xfs_bmap_btree.c 19.9 KB
Newer Older
L
Linus Torvalds 已提交
1
/*
2 3
 * Copyright (c) 2000-2003,2005 Silicon Graphics, Inc.
 * 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
#include "xfs_format.h"
22 23
#include "xfs_log_format.h"
#include "xfs_trans_resv.h"
24
#include "xfs_bit.h"
L
Linus Torvalds 已提交
25
#include "xfs_mount.h"
26
#include "xfs_defer.h"
L
Linus Torvalds 已提交
27
#include "xfs_inode.h"
28
#include "xfs_trans.h"
29
#include "xfs_inode_item.h"
L
Linus Torvalds 已提交
30
#include "xfs_alloc.h"
31
#include "xfs_btree.h"
32
#include "xfs_bmap_btree.h"
L
Linus Torvalds 已提交
33 34 35
#include "xfs_bmap.h"
#include "xfs_error.h"
#include "xfs_quota.h"
36
#include "xfs_trace.h"
37
#include "xfs_cksum.h"
38
#include "xfs_rmap.h"
L
Linus Torvalds 已提交
39 40 41 42 43 44

/*
 * Convert on-disk form of btree root to in-memory form.
 */
void
xfs_bmdr_to_bmbt(
45
	struct xfs_inode	*ip,
L
Linus Torvalds 已提交
46 47
	xfs_bmdr_block_t	*dblock,
	int			dblocklen,
48
	struct xfs_btree_block	*rblock,
L
Linus Torvalds 已提交
49 50
	int			rblocklen)
{
51
	struct xfs_mount	*mp = ip->i_mount;
L
Linus Torvalds 已提交
52 53
	int			dmxr;
	xfs_bmbt_key_t		*fkp;
54
	__be64			*fpp;
L
Linus Torvalds 已提交
55
	xfs_bmbt_key_t		*tkp;
56
	__be64			*tpp;
L
Linus Torvalds 已提交
57

58 59
	xfs_btree_init_block_int(mp, rblock, XFS_BUF_DADDR_NULL,
				 XFS_BTNUM_BMAP, 0, 0, ip->i_ino,
60
				 XFS_BTREE_LONG_PTRS);
61 62 63
	rblock->bb_level = dblock->bb_level;
	ASSERT(be16_to_cpu(rblock->bb_level) > 0);
	rblock->bb_numrecs = dblock->bb_numrecs;
64
	dmxr = xfs_bmdr_maxrecs(dblocklen, 0);
65 66 67
	fkp = XFS_BMDR_KEY_ADDR(dblock, 1);
	tkp = XFS_BMBT_KEY_ADDR(mp, rblock, 1);
	fpp = XFS_BMDR_PTR_ADDR(dblock, 1, dmxr);
68
	tpp = XFS_BMAP_BROOT_PTR_ADDR(mp, rblock, 1, rblocklen);
69
	dmxr = be16_to_cpu(dblock->bb_numrecs);
L
Linus Torvalds 已提交
70
	memcpy(tkp, fkp, sizeof(*fkp) * dmxr);
71
	memcpy(tpp, fpp, sizeof(*fpp) * dmxr);
L
Linus Torvalds 已提交
72 73 74 75 76
}

/*
 * Convert a compressed bmap extent record to an uncompressed form.
 * This code must be in sync with the routines xfs_bmbt_get_startoff,
C
Christoph Hellwig 已提交
77
 * xfs_bmbt_get_startblock and xfs_bmbt_get_blockcount.
L
Linus Torvalds 已提交
78
 */
79
STATIC void
L
Linus Torvalds 已提交
80
__xfs_bmbt_get_all(
81 82
		uint64_t l0,
		uint64_t l1,
L
Linus Torvalds 已提交
83 84 85 86 87 88 89
		xfs_bmbt_irec_t *s)
{
	int	ext_flag;
	xfs_exntst_t st;

	ext_flag = (int)(l0 >> (64 - BMBT_EXNTFLAG_BITLEN));
	s->br_startoff = ((xfs_fileoff_t)l0 &
90 91
			   xfs_mask64lo(64 - BMBT_EXNTFLAG_BITLEN)) >> 9;
	s->br_startblock = (((xfs_fsblock_t)l0 & xfs_mask64lo(9)) << 43) |
L
Linus Torvalds 已提交
92
			   (((xfs_fsblock_t)l1) >> 21);
93
	s->br_blockcount = (xfs_filblks_t)(l1 & xfs_mask64lo(21));
L
Linus Torvalds 已提交
94 95 96 97 98 99 100 101 102 103 104
	/* This is xfs_extent_state() in-line */
	if (ext_flag) {
		ASSERT(s->br_blockcount != 0);	/* saved for DMIG */
		st = XFS_EXT_UNWRITTEN;
	} else
		st = XFS_EXT_NORM;
	s->br_state = st;
}

