 已提交
5155
		return -EIO;
L
Linus Torvalds 已提交
5156

5157
	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
5158 5159
	ASSERT(len > 0);
	ASSERT(nexts >= 0);
L
Linus Torvalds 已提交
5160

5161 5162 5163 5164 5165 5166 5167 5168
	if (!(ifp->if_flags & XFS_IFEXTENTS) &&
	    (error = xfs_iread_extents(tp, ip, whichfork)))
		return error;
	nextents = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t);
	if (nextents == 0) {
		*done = 1;
		return 0;
	}
5169
	XFS_STATS_INC(mp, xs_blk_unmap);
5170 5171 5172 5173 5174
	isrt = (whichfork == XFS_DATA_FORK) && XFS_IS_REALTIME_INODE(ip);
	start = bno;
	bno = start + len - 1;
	ep = xfs_bmap_search_extents(ip, bno, whichfork, &eof, &lastx, &got,
		&prev);
L
Linus Torvalds 已提交
5175

5176 5177 5178 5179 5180 5181 5182 5183 5184 5185 5186 5187 5188 5189 5190 5191 5192 5193 5194 5195 5196 5197 5198 5199 5200 5201
	/*
	 * Check to see if the given block number is past the end of the
	 * file, back up to the last block if so...
	 */
	if (eof) {
		ep = xfs_iext_get_ext(ifp, --lastx);
		xfs_bmbt_get_all(ep, &got);
		bno = got.br_startoff + got.br_blockcount - 1;
	}
	logflags = 0;
	if (ifp->if_flags & XFS_IFBROOT) {
		ASSERT(XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_BTREE);
		cur = xfs_bmbt_init_cursor(mp, tp, ip, whichfork);
		cur->bc_private.b.firstblock = *firstblock;
		cur->bc_private.b.flist = flist;
		cur->bc_private.b.flags = 0;
	} else
		cur = NULL;

	if (isrt) {
		/*
		 * Synchronize by locking the bitmap inode.
		 */
		xfs_ilock(mp->m_rbmip, XFS_ILOCK_EXCL);
		xfs_trans_ijoin(tp, mp->m_rbmip, XFS_ILOCK_EXCL);
	}
5202

