xref: /openbmc/linux/fs/xfs/libxfs/xfs_ialloc_btree.c (revision 93f5715e)
1 // SPDX-License-Identifier: GPL-2.0
2 /*
3  * Copyright (c) 2000-2001,2005 Silicon Graphics, Inc.
4  * All Rights Reserved.
5  */
6 #include "xfs.h"
7 #include "xfs_fs.h"
8 #include "xfs_shared.h"
9 #include "xfs_format.h"
10 #include "xfs_log_format.h"
11 #include "xfs_trans_resv.h"
12 #include "xfs_bit.h"
13 #include "xfs_mount.h"
14 #include "xfs_inode.h"
15 #include "xfs_btree.h"
16 #include "xfs_ialloc.h"
17 #include "xfs_ialloc_btree.h"
18 #include "xfs_alloc.h"
19 #include "xfs_error.h"
20 #include "xfs_trace.h"
21 #include "xfs_cksum.h"
22 #include "xfs_trans.h"
23 #include "xfs_rmap.h"
24 
25 
26 STATIC int
27 xfs_inobt_get_minrecs(
28 	struct xfs_btree_cur	*cur,
29 	int			level)
30 {
31 	return cur->bc_mp->m_inobt_mnr[level != 0];
32 }
33 
34 STATIC struct xfs_btree_cur *
35 xfs_inobt_dup_cursor(
36 	struct xfs_btree_cur	*cur)
37 {
38 	return xfs_inobt_init_cursor(cur->bc_mp, cur->bc_tp,
39 			cur->bc_private.a.agbp, cur->bc_private.a.agno,
40 			cur->bc_btnum);
41 }
42 
43 STATIC void
44 xfs_inobt_set_root(
45 	struct xfs_btree_cur	*cur,
46 	union xfs_btree_ptr	*nptr,
47 	int			inc)	/* level change */
48 {
49 	struct xfs_buf		*agbp = cur->bc_private.a.agbp;
50 	struct xfs_agi		*agi = XFS_BUF_TO_AGI(agbp);
51 
52 	agi->agi_root = nptr->s;
53 	be32_add_cpu(&agi->agi_level, inc);
54 	xfs_ialloc_log_agi(cur->bc_tp, agbp, XFS_AGI_ROOT | XFS_AGI_LEVEL);
55 }
56 
57 STATIC void
58 xfs_finobt_set_root(
59 	struct xfs_btree_cur	*cur,
60 	union xfs_btree_ptr	*nptr,
61 	int			inc)	/* level change */
62 {
63 	struct xfs_buf		*agbp = cur->bc_private.a.agbp;
64 	struct xfs_agi		*agi = XFS_BUF_TO_AGI(agbp);
65 
66 	agi->agi_free_root = nptr->s;
67 	be32_add_cpu(&agi->agi_free_level, inc);
68 	xfs_ialloc_log_agi(cur->bc_tp, agbp,
69 			   XFS_AGI_FREE_ROOT | XFS_AGI_FREE_LEVEL);
70 }
71 
72 STATIC int
73 __xfs_inobt_alloc_block(
74 	struct xfs_btree_cur	*cur,
75 	union xfs_btree_ptr	*start,
76 	union xfs_btree_ptr	*new,
77 	int			*stat,
78 	enum xfs_ag_resv_type	resv)
79 {
80 	xfs_alloc_arg_t		args;		/* block allocation args */
81 	int			error;		/* error return value */
82 	xfs_agblock_t		sbno = be32_to_cpu(start->s);
83 
84 	memset(&args, 0, sizeof(args));
85 	args.tp = cur->bc_tp;
86 	args.mp = cur->bc_mp;
87 	xfs_rmap_ag_owner(&args.oinfo, XFS_RMAP_OWN_INOBT);
88 	args.fsbno = XFS_AGB_TO_FSB(args.mp, cur->bc_private.a.agno, sbno);
89 	args.minlen = 1;
90 	args.maxlen = 1;
91 	args.prod = 1;
92 	args.type = XFS_ALLOCTYPE_NEAR_BNO;
93 	args.resv = resv;
94 
95 	error = xfs_alloc_vextent(&args);
96 	if (error)
97 		return error;
98 
99 	if (args.fsbno == NULLFSBLOCK) {
100 		*stat = 0;
101 		return 0;
102 	}
103 	ASSERT(args.len == 1);
104 
105 	new->s = cpu_to_be32(XFS_FSB_TO_AGBNO(args.mp, args.fsbno));