void
xfs_bmbt_get_all(
105
	xfs_bmbt_rec_host_t *r,
L
Linus Torvalds 已提交
106 107 108 109 110 111 112 113 114 115
	xfs_bmbt_irec_t *s)
{
	__xfs_bmbt_get_all(r->l0, r->l1, s);
}

/*
 * Extract the blockcount field from an in memory bmap extent record.
 */
xfs_filblks_t
xfs_bmbt_get_blockcount(
116
	xfs_bmbt_rec_host_t	*r)
L
Linus Torvalds 已提交
117
{
118
	return (xfs_filblks_t)(r->l1 & xfs_mask64lo(21));
L
Linus Torvalds 已提交
119 120 121 122 123 124 125
}

/*
 * Extract the startblock field from an in memory bmap extent record.
 */
xfs_fsblock_t
xfs_bmbt_get_startblock(
126
	xfs_bmbt_rec_host_t	*r)
L
Linus Torvalds 已提交
127
{
128
	return (((xfs_fsblock_t)r->l0 & xfs_mask64lo(9)) << 43) |
L
Linus Torvalds 已提交
129 130 131 132 133 134 135 136
	       (((xfs_fsblock_t)r->l1) >> 21);
}

/*
 * Extract the startoff field from an in memory bmap extent record.
 */
xfs_fileoff_t
xfs_bmbt_get_startoff(
137
	xfs_bmbt_rec_host_t	*r)
L
Linus Torvalds 已提交
138 139
{
	return ((xfs_fileoff_t)r->l0 &
140
		 xfs_mask64lo(64 - BMBT_EXNTFLAG_BITLEN)) >> 9;
L
Linus Torvalds 已提交
141 142 143 144 145 146 147 148 149
}

/*
 * Extract the blockcount field from an on disk bmap extent record.
 */
xfs_filblks_t
xfs_bmbt_disk_get_blockcount(
	xfs_bmbt_rec_t	*r)
{
150
	return (xfs_filblks_t)(be64_to_cpu(r->l1) & xfs_mask64lo(21));
L
Linus Torvalds 已提交
151 152 153 154 155 156 157 158 159
}

/*
 * Extract the startoff field from a disk format bmap extent record.
 */
xfs_fileoff_t
xfs_bmbt_disk_get_startoff(
	xfs_bmbt_rec_t	*r)
{
160
	return ((xfs_fileoff_t)be64_to_cpu(r->l0) &
161
		 xfs_mask64lo(64 - BMBT_EXNTFLAG_BITLEN)) >> 9;
L
Linus Torvalds 已提交
162 163 164
}

/*
165
 * Set all the fields in a bmap extent record from the uncompressed form.
L
Linus Torvalds 已提交
166 167
 */
void
168 169 170
xfs_bmbt_set_all(
	struct xfs_bmbt_rec_host *r,
	struct xfs_bmbt_irec	*s)
L
Linus Torvalds 已提交
171
{
172
	int			extent_flag = (s->br_state != XFS_EXT_NORM);
L
Linus Torvalds 已提交
173

174 175 176 177
	ASSERT(s->br_state == XFS_EXT_NORM || s->br_state == XFS_EXT_UNWRITTEN);
	ASSERT(!(s->br_startoff & xfs_mask64hi(64-BMBT_STARTOFF_BITLEN)));
	ASSERT(!(s->br_blockcount & xfs_mask64hi(64-BMBT_BLOCKCOUNT_BITLEN)));
	ASSERT(!(s->br_startblock & xfs_mask64hi(64-BMBT_STARTBLOCK_BITLEN)));
178

L
Linus Torvalds 已提交
179
	r->l0 = ((xfs_bmbt_rec_base_t)extent_flag << 63) |
180 181 182 183 184
		 ((xfs_bmbt_rec_base_t)s->br_startoff << 9) |
		 ((xfs_bmbt_rec_base_t)s->br_startblock >> 43);
	r->l1 = ((xfs_bmbt_rec_base_t)s->br_startblock << 21) |
		 ((xfs_bmbt_rec_base_t)s->br_blockcount &
		  (xfs_bmbt_rec_base_t)xfs_mask64lo(21));
L
Linus Torvalds 已提交
185 186 187
}