5203 5204 5205 5206 5207 5208 5209 5210 5211 5212 5213 5214 5215 5216 5217 5218 5219 5220 5221 5222 5223 5224 5225 5226 5227 5228 5229 5230 5231 5232 5233 5234 5235 5236 5237 5238 5239 5240 5241
	extno = 0;
	while (bno != (xfs_fileoff_t)-1 && bno >= start && lastx >= 0 &&
	       (nexts == 0 || extno < nexts)) {
		/*
		 * Is the found extent after a hole in which bno lives?
		 * Just back up to the previous extent, if so.
		 */
		if (got.br_startoff > bno) {
			if (--lastx < 0)
				break;
			ep = xfs_iext_get_ext(ifp, lastx);
			xfs_bmbt_get_all(ep, &got);
		}
		/*
		 * Is the last block of this extent before the range
		 * we're supposed to delete?  If so, we're done.
		 */
		bno = XFS_FILEOFF_MIN(bno,
			got.br_startoff + got.br_blockcount - 1);
		if (bno < start)
			break;
		/*
		 * Then deal with the (possibly delayed) allocated space
		 * we found.
		 */
		ASSERT(ep != NULL);
		del = got;
		wasdel = isnullstartblock(del.br_startblock);
		if (got.br_startoff < start) {
			del.br_startoff = start;
			del.br_blockcount -= start - got.br_startoff;
			if (!wasdel)
				del.br_startblock += start - got.br_startoff;
		}
		if (del.br_startoff + del.br_blockcount > bno + 1)
			del.br_blockcount = bno + 1 - del.br_startoff;
		sum = del.br_startblock + del.br_blockcount;
		if (isrt &&
		    (mod = do_mod(sum, mp->m_sb.sb_rextsize))) {
5242
			/*
5243 5244 5245 5246 5247
			 * Realtime extent not lined up at the end.
			 * The extent could have been split into written
			 * and unwritten pieces, or we could just be
			 * unmapping part of it.  But we can't really
			 * get rid of part of a realtime extent.
5248
			 */
5249 5250 5251 5252 5253 5254 5255 5256 5257 5258 5259 5260 5261 5262 5263
			if (del.br_state == XFS_EXT_UNWRITTEN ||
			    !xfs_sb_version_hasextflgbit(&mp->m_sb)) {
				/*
				 * This piece is unwritten, or we're not
				 * using unwritten extents.  Skip over it.
				 */
				ASSERT(bno >= mod);
				bno -= mod > del.br_blockcount ?
					del.br_blockcount : mod;
				if (bno < got.br_startoff) {
					if (--lastx >= 0)
						xfs_bmbt_get_all(xfs_iext_get_ext(
							ifp, lastx), &got);
				}
				continue;
L
Linus Torvalds 已提交
5264
			}
T
Tao Ma 已提交
5265
			/*
5266 5267
			 * It's written, turn it unwritten.
			 * This is better than zeroing it.
T
Tao Ma 已提交
5268
			 */
5269
			ASSERT(del.br_state == XFS_EXT_NORM);
5270
			ASSERT(tp->t_blk_res > 0);
5271 5272 5273 5274 5275 5276 5277 5278 5279 5280 5281 5282 5283 5284 5285 5286 5287 5288 5289 5290 5291 5292 5293 5294 5295 5296 5297 5298 5299 5300
			/*
			 * If this spans a realtime extent boundary,
			 * chop it back to the start of the one we end at.
			 */
			if (del.br_blockcount > mod) {
				del.br_startoff += del.br_blockcount - mod;
				del.br_startblock += del.br_blockcount - mod;
				del.br_blockcount = mod;
			}
			del.br_state = XFS_EXT_UNWRITTEN;
			error = xfs_bmap_add_extent_unwritten_real(tp, ip,
					&lastx, &cur, &del, firstblock, flist,
					&logflags);
			if (error)
				goto error0;
			goto nodelete;
		}
		if (isrt && (mod = do_mod(del.br_startblock, mp->m_sb.sb_rextsize))) {
			/*
			 * Realtime extent is lined up at the end but not
			 * at the front.  We'll get rid of full extents if
			 * we can.
			 */
			mod = mp->m_sb.sb_rextsize - mod;
			if (del.br_blockcount > mod) {
				del.br_blockcount -= mod;
				del.br_startoff += mod;
				del.br_startblock += mod;
			} else if ((del.br_startoff == start &&
				    (del.br_state == XFS_EXT_UNWRITTEN ||
5301
				     tp->t_blk_res == 0)) ||
5302 5303 5304 5305 5306 5307 5308 5309 5310 5311 5312 5313 5314 5315
				   !xfs_sb_version_hasextflgbit(&mp->m_sb)) {
				/*
				 * Can't make it unwritten.  There isn't
				 * a full extent here so just skip it.
				 */
				ASSERT(bno >= del.br_blockcount);
				bno -= del.br_blockcount;
				if (got.br_startoff > bno) {
					if (--lastx >= 0) {
						ep = xfs_iext_get_ext(ifp,
								      lastx);
						xfs_bmbt_get_all(ep, &got);
					}
				}
T
Tao Ma 已提交
5316
				continue;
5317 5318 5319 5320 5321 5322 5323 5324 5325 5326 5327 5328 5329 5330 5331 5332 5333 5334 5335 5336 5337 5338 5339 5340 5341 5342 5343 5344 5345 5346 5347 5348 5349 5350 5351 5352 5353
			} else if (del.br_state == XFS_EXT_UNWRITTEN) {
				/*
				 * This one is already unwritten.
				 * It must have a written left neighbor.
				 * Unwrite the killed part of that one and
				 * try again.
				 */
				ASSERT(lastx > 0);
				xfs_bmbt_get_all(xfs_iext_get_ext(ifp,
						lastx - 1), &prev);
				ASSERT(prev.br_state == XFS_EXT_NORM);
				ASSERT(!isnullstartblock(prev.br_startblock));
				ASSERT(del.br_startblock ==
				       prev.br_startblock + prev.br_blockcount);
				if (prev.br_startoff < start) {
					mod = start - prev.br_startoff;
					prev.br_blockcount -= mod;
					prev.br_startblock += mod;
					prev.br_startoff = start;
				}
				prev.br_state = XFS_EXT_UNWRITTEN;
				lastx--;
				error = xfs_bmap_add_extent_unwritten_real(tp,
						ip, &lastx, &cur, &prev,
						firstblock, flist, &logflags);
				if (error)
					goto error0;
				goto nodelete;
			} else {
				ASSERT(del.br_state == XFS_EXT_NORM);
				del.br_state = XFS_EXT_UNWRITTEN;
				error = xfs_bmap_add_extent_unwritten_real(tp,
						ip, &lastx, &cur, &del,
						firstblock, flist, &logflags);
				if (error)
					goto error0;
				goto nodelete;
T
Tao Ma 已提交
5354
			}
L
Linus Torvalds 已提交
5355 5356
		}

5357 5358 5359 5360 5361 5362 5363 5364 5365 5366 5367
		/*
		 * If it's the case where the directory code is running
		 * with no block reservation, and the deleted block is in
		 * the middle of its extent, and the resulting insert
		 * of an extent would cause transformation to btree format,
		 * then reject it.  The calling code will then swap
		 * blocks around instead.
		 * We have to do this now, rather than waiting for the
		 * conversion to btree format, since the transaction
		 * will be dirty.
		 */
5368
		if (!wasdel && tp->t_blk_res == 0 &&
5369 5370 5371 5372 5373 5374
		    XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_EXTENTS &&
		    XFS_IFORK_NEXTENTS(ip, whichfork) >= /* Note the >= */
			XFS_IFORK_MAXEXT(ip, whichfork) &&
		    del.br_startoff > got.br_startoff &&
		    del.br_startoff + del.br_blockcount <
		    got.br_startoff + got.br_blockcount) {
D
Dave Chinner 已提交
5375
			error = -ENOSPC;
5376
			goto error0;
L
Linus Torvalds 已提交
5377
		}
5378 5379 5380 5381 5382 5383 5384 5385 5386 5387 5388 5389 5390 5391 5392 5393 5394 5395 5396 5397 5398 5399 5400 5401 5402 5403 5404 5405 5406 5407

