xfs_ialloc_btree.c 18.5 KB
Newer Older
D
Dave Chinner 已提交
1
// SPDX-License-Identifier: GPL-2.0
L
Linus Torvalds 已提交
2
/*
3 4
 * Copyright (c) 2000-2001,2005 Silicon Graphics, Inc.
 * All Rights Reserved.
L
Linus Torvalds 已提交
5 6
 */
#include "xfs.h"
7
#include "xfs_fs.h"
8
#include "xfs_shared.h"
9
#include "xfs_format.h"
10 11
#include "xfs_log_format.h"
#include "xfs_trans_resv.h"
12
#include "xfs_bit.h"
L
Linus Torvalds 已提交
13 14
#include "xfs_mount.h"
#include "xfs_btree.h"
15
#include "xfs_btree_staging.h"
L
Linus Torvalds 已提交
16
#include "xfs_ialloc.h"
17
#include "xfs_ialloc_btree.h"
L
Linus Torvalds 已提交
18 19
#include "xfs_alloc.h"
#include "xfs_error.h"
20
#include "xfs_trace.h"
21
#include "xfs_trans.h"
22
#include "xfs_rmap.h"
23
#include "xfs_ag.h"
L
Linus Torvalds 已提交
24

25 26 27 28 29
STATIC int
xfs_inobt_get_minrecs(
	struct xfs_btree_cur	*cur,
	int			level)
{
D
Darrick J. Wong 已提交
30
	return M_IGEO(cur->bc_mp)->inobt_mnr[level != 0];
31
}
L
Linus Torvalds 已提交
32

33 34 35 36 37
STATIC struct xfs_btree_cur *
xfs_inobt_dup_cursor(
	struct xfs_btree_cur	*cur)
{
	return xfs_inobt_init_cursor(cur->bc_mp, cur->bc_tp,
38
			cur->bc_ag.agbp, cur->bc_ag.agno,
39
			cur->bc_ag.pag, cur->bc_btnum);
40 41
}

42 43 44 45 46 47
STATIC void
xfs_inobt_set_root(
	struct xfs_btree_cur	*cur,
	union xfs_btree_ptr	*nptr,
	int			inc)	/* level change */
{
48
	struct xfs_buf		*agbp = cur->bc_ag.agbp;
C
Christoph Hellwig 已提交
49
	struct xfs_agi		*agi = agbp->b_addr;
50 51 52 53 54 55

	agi->agi_root = nptr->s;
	be32_add_cpu(&agi->agi_level, inc);
	xfs_ialloc_log_agi(cur->bc_tp, agbp, XFS_AGI_ROOT | XFS_AGI_LEVEL);
}

56 57 58 59 60 61
STATIC void
xfs_finobt_set_root(
	struct xfs_btree_cur	*cur,
	union xfs_btree_ptr	*nptr,
	int			inc)	/* level change */
{
62
	struct xfs_buf		*agbp = cur->bc_ag.agbp;
C
Christoph Hellwig 已提交
63
	struct xfs_agi		*agi = agbp->b_addr;
64 65 66 67 68 69 70

	agi->agi_free_root = nptr->s;
	be32_add_cpu(&agi->agi_free_level, inc);
	xfs_ialloc_log_agi(cur->bc_tp, agbp,
			   XFS_AGI_FREE_ROOT | XFS_AGI_FREE_LEVEL);
}

71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89
/* Update the inode btree block counter for this btree. */
static inline void
xfs_inobt_mod_blockcount(
	struct xfs_btree_cur	*cur,
	int			howmuch)
{
	struct xfs_buf		*agbp = cur->bc_ag.agbp;
	struct xfs_agi		*agi = agbp->b_addr;

	if (!xfs_sb_version_hasinobtcounts(&cur->bc_mp->m_sb))
		return;

	if (cur->bc_btnum == XFS_BTNUM_FINO)
		be32_add_cpu(&agi->agi_fblocks, howmuch);
	else if (cur->bc_btnum == XFS_BTNUM_INO)
		be32_add_cpu(&agi->agi_iblocks, howmuch);
	xfs_ialloc_log_agi(cur->bc_tp, agbp, XFS_AGI_IBLOCKS);
}