/*
188
 * Set all the fields in a bmap extent record from the uncompressed form.
L
Linus Torvalds 已提交
189 190
 */
void
191 192 193
xfs_bmbt_disk_set_all(
	struct xfs_bmbt_rec	*r,
	struct xfs_bmbt_irec	*s)
L
Linus Torvalds 已提交
194
{
195
	int			extent_flag = (s->br_state != XFS_EXT_NORM);
196

197 198 199 200
	ASSERT(s->br_state == XFS_EXT_NORM || s->br_state == XFS_EXT_UNWRITTEN);
	ASSERT(!(s->br_startoff & xfs_mask64hi(64-BMBT_STARTOFF_BITLEN)));
	ASSERT(!(s->br_blockcount & xfs_mask64hi(64-BMBT_BLOCKCOUNT_BITLEN)));
	ASSERT(!(s->br_startblock & xfs_mask64hi(64-BMBT_STARTBLOCK_BITLEN)));
201

202
	r->l0 = cpu_to_be64(
203
		((xfs_bmbt_rec_base_t)extent_flag << 63) |
204 205
		 ((xfs_bmbt_rec_base_t)s->br_startoff << 9) |
		 ((xfs_bmbt_rec_base_t)s->br_startblock >> 43));
206
	r->l1 = cpu_to_be64(
207 208
		((xfs_bmbt_rec_base_t)s->br_startblock << 21) |
		 ((xfs_bmbt_rec_base_t)s->br_blockcount &
209
		  (xfs_bmbt_rec_base_t)xfs_mask64lo(21)));
L
Linus Torvalds 已提交
210
}
211

L
Linus Torvalds 已提交
212 213 214 215 216
/*
 * Convert in-memory form of btree root to on-disk form.
 */
void
xfs_bmbt_to_bmdr(
217
	struct xfs_mount	*mp,
218
	struct xfs_btree_block	*rblock,
L
Linus Torvalds 已提交
219 220 221 222 223 224
	int			rblocklen,
	xfs_bmdr_block_t	*dblock,
	int			dblocklen)
{
	int			dmxr;
	xfs_bmbt_key_t		*fkp;
225
	__be64			*fpp;
L
Linus Torvalds 已提交
226
	xfs_bmbt_key_t		*tkp;
227
	__be64			*tpp;
L
Linus Torvalds 已提交
228

229 230
	if (xfs_sb_version_hascrc(&mp->m_sb)) {
		ASSERT(rblock->bb_magic == cpu_to_be32(XFS_BMAP_CRC_MAGIC));
231 232
		ASSERT(uuid_equal(&rblock->bb_u.l.bb_uuid,
		       &mp->m_sb.sb_meta_uuid));
233 234 235 236
		ASSERT(rblock->bb_u.l.bb_blkno ==
		       cpu_to_be64(XFS_BUF_DADDR_NULL));
	} else
		ASSERT(rblock->bb_magic == cpu_to_be32(XFS_BMAP_MAGIC));
C
Christoph Hellwig 已提交
237 238
	ASSERT(rblock->bb_u.l.bb_leftsib == cpu_to_be64(NULLFSBLOCK));
	ASSERT(rblock->bb_u.l.bb_rightsib == cpu_to_be64(NULLFSBLOCK));
239
	ASSERT(rblock->bb_level != 0);
240 241
	dblock->bb_level = rblock->bb_level;
	dblock->bb_numrecs = rblock->bb_numrecs;
242
	dmxr = xfs_bmdr_maxrecs(dblocklen, 0);
243 244
	fkp = XFS_BMBT_KEY_ADDR(mp, rblock, 1);
	tkp = XFS_BMDR_KEY_ADDR(dblock, 1);
245
	fpp = XFS_BMAP_BROOT_PTR_ADDR(mp, rblock, 1, rblocklen);
246
	tpp = XFS_BMDR_PTR_ADDR(dblock, 1, dmxr);
247
	dmxr = be16_to_cpu(dblock->bb_numrecs);
L
Linus Torvalds 已提交
248
	memcpy(tkp, fkp, sizeof(*fkp) * dmxr);
249
	memcpy(tpp, fpp, sizeof(*fpp) * dmxr);
L
Linus Torvalds 已提交
250 251
}