		/*
		 * Unreserve quota and update realtime free space, if
		 * appropriate. If delayed allocation, update the inode delalloc
		 * counter now and wait to update the sb counters as
		 * xfs_bmap_del_extent() might need to borrow some blocks.
		 */
		if (wasdel) {
			ASSERT(startblockval(del.br_startblock) > 0);
			if (isrt) {
				xfs_filblks_t rtexts;

				rtexts = XFS_FSB_TO_B(mp, del.br_blockcount);
				do_div(rtexts, mp->m_sb.sb_rextsize);
				xfs_mod_frextents(mp, (int64_t)rtexts);
				(void)xfs_trans_reserve_quota_nblks(NULL,
					ip, -((long)del.br_blockcount), 0,
					XFS_QMOPT_RES_RTBLKS);
			} else {
				(void)xfs_trans_reserve_quota_nblks(NULL,
					ip, -((long)del.br_blockcount), 0,
					XFS_QMOPT_RES_REGBLKS);
			}
			ip->i_delayed_blks -= del.br_blockcount;
			if (cur)
				cur->bc_private.b.flags |=
					XFS_BTCUR_BPRV_WASDEL;
		} else if (cur)
			cur->bc_private.b.flags &= ~XFS_BTCUR_BPRV_WASDEL;

5408 5409 5410 5411 5412
		error = xfs_bmap_del_extent(ip, tp, &lastx, flist, cur, &del,
				&tmp_logflags, whichfork);
		logflags |= tmp_logflags;
		if (error)
			goto error0;
5413 5414 5415 5416

		if (!isrt && wasdel)
			xfs_mod_fdblocks(mp, (int64_t)del.br_blockcount, false);

5417 5418
		bno = del.br_startoff - 1;
nodelete:
L
Linus Torvalds 已提交
5419
		/*
5420
		 * If not done go on to the next (previous) record.
L
Linus Torvalds 已提交
5421
		 */
5422 5423 5424 5425 5426 5427 5428 5429 5430
		if (bno != (xfs_fileoff_t)-1 && bno >= start) {
			if (lastx >= 0) {
				ep = xfs_iext_get_ext(ifp, lastx);
				if (xfs_bmbt_get_startoff(ep) > bno) {
					if (--lastx >= 0)
						ep = xfs_iext_get_ext(ifp,
								      lastx);
				}
				xfs_bmbt_get_all(ep, &got);
L
Linus Torvalds 已提交
5431
			}
5432
			extno++;
L
Linus Torvalds 已提交
5433 5434
		}
	}
5435
	*done = bno == (xfs_fileoff_t)-1 || bno < start || lastx < 0;
5436

L
Linus Torvalds 已提交
5437
	/*
5438
	 * Convert to a btree if necessary.
L
Linus Torvalds 已提交
5439
	 */
5440 5441 5442 5443 5444 5445 5446
	if (xfs_bmap_needs_btree(ip, whichfork)) {
		ASSERT(cur == NULL);
		error = xfs_bmap_extents_to_btree(tp, ip, firstblock, flist,
			&cur, 0, &tmp_logflags, whichfork);
		logflags |= tmp_logflags;
		if (error)
			goto error0;
L
Linus Torvalds 已提交
5447 5448
	}
	/*
5449
	 * transform from btree to extents, give it cur
L
Linus Torvalds 已提交
5450
	 */
5451 5452 5453 5454 5455 5456 5457 5458
	else if (xfs_bmap_wants_extents(ip, whichfork)) {
		ASSERT(cur != NULL);
		error = xfs_bmap_btree_to_extents(tp, ip, cur, &tmp_logflags,
			whichfork);
		logflags |= tmp_logflags;
		if (error)
			goto error0;
	}
L
Linus Torvalds 已提交
5459
	/*
5460
	 * transform from extents to local?
L
Linus Torvalds 已提交
5461
	 */
5462 5463 5464 5465 5466 5467 5468 5469 5470 5471 5472 5473 5474 5475 5476 5477 5478 5479 5480 5481 5482 5483 5484 5485 5486 5487 5488 5489
	error = 0;
error0:
	/*
	 * Log everything.  Do this after conversion, there's no point in
	 * logging the extent records if we've converted to btree format.
	 */
	if ((logflags & xfs_ilog_fext(whichfork)) &&
	    XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS)
		logflags &= ~xfs_ilog_fext(whichfork);
	else if ((logflags & xfs_ilog_fbroot(whichfork)) &&
		 XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE)
		logflags &= ~xfs_ilog_fbroot(whichfork);
	/*
	 * Log inode even in the error case, if the transaction
	 * is dirty we'll need to shut down the filesystem.
	 */
	if (logflags)
		xfs_trans_log_inode(tp, ip, logflags);
	if (cur) {
		if (!error) {
			*firstblock = cur->bc_private.b.firstblock;
			cur->bc_private.b.allocated = 0;
		}
		xfs_btree_del_cursor(cur,
			error ? XFS_BTREE_ERROR : XFS_BTREE_NOERROR);
	}
	return error;
}
5490