90
STATIC int
91
__xfs_inobt_alloc_block(
92 93 94
	struct xfs_btree_cur	*cur,
	union xfs_btree_ptr	*start,
	union xfs_btree_ptr	*new,
95 96
	int			*stat,
	enum xfs_ag_resv_type	resv)
97 98 99 100 101 102 103 104
{
	xfs_alloc_arg_t		args;		/* block allocation args */
	int			error;		/* error return value */
	xfs_agblock_t		sbno = be32_to_cpu(start->s);

	memset(&args, 0, sizeof(args));
	args.tp = cur->bc_tp;
	args.mp = cur->bc_mp;
105
	args.oinfo = XFS_RMAP_OINFO_INOBT;
106
	args.fsbno = XFS_AGB_TO_FSB(args.mp, cur->bc_ag.agno, sbno);
107 108 109 110
	args.minlen = 1;
	args.maxlen = 1;
	args.prod = 1;
	args.type = XFS_ALLOCTYPE_NEAR_BNO;
111
	args.resv = resv;
112 113

	error = xfs_alloc_vextent(&args);
C
Carlos Maiolino 已提交
114
	if (error)
115
		return error;
C
Carlos Maiolino 已提交
116

117 118 119 120 121 122 123 124
	if (args.fsbno == NULLFSBLOCK) {
		*stat = 0;
		return 0;
	}
	ASSERT(args.len == 1);

	new->s = cpu_to_be32(XFS_FSB_TO_AGBNO(args.mp, args.fsbno));
	*stat = 1;
125
	xfs_inobt_mod_blockcount(cur, 1);
126 127 128
	return 0;
}

129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145
STATIC int
xfs_inobt_alloc_block(
	struct xfs_btree_cur	*cur,
	union xfs_btree_ptr	*start,
	union xfs_btree_ptr	*new,
	int			*stat)
{
	return __xfs_inobt_alloc_block(cur, start, new, stat, XFS_AG_RESV_NONE);
}

STATIC int
xfs_finobt_alloc_block(
	struct xfs_btree_cur	*cur,
	union xfs_btree_ptr	*start,
	union xfs_btree_ptr	*new,
	int			*stat)
{
146
	if (cur->bc_mp->m_finobt_nores)
147
		return xfs_inobt_alloc_block(cur, start, new, stat);
148 149 150 151
	return __xfs_inobt_alloc_block(cur, start, new, stat,
			XFS_AG_RESV_METADATA);
}

152
STATIC int
153
__xfs_inobt_free_block(
154
	struct xfs_btree_cur	*cur,
155 156
	struct xfs_buf		*bp,
	enum xfs_ag_resv_type	resv)
157
{
158
	xfs_inobt_mod_blockcount(cur, -1);
159
	return xfs_free_extent(cur->bc_tp,
160
			XFS_DADDR_TO_FSB(cur->bc_mp, XFS_BUF_ADDR(bp)), 1,
161
			&XFS_RMAP_OINFO_INOBT, resv);
162 163 164 165 166 167 168 169 170 171 172 173 174 175 176
}

STATIC int
xfs_inobt_free_block(
	struct xfs_btree_cur	*cur,
	struct xfs_buf		*bp)
{
	return __xfs_inobt_free_block(cur, bp, XFS_AG_RESV_NONE);
}

STATIC int
xfs_finobt_free_block(
	struct xfs_btree_cur	*cur,
	struct xfs_buf		*bp)
{
177
	if (cur->bc_mp->m_finobt_nores)
178 179
		return xfs_inobt_free_block(cur, bp);
	return __xfs_inobt_free_block(cur, bp, XFS_AG_RESV_METADATA);
180
}
181

182 183 184 185 186
STATIC int
xfs_inobt_get_maxrecs(
	struct xfs_btree_cur	*cur,
	int			level)
{
D
Darrick J. Wong 已提交
187
	return M_IGEO(cur->bc_mp)->inobt_mxr[level != 0];
188 189
}