252 253 254 255 256 257 258 259 260 261
STATIC struct xfs_btree_cur *
xfs_bmbt_dup_cursor(
	struct xfs_btree_cur	*cur)
{
	struct xfs_btree_cur	*new;

	new = xfs_bmbt_init_cursor(cur->bc_mp, cur->bc_tp,
			cur->bc_private.b.ip, cur->bc_private.b.whichfork);

	/*
262
	 * Copy the firstblock, dfops, and flags values,
263 264 265
	 * since init cursor doesn't get them.
	 */
	new->bc_private.b.firstblock = cur->bc_private.b.firstblock;
266
	new->bc_private.b.dfops = cur->bc_private.b.dfops;
267 268 269 270 271
	new->bc_private.b.flags = cur->bc_private.b.flags;

	return new;
}

272 273 274 275 276 277 278
STATIC void
xfs_bmbt_update_cursor(
	struct xfs_btree_cur	*src,
	struct xfs_btree_cur	*dst)
{
	ASSERT((dst->bc_private.b.firstblock != NULLFSBLOCK) ||
	       (dst->bc_private.b.ip->i_d.di_flags & XFS_DIFLAG_REALTIME));
279
	ASSERT(dst->bc_private.b.dfops == src->bc_private.b.dfops);
280 281 282 283 284 285 286

	dst->bc_private.b.allocated += src->bc_private.b.allocated;
	dst->bc_private.b.firstblock = src->bc_private.b.firstblock;

	src->bc_private.b.allocated = 0;
}

287 288 289 290 291 292 293 294 295 296 297 298 299 300 301
STATIC int
xfs_bmbt_alloc_block(
	struct xfs_btree_cur	*cur,
	union xfs_btree_ptr	*start,
	union xfs_btree_ptr	*new,
	int			*stat)
{
	xfs_alloc_arg_t		args;		/* block allocation args */
	int			error;		/* error return value */

	memset(&args, 0, sizeof(args));
	args.tp = cur->bc_tp;
	args.mp = cur->bc_mp;
	args.fsbno = cur->bc_private.b.firstblock;
	args.firstblock = args.fsbno;
302 303
	xfs_rmap_ino_bmbt_owner(&args.oinfo, cur->bc_private.b.ip->i_ino,
			cur->bc_private.b.whichfork);
304 305 306 307 308 309 310 311 312 313 314 315 316 317 318

	if (args.fsbno == NULLFSBLOCK) {
		args.fsbno = be64_to_cpu(start->l);
		args.type = XFS_ALLOCTYPE_START_BNO;
		/*
		 * Make sure there is sufficient room left in the AG to
		 * complete a full tree split for an extent insert.  If
		 * we are converting the middle part of an extent then
		 * we may need space for two tree splits.
		 *
		 * We are relying on the caller to make the correct block
		 * reservation for this operation to succeed.  If the
		 * reservation amount is insufficient then we may fail a
		 * block allocation here and corrupt the filesystem.
		 */
319
		args.minleft = args.tp->t_blk_res;
320
	} else if (cur->bc_private.b.dfops->dop_low) {
321 322 323 324 325 326 327
		args.type = XFS_ALLOCTYPE_START_BNO;
	} else {
		args.type = XFS_ALLOCTYPE_NEAR_BNO;
	}

	args.minlen = args.maxlen = args.prod = 1;
	args.wasdel = cur->bc_private.b.flags & XFS_BTCUR_BPRV_WASDEL;
328
	if (!args.wasdel && args.tp->t_blk_res == 0) {
D
Dave Chinner 已提交
329
		error = -ENOSPC;
330 331 332 333 334 335 336 337 338
		goto error0;
	}
	error = xfs_alloc_vextent(&args);
	if (error)
		goto error0;