5491 5492 5493 5494 5495 5496 5497 5498 5499 5500 5501 5502 5503 5504 5505 5506 5507 5508 5509 5510 5511 5512 5513 5514 5515 5516 5517 5518 5519 5520 5521 5522 5523 5524 5525 5526 5527 5528 5529 5530 5531 5532 5533 5534 5535 5536 5537 5538 5539 5540 5541
/*
 * Determine whether an extent shift can be accomplished by a merge with the
 * extent that precedes the target hole of the shift.
 */
STATIC bool
xfs_bmse_can_merge(
	struct xfs_bmbt_irec	*left,	/* preceding extent */
	struct xfs_bmbt_irec	*got,	/* current extent to shift */
	xfs_fileoff_t		shift)	/* shift fsb */
{
	xfs_fileoff_t		startoff;

	startoff = got->br_startoff - shift;

	/*
	 * The extent, once shifted, must be adjacent in-file and on-disk with
	 * the preceding extent.
	 */
	if ((left->br_startoff + left->br_blockcount != startoff) ||
	    (left->br_startblock + left->br_blockcount != got->br_startblock) ||
	    (left->br_state != got->br_state) ||
	    (left->br_blockcount + got->br_blockcount > MAXEXTLEN))
		return false;

	return true;
}

/*
 * A bmap extent shift adjusts the file offset of an extent to fill a preceding
 * hole in the file. If an extent shift would result in the extent being fully
 * adjacent to the extent that currently precedes the hole, we can merge with
 * the preceding extent rather than do the shift.
 *
 * This function assumes the caller has verified a shift-by-merge is possible
 * with the provided extents via xfs_bmse_can_merge().
 */
STATIC int
xfs_bmse_merge(
	struct xfs_inode		*ip,
	int				whichfork,
	xfs_fileoff_t			shift,		/* shift fsb */
	int				current_ext,	/* idx of gotp */
	struct xfs_bmbt_rec_host	*gotp,		/* extent to shift */
	struct xfs_bmbt_rec_host	*leftp,		/* preceding extent */
	struct xfs_btree_cur		*cur,
	int				*logflags)	/* output */
{
	struct xfs_bmbt_irec		got;
	struct xfs_bmbt_irec		left;
	xfs_filblks_t			blockcount;
	int				error, i;
5542
	struct xfs_mount		*mp = ip->i_mount;
5543 5544 5545 5546 5547 5548 5549 5550 5551 5552 5553 5554 5555 5556 5557 5558 5559 5560 5561 5562 5563 5564 5565 5566 5567 5568 5569 5570 5571 5572 5573 5574 5575

	xfs_bmbt_get_all(gotp, &got);
	xfs_bmbt_get_all(leftp, &left);
	blockcount = left.br_blockcount + got.br_blockcount;

	ASSERT(xfs_isilocked(ip, XFS_IOLOCK_EXCL));
	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
	ASSERT(xfs_bmse_can_merge(&left, &got, shift));

	/*
	 * Merge the in-core extents. Note that the host record pointers and
	 * current_ext index are invalid once the extent has been removed via
	 * xfs_iext_remove().
	 */
	xfs_bmbt_set_blockcount(leftp, blockcount);
	xfs_iext_remove(ip, current_ext, 1, 0);

	/*
	 * Update the on-disk extent count, the btree if necessary and log the
	 * inode.
	 */
	XFS_IFORK_NEXT_SET(ip, whichfork,
			   XFS_IFORK_NEXTENTS(ip, whichfork) - 1);
	*logflags |= XFS_ILOG_CORE;
	if (!cur) {
		*logflags |= XFS_ILOG_DEXT;
		return 0;
	}

	/* lookup and remove the extent to merge */
	error = xfs_bmbt_lookup_eq(cur, got.br_startoff, got.br_startblock,
				   got.br_blockcount, &i);
	if (error)
D
Dave Chinner 已提交
5576
		return error;
5577
	XFS_WANT_CORRUPTED_RETURN(mp, i == 1);
5578 5579 5580

	error = xfs_btree_delete(cur, &i);
	if (error)
D
Dave Chinner 已提交
5581
		return error;
5582
	XFS_WANT_CORRUPTED_RETURN(mp, i == 1);
5583 5584 5585 5586 5587

	/* lookup and update size of the previous extent */
	error = xfs_bmbt_lookup_eq(cur, left.br_startoff, left.br_startblock,
				   left.br_blockcount, &i);
	if (error)
D
Dave Chinner 已提交
5588
		return error;
5589
	XFS_WANT_CORRUPTED_RETURN(mp, i == 1);
5590 5591 5592

	left.br_blockcount = blockcount;