190 191 192 193 194 195 196 197
STATIC void
xfs_inobt_init_key_from_rec(
	union xfs_btree_key	*key,
	union xfs_btree_rec	*rec)
{
	key->inobt.ir_startino = rec->inobt.ir_startino;
}

198 199 200 201 202 203 204 205 206 207 208 209
STATIC void
xfs_inobt_init_high_key_from_rec(
	union xfs_btree_key	*key,
	union xfs_btree_rec	*rec)
{
	__u32			x;

	x = be32_to_cpu(rec->inobt.ir_startino);
	x += XFS_INODES_PER_CHUNK - 1;
	key->inobt.ir_startino = cpu_to_be32(x);
}

210 211 212 213 214 215
STATIC void
xfs_inobt_init_rec_from_cur(
	struct xfs_btree_cur	*cur,
	union xfs_btree_rec	*rec)
{
	rec->inobt.ir_startino = cpu_to_be32(cur->bc_rec.i.ir_startino);
216 217 218 219 220 221 222 223 224 225
	if (xfs_sb_version_hassparseinodes(&cur->bc_mp->m_sb)) {
		rec->inobt.ir_u.sp.ir_holemask =
					cpu_to_be16(cur->bc_rec.i.ir_holemask);
		rec->inobt.ir_u.sp.ir_count = cur->bc_rec.i.ir_count;
		rec->inobt.ir_u.sp.ir_freecount = cur->bc_rec.i.ir_freecount;
	} else {
		/* ir_holemask/ir_count not supported on-disk */
		rec->inobt.ir_u.f.ir_freecount =
					cpu_to_be32(cur->bc_rec.i.ir_freecount);
	}
226 227 228
	rec->inobt.ir_free = cpu_to_be64(cur->bc_rec.i.ir_free);
}

229
/*
M
Malcolm Parsons 已提交
230
 * initial value of ptr for lookup
231 232 233 234 235 236
 */
STATIC void
xfs_inobt_init_ptr_from_cur(
	struct xfs_btree_cur	*cur,
	union xfs_btree_ptr	*ptr)
{
237
	struct xfs_agi		*agi = cur->bc_ag.agbp->b_addr;
238

239
	ASSERT(cur->bc_ag.agno == be32_to_cpu(agi->agi_seqno));
240 241 242 243

	ptr->s = agi->agi_root;
}

244 245 246 247 248
STATIC void
xfs_finobt_init_ptr_from_cur(
	struct xfs_btree_cur	*cur,
	union xfs_btree_ptr	*ptr)
{
249
	struct xfs_agi		*agi = cur->bc_ag.agbp->b_addr;
250

251
	ASSERT(cur->bc_ag.agno == be32_to_cpu(agi->agi_seqno));
252 253 254
	ptr->s = agi->agi_free_root;
}

255
STATIC int64_t
256 257 258 259
xfs_inobt_key_diff(
	struct xfs_btree_cur	*cur,
	union xfs_btree_key	*key)
{
260
	return (int64_t)be32_to_cpu(key->inobt.ir_startino) -
261 262 263
			  cur->bc_rec.i.ir_startino;
}

264 265 266 267 268 269 270 271 272 273
STATIC int64_t
xfs_inobt_diff_two_keys(
	struct xfs_btree_cur	*cur,
	union xfs_btree_key	*k1,
	union xfs_btree_key	*k2)
{
	return (int64_t)be32_to_cpu(k1->inobt.ir_startino) -
			  be32_to_cpu(k2->inobt.ir_startino);
}

274
static xfs_failaddr_t
275
xfs_inobt_verify(
276 277
	struct xfs_buf		*bp)
{
278
	struct xfs_mount	*mp = bp->b_mount;
279
	struct xfs_btree_block	*block = XFS_BUF_TO_BLOCK(bp);
280
	xfs_failaddr_t		fa;
281 282
	unsigned int		level;

283 284 285
	if (!xfs_verify_magic(bp, block->bb_magic))
		return __this_address;

