xfs_rmap_btree.c revision 38c26bfd
1// SPDX-License-Identifier: GPL-2.0
2/*
3 * Copyright (c) 2014 Red Hat, 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_mount.h"
13#include "xfs_trans.h"
14#include "xfs_alloc.h"
15#include "xfs_btree.h"
16#include "xfs_btree_staging.h"
17#include "xfs_rmap.h"
18#include "xfs_rmap_btree.h"
19#include "xfs_trace.h"
20#include "xfs_error.h"
21#include "xfs_extent_busy.h"
22#include "xfs_ag.h"
23#include "xfs_ag_resv.h"
24
25/*
26 * Reverse map btree.
27 *
28 * This is a per-ag tree used to track the owner(s) of a given extent. With
29 * reflink it is possible for there to be multiple owners, which is a departure
30 * from classic XFS. Owner records for data extents are inserted when the
31 * extent is mapped and removed when an extent is unmapped.  Owner records for
32 * all other block types (i.e. metadata) are inserted when an extent is
33 * allocated and removed when an extent is freed. There can only be one owner
34 * of a metadata extent, usually an inode or some other metadata structure like
35 * an AG btree.
36 *
37 * The rmap btree is part of the free space management, so blocks for the tree
38 * are sourced from the agfl. Hence we need transaction reservation support for
39 * this tree so that the freelist is always large enough. This also impacts on
40 * the minimum space we need to leave free in the AG.
41 *
42 * The tree is ordered by [ag block, owner, offset]. This is a large key size,
43 * but it is the only way to enforce unique keys when a block can be owned by
44 * multiple files at any offset. There's no need to order/search by extent
45 * size for online updating/management of the tree. It is intended that most
46 * reverse lookups will be to find the owner(s) of a particular block, or to
47 * try to recover tree and file data from corrupt primary metadata.
48 */
49
50static struct xfs_btree_cur *
51xfs_rmapbt_dup_cursor(
52	struct xfs_btree_cur	*cur)
53{
54	return xfs_rmapbt_init_cursor(cur->bc_mp, cur->bc_tp,
55				cur->bc_ag.agbp, cur->bc_ag.pag);
56}
57
58STATIC void
59xfs_rmapbt_set_root(
60	struct xfs_btree_cur		*cur,
61	const union xfs_btree_ptr	*ptr,
62	int				inc)
63{
64	struct xfs_buf		*agbp = cur->bc_ag.agbp;
65	struct xfs_agf		*agf = agbp->b_addr;
66	int			btnum = cur->bc_btnum;
67
68	ASSERT(ptr->s != 0);
69
70	agf->agf_roots[btnum] = ptr->s;
71	be32_add_cpu(&agf->agf_levels[btnum], inc);
72	cur->bc_ag.pag->pagf_levels[btnum] += inc;
73
74	xfs_alloc_log_agf(cur->bc_tp, agbp, XFS_AGF_ROOTS | XFS_AGF_LEVELS);
75}
76
77STATIC int
78xfs_rmapbt_alloc_block(
79	struct xfs_btree_cur		*cur,
80	const union xfs_btree_ptr	*start,
81	union xfs_btree_ptr		*new,
82	int				*stat)
83{
84	struct xfs_buf		*agbp = cur->bc_ag.agbp;
85	struct xfs_agf		*agf = agbp->b_addr;
86	struct xfs_perag	*pag = cur->bc_ag.pag;
87	int			error;
88	xfs_agblock_t		bno;
89
90	/* Allocate the new block from the freelist. If we can't, give up.  */