	if (args.fsbno == NULLFSBLOCK && args.minleft) {
		/*
		 * Could not find an AG with enough free space to satisfy
339
		 * a full btree split.  Try again and if
340 341 342 343 344 345 346
		 * successful activate the lowspace algorithm.
		 */
		args.fsbno = 0;
		args.type = XFS_ALLOCTYPE_FIRST_AG;
		error = xfs_alloc_vextent(&args);
		if (error)
			goto error0;
347
		cur->bc_private.b.dfops->dop_low = true;
348
	}
349
	if (WARN_ON_ONCE(args.fsbno == NULLFSBLOCK)) {
350 351 352 353 354 355 356 357 358
		XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT);
		*stat = 0;
		return 0;
	}
	ASSERT(args.len == 1);
	cur->bc_private.b.firstblock = args.fsbno;
	cur->bc_private.b.allocated++;
	cur->bc_private.b.ip->i_d.di_nblocks++;
	xfs_trans_log_inode(args.tp, cur->bc_private.b.ip, XFS_ILOG_CORE);
C
Christoph Hellwig 已提交
359
	xfs_trans_mod_dquot_byino(args.tp, cur->bc_private.b.ip,
360 361 362 363 364 365 366 367 368 369 370 371 372
			XFS_TRANS_DQ_BCOUNT, 1L);

	new->l = cpu_to_be64(args.fsbno);

	XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT);
	*stat = 1;
	return 0;

 error0:
	XFS_BTREE_TRACE_CURSOR(cur, XBT_ERROR);
	return error;
}

373 374 375 376 377 378 379 380 381
STATIC int
xfs_bmbt_free_block(
	struct xfs_btree_cur	*cur,
	struct xfs_buf		*bp)
{
	struct xfs_mount	*mp = cur->bc_mp;
	struct xfs_inode	*ip = cur->bc_private.b.ip;
	struct xfs_trans	*tp = cur->bc_tp;
	xfs_fsblock_t		fsbno = XFS_DADDR_TO_FSB(mp, XFS_BUF_ADDR(bp));
382
	struct xfs_owner_info	oinfo;
383

384 385
	xfs_rmap_ino_bmbt_owner(&oinfo, ip->i_ino, cur->bc_private.b.whichfork);
	xfs_bmap_add_free(mp, cur->bc_private.b.dfops, fsbno, 1, &oinfo);
386 387 388
	ip->i_d.di_nblocks--;

	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
C
Christoph Hellwig 已提交
389
	xfs_trans_mod_dquot_byino(tp, ip, XFS_TRANS_DQ_BCOUNT, -1L);
390 391 392
	return 0;
}

393 394 395 396 397
STATIC int
xfs_bmbt_get_minrecs(
	struct xfs_btree_cur	*cur,
	int			level)
{
398 399 400 401 402 403 404 405 406 407 408
	if (level == cur->bc_nlevels - 1) {
		struct xfs_ifork	*ifp;

		ifp = XFS_IFORK_PTR(cur->bc_private.b.ip,
				    cur->bc_private.b.whichfork);

		return xfs_bmbt_maxrecs(cur->bc_mp,
					ifp->if_broot_bytes, level == 0) / 2;
	}

	return cur->bc_mp->m_bmap_dmnr[level != 0];
409 410
}

411
int
412 413 414 415
xfs_bmbt_get_maxrecs(
	struct xfs_btree_cur	*cur,
	int			level)
{
416 417 418 419 420 421 422 423 424 425 426 427
	if (level == cur->bc_nlevels - 1) {
		struct xfs_ifork	*ifp;

		ifp = XFS_IFORK_PTR(cur->bc_private.b.ip,
				    cur->bc_private.b.whichfork);

		return xfs_bmbt_maxrecs(cur->bc_mp,
					ifp->if_broot_bytes, level == 0);
	}

	return cur->bc_mp->m_bmap_dmxr[level != 0];

428 429
}

430 431 432 433 434 435 436 437 438 439 440 441 442 443
/*
 * Get the maximum records we could store in the on-disk format.
 *
 * For non-root nodes this is equivalent to xfs_bmbt_get_maxrecs, but
 * for the root node this checks the available space in the dinode fork
 * so that we can resize the in-memory buffer to match it.  After a
 * resize to the maximum size this function returns the same value
 * as xfs_bmbt_get_maxrecs for the root node, too.
 */
STATIC int
xfs_bmbt_get_dmaxrecs(
	struct xfs_btree_cur	*cur,
	int			level)
{
444 445
	if (level != cur->bc_nlevels - 1)
		return cur->bc_mp->m_bmap_dmxr[level != 0];
446
	return xfs_bmdr_maxrecs(cur->bc_private.b.forksize, level == 0);
447 448
}

449 450 451 452 453 454 455 456 457
STATIC void
xfs_bmbt_init_key_from_rec(
	union xfs_btree_key	*key,
	union xfs_btree_rec	*rec)
{
	key->bmbt.br_startoff =
		cpu_to_be64(xfs_bmbt_disk_get_startoff(&rec->bmbt));
}