286 287 288 289 290 291 292 293 294 295
	/*
	 * During growfs operations, we can't verify the exact owner as the
	 * perag is not fully initialised and hence not attached to the buffer.
	 *
	 * Similarly, during log recovery we will have a perag structure
	 * attached, but the agi information will not yet have been initialised
	 * from the on disk AGI. We don't currently use any of this information,
	 * but beware of the landmine (i.e. need to check pag->pagi_init) if we
	 * ever do.
	 */
296
	if (xfs_sb_version_hascrc(&mp->m_sb)) {
297 298 299
		fa = xfs_btree_sblock_v5hdr_verify(bp);
		if (fa)
			return fa;
300
	}
301

302
	/* level verification */
303
	level = be16_to_cpu(block->bb_level);
D
Darrick J. Wong 已提交
304
	if (level >= M_IGEO(mp)->inobt_maxlevels)
305
		return __this_address;
306

D
Darrick J. Wong 已提交
307 308
	return xfs_btree_sblock_verify(bp,
			M_IGEO(mp)->inobt_mxr[level != 0]);
309 310 311
}

static void
312
xfs_inobt_read_verify(
313 314
	struct xfs_buf	*bp)
{
315 316
	xfs_failaddr_t	fa;

317
	if (!xfs_btree_sblock_verify_crc(bp))
318 319 320 321 322 323
		xfs_verifier_error(bp, -EFSBADCRC, __this_address);
	else {
		fa = xfs_inobt_verify(bp);
		if (fa)
			xfs_verifier_error(bp, -EFSCORRUPTED, fa);
	}
324

325
	if (bp->b_error)
326
		trace_xfs_btree_corrupt(bp, _RET_IP_);
327
}
328

329 330
static void
xfs_inobt_write_verify(
331 332
	struct xfs_buf	*bp)
{
333 334 335 336
	xfs_failaddr_t	fa;

	fa = xfs_inobt_verify(bp);
	if (fa) {
337
		trace_xfs_btree_corrupt(bp, _RET_IP_);
338
		xfs_verifier_error(bp, -EFSCORRUPTED, fa);
339
		return;
340 341 342
	}
	xfs_btree_sblock_calc_crc(bp);

343 344
}

345
const struct xfs_buf_ops xfs_inobt_buf_ops = {
346
	.name = "xfs_inobt",
347
	.magic = { cpu_to_be32(XFS_IBT_MAGIC), cpu_to_be32(XFS_IBT_CRC_MAGIC) },
348 349
	.verify_read = xfs_inobt_read_verify,
	.verify_write = xfs_inobt_write_verify,
350
	.verify_struct = xfs_inobt_verify,
351 352
};

353 354
const struct xfs_buf_ops xfs_finobt_buf_ops = {
	.name = "xfs_finobt",
355 356
	.magic = { cpu_to_be32(XFS_FIBT_MAGIC),
		   cpu_to_be32(XFS_FIBT_CRC_MAGIC) },
357 358 359 360 361
	.verify_read = xfs_inobt_read_verify,
	.verify_write = xfs_inobt_write_verify,
	.verify_struct = xfs_inobt_verify,
};

362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381
STATIC int
xfs_inobt_keys_inorder(
	struct xfs_btree_cur	*cur,
	union xfs_btree_key	*k1,
	union xfs_btree_key	*k2)
{
	return be32_to_cpu(k1->inobt.ir_startino) <
		be32_to_cpu(k2->inobt.ir_startino);
}

STATIC int
xfs_inobt_recs_inorder(
	struct xfs_btree_cur	*cur,
	union xfs_btree_rec	*r1,
	union xfs_btree_rec	*r2)
{
	return be32_to_cpu(r1->inobt.ir_startino) + XFS_INODES_PER_CHUNK <=
		be32_to_cpu(r2->inobt.ir_startino);
}

382
static const struct xfs_btree_ops xfs_inobt_ops = {
383 384 385
	.rec_len		= sizeof(xfs_inobt_rec_t),
	.key_len		= sizeof(xfs_inobt_key_t),