91	error = xfs_alloc_get_freelist(cur->bc_tp, cur->bc_ag.agbp,
92				       &bno, 1);
93	if (error)
94		return error;
95
96	trace_xfs_rmapbt_alloc_block(cur->bc_mp, pag->pag_agno, bno, 1);
97	if (bno == NULLAGBLOCK) {
98		*stat = 0;
99		return 0;
100	}
101
102	xfs_extent_busy_reuse(cur->bc_mp, pag, bno, 1, false);
103
104	new->s = cpu_to_be32(bno);
105	be32_add_cpu(&agf->agf_rmap_blocks, 1);
106	xfs_alloc_log_agf(cur->bc_tp, agbp, XFS_AGF_RMAP_BLOCKS);
107
108	xfs_ag_resv_rmapbt_alloc(cur->bc_mp, pag->pag_agno);
109
110	*stat = 1;
111	return 0;
112}
113
114STATIC int
115xfs_rmapbt_free_block(
116	struct xfs_btree_cur	*cur,
117	struct xfs_buf		*bp)
118{
119	struct xfs_buf		*agbp = cur->bc_ag.agbp;
120	struct xfs_agf		*agf = agbp->b_addr;
121	struct xfs_perag	*pag = cur->bc_ag.pag;
122	xfs_agblock_t		bno;
123	int			error;
124
125	bno = xfs_daddr_to_agbno(cur->bc_mp, XFS_BUF_ADDR(bp));
126	trace_xfs_rmapbt_free_block(cur->bc_mp, pag->pag_agno,
127			bno, 1);
128	be32_add_cpu(&agf->agf_rmap_blocks, -1);
129	xfs_alloc_log_agf(cur->bc_tp, agbp, XFS_AGF_RMAP_BLOCKS);
130	error = xfs_alloc_put_freelist(cur->bc_tp, agbp, NULL, bno, 1);
131	if (error)
132		return error;
133
134	xfs_extent_busy_insert(cur->bc_tp, pag, bno, 1,
135			      XFS_EXTENT_BUSY_SKIP_DISCARD);
136
137	xfs_ag_resv_free_extent(pag, XFS_AG_RESV_RMAPBT, NULL, 1);
138	return 0;
139}
140
141STATIC int
142xfs_rmapbt_get_minrecs(
143	struct xfs_btree_cur	*cur,
144	int			level)
145{
146	return cur->bc_mp->m_rmap_mnr[level != 0];
147}
148
149STATIC int
150xfs_rmapbt_get_maxrecs(
151	struct xfs_btree_cur	*cur,
152	int			level)
153{
154	return cur->bc_mp->m_rmap_mxr[level != 0];
155}
156
157STATIC void
158xfs_rmapbt_init_key_from_rec(
159	union xfs_btree_key		*key,
160	const union xfs_btree_rec	*rec)
161{
162	key->rmap.rm_startblock = rec->rmap.rm_startblock;
163	key->rmap.rm_owner = rec->rmap.rm_owner;
164	key->rmap.rm_offset = rec->rmap.rm_offset;
165}
166
167/*
168 * The high key for a reverse mapping record can be computed by shifting
169 * the startblock and offset to the highest value that would still map
170 * to that record.  In practice this means that we add blockcount-1 to
171 * the startblock for all records, and if the record is for a data/attr
172 * fork mapping, we add blockcount-1 to the offset too.
173 */
174STATIC void
175xfs_rmapbt_init_high_key_from_rec(
176	union xfs_btree_key		*key,
177	const union xfs_btree_rec	*rec)
178{
179	uint64_t			off;
180	int				adj;
181
182	adj = be32_to_cpu(rec->rmap.rm_blockcount) - 1;
183
184	key->rmap.rm_startblock = rec->rmap.rm_startblock;
185	be32_add_cpu(&key->rmap.rm_startblock, adj);
186	key->rmap.rm_owner = rec->rmap.rm_owner;
187	key->rmap.rm_offset = rec->rmap.rm_offset;
188	if (XFS_RMAP_NON_INODE_OWNER(be64_to_cpu(rec->rmap.rm_owner)) ||
189	    XFS_RMAP_IS_BMBT_BLOCK(be64_to_cpu(rec->rmap.rm_offset)))
190		return;
191	off = be64_to_cpu(key->rmap.rm_offset);
192	off = (XFS_RMAP_OFF(off) + adj) | (off & ~XFS_RMAP_OFF_MASK);