106 	*stat = 1;
107 	return 0;
108 }
109 
110 STATIC int
111 xfs_inobt_alloc_block(
112 	struct xfs_btree_cur	*cur,
113 	union xfs_btree_ptr	*start,
114 	union xfs_btree_ptr	*new,
115 	int			*stat)
116 {
117 	return __xfs_inobt_alloc_block(cur, start, new, stat, XFS_AG_RESV_NONE);
118 }
119 
120 STATIC int
121 xfs_finobt_alloc_block(
122 	struct xfs_btree_cur	*cur,
123 	union xfs_btree_ptr	*start,
124 	union xfs_btree_ptr	*new,
125 	int			*stat)
126 {
127 	if (cur->bc_mp->m_inotbt_nores)
128 		return xfs_inobt_alloc_block(cur, start, new, stat);
129 	return __xfs_inobt_alloc_block(cur, start, new, stat,
130 			XFS_AG_RESV_METADATA);
131 }
132 
133 STATIC int
134 __xfs_inobt_free_block(
135 	struct xfs_btree_cur	*cur,
136 	struct xfs_buf		*bp,
137 	enum xfs_ag_resv_type	resv)
138 {
139 	struct xfs_owner_info	oinfo;
140 
141 	xfs_rmap_ag_owner(&oinfo, XFS_RMAP_OWN_INOBT);
142 	return xfs_free_extent(cur->bc_tp,
143 			XFS_DADDR_TO_FSB(cur->bc_mp, XFS_BUF_ADDR(bp)), 1,
144 			&oinfo, resv);
145 }
146 
147 STATIC int
148 xfs_inobt_free_block(
149 	struct xfs_btree_cur	*cur,
150 	struct xfs_buf		*bp)
151 {
152 	return __xfs_inobt_free_block(cur, bp, XFS_AG_RESV_NONE);
153 }
154 
155 STATIC int
156 xfs_finobt_free_block(
157 	struct xfs_btree_cur	*cur,
158 	struct xfs_buf		*bp)
159 {
160 	if (cur->bc_mp->m_inotbt_nores)
161 		return xfs_inobt_free_block(cur, bp);
162 	return __xfs_inobt_free_block(cur, bp, XFS_AG_RESV_METADATA);
163 }
164 
165 STATIC int
166 xfs_inobt_get_maxrecs(
167 	struct xfs_btree_cur	*cur,
168 	int			level)
169 {
170 	return cur->bc_mp->m_inobt_mxr[level != 0];
171 }
172 
173 STATIC void
174 xfs_inobt_init_key_from_rec(
175 	union xfs_btree_key	*key,
176 	union xfs_btree_rec	*rec)
177 {
178 	key->inobt.ir_startino = rec->inobt.ir_startino;
179 }
180 
181 STATIC void
182 xfs_inobt_init_high_key_from_rec(
183 	union xfs_btree_key	*key,
184 	union xfs_btree_rec	*rec)
185 {
186 	__u32			x;
187 
188 	x = be32_to_cpu(rec->inobt.ir_startino);
189 	x += XFS_INODES_PER_CHUNK - 1;
190 	key->inobt.ir_startino = cpu_to_be32(x);
191 }
192 
193 STATIC void
194 xfs_inobt_init_rec_from_cur(
195 	struct xfs_btree_cur	*cur,
196 	union xfs_btree_rec	*rec)
197 {
198 	rec->inobt.ir_startino = cpu_to_be32(cur->bc_rec.i.ir_startino);
199 	if (xfs_sb_version_hassparseinodes(&cur->bc_mp->m_sb)) {
200 		rec->inobt.ir_u.sp.ir_holemask =
201 					cpu_to_be16(cur->bc_rec.i.ir_holemask);
202 		rec->inobt.ir_u.sp.ir_count = cur->bc_rec.i.ir_count;
203 		rec->inobt.ir_u.sp.ir_freecount = cur->bc_rec.i.ir_freecount;
204 	} else {
205 		/* ir_holemask/ir_count not supported on-disk */
206 		rec->inobt.ir_u.f.ir_freecount =
207 					cpu_to_be32(cur->bc_rec.i.ir_freecount);
208 	}
209 	rec->inobt.ir_free = cpu_to_be64(cur->bc_rec.i.ir_free);
210 }
211 
212 /*
213  * initial value of ptr for lookup