386
	.dup_cursor		= xfs_inobt_dup_cursor,
387
	.set_root		= xfs_inobt_set_root,
388
	.alloc_block		= xfs_inobt_alloc_block,
389
	.free_block		= xfs_inobt_free_block,
390
	.get_minrecs		= xfs_inobt_get_minrecs,
391
	.get_maxrecs		= xfs_inobt_get_maxrecs,
392
	.init_key_from_rec	= xfs_inobt_init_key_from_rec,
393
	.init_high_key_from_rec	= xfs_inobt_init_high_key_from_rec,
394
	.init_rec_from_cur	= xfs_inobt_init_rec_from_cur,
395 396
	.init_ptr_from_cur	= xfs_inobt_init_ptr_from_cur,
	.key_diff		= xfs_inobt_key_diff,
397
	.buf_ops		= &xfs_inobt_buf_ops,
398
	.diff_two_keys		= xfs_inobt_diff_two_keys,
399 400
	.keys_inorder		= xfs_inobt_keys_inorder,
	.recs_inorder		= xfs_inobt_recs_inorder,
401 402
};

403 404 405 406 407 408
static const struct xfs_btree_ops xfs_finobt_ops = {
	.rec_len		= sizeof(xfs_inobt_rec_t),
	.key_len		= sizeof(xfs_inobt_key_t),

	.dup_cursor		= xfs_inobt_dup_cursor,
	.set_root		= xfs_finobt_set_root,
409
	.alloc_block		= xfs_finobt_alloc_block,
410
	.free_block		= xfs_finobt_free_block,
411 412 413
	.get_minrecs		= xfs_inobt_get_minrecs,
	.get_maxrecs		= xfs_inobt_get_maxrecs,
	.init_key_from_rec	= xfs_inobt_init_key_from_rec,
414
	.init_high_key_from_rec	= xfs_inobt_init_high_key_from_rec,
415 416 417
	.init_rec_from_cur	= xfs_inobt_init_rec_from_cur,
	.init_ptr_from_cur	= xfs_finobt_init_ptr_from_cur,
	.key_diff		= xfs_inobt_key_diff,
418
	.buf_ops		= &xfs_finobt_buf_ops,
419
	.diff_two_keys		= xfs_inobt_diff_two_keys,
420 421 422 423
	.keys_inorder		= xfs_inobt_keys_inorder,
	.recs_inorder		= xfs_inobt_recs_inorder,
};

424
/*
425
 * Initialize a new inode btree cursor.
426
 */
427 428
static struct xfs_btree_cur *
xfs_inobt_init_common(
429 430
	struct xfs_mount	*mp,		/* file system mount point */
	struct xfs_trans	*tp,		/* transaction pointer */
431
	xfs_agnumber_t		agno,		/* allocation group number */
432
	struct xfs_perag	*pag,
433
	xfs_btnum_t		btnum)		/* ialloc or free ino btree */
434 435 436
{
	struct xfs_btree_cur	*cur;

437
	cur = kmem_cache_zalloc(xfs_btree_cur_zone, GFP_NOFS | __GFP_NOFAIL);
438 439
	cur->bc_tp = tp;
	cur->bc_mp = mp;
440
	cur->bc_btnum = btnum;
441
	if (btnum == XFS_BTNUM_INO) {
D
Dave Chinner 已提交
442
		cur->bc_statoff = XFS_STATS_CALC_INDEX(xs_ibt_2);
443
		cur->bc_ops = &xfs_inobt_ops;
444
	} else {
D
Dave Chinner 已提交
445
		cur->bc_statoff = XFS_STATS_CALC_INDEX(xs_fibt_2);
446
		cur->bc_ops = &xfs_finobt_ops;
447 448
	}

449 450
	cur->bc_blocklog = mp->m_sb.sb_blocklog;

451 452
	if (xfs_sb_version_hascrc(&mp->m_sb))
		cur->bc_flags |= XFS_BTREE_CRC_BLOCKS;
453

454
	cur->bc_ag.agno = agno;
455 456 457 458 459
	if (pag) {
		/* take a reference for the cursor */
		atomic_inc(&pag->pag_ref);
	}
	cur->bc_ag.pag = pag;
460 461 462 463 464 465 466 467 468 469
	return cur;
}