193	key->rmap.rm_offset = cpu_to_be64(off);
194}
195
196STATIC void
197xfs_rmapbt_init_rec_from_cur(
198	struct xfs_btree_cur	*cur,
199	union xfs_btree_rec	*rec)
200{
201	rec->rmap.rm_startblock = cpu_to_be32(cur->bc_rec.r.rm_startblock);
202	rec->rmap.rm_blockcount = cpu_to_be32(cur->bc_rec.r.rm_blockcount);
203	rec->rmap.rm_owner = cpu_to_be64(cur->bc_rec.r.rm_owner);
204	rec->rmap.rm_offset = cpu_to_be64(
205			xfs_rmap_irec_offset_pack(&cur->bc_rec.r));
206}
207
208STATIC void
209xfs_rmapbt_init_ptr_from_cur(
210	struct xfs_btree_cur	*cur,
211	union xfs_btree_ptr	*ptr)
212{
213	struct xfs_agf		*agf = cur->bc_ag.agbp->b_addr;
214
215	ASSERT(cur->bc_ag.pag->pag_agno == be32_to_cpu(agf->agf_seqno));
216
217	ptr->s = agf->agf_roots[cur->bc_btnum];
218}
219
220STATIC int64_t
221xfs_rmapbt_key_diff(
222	struct xfs_btree_cur		*cur,
223	const union xfs_btree_key	*key)
224{
225	struct xfs_rmap_irec		*rec = &cur->bc_rec.r;
226	const struct xfs_rmap_key	*kp = &key->rmap;
227	__u64				x, y;
228	int64_t				d;
229
230	d = (int64_t)be32_to_cpu(kp->rm_startblock) - rec->rm_startblock;
231	if (d)
232		return d;
233
234	x = be64_to_cpu(kp->rm_owner);
235	y = rec->rm_owner;
236	if (x > y)
237		return 1;
238	else if (y > x)
239		return -1;
240
241	x = XFS_RMAP_OFF(be64_to_cpu(kp->rm_offset));
242	y = rec->rm_offset;
243	if (x > y)
244		return 1;
245	else if (y > x)
246		return -1;
247	return 0;
248}
249
250STATIC int64_t
251xfs_rmapbt_diff_two_keys(
252	struct xfs_btree_cur		*cur,
253	const union xfs_btree_key	*k1,
254	const union xfs_btree_key	*k2)
255{
256	const struct xfs_rmap_key	*kp1 = &k1->rmap;
257	const struct xfs_rmap_key	*kp2 = &k2->rmap;
258	int64_t				d;
259	__u64				x, y;
260
261	d = (int64_t)be32_to_cpu(kp1->rm_startblock) -
262		       be32_to_cpu(kp2->rm_startblock);
263	if (d)
264		return d;
265
266	x = be64_to_cpu(kp1->rm_owner);
267	y = be64_to_cpu(kp2->rm_owner);
268	if (x > y)
269		return 1;
270	else if (y > x)
271		return -1;
272
273	x = XFS_RMAP_OFF(be64_to_cpu(kp1->rm_offset));
274	y = XFS_RMAP_OFF(be64_to_cpu(kp2->rm_offset));
275	if (x > y)
276		return 1;
277	else if (y > x)
278		return -1;
279	return 0;
280}
281
282static xfs_failaddr_t
283xfs_rmapbt_verify(
284	struct xfs_buf		*bp)
285{
286	struct xfs_mount	*mp = bp->b_mount;
287	struct xfs_btree_block	*block = XFS_BUF_TO_BLOCK(bp);
288	struct xfs_perag	*pag = bp->b_pag;
289	xfs_failaddr_t		fa;
290	unsigned int		level;
291
292	/*
293	 * magic number and level verification
294	 *
295	 * During growfs operations, we can't verify the exact level or owner as
296	 * the perag is not fully initialised and hence not attached to the
297	 * buffer.  In this case, check against the maximum tree depth.
298	 *
299	 * Similarly, during log recovery we will have a perag structure
300	 * attached, but the agf information will not yet have been initialised
301	 * from the on disk AGF. Again, we can only check against maximum limits
302	 * in this case.