214  */
215 STATIC void
216 xfs_inobt_init_ptr_from_cur(
217 	struct xfs_btree_cur	*cur,
218 	union xfs_btree_ptr	*ptr)
219 {
220 	struct xfs_agi		*agi = XFS_BUF_TO_AGI(cur->bc_private.a.agbp);
221 
222 	ASSERT(cur->bc_private.a.agno == be32_to_cpu(agi->agi_seqno));
223 
224 	ptr->s = agi->agi_root;
225 }
226 
227 STATIC void
228 xfs_finobt_init_ptr_from_cur(
229 	struct xfs_btree_cur	*cur,
230 	union xfs_btree_ptr	*ptr)
231 {
232 	struct xfs_agi		*agi = XFS_BUF_TO_AGI(cur->bc_private.a.agbp);
233 
234 	ASSERT(cur->bc_private.a.agno == be32_to_cpu(agi->agi_seqno));
235 	ptr->s = agi->agi_free_root;
236 }
237 
238 STATIC int64_t
239 xfs_inobt_key_diff(
240 	struct xfs_btree_cur	*cur,
241 	union xfs_btree_key	*key)
242 {
243 	return (int64_t)be32_to_cpu(key->inobt.ir_startino) -
244 			  cur->bc_rec.i.ir_startino;
245 }
246 
247 STATIC int64_t
248 xfs_inobt_diff_two_keys(
249 	struct xfs_btree_cur	*cur,
250 	union xfs_btree_key	*k1,
251 	union xfs_btree_key	*k2)
252 {
253 	return (int64_t)be32_to_cpu(k1->inobt.ir_startino) -
254 			  be32_to_cpu(k2->inobt.ir_startino);
255 }
256 
257 static xfs_failaddr_t
258 xfs_inobt_verify(
259 	struct xfs_buf		*bp)
260 {
261 	struct xfs_mount	*mp = bp->b_target->bt_mount;
262 	struct xfs_btree_block	*block = XFS_BUF_TO_BLOCK(bp);
263 	xfs_failaddr_t		fa;
264 	unsigned int		level;
265 
266 	/*
267 	 * During growfs operations, we can't verify the exact owner as the
268 	 * perag is not fully initialised and hence not attached to the buffer.
269 	 *
270 	 * Similarly, during log recovery we will have a perag structure
271 	 * attached, but the agi information will not yet have been initialised
272 	 * from the on disk AGI. We don't currently use any of this information,
273 	 * but beware of the landmine (i.e. need to check pag->pagi_init) if we
274 	 * ever do.
275 	 */
276 	switch (block->bb_magic) {
277 	case cpu_to_be32(XFS_IBT_CRC_MAGIC):
278 	case cpu_to_be32(XFS_FIBT_CRC_MAGIC):
279 		fa = xfs_btree_sblock_v5hdr_verify(bp);
280 		if (fa)
281 			return fa;
282 		/* fall through */
283 	case cpu_to_be32(XFS_IBT_MAGIC):
284 	case cpu_to_be32(XFS_FIBT_MAGIC):
285 		break;
286 	default:
287 		return __this_address;
288 	}
289 
290 	/* level verification */
291 	level = be16_to_cpu(block->bb_level);
292 	if (level >= mp->m_in_maxlevels)
293 		return __this_address;
294 
295 	return xfs_btree_sblock_verify(bp, mp->m_inobt_mxr[level != 0]);
296 }
297 
298 static void
299 xfs_inobt_read_verify(
300 	struct xfs_buf	*bp)
301 {
302 	xfs_failaddr_t	fa;
303 
304 	if (!xfs_btree_sblock_verify_crc(bp))
305 		xfs_verifier_error(bp, -EFSBADCRC, __this_address);
306 	else {
307 		fa = xfs_inobt_verify(bp);
308 		if (fa)
309 			xfs_verifier_error(bp, -EFSCORRUPTED, fa);
310 	}
311 
312 	if (bp->b_error)
313 		trace_xfs_btree_corrupt(bp, _RET_IP_);