/* Create an inode btree cursor. */
struct xfs_btree_cur *
xfs_inobt_init_cursor(
	struct xfs_mount	*mp,
	struct xfs_trans	*tp,
	struct xfs_buf		*agbp,
	xfs_agnumber_t		agno,
470
	struct xfs_perag	*pag,
471 472 473 474 475
	xfs_btnum_t		btnum)
{
	struct xfs_btree_cur	*cur;
	struct xfs_agi		*agi = agbp->b_addr;

476
	cur = xfs_inobt_init_common(mp, tp, agno, pag, btnum);
477 478 479 480 481 482 483
	if (btnum == XFS_BTNUM_INO)
		cur->bc_nlevels = be32_to_cpu(agi->agi_level);
	else
		cur->bc_nlevels = be32_to_cpu(agi->agi_free_level);
	cur->bc_ag.agbp = agbp;
	return cur;
}
484

485 486 487 488 489 490 491 492 493 494
/* Create an inode btree cursor with a fake root for staging. */
struct xfs_btree_cur *
xfs_inobt_stage_cursor(
	struct xfs_mount	*mp,
	struct xbtree_afakeroot	*afake,
	xfs_agnumber_t		agno,
	xfs_btnum_t		btnum)
{
	struct xfs_btree_cur	*cur;

495
	cur = xfs_inobt_init_common(mp, NULL, agno, NULL, btnum);
496
	xfs_btree_stage_afakeroot(cur, afake);
497 498
	return cur;
}
499

500 501 502 503 504 505 506 507 508 509 510 511
/*
 * Install a new inobt btree root.  Caller is responsible for invalidating
 * and freeing the old btree blocks.
 */
void
xfs_inobt_commit_staged_btree(
	struct xfs_btree_cur	*cur,
	struct xfs_trans	*tp,
	struct xfs_buf		*agbp)
{
	struct xfs_agi		*agi = agbp->b_addr;
	struct xbtree_afakeroot	*afake = cur->bc_ag.afake;
512
	int			fields;
513 514 515 516

	ASSERT(cur->bc_flags & XFS_BTREE_STAGING);

	if (cur->bc_btnum == XFS_BTNUM_INO) {
517
		fields = XFS_AGI_ROOT | XFS_AGI_LEVEL;
518 519
		agi->agi_root = cpu_to_be32(afake->af_root);
		agi->agi_level = cpu_to_be32(afake->af_levels);
520 521 522 523 524
		if (xfs_sb_version_hasinobtcounts(&cur->bc_mp->m_sb)) {
			agi->agi_iblocks = cpu_to_be32(afake->af_blocks);
			fields |= XFS_AGI_IBLOCKS;
		}
		xfs_ialloc_log_agi(tp, agbp, fields);
525 526
		xfs_btree_commit_afakeroot(cur, tp, agbp, &xfs_inobt_ops);
	} else {
527
		fields = XFS_AGI_FREE_ROOT | XFS_AGI_FREE_LEVEL;
528 529
		agi->agi_free_root = cpu_to_be32(afake->af_root);
		agi->agi_free_level = cpu_to_be32(afake->af_levels);
530 531 532 533 534
		if (xfs_sb_version_hasinobtcounts(&cur->bc_mp->m_sb)) {
			agi->agi_fblocks = cpu_to_be32(afake->af_blocks);
			fields |= XFS_AGI_IBLOCKS;
		}
		xfs_ialloc_log_agi(tp, agbp, fields);
535 536 537 538
		xfs_btree_commit_afakeroot(cur, tp, agbp, &xfs_finobt_ops);
	}
}

539 540 541 542 543 544 545 546 547
/*
 * Calculate number of records in an inobt btree block.
 */
int
xfs_inobt_maxrecs(
	struct xfs_mount	*mp,
	int			blocklen,
	int			leaf)
{
548
	blocklen -= XFS_INOBT_BLOCK_LEN(mp);
549 550 551 552 553

	if (leaf)
		return blocklen / sizeof(xfs_inobt_rec_t);
	return blocklen / (sizeof(xfs_inobt_key_t) + sizeof(xfs_inobt_ptr_t));
}
554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604