303	 */
304	if (!xfs_verify_magic(bp, block->bb_magic))
305		return __this_address;
306
307	if (!xfs_has_rmapbt(mp))
308		return __this_address;
309	fa = xfs_btree_sblock_v5hdr_verify(bp);
310	if (fa)
311		return fa;
312
313	level = be16_to_cpu(block->bb_level);
314	if (pag && pag->pagf_init) {
315		if (level >= pag->pagf_levels[XFS_BTNUM_RMAPi])
316			return __this_address;
317	} else if (level >= mp->m_rmap_maxlevels)
318		return __this_address;
319
320	return xfs_btree_sblock_verify(bp, mp->m_rmap_mxr[level != 0]);
321}
322
323static void
324xfs_rmapbt_read_verify(
325	struct xfs_buf	*bp)
326{
327	xfs_failaddr_t	fa;
328
329	if (!xfs_btree_sblock_verify_crc(bp))
330		xfs_verifier_error(bp, -EFSBADCRC, __this_address);
331	else {
332		fa = xfs_rmapbt_verify(bp);
333		if (fa)
334			xfs_verifier_error(bp, -EFSCORRUPTED, fa);
335	}
336
337	if (bp->b_error)
338		trace_xfs_btree_corrupt(bp, _RET_IP_);
339}
340
341static void
342xfs_rmapbt_write_verify(
343	struct xfs_buf	*bp)
344{
345	xfs_failaddr_t	fa;
346
347	fa = xfs_rmapbt_verify(bp);
348	if (fa) {
349		trace_xfs_btree_corrupt(bp, _RET_IP_);
350		xfs_verifier_error(bp, -EFSCORRUPTED, fa);
351		return;
352	}
353	xfs_btree_sblock_calc_crc(bp);
354
355}
356
357const struct xfs_buf_ops xfs_rmapbt_buf_ops = {
358	.name			= "xfs_rmapbt",
359	.magic			= { 0, cpu_to_be32(XFS_RMAP_CRC_MAGIC) },
360	.verify_read		= xfs_rmapbt_read_verify,
361	.verify_write		= xfs_rmapbt_write_verify,
362	.verify_struct		= xfs_rmapbt_verify,
363};
364
365STATIC int
366xfs_rmapbt_keys_inorder(
367	struct xfs_btree_cur		*cur,
368	const union xfs_btree_key	*k1,
369	const union xfs_btree_key	*k2)
370{
371	uint32_t		x;
372	uint32_t		y;
373	uint64_t		a;
374	uint64_t		b;
375
376	x = be32_to_cpu(k1->rmap.rm_startblock);
377	y = be32_to_cpu(k2->rmap.rm_startblock);
378	if (x < y)
379		return 1;
380	else if (x > y)
381		return 0;
382	a = be64_to_cpu(k1->rmap.rm_owner);
383	b = be64_to_cpu(k2->rmap.rm_owner);
384	if (a < b)
385		return 1;
386	else if (a > b)
387		return 0;
388	a = XFS_RMAP_OFF(be64_to_cpu(k1->rmap.rm_offset));
389	b = XFS_RMAP_OFF(be64_to_cpu(k2->rmap.rm_offset));
390	if (a <= b)
391		return 1;
392	return 0;
393}
394
395STATIC int
396xfs_rmapbt_recs_inorder(
397	struct xfs_btree_cur		*cur,
398	const union xfs_btree_rec	*r1,
399	const union xfs_btree_rec	*r2)
400{
401	uint32_t		x;
402	uint32_t		y;
403	uint64_t		a;
404	uint64_t		b;
405
406	x = be32_to_cpu(r1->rmap.rm_startblock);
407	y = be32_to_cpu(r2->rmap.rm_startblock);
408	if (x < y)
409		return 1;
410	else if (x > y)
411		return 0;
412	a = be64_to_cpu(r1->rmap.rm_owner);
413	b = be64_to_cpu(r2->rmap.rm_owner);
414	if (a < b)
415		return 1;
416	else if (a > b)
417		return 0;
418	a = XFS_RMAP_OFF(be64_to_cpu(r1->rmap.rm_offset));
419	b = XFS_RMAP_OFF(be64_to_cpu(r2->rmap.rm_offset));
420	if (a <= b)
421		return 1;
422	return 0;
423}
424
425static const struct xfs_btree_ops xfs_rmapbt_ops = {
426	.rec_len		= sizeof(struct xfs_rmap_rec),
427	.key_len		= 2 * sizeof(struct xfs_rmap_key),
428
429	.dup_cursor		= xfs_rmapbt_dup_cursor,