458 459 460 461 462 463 464 465 466 467
STATIC void
xfs_bmbt_init_high_key_from_rec(
	union xfs_btree_key	*key,
	union xfs_btree_rec	*rec)
{
	key->bmbt.br_startoff = cpu_to_be64(
			xfs_bmbt_disk_get_startoff(&rec->bmbt) +
			xfs_bmbt_disk_get_blockcount(&rec->bmbt) - 1);
}

468 469 470 471 472 473 474 475
STATIC void
xfs_bmbt_init_rec_from_cur(
	struct xfs_btree_cur	*cur,
	union xfs_btree_rec	*rec)
{
	xfs_bmbt_disk_set_all(&rec->bmbt, &cur->bc_rec.b);
}

476 477 478 479 480 481 482 483
STATIC void
xfs_bmbt_init_ptr_from_cur(
	struct xfs_btree_cur	*cur,
	union xfs_btree_ptr	*ptr)
{
	ptr->l = 0;
}

484
STATIC int64_t
485 486 487 488
xfs_bmbt_key_diff(
	struct xfs_btree_cur	*cur,
	union xfs_btree_key	*key)
{
489
	return (int64_t)be64_to_cpu(key->bmbt.br_startoff) -
490 491 492
				      cur->bc_rec.b.br_startoff;
}

493 494 495 496 497 498 499 500 501 502
STATIC int64_t
xfs_bmbt_diff_two_keys(
	struct xfs_btree_cur	*cur,
	union xfs_btree_key	*k1,
	union xfs_btree_key	*k2)
{
	return (int64_t)be64_to_cpu(k1->bmbt.br_startoff) -
			  be64_to_cpu(k2->bmbt.br_startoff);
}

D
Dave Chinner 已提交
503
static bool
504
xfs_bmbt_verify(
505 506 507 508 509 510
	struct xfs_buf		*bp)
{
	struct xfs_mount	*mp = bp->b_target->bt_mount;
	struct xfs_btree_block	*block = XFS_BUF_TO_BLOCK(bp);
	unsigned int		level;

511 512 513 514
	switch (block->bb_magic) {
	case cpu_to_be32(XFS_BMAP_CRC_MAGIC):
		if (!xfs_sb_version_hascrc(&mp->m_sb))
			return false;
515
		if (!uuid_equal(&block->bb_u.l.bb_uuid, &mp->m_sb.sb_meta_uuid))
516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533
			return false;
		if (be64_to_cpu(block->bb_u.l.bb_blkno) != bp->b_bn)
			return false;
		/*
		 * XXX: need a better way of verifying the owner here. Right now
		 * just make sure there has been one set.
		 */
		if (be64_to_cpu(block->bb_u.l.bb_owner) == 0)
			return false;
		/* fall through */
	case cpu_to_be32(XFS_BMAP_MAGIC):
		break;
	default:
		return false;
	}

	/*
	 * numrecs and level verification.
534
	 *
535
	 * We don't know what fork we belong to, so just verify that the level
536 537 538 539
	 * is less than the maximum of the two. Later checks will be more
	 * precise.
	 */
	level = be16_to_cpu(block->bb_level);
540 541 542 543
	if (level > max(mp->m_bm_maxlevels[0], mp->m_bm_maxlevels[1]))
		return false;
	if (be16_to_cpu(block->bb_numrecs) > mp->m_bmap_dmxr[level != 0])
		return false;
544 545

	/* sibling pointer verification */
546
	if (!block->bb_u.l.bb_leftsib ||
C
Christoph Hellwig 已提交
547
	    (block->bb_u.l.bb_leftsib != cpu_to_be64(NULLFSBLOCK) &&
548 549 550
	     !XFS_FSB_SANITY_CHECK(mp, be64_to_cpu(block->bb_u.l.bb_leftsib))))
		return false;
	if (!block->bb_u.l.bb_rightsib ||
C
Christoph Hellwig 已提交
551
	    (block->bb_u.l.bb_rightsib != cpu_to_be64(NULLFSBLOCK) &&
552 553 554 555
	     !XFS_FSB_SANITY_CHECK(mp, be64_to_cpu(block->bb_u.l.bb_rightsib))))
		return false;

	return true;
556
}
557

558 559
static void
xfs_bmbt_read_verify(
560 561
	struct xfs_buf	*bp)
{
562
	if (!xfs_btree_lblock_verify_crc(bp))
D
Dave Chinner 已提交
563
		xfs_buf_ioerror(bp, -EFSBADCRC);
564
	else if (!xfs_bmbt_verify(bp))
D
Dave Chinner 已提交
565
		xfs_buf_ioerror(bp, -EFSCORRUPTED);
566 567 568 569

	if (bp->b_error) {
		trace_xfs_btree_corrupt(bp, _RET_IP_);
		xfs_verifier_error(bp);
570
	}
571 572
}