/*
 * Convert the inode record holemask to an inode allocation bitmap. The inode
 * allocation bitmap is inode granularity and specifies whether an inode is
 * physically allocated on disk (not whether the inode is considered allocated
 * or free by the fs).
 *
 * A bit value of 1 means the inode is allocated, a value of 0 means it is free.
 */
uint64_t
xfs_inobt_irec_to_allocmask(
	struct xfs_inobt_rec_incore	*rec)
{
	uint64_t			bitmap = 0;
	uint64_t			inodespbit;
	int				nextbit;
	uint				allocbitmap;

	/*
	 * The holemask has 16-bits for a 64 inode record. Therefore each
	 * holemask bit represents multiple inodes. Create a mask of bits to set
	 * in the allocmask for each holemask bit.
	 */
	inodespbit = (1 << XFS_INODES_PER_HOLEMASK_BIT) - 1;

	/*
	 * Allocated inodes are represented by 0 bits in holemask. Invert the 0
	 * bits to 1 and convert to a uint so we can use xfs_next_bit(). Mask
	 * anything beyond the 16 holemask bits since this casts to a larger
	 * type.
	 */
	allocbitmap = ~rec->ir_holemask & ((1 << XFS_INOBT_HOLEMASK_BITS) - 1);

	/*
	 * allocbitmap is the inverted holemask so every set bit represents
	 * allocated inodes. To expand from 16-bit holemask granularity to
	 * 64-bit (e.g., bit-per-inode), set inodespbit bits in the target
	 * bitmap for every holemask bit.
	 */
	nextbit = xfs_next_bit(&allocbitmap, 1, 0);
	while (nextbit != -1) {
		ASSERT(nextbit < (sizeof(rec->ir_holemask) * NBBY));

		bitmap |= (inodespbit <<
			   (nextbit * XFS_INODES_PER_HOLEMASK_BIT));

		nextbit = xfs_next_bit(&allocbitmap, 1, nextbit + 1);
	}

	return bitmap;
}
605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635

#if defined(DEBUG) || defined(XFS_WARN)
/*
 * Verify that an in-core inode record has a valid inode count.
 */
int
xfs_inobt_rec_check_count(
	struct xfs_mount		*mp,
	struct xfs_inobt_rec_incore	*rec)
{
	int				inocount = 0;
	int				nextbit = 0;
	uint64_t			allocbmap;
	int				wordsz;

	wordsz = sizeof(allocbmap) / sizeof(unsigned int);
	allocbmap = xfs_inobt_irec_to_allocmask(rec);

	nextbit = xfs_next_bit((uint *) &allocbmap, wordsz, nextbit);
	while (nextbit != -1) {
		inocount++;
		nextbit = xfs_next_bit((uint *) &allocbmap, wordsz,
				       nextbit + 1);
	}

	if (inocount != rec->ir_count)
		return -EFSCORRUPTED;

	return 0;
}
#endif	/* DEBUG */
636 637 638

static xfs_extlen_t
xfs_inobt_max_size(
639 640
	struct xfs_mount	*mp,
	xfs_agnumber_t		agno)
641
{
642 643
	xfs_agblock_t		agblocks = xfs_ag_block_count(mp, agno);

644
	/* Bail out if we're uninitialized, which can happen in mkfs. */
D
Darrick J. Wong 已提交
645
	if (M_IGEO(mp)->inobt_mxr[0] == 0)
646 647
		return 0;

648 649 650 651 652 653 654 655 656
	/*
	 * The log is permanently allocated, so the space it occupies will
	 * never be available for the kinds of things that would require btree
	 * expansion.  We therefore can pretend the space isn't there.
	 */
	if (mp->m_sb.sb_logstart &&
	    XFS_FSB_TO_AGNO(mp, mp->m_sb.sb_logstart) == agno)
		agblocks -= mp->m_sb.sb_logblocks;