430	.set_root		= xfs_rmapbt_set_root,
431	.alloc_block		= xfs_rmapbt_alloc_block,
432	.free_block		= xfs_rmapbt_free_block,
433	.get_minrecs		= xfs_rmapbt_get_minrecs,
434	.get_maxrecs		= xfs_rmapbt_get_maxrecs,
435	.init_key_from_rec	= xfs_rmapbt_init_key_from_rec,
436	.init_high_key_from_rec	= xfs_rmapbt_init_high_key_from_rec,
437	.init_rec_from_cur	= xfs_rmapbt_init_rec_from_cur,
438	.init_ptr_from_cur	= xfs_rmapbt_init_ptr_from_cur,
439	.key_diff		= xfs_rmapbt_key_diff,
440	.buf_ops		= &xfs_rmapbt_buf_ops,
441	.diff_two_keys		= xfs_rmapbt_diff_two_keys,
442	.keys_inorder		= xfs_rmapbt_keys_inorder,
443	.recs_inorder		= xfs_rmapbt_recs_inorder,
444};
445
446static struct xfs_btree_cur *
447xfs_rmapbt_init_common(
448	struct xfs_mount	*mp,
449	struct xfs_trans	*tp,
450	struct xfs_perag	*pag)
451{
452	struct xfs_btree_cur	*cur;
453
454	cur = kmem_cache_zalloc(xfs_btree_cur_zone, GFP_NOFS | __GFP_NOFAIL);
455	cur->bc_tp = tp;
456	cur->bc_mp = mp;
457	/* Overlapping btree; 2 keys per pointer. */
458	cur->bc_btnum = XFS_BTNUM_RMAP;
459	cur->bc_flags = XFS_BTREE_CRC_BLOCKS | XFS_BTREE_OVERLAPPING;
460	cur->bc_blocklog = mp->m_sb.sb_blocklog;
461	cur->bc_statoff = XFS_STATS_CALC_INDEX(xs_rmap_2);
462	cur->bc_ops = &xfs_rmapbt_ops;
463
464	/* take a reference for the cursor */
465	atomic_inc(&pag->pag_ref);
466	cur->bc_ag.pag = pag;
467
468	return cur;
469}
470
471/* Create a new reverse mapping btree cursor. */
472struct xfs_btree_cur *
473xfs_rmapbt_init_cursor(
474	struct xfs_mount	*mp,
475	struct xfs_trans	*tp,
476	struct xfs_buf		*agbp,
477	struct xfs_perag	*pag)
478{
479	struct xfs_agf		*agf = agbp->b_addr;
480	struct xfs_btree_cur	*cur;
481
482	cur = xfs_rmapbt_init_common(mp, tp, pag);
483	cur->bc_nlevels = be32_to_cpu(agf->agf_levels[XFS_BTNUM_RMAP]);
484	cur->bc_ag.agbp = agbp;
485	return cur;
486}
487
488/* Create a new reverse mapping btree cursor with a fake root for staging. */
489struct xfs_btree_cur *
490xfs_rmapbt_stage_cursor(
491	struct xfs_mount	*mp,
492	struct xbtree_afakeroot	*afake,
493	struct xfs_perag	*pag)
494{
495	struct xfs_btree_cur	*cur;
496
497	cur = xfs_rmapbt_init_common(mp, NULL, pag);
498	xfs_btree_stage_afakeroot(cur, afake);
499	return cur;
500}
501
502/*
503 * Install a new reverse mapping btree root.  Caller is responsible for
504 * invalidating and freeing the old btree blocks.
505 */
506void
507xfs_rmapbt_commit_staged_btree(
508	struct xfs_btree_cur	*cur,
509	struct xfs_trans	*tp,
510	struct xfs_buf		*agbp)
511{
512	struct xfs_agf		*agf = agbp->b_addr;
513	struct xbtree_afakeroot	*afake = cur->bc_ag.afake;
514
515	ASSERT(cur->bc_flags & XFS_BTREE_STAGING);
516
517	agf->agf_roots[cur->bc_btnum] = cpu_to_be32(afake->af_root);
518	agf->agf_levels[cur->bc_btnum] = cpu_to_be32(afake->af_levels);
519	agf->agf_rmap_blocks = cpu_to_be32(afake->af_blocks);
520	xfs_alloc_log_agf(tp, agbp, XFS_AGF_ROOTS | XFS_AGF_LEVELS |
521				    XFS_AGF_RMAP_BLOCKS);
522	xfs_btree_commit_afakeroot(cur, tp, agbp, &xfs_rmapbt_ops);
523}
524
525/*
526 * Calculate number of records in an rmap btree block.