573 574
static void
xfs_bmbt_write_verify(
575 576
	struct xfs_buf	*bp)
{
577 578
	if (!xfs_bmbt_verify(bp)) {
		trace_xfs_btree_corrupt(bp, _RET_IP_);
D
Dave Chinner 已提交
579
		xfs_buf_ioerror(bp, -EFSCORRUPTED);
580
		xfs_verifier_error(bp);
581 582 583
		return;
	}
	xfs_btree_lblock_calc_crc(bp);
584 585
}

586
const struct xfs_buf_ops xfs_bmbt_buf_ops = {
587
	.name = "xfs_bmbt",
588 589 590 591 592
	.verify_read = xfs_bmbt_read_verify,
	.verify_write = xfs_bmbt_write_verify,
};


593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613
STATIC int
xfs_bmbt_keys_inorder(
	struct xfs_btree_cur	*cur,
	union xfs_btree_key	*k1,
	union xfs_btree_key	*k2)
{
	return be64_to_cpu(k1->bmbt.br_startoff) <
		be64_to_cpu(k2->bmbt.br_startoff);
}

STATIC int
xfs_bmbt_recs_inorder(
	struct xfs_btree_cur	*cur,
	union xfs_btree_rec	*r1,
	union xfs_btree_rec	*r2)
{
	return xfs_bmbt_disk_get_startoff(&r1->bmbt) +
		xfs_bmbt_disk_get_blockcount(&r1->bmbt) <=
		xfs_bmbt_disk_get_startoff(&r2->bmbt);
}

614
static const struct xfs_btree_ops xfs_bmbt_ops = {
615 616 617
	.rec_len		= sizeof(xfs_bmbt_rec_t),
	.key_len		= sizeof(xfs_bmbt_key_t),

618
	.dup_cursor		= xfs_bmbt_dup_cursor,
619
	.update_cursor		= xfs_bmbt_update_cursor,
620
	.alloc_block		= xfs_bmbt_alloc_block,
621
	.free_block		= xfs_bmbt_free_block,
622
	.get_maxrecs		= xfs_bmbt_get_maxrecs,
623
	.get_minrecs		= xfs_bmbt_get_minrecs,
624
	.get_dmaxrecs		= xfs_bmbt_get_dmaxrecs,
625
	.init_key_from_rec	= xfs_bmbt_init_key_from_rec,
626
	.init_high_key_from_rec	= xfs_bmbt_init_high_key_from_rec,
627
	.init_rec_from_cur	= xfs_bmbt_init_rec_from_cur,
628 629
	.init_ptr_from_cur	= xfs_bmbt_init_ptr_from_cur,
	.key_diff		= xfs_bmbt_key_diff,
630
	.diff_two_keys		= xfs_bmbt_diff_two_keys,
631
	.buf_ops		= &xfs_bmbt_buf_ops,
632 633
	.keys_inorder		= xfs_bmbt_keys_inorder,
	.recs_inorder		= xfs_bmbt_recs_inorder,
634 635 636 637 638 639 640 641 642 643 644 645 646 647
};

/*
 * Allocate a new bmap btree cursor.
 */
struct xfs_btree_cur *				/* new bmap btree cursor */
xfs_bmbt_init_cursor(
	struct xfs_mount	*mp,		/* file system mount point */
	struct xfs_trans	*tp,		/* transaction pointer */
	struct xfs_inode	*ip,		/* inode owning the btree */
	int			whichfork)	/* data or attr fork */
{
	struct xfs_ifork	*ifp = XFS_IFORK_PTR(ip, whichfork);
	struct xfs_btree_cur	*cur;
D
Darrick J. Wong 已提交
648
	ASSERT(whichfork != XFS_COW_FORK);
649