314 }
315 
316 static void
317 xfs_inobt_write_verify(
318 	struct xfs_buf	*bp)
319 {
320 	xfs_failaddr_t	fa;
321 
322 	fa = xfs_inobt_verify(bp);
323 	if (fa) {
324 		trace_xfs_btree_corrupt(bp, _RET_IP_);
325 		xfs_verifier_error(bp, -EFSCORRUPTED, fa);
326 		return;
327 	}
328 	xfs_btree_sblock_calc_crc(bp);
329 
330 }
331 
332 const struct xfs_buf_ops xfs_inobt_buf_ops = {
333 	.name = "xfs_inobt",
334 	.verify_read = xfs_inobt_read_verify,
335 	.verify_write = xfs_inobt_write_verify,
336 	.verify_struct = xfs_inobt_verify,
337 };
338 
339 STATIC int
340 xfs_inobt_keys_inorder(
341 	struct xfs_btree_cur	*cur,
342 	union xfs_btree_key	*k1,
343 	union xfs_btree_key	*k2)
344 {
345 	return be32_to_cpu(k1->inobt.ir_startino) <
346 		be32_to_cpu(k2->inobt.ir_startino);
347 }
348 
349 STATIC int
350 xfs_inobt_recs_inorder(
351 	struct xfs_btree_cur	*cur,
352 	union xfs_btree_rec	*r1,
353 	union xfs_btree_rec	*r2)
354 {
355 	return be32_to_cpu(r1->inobt.ir_startino) + XFS_INODES_PER_CHUNK <=
356 		be32_to_cpu(r2->inobt.ir_startino);
357 }
358 
359 static const struct xfs_btree_ops xfs_inobt_ops = {
360 	.rec_len		= sizeof(xfs_inobt_rec_t),
361 	.key_len		= sizeof(xfs_inobt_key_t),
362 
363 	.dup_cursor		= xfs_inobt_dup_cursor,
364 	.set_root		= xfs_inobt_set_root,
365 	.alloc_block		= xfs_inobt_alloc_block,
366 	.free_block		= xfs_inobt_free_block,
367 	.get_minrecs		= xfs_inobt_get_minrecs,
368 	.get_maxrecs		= xfs_inobt_get_maxrecs,
369 	.init_key_from_rec	= xfs_inobt_init_key_from_rec,
370 	.init_high_key_from_rec	= xfs_inobt_init_high_key_from_rec,
371 	.init_rec_from_cur	= xfs_inobt_init_rec_from_cur,
372 	.init_ptr_from_cur	= xfs_inobt_init_ptr_from_cur,
373 	.key_diff		= xfs_inobt_key_diff,
374 	.buf_ops		= &xfs_inobt_buf_ops,
375 	.diff_two_keys		= xfs_inobt_diff_two_keys,
376 	.keys_inorder		= xfs_inobt_keys_inorder,
377 	.recs_inorder		= xfs_inobt_recs_inorder,
378 };
379 
380 static const struct xfs_btree_ops xfs_finobt_ops = {
381 	.rec_len		= sizeof(xfs_inobt_rec_t),
382 	.key_len		= sizeof(xfs_inobt_key_t),
383 
384 	.dup_cursor		= xfs_inobt_dup_cursor,
385 	.set_root		= xfs_finobt_set_root,
386 	.alloc_block		= xfs_finobt_alloc_block,
387 	.free_block		= xfs_finobt_free_block,
388 	.get_minrecs		= xfs_inobt_get_minrecs,
389 	.get_maxrecs		= xfs_inobt_get_maxrecs,
390 	.init_key_from_rec	= xfs_inobt_init_key_from_rec,
391 	.init_high_key_from_rec	= xfs_inobt_init_high_key_from_rec,
392 	.init_rec_from_cur	= xfs_inobt_init_rec_from_cur,
393 	.init_ptr_from_cur	= xfs_finobt_init_ptr_from_cur,
394 	.key_diff		= xfs_inobt_key_diff,
395 	.buf_ops		= &xfs_inobt_buf_ops,
396 	.diff_two_keys		= xfs_inobt_diff_two_keys,
397 	.keys_inorder		= xfs_inobt_keys_inorder,
398 	.recs_inorder		= xfs_inobt_recs_inorder,
399 };
400 
401 /*
402  * Allocate a new inode btree cursor.