D
Dave Chinner 已提交
5593 5594
	return xfs_bmbt_update(cur, left.br_startoff, left.br_startblock,
			       left.br_blockcount, left.br_state);
5595 5596
}

5597 5598 5599 5600 5601 5602 5603 5604 5605 5606 5607
/*
 * Shift a single extent.
 */
STATIC int
xfs_bmse_shift_one(
	struct xfs_inode		*ip,
	int				whichfork,
	xfs_fileoff_t			offset_shift_fsb,
	int				*current_ext,
	struct xfs_bmbt_rec_host	*gotp,
	struct xfs_btree_cur		*cur,
5608 5609
	int				*logflags,
	enum shift_direction		direction)
5610 5611
{
	struct xfs_ifork		*ifp;
5612
	struct xfs_mount		*mp;
5613
	xfs_fileoff_t			startoff;
5614
	struct xfs_bmbt_rec_host	*adj_irecp;
5615
	struct xfs_bmbt_irec		got;
5616
	struct xfs_bmbt_irec		adj_irec;
5617 5618
	int				error;
	int				i;
5619
	int				total_extents;
5620

5621
	mp = ip->i_mount;
5622
	ifp = XFS_IFORK_PTR(ip, whichfork);
5623
	total_extents = ifp->if_bytes / sizeof(xfs_bmbt_rec_t);
5624 5625 5626

	xfs_bmbt_get_all(gotp, &got);

5627
	/* delalloc extents should be prevented by caller */
5628
	XFS_WANT_CORRUPTED_RETURN(mp, !isnullstartblock(got.br_startblock));
5629

5630 5631 5632 5633 5634 5635 5636 5637 5638 5639 5640 5641 5642 5643 5644 5645 5646 5647 5648
	if (direction == SHIFT_LEFT) {
		startoff = got.br_startoff - offset_shift_fsb;

		/*
		 * Check for merge if we've got an extent to the left,
		 * otherwise make sure there's enough room at the start
		 * of the file for the shift.
		 */
		if (!*current_ext) {
			if (got.br_startoff < offset_shift_fsb)
				return -EINVAL;
			goto update_current_ext;
		}
		/*
		 * grab the left extent and check for a large
		 * enough hole.
		 */
		adj_irecp = xfs_iext_get_ext(ifp, *current_ext - 1);
		xfs_bmbt_get_all(adj_irecp, &adj_irec);
5649

5650 5651
		if (startoff <
		    adj_irec.br_startoff + adj_irec.br_blockcount)
5652
			return -EINVAL;
5653

5654
		/* check whether to merge the extent or shift it down */
5655 5656
		if (xfs_bmse_can_merge(&adj_irec, &got,
				       offset_shift_fsb)) {
5657
			return xfs_bmse_merge(ip, whichfork, offset_shift_fsb,
5658 5659
					      *current_ext, gotp, adj_irecp,
					      cur, logflags);
5660
		}
5661 5662 5663 5664 5665 5666 5667 5668 5669 5670 5671 5672 5673 5674 5675 5676 5677 5678 5679 5680 5681 5682 5683 5684
	} else {
		startoff = got.br_startoff + offset_shift_fsb;
		/* nothing to move if this is the last extent */
		if (*current_ext >= (total_extents - 1))
			goto update_current_ext;
		/*
		 * If this is not the last extent in the file, make sure there
		 * is enough room between current extent and next extent for
		 * accommodating the shift.
		 */
		adj_irecp = xfs_iext_get_ext(ifp, *current_ext + 1);
		xfs_bmbt_get_all(adj_irecp, &adj_irec);
		if (startoff + got.br_blockcount > adj_irec.br_startoff)
			return -EINVAL;
		/*
		 * Unlike a left shift (which involves a hole punch),
		 * a right shift does not modify extent neighbors
		 * in any way. We should never find mergeable extents
		 * in this scenario. Check anyways and warn if we
		 * encounter two extents that could be one.
		 */
		if (xfs_bmse_can_merge(&got, &adj_irec, offset_shift_fsb))
			WARN_ON_ONCE(1);
	}
5685 5686 5687 5688
	/*
	 * Increment the extent index for the next iteration, update the start
	 * offset of the in-core extent and update the btree if applicable.
	 */
5689 5690 5691 5692 5693
update_current_ext:
	if (direction == SHIFT_LEFT)
		(*current_ext)++;
	else
		(*current_ext)--;
5694 5695 5696 5697 5698 5699 5700 5701 5702 5703 5704
	xfs_bmbt_set_startoff(gotp, startoff);
	*logflags |= XFS_ILOG_CORE;
	if (!cur) {
		*logflags |= XFS_ILOG_DEXT;
		return 0;
	}

	error = xfs_bmbt_lookup_eq(cur, got.br_startoff, got.br_startblock,
				   got.br_blockcount, &i);
	if (error)
		return error;
5705
	XFS_WANT_CORRUPTED_RETURN(mp, i == 1);
5706 5707

	got.br_startoff = startoff;
5708
	return xfs_bmbt_update(cur, got.br_startoff, got.br_startblock,
5709
			       got.br_blockcount, got.br_state);
5710 5711
}