650
	cur = kmem_zone_zalloc(xfs_btree_cur_zone, KM_NOFS);
651 652 653 654 655 656

	cur->bc_tp = tp;
	cur->bc_mp = mp;
	cur->bc_nlevels = be16_to_cpu(ifp->if_broot->bb_level) + 1;
	cur->bc_btnum = XFS_BTNUM_BMAP;
	cur->bc_blocklog = mp->m_sb.sb_blocklog;
D
Dave Chinner 已提交
657
	cur->bc_statoff = XFS_STATS_CALC_INDEX(xs_bmbt_2);
658 659

	cur->bc_ops = &xfs_bmbt_ops;
660
	cur->bc_flags = XFS_BTREE_LONG_PTRS | XFS_BTREE_ROOT_IN_INODE;
661 662
	if (xfs_sb_version_hascrc(&mp->m_sb))
		cur->bc_flags |= XFS_BTREE_CRC_BLOCKS;
663 664 665 666

	cur->bc_private.b.forksize = XFS_IFORK_SIZE(ip, whichfork);
	cur->bc_private.b.ip = ip;
	cur->bc_private.b.firstblock = NULLFSBLOCK;
667
	cur->bc_private.b.dfops = NULL;
668 669 670 671 672 673
	cur->bc_private.b.allocated = 0;
	cur->bc_private.b.flags = 0;
	cur->bc_private.b.whichfork = whichfork;

	return cur;
}
674 675 676 677 678 679 680 681 682 683

/*
 * Calculate number of records in a bmap btree block.
 */
int
xfs_bmbt_maxrecs(
	struct xfs_mount	*mp,
	int			blocklen,
	int			leaf)
{
684
	blocklen -= XFS_BMBT_BLOCK_LEN(mp);
685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704

	if (leaf)
		return blocklen / sizeof(xfs_bmbt_rec_t);
	return blocklen / (sizeof(xfs_bmbt_key_t) + sizeof(xfs_bmbt_ptr_t));
}

/*
 * Calculate number of records in a bmap btree inode root.
 */
int
xfs_bmdr_maxrecs(
	int			blocklen,
	int			leaf)
{
	blocklen -= sizeof(xfs_bmdr_block_t);

	if (leaf)
		return blocklen / sizeof(xfs_bmdr_rec_t);
	return blocklen / (sizeof(xfs_bmdr_key_t) + sizeof(xfs_bmdr_ptr_t));
}
705 706 707 708 709 710 711

/*
 * Change the owner of a btree format fork fo the inode passed in. Change it to
 * the owner of that is passed in so that we can change owners before or after
 * we switch forks between inodes. The operation that the caller is doing will
 * determine whether is needs to change owner before or after the switch.
 *
712 713 714 715 716 717 718 719 720 721
 * For demand paged transactional modification, the fork switch should be done
 * after reading in all the blocks, modifying them and pinning them in the
 * transaction. For modification when the buffers are already pinned in memory,
 * the fork switch can be done before changing the owner as we won't need to
 * validate the owner until the btree buffers are unpinned and writes can occur
 * again.
 *
 * For recovery based ownership change, there is no transactional context and
 * so a buffer list must be supplied so that we can record the buffers that we
 * modified for the caller to issue IO on.
722 723 724 725 726 727
 */
int
xfs_bmbt_change_owner(
	struct xfs_trans	*tp,
	struct xfs_inode	*ip,
	int			whichfork,
728 729
	xfs_ino_t		new_owner,
	struct list_head	*buffer_list)
730 731 732 733
{
	struct xfs_btree_cur	*cur;
	int			error;

734 735
	ASSERT(tp || buffer_list);
	ASSERT(!(tp && buffer_list));
736
	if (whichfork == XFS_DATA_FORK)
D
Dan Carpenter 已提交
737
		ASSERT(ip->i_d.di_format == XFS_DINODE_FMT_BTREE);
738
	else
D
Dan Carpenter 已提交
739
		ASSERT(ip->i_d.di_aformat == XFS_DINODE_FMT_BTREE);
740 741

	cur = xfs_bmbt_init_cursor(ip->i_mount, tp, ip, whichfork);
742
	if (!cur)
D
Dave Chinner 已提交
743
		return -ENOMEM;
744
	cur->bc_private.b.flags |= XFS_BTCUR_BPRV_INVALID_OWNER;
745 746

	error = xfs_btree_change_owner(cur, new_owner, buffer_list);
747 748 749
	xfs_btree_del_cursor(cur, error ? XFS_BTREE_ERROR : XFS_BTREE_NOERROR);
	return error;
}