403  */
404 struct xfs_btree_cur *				/* new inode btree cursor */
405 xfs_inobt_init_cursor(
406 	struct xfs_mount	*mp,		/* file system mount point */
407 	struct xfs_trans	*tp,		/* transaction pointer */
408 	struct xfs_buf		*agbp,		/* buffer for agi structure */
409 	xfs_agnumber_t		agno,		/* allocation group number */
410 	xfs_btnum_t		btnum)		/* ialloc or free ino btree */
411 {
412 	struct xfs_agi		*agi = XFS_BUF_TO_AGI(agbp);
413 	struct xfs_btree_cur	*cur;
414 
415 	cur = kmem_zone_zalloc(xfs_btree_cur_zone, KM_NOFS);
416 
417 	cur->bc_tp = tp;
418 	cur->bc_mp = mp;
419 	cur->bc_btnum = btnum;
420 	if (btnum == XFS_BTNUM_INO) {
421 		cur->bc_nlevels = be32_to_cpu(agi->agi_level);
422 		cur->bc_ops = &xfs_inobt_ops;
423 		cur->bc_statoff = XFS_STATS_CALC_INDEX(xs_ibt_2);
424 	} else {
425 		cur->bc_nlevels = be32_to_cpu(agi->agi_free_level);
426 		cur->bc_ops = &xfs_finobt_ops;
427 		cur->bc_statoff = XFS_STATS_CALC_INDEX(xs_fibt_2);
428 	}
429 
430 	cur->bc_blocklog = mp->m_sb.sb_blocklog;
431 
432 	if (xfs_sb_version_hascrc(&mp->m_sb))
433 		cur->bc_flags |= XFS_BTREE_CRC_BLOCKS;
434 
435 	cur->bc_private.a.agbp = agbp;
436 	cur->bc_private.a.agno = agno;
437 
438 	return cur;
439 }
440 
441 /*
442  * Calculate number of records in an inobt btree block.
443  */
444 int
445 xfs_inobt_maxrecs(
446 	struct xfs_mount	*mp,
447 	int			blocklen,
448 	int			leaf)
449 {
450 	blocklen -= XFS_INOBT_BLOCK_LEN(mp);
451 
452 	if (leaf)
453 		return blocklen / sizeof(xfs_inobt_rec_t);
454 	return blocklen / (sizeof(xfs_inobt_key_t) + sizeof(xfs_inobt_ptr_t));
455 }
456 
457 /*
458  * Convert the inode record holemask to an inode allocation bitmap. The inode
459  * allocation bitmap is inode granularity and specifies whether an inode is
460  * physically allocated on disk (not whether the inode is considered allocated
461  * or free by the fs).
462  *
463  * A bit value of 1 means the inode is allocated, a value of 0 means it is free.
464  */
465 uint64_t
466 xfs_inobt_irec_to_allocmask(
467 	struct xfs_inobt_rec_incore	*rec)
468 {
469 	uint64_t			bitmap = 0;
470 	uint64_t			inodespbit;
471 	int				nextbit;
472 	uint				allocbitmap;
473 
474 	/*
475 	 * The holemask has 16-bits for a 64 inode record. Therefore each
476 	 * holemask bit represents multiple inodes. Create a mask of bits to set
477 	 * in the allocmask for each holemask bit.
478 	 */
479 	inodespbit = (1 << XFS_INODES_PER_HOLEMASK_BIT) - 1;
480 
481 	/*
482 	 * Allocated inodes are represented by 0 bits in holemask. Invert the 0
483 	 * bits to 1 and convert to a uint so we can use xfs_next_bit(). Mask
484 	 * anything beyond the 16 holemask bits since this casts to a larger
485 	 * type.
486 	 */
487 	allocbitmap = ~rec->ir_holemask & ((1 << XFS_INOBT_HOLEMASK_BITS) - 1);
488 
489 	/*
490 	 * allocbitmap is the inverted holemask so every set bit represents
491 	 * allocated inodes. To expand from 16-bit holemask granularity to
492 	 * 64-bit (e.g., bit-per-inode), set inodespbit bits in the target
493 	 * bitmap for every holemask bit.