527 */
528int
529xfs_rmapbt_maxrecs(
530	int			blocklen,
531	int			leaf)
532{
533	blocklen -= XFS_RMAP_BLOCK_LEN;
534
535	if (leaf)
536		return blocklen / sizeof(struct xfs_rmap_rec);
537	return blocklen /
538		(2 * sizeof(struct xfs_rmap_key) + sizeof(xfs_rmap_ptr_t));
539}
540
541/* Compute the maximum height of an rmap btree. */
542void
543xfs_rmapbt_compute_maxlevels(
544	struct xfs_mount		*mp)
545{
546	/*
547	 * On a non-reflink filesystem, the maximum number of rmap
548	 * records is the number of blocks in the AG, hence the max
549	 * rmapbt height is log_$maxrecs($agblocks).  However, with
550	 * reflink each AG block can have up to 2^32 (per the refcount
551	 * record format) owners, which means that theoretically we
552	 * could face up to 2^64 rmap records.
553	 *
554	 * That effectively means that the max rmapbt height must be
555	 * XFS_BTREE_MAXLEVELS.  "Fortunately" we'll run out of AG
556	 * blocks to feed the rmapbt long before the rmapbt reaches
557	 * maximum height.  The reflink code uses ag_resv_critical to
558	 * disallow reflinking when less than 10% of the per-AG metadata
559	 * block reservation since the fallback is a regular file copy.
560	 */
561	if (xfs_has_reflink(mp))
562		mp->m_rmap_maxlevels = XFS_BTREE_MAXLEVELS;
563	else
564		mp->m_rmap_maxlevels = xfs_btree_compute_maxlevels(
565				mp->m_rmap_mnr, mp->m_sb.sb_agblocks);
566}
567
568/* Calculate the refcount btree size for some records. */
569xfs_extlen_t
570xfs_rmapbt_calc_size(
571	struct xfs_mount	*mp,
572	unsigned long long	len)
573{
574	return xfs_btree_calc_size(mp->m_rmap_mnr, len);
575}
576
577/*
578 * Calculate the maximum refcount btree size.
579 */
580xfs_extlen_t
581xfs_rmapbt_max_size(
582	struct xfs_mount	*mp,
583	xfs_agblock_t		agblocks)
584{
585	/* Bail out if we're uninitialized, which can happen in mkfs. */
586	if (mp->m_rmap_mxr[0] == 0)
587		return 0;
588
589	return xfs_rmapbt_calc_size(mp, agblocks);
590}
591
592/*
593 * Figure out how many blocks to reserve and how many are used by this btree.
594 */
595int
596xfs_rmapbt_calc_reserves(
597	struct xfs_mount	*mp,
598	struct xfs_trans	*tp,
599	struct xfs_perag	*pag,
600	xfs_extlen_t		*ask,
601	xfs_extlen_t		*used)
602{
603	struct xfs_buf		*agbp;
604	struct xfs_agf		*agf;
605	xfs_agblock_t		agblocks;
606	xfs_extlen_t		tree_len;
607	int			error;
608
609	if (!xfs_has_rmapbt(mp))
610		return 0;
611
612	error = xfs_alloc_read_agf(mp, tp, pag->pag_agno, 0, &agbp);
613	if (error)
614		return error;
615
616	agf = agbp->b_addr;
617	agblocks = be32_to_cpu(agf->agf_length);
618	tree_len = be32_to_cpu(agf->agf_rmap_blocks);
619	xfs_trans_brelse(tp, agbp);
620
621	/*
622	 * The log is permanently allocated, so the space it occupies will
623	 * never be available for the kinds of things that would require btree
624	 * expansion.  We therefore can pretend the space isn't there.
625	 */
626	if (mp->m_sb.sb_logstart &&
627	    XFS_FSB_TO_AGNO(mp, mp->m_sb.sb_logstart) == pag->pag_agno)
628		agblocks -= mp->m_sb.sb_logblocks;
629
630	/* Reserve 1% of the AG or enough for 1 block per record. */
631	*ask += max(agblocks / 100, xfs_rmapbt_max_size(mp, agblocks));
632	*used += tree_len;
633
634	return error;
635}
636