5712
/*
5713
 * Shift extent records to the left/right to cover/create a hole.
5714
 *
5715
 * The maximum number of extents to be shifted in a single operation is
5716
 * @num_exts. @stop_fsb specifies the file offset at which to stop shift and the
5717 5718 5719 5720
 * file offset where we've left off is returned in @next_fsb. @offset_shift_fsb
 * is the length by which each extent is shifted. If there is no hole to shift
 * the extents into, this will be considered invalid operation and we abort
 * immediately.
5721 5722 5723 5724 5725
 */
int
xfs_bmap_shift_extents(
	struct xfs_trans	*tp,
	struct xfs_inode	*ip,
5726
	xfs_fileoff_t		*next_fsb,
5727
	xfs_fileoff_t		offset_shift_fsb,
5728
	int			*done,
5729
	xfs_fileoff_t		stop_fsb,
5730 5731
	xfs_fsblock_t		*firstblock,
	struct xfs_bmap_free	*flist,
5732
	enum shift_direction	direction,
5733 5734
	int			num_exts)
{
5735
	struct xfs_btree_cur		*cur = NULL;
5736 5737 5738 5739 5740
	struct xfs_bmbt_rec_host	*gotp;
	struct xfs_bmbt_irec            got;
	struct xfs_mount		*mp = ip->i_mount;
	struct xfs_ifork		*ifp;
	xfs_extnum_t			nexts = 0;
5741
	xfs_extnum_t			current_ext;
5742 5743
	xfs_extnum_t			total_extents;
	xfs_extnum_t			stop_extent;
5744 5745
	int				error = 0;
	int				whichfork = XFS_DATA_FORK;
5746
	int				logflags = 0;
5747 5748 5749 5750 5751 5752 5753

	if (unlikely(XFS_TEST_ERROR(
	    (XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS &&
	     XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE),
	     mp, XFS_ERRTAG_BMAPIFORMAT, XFS_RANDOM_BMAPIFORMAT))) {
		XFS_ERROR_REPORT("xfs_bmap_shift_extents",
				 XFS_ERRLEVEL_LOW, mp);
D
Dave Chinner 已提交
5754
		return -EFSCORRUPTED;
5755 5756 5757
	}

	if (XFS_FORCED_SHUTDOWN(mp))
D
Dave Chinner 已提交
5758
		return -EIO;
5759

5760 5761
	ASSERT(xfs_isilocked(ip, XFS_IOLOCK_EXCL));
	ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
5762 5763
	ASSERT(direction == SHIFT_LEFT || direction == SHIFT_RIGHT);
	ASSERT(*next_fsb != NULLFSBLOCK || direction == SHIFT_RIGHT);
5764 5765 5766 5767 5768 5769 5770 5771 5772

	ifp = XFS_IFORK_PTR(ip, whichfork);
	if (!(ifp->if_flags & XFS_IFEXTENTS)) {
		/* Read in all the extents */
		error = xfs_iread_extents(tp, ip, whichfork);
		if (error)
			return error;
	}

5773 5774 5775 5776 5777 5778 5779
	if (ifp->if_flags & XFS_IFBROOT) {
		cur = xfs_bmbt_init_cursor(mp, tp, ip, whichfork);
		cur->bc_private.b.firstblock = *firstblock;
		cur->bc_private.b.flist = flist;
		cur->bc_private.b.flags = 0;
	}

5780 5781 5782 5783 5784 5785 5786 5787 5788 5789 5790 5791 5792 5793 5794 5795 5796 5797 5798 5799 5800 5801 5802 5803 5804 5805 5806 5807 5808 5809 5810 5811
	/*
	 * There may be delalloc extents in the data fork before the range we
	 * are collapsing out, so we cannot use the count of real extents here.
	 * Instead we have to calculate it from the incore fork.
	 */
	total_extents = ifp->if_bytes / sizeof(xfs_bmbt_rec_t);
	if (total_extents == 0) {
		*done = 1;
		goto del_cursor;
	}

	/*
	 * In case of first right shift, we need to initialize next_fsb
	 */
	if (*next_fsb == NULLFSBLOCK) {
		gotp = xfs_iext_get_ext(ifp, total_extents - 1);
		xfs_bmbt_get_all(gotp, &got);
		*next_fsb = got.br_startoff;
		if (stop_fsb > *next_fsb) {
			*done = 1;
			goto del_cursor;
		}
	}

	/* Lookup the extent index at which we have to stop */
	if (direction == SHIFT_RIGHT) {
		gotp = xfs_iext_bno_to_ext(ifp, stop_fsb, &stop_extent);
		/* Make stop_extent exclusive of shift range */
		stop_extent--;
	} else
		stop_extent = total_extents;