D
Darrick J. Wong 已提交
657
	return xfs_btree_calc_size(M_IGEO(mp)->inobt_mnr,
658 659
				(uint64_t)agblocks * mp->m_sb.sb_inopblock /
					XFS_INODES_PER_CHUNK);
660 661
}

662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681
/* Read AGI and create inobt cursor. */
int
xfs_inobt_cur(
	struct xfs_mount	*mp,
	struct xfs_trans	*tp,
	xfs_agnumber_t		agno,
	xfs_btnum_t		which,
	struct xfs_btree_cur	**curpp,
	struct xfs_buf		**agi_bpp)
{
	struct xfs_btree_cur	*cur;
	int			error;

	ASSERT(*agi_bpp == NULL);
	ASSERT(*curpp == NULL);

	error = xfs_ialloc_read_agi(mp, tp, agno, agi_bpp);
	if (error)
		return error;

682
	cur = xfs_inobt_init_cursor(mp, tp, *agi_bpp, agno, NULL, which);
683 684 685 686
	*curpp = cur;
	return 0;
}

687 688 689
static int
xfs_inobt_count_blocks(
	struct xfs_mount	*mp,
690
	struct xfs_trans	*tp,
691
	struct xfs_perag	*pag,
692 693 694
	xfs_btnum_t		btnum,
	xfs_extlen_t		*tree_blocks)
{
695 696
	struct xfs_buf		*agbp = NULL;
	struct xfs_btree_cur	*cur = NULL;
697 698
	int			error;

699
	error = xfs_inobt_cur(mp, tp, pag->pag_agno, btnum, &cur, &agbp);
700 701 702 703
	if (error)
		return error;

	error = xfs_btree_count_blocks(cur, tree_blocks);
704
	xfs_btree_del_cursor(cur, error);
705
	xfs_trans_brelse(tp, agbp);
706 707 708 709

	return error;
}

710 711 712 713 714
/* Read finobt block count from AGI header. */
static int
xfs_finobt_read_blocks(
	struct xfs_mount	*mp,
	struct xfs_trans	*tp,
715
	struct xfs_perag	*pag,
716 717 718 719 720 721
	xfs_extlen_t		*tree_blocks)
{
	struct xfs_buf		*agbp;
	struct xfs_agi		*agi;
	int			error;

722
	error = xfs_ialloc_read_agi(mp, tp, pag->pag_agno, &agbp);
723 724 725 726 727 728 729 730 731
	if (error)
		return error;

	agi = agbp->b_addr;
	*tree_blocks = be32_to_cpu(agi->agi_fblocks);
	xfs_trans_brelse(tp, agbp);
	return 0;
}

732 733 734 735 736 737
/*
 * Figure out how many blocks to reserve and how many are used by this btree.
 */
int
xfs_finobt_calc_reserves(
	struct xfs_mount	*mp,
738
	struct xfs_trans	*tp,
739
	struct xfs_perag	*pag,
740 741 742 743 744 745 746 747 748
	xfs_extlen_t		*ask,
	xfs_extlen_t		*used)
{
	xfs_extlen_t		tree_len = 0;
	int			error;

	if (!xfs_sb_version_hasfinobt(&mp->m_sb))
		return 0;

749
	if (xfs_sb_version_hasinobtcounts(&mp->m_sb))
750
		error = xfs_finobt_read_blocks(mp, tp, pag, &tree_len);
751
	else
752
		error = xfs_inobt_count_blocks(mp, tp, pag, XFS_BTNUM_FINO,
753
				&tree_len);
754 755 756
	if (error)
		return error;

757
	*ask += xfs_inobt_max_size(mp, pag->pag_agno);
758 759 760
	*used += tree_len;
	return 0;
}
761 762 763 764 765 766 767

/* Calculate the inobt btree size for some records. */
xfs_extlen_t
xfs_iallocbt_calc_size(
	struct xfs_mount	*mp,
	unsigned long long	len)
{
D
Darrick J. Wong 已提交
768
	return xfs_btree_calc_size(M_IGEO(mp)->inobt_mnr, len);
769
}