494 	 */
495 	nextbit = xfs_next_bit(&allocbitmap, 1, 0);
496 	while (nextbit != -1) {
497 		ASSERT(nextbit < (sizeof(rec->ir_holemask) * NBBY));
498 
499 		bitmap |= (inodespbit <<
500 			   (nextbit * XFS_INODES_PER_HOLEMASK_BIT));
501 
502 		nextbit = xfs_next_bit(&allocbitmap, 1, nextbit + 1);
503 	}
504 
505 	return bitmap;
506 }
507 
508 #if defined(DEBUG) || defined(XFS_WARN)
509 /*
510  * Verify that an in-core inode record has a valid inode count.
511  */
512 int
513 xfs_inobt_rec_check_count(
514 	struct xfs_mount		*mp,
515 	struct xfs_inobt_rec_incore	*rec)
516 {
517 	int				inocount = 0;
518 	int				nextbit = 0;
519 	uint64_t			allocbmap;
520 	int				wordsz;
521 
522 	wordsz = sizeof(allocbmap) / sizeof(unsigned int);
523 	allocbmap = xfs_inobt_irec_to_allocmask(rec);
524 
525 	nextbit = xfs_next_bit((uint *) &allocbmap, wordsz, nextbit);
526 	while (nextbit != -1) {
527 		inocount++;
528 		nextbit = xfs_next_bit((uint *) &allocbmap, wordsz,
529 				       nextbit + 1);
530 	}
531 
532 	if (inocount != rec->ir_count)
533 		return -EFSCORRUPTED;
534 
535 	return 0;
536 }
537 #endif	/* DEBUG */
538 
539 static xfs_extlen_t
540 xfs_inobt_max_size(
541 	struct xfs_mount	*mp)
542 {
543 	/* Bail out if we're uninitialized, which can happen in mkfs. */
544 	if (mp->m_inobt_mxr[0] == 0)
545 		return 0;
546 
547 	return xfs_btree_calc_size(mp->m_inobt_mnr,
548 		(uint64_t)mp->m_sb.sb_agblocks * mp->m_sb.sb_inopblock /
549 				XFS_INODES_PER_CHUNK);
550 }
551 
552 static int
553 xfs_inobt_count_blocks(
554 	struct xfs_mount	*mp,
555 	struct xfs_trans	*tp,
556 	xfs_agnumber_t		agno,
557 	xfs_btnum_t		btnum,
558 	xfs_extlen_t		*tree_blocks)
559 {
560 	struct xfs_buf		*agbp;
561 	struct xfs_btree_cur	*cur;
562 	int			error;
563 
564 	error = xfs_ialloc_read_agi(mp, tp, agno, &agbp);
565 	if (error)
566 		return error;
567 
568 	cur = xfs_inobt_init_cursor(mp, tp, agbp, agno, btnum);
569 	error = xfs_btree_count_blocks(cur, tree_blocks);
570 	xfs_btree_del_cursor(cur, error);
571 	xfs_trans_brelse(tp, agbp);
572 
573 	return error;
574 }
575 
576 /*
577  * Figure out how many blocks to reserve and how many are used by this btree.
578  */
579 int
580 xfs_finobt_calc_reserves(
581 	struct xfs_mount	*mp,
582 	struct xfs_trans	*tp,
583 	xfs_agnumber_t		agno,
584 	xfs_extlen_t		*ask,
585 	xfs_extlen_t		*used)
586 {
587 	xfs_extlen_t		tree_len = 0;
588 	int			error;
589 
590 	if (!xfs_sb_version_hasfinobt(&mp->m_sb))
591 		return 0;
592 
593 	error = xfs_inobt_count_blocks(mp, tp, agno, XFS_BTNUM_FINO, &tree_len);
594 	if (error)
595 		return error;
596 
597 	*ask += xfs_inobt_max_size(mp);
598 	*used += tree_len;
599 	return 0;
600 }
601 
602 /* Calculate the inobt btree size for some records. */
603 xfs_extlen_t
604 xfs_iallocbt_calc_size(
605 	struct xfs_mount	*mp,
606 	unsigned long long	len)
607 {
608 	return xfs_btree_calc_size(mp->m_inobt_mnr, len);
609 }
610