5812
	/*
5813 5814 5815 5816 5817
	 * Look up the extent index for the fsb where we start shifting. We can
	 * henceforth iterate with current_ext as extent list changes are locked
	 * out via ilock.
	 *
	 * gotp can be null in 2 cases: 1) if there are no extents or 2)
5818
	 * *next_fsb lies in a hole beyond which there are no extents. Either
5819
	 * way, we are done.
5820
	 */
5821
	gotp = xfs_iext_bno_to_ext(ifp, *next_fsb, &current_ext);
5822 5823
	if (!gotp) {
		*done = 1;
5824
		goto del_cursor;
5825 5826
	}

5827 5828 5829 5830 5831 5832 5833 5834
	/* some sanity checking before we finally start shifting extents */
	if ((direction == SHIFT_LEFT && current_ext >= stop_extent) ||
	     (direction == SHIFT_RIGHT && current_ext <= stop_extent)) {
		error = -EIO;
		goto del_cursor;
	}

	while (nexts++ < num_exts) {
5835
		error = xfs_bmse_shift_one(ip, whichfork, offset_shift_fsb,
5836 5837
					   &current_ext, gotp, cur, &logflags,
					   direction);
5838 5839
		if (error)
			goto del_cursor;
5840 5841 5842 5843 5844 5845 5846 5847
		/*
		 * If there was an extent merge during the shift, the extent
		 * count can change. Update the total and grade the next record.
		 */
		if (direction == SHIFT_LEFT) {
			total_extents = ifp->if_bytes / sizeof(xfs_bmbt_rec_t);
			stop_extent = total_extents;
		}
5848

5849 5850 5851
		if (current_ext == stop_extent) {
			*done = 1;
			*next_fsb = NULLFSBLOCK;
5852
			break;
5853
		}
5854
		gotp = xfs_iext_get_ext(ifp, current_ext);
5855 5856
	}

5857
	if (!*done) {
5858
		xfs_bmbt_get_all(gotp, &got);
5859
		*next_fsb = got.br_startoff;
5860
	}
5861 5862 5863 5864 5865 5866

del_cursor:
	if (cur)
		xfs_btree_del_cursor(cur,
			error ? XFS_BTREE_ERROR : XFS_BTREE_NOERROR);

5867 5868
	if (logflags)
		xfs_trans_log_inode(tp, ip, logflags);
5869

5870 5871
	return error;
}
5872 5873 5874 5875 5876 5877 5878 5879 5880 5881 5882 5883 5884 5885 5886 5887 5888 5889 5890 5891 5892 5893 5894 5895 5896 5897 5898 5899 5900 5901 5902 5903 5904 5905 5906 5907 5908 5909 5910 5911 5912 5913 5914 5915 5916 5917 5918 5919 5920 5921 5922 5923 5924 5925 5926 5927 5928 5929 5930 5931 5932 5933 5934 5935 5936 5937 5938 5939 5940 5941 5942 5943 5944 5945 5946 5947 5948 5949 5950 5951 5952 5953 5954 5955 5956 5957 5958 5959 5960 5961 5962 5963 5964 5965 5966 5967 5968 5969 5970 5971 5972 5973 5974 5975 5976 5977 5978 5979 5980 5981 5982 5983 5984 5985 5986 5987 5988 5989 5990 5991 5992 5993 5994 5995 5996 5997 5998 5999 6000 6001 6002 6003 6004 6005 6006 6007 6008 6009 6010 6011 6012 6013 6014 6015 6016 6017 6018 6019 6020 6021 6022 6023 6024 6025 6026 6027

/*
 * Splits an extent into two extents at split_fsb block such that it is
 * the first block of the current_ext. @current_ext is a target extent
 * to be split. @split_fsb is a block where the extents is split.
 * If split_fsb lies in a hole or the first block of extents, just return 0.
 */
STATIC int
xfs_bmap_split_extent_at(
	struct xfs_trans	*tp,
	struct xfs_inode	*ip,
	xfs_fileoff_t		split_fsb,
	xfs_fsblock_t		*firstfsb,
	struct xfs_bmap_free	*free_list)
{
	int				whichfork = XFS_DATA_FORK;
	struct xfs_btree_cur		*cur = NULL;
	struct xfs_bmbt_rec_host	*gotp;
	struct xfs_bmbt_irec		got;
	struct xfs_bmbt_irec		new; /* split extent */
	struct xfs_mount		*mp = ip->i_mount;
	struct xfs_ifork		*ifp;
	xfs_fsblock_t			gotblkcnt; /* new block count for got */
	xfs_extnum_t			current_ext;
	int				error = 0;
	int				logflags = 0;
	int				i = 0;

	if (unlikely(XFS_TEST_ERROR(
	    (XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS &&
	     XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE),
	     mp, XFS_ERRTAG_BMAPIFORMAT, XFS_RANDOM_BMAPIFORMAT))) {
		XFS_ERROR_REPORT("xfs_bmap_split_extent_at",
				 XFS_ERRLEVEL_LOW, mp);
		return -EFSCORRUPTED;
	}

	if (XFS_FORCED_SHUTDOWN(mp))
		return -EIO;

	ifp = XFS_IFORK_PTR(ip, whichfork);
	if (!(ifp->if_flags & XFS_IFEXTENTS)) {
		/* Read in all the extents */
		error = xfs_iread_extents(tp, ip, whichfork);
		if (error)
			return error;
	}

	/*
	 * gotp can be null in 2 cases: 1) if there are no extents
	 * or 2) split_fsb lies in a hole beyond which there are
	 * no extents. Either way, we are done.
	 */
	gotp = xfs_iext_bno_to_ext(ifp, split_fsb, &current_ext);
	if (!gotp)
		return 0;

	xfs_bmbt_get_all(gotp, &got);

	/*
	 * Check split_fsb lies in a hole or the start boundary offset
	 * of the extent.
	 */
	if (got.br_startoff >= split_fsb)
		return 0;

	gotblkcnt = split_fsb - got.br_startoff;
	new.br_startoff = split_fsb;
	new.br_startblock = got.br_startblock + gotblkcnt;
	new.br_blockcount = got.br_blockcount - gotblkcnt;
	new.br_state = got.br_state;

	if (ifp->if_flags & XFS_IFBROOT) {
		cur = xfs_bmbt_init_cursor(mp, tp, ip, whichfork);
		cur->bc_private.b.firstblock = *firstfsb;
		cur->bc_private.b.flist = free_list;
		cur->bc_private.b.flags = 0;
		error = xfs_bmbt_lookup_eq(cur, got.br_startoff,
				got.br_startblock,
				got.br_blockcount,
				&i);
		if (error)
			goto del_cursor;
		XFS_WANT_CORRUPTED_GOTO(mp, i == 1, del_cursor);
	}

	xfs_bmbt_set_blockcount(gotp, gotblkcnt);
	got.br_blockcount = gotblkcnt;

	logflags = XFS_ILOG_CORE;
	if (cur) {
		error = xfs_bmbt_update(cur, got.br_startoff,
				got.br_startblock,
				got.br_blockcount,
				got.br_state);
		if (error)
			goto del_cursor;
	} else
		logflags |= XFS_ILOG_DEXT;

	/* Add new extent */
	current_ext++;
	xfs_iext_insert(ip, current_ext, 1, &new, 0);
	XFS_IFORK_NEXT_SET(ip, whichfork,
			   XFS_IFORK_NEXTENTS(ip, whichfork) + 1);

	if (cur) {
		error = xfs_bmbt_lookup_eq(cur, new.br_startoff,
				new.br_startblock, new.br_blockcount,
				&i);
		if (error)
			goto del_cursor;
		XFS_WANT_CORRUPTED_GOTO(mp, i == 0, del_cursor);
		cur->bc_rec.b.br_state = new.br_state;

		error = xfs_btree_insert(cur, &i);
		if (error)
			goto del_cursor;
		XFS_WANT_CORRUPTED_GOTO(mp, i == 1, del_cursor);
	}

	/*
	 * Convert to a btree if necessary.
	 */
	if (xfs_bmap_needs_btree(ip, whichfork)) {
		int tmp_logflags; /* partial log flag return val */

		ASSERT(cur == NULL);
		error = xfs_bmap_extents_to_btree(tp, ip, firstfsb, free_list,
				&cur, 0, &tmp_logflags, whichfork);
		logflags |= tmp_logflags;
	}

del_cursor:
	if (cur) {
		cur->bc_private.b.allocated = 0;
		xfs_btree_del_cursor(cur,
				error ? XFS_BTREE_ERROR : XFS_BTREE_NOERROR);
	}

	if (logflags)
		xfs_trans_log_inode(tp, ip, logflags);
	return error;
}

int
xfs_bmap_split_extent(
	struct xfs_inode        *ip,
	xfs_fileoff_t           split_fsb)
{
	struct xfs_mount        *mp = ip->i_mount;
	struct xfs_trans        *tp;
	struct xfs_bmap_free    free_list;
	xfs_fsblock_t           firstfsb;
	int                     error;

6028 6029 6030
	error = xfs_trans_alloc(mp, &M_RES(mp)->tr_write,
			XFS_DIOSTRAT_SPACE_RES(mp, 0), 0, 0, &tp);
	if (error)
6031 6032 6033 6034 6035 6036 6037 6038 6039 6040 6041 6042
		return error;

	xfs_ilock(ip, XFS_ILOCK_EXCL);
	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);

	xfs_bmap_init(&free_list, &firstfsb);

	error = xfs_bmap_split_extent_at(tp, ip, split_fsb,
			&firstfsb, &free_list);
	if (error)
		goto out;

6043
	error = xfs_bmap_finish(&tp, &free_list, NULL);
6044 6045 6046
	if (error)
		goto out;

6047
	return xfs_trans_commit(tp);
6048 6049

out:
6050
	xfs_bmap_cancel(&free_list);
6051
	xfs_trans_cancel(tp);
6052 6053
	return error;
}