1 // SPDX-License-Identifier: GPL-2.0
3 * Copyright (c) 2000-2002,2005 Silicon Graphics, Inc.
8 #include "xfs_format.h"
9 #include "xfs_log_format.h"
10 #include "xfs_shared.h"
11 #include "xfs_trans_resv.h"
14 #include "xfs_mount.h"
15 #include "xfs_defer.h"
16 #include "xfs_inode.h"
17 #include "xfs_btree.h"
19 #include "xfs_alloc_btree.h"
20 #include "xfs_alloc.h"
21 #include "xfs_extent_busy.h"
22 #include "xfs_errortag.h"
23 #include "xfs_error.h"
24 #include "xfs_cksum.h"
25 #include "xfs_trace.h"
26 #include "xfs_trans.h"
27 #include "xfs_buf_item.h"
29 #include "xfs_ag_resv.h"
32 extern kmem_zone_t *xfs_bmap_free_item_zone;
34 struct workqueue_struct *xfs_alloc_wq;
36 #define XFS_ABSDIFF(a,b) (((a) <= (b)) ? ((b) - (a)) : ((a) - (b)))
38 #define XFSA_FIXUP_BNO_OK 1
39 #define XFSA_FIXUP_CNT_OK 2
41 STATIC int xfs_alloc_ag_vextent_exact(xfs_alloc_arg_t *);
42 STATIC int xfs_alloc_ag_vextent_near(xfs_alloc_arg_t *);
43 STATIC int xfs_alloc_ag_vextent_size(xfs_alloc_arg_t *);
44 STATIC int xfs_alloc_ag_vextent_small(xfs_alloc_arg_t *,
45 xfs_btree_cur_t *, xfs_agblock_t *, xfs_extlen_t *, int *);
48 * Size of the AGFL. For CRC-enabled filesystes we steal a couple of slots in
49 * the beginning of the block for a proper header with the location information
56 unsigned int size = mp->m_sb.sb_sectsize;
58 if (xfs_sb_version_hascrc(&mp->m_sb))
59 size -= sizeof(struct xfs_agfl);
61 return size / sizeof(xfs_agblock_t);
68 if (xfs_sb_version_hasrmapbt(&mp->m_sb))
69 return XFS_RMAP_BLOCK(mp) + 1;
70 if (xfs_sb_version_hasfinobt(&mp->m_sb))
71 return XFS_FIBT_BLOCK(mp) + 1;
72 return XFS_IBT_BLOCK(mp) + 1;
79 if (xfs_sb_version_hasreflink(&mp->m_sb))
80 return xfs_refc_block(mp) + 1;
81 if (xfs_sb_version_hasrmapbt(&mp->m_sb))
82 return XFS_RMAP_BLOCK(mp) + 1;
83 if (xfs_sb_version_hasfinobt(&mp->m_sb))
84 return XFS_FIBT_BLOCK(mp) + 1;
85 return XFS_IBT_BLOCK(mp) + 1;
89 * In order to avoid ENOSPC-related deadlock caused by out-of-order locking of
90 * AGF buffer (PV 947395), we place constraints on the relationship among
91 * actual allocations for data blocks, freelist blocks, and potential file data
92 * bmap btree blocks. However, these restrictions may result in no actual space
93 * allocated for a delayed extent, for example, a data block in a certain AG is
94 * allocated but there is no additional block for the additional bmap btree
95 * block due to a split of the bmap btree of the file. The result of this may
96 * lead to an infinite loop when the file gets flushed to disk and all delayed
97 * extents need to be actually allocated. To get around this, we explicitly set
98 * aside a few blocks which will not be reserved in delayed allocation.
100 * We need to reserve 4 fsbs _per AG_ for the freelist and 4 more to handle a
101 * potential split of the file's bmap btree.
105 struct xfs_mount *mp)
107 return mp->m_sb.sb_agcount * (XFS_ALLOC_AGFL_RESERVE + 4);
111 * When deciding how much space to allocate out of an AG, we limit the
112 * allocation maximum size to the size the AG. However, we cannot use all the
113 * blocks in the AG - some are permanently used by metadata. These
114 * blocks are generally:
115 * - the AG superblock, AGF, AGI and AGFL
116 * - the AGF (bno and cnt) and AGI btree root blocks, and optionally
117 * the AGI free inode and rmap btree root blocks.
118 * - blocks on the AGFL according to xfs_alloc_set_aside() limits
119 * - the rmapbt root block
121 * The AG headers are sector sized, so the amount of space they take up is
122 * dependent on filesystem geometry. The others are all single blocks.
125 xfs_alloc_ag_max_usable(
126 struct xfs_mount *mp)
130 blocks = XFS_BB_TO_FSB(mp, XFS_FSS_TO_BB(mp, 4)); /* ag headers */
131 blocks += XFS_ALLOC_AGFL_RESERVE;
132 blocks += 3; /* AGF, AGI btree root blocks */
133 if (xfs_sb_version_hasfinobt(&mp->m_sb))
134 blocks++; /* finobt root block */
135 if (xfs_sb_version_hasrmapbt(&mp->m_sb))
136 blocks++; /* rmap root block */
137 if (xfs_sb_version_hasreflink(&mp->m_sb))
138 blocks++; /* refcount root block */
140 return mp->m_sb.sb_agblocks - blocks;
144 * Lookup the record equal to [bno, len] in the btree given by cur.
146 STATIC int /* error */
148 struct xfs_btree_cur *cur, /* btree cursor */
149 xfs_agblock_t bno, /* starting block of extent */
150 xfs_extlen_t len, /* length of extent */
151 int *stat) /* success/failure */
153 cur->bc_rec.a.ar_startblock = bno;
154 cur->bc_rec.a.ar_blockcount = len;
155 return xfs_btree_lookup(cur, XFS_LOOKUP_EQ, stat);
159 * Lookup the first record greater than or equal to [bno, len]
160 * in the btree given by cur.
164 struct xfs_btree_cur *cur, /* btree cursor */
165 xfs_agblock_t bno, /* starting block of extent */
166 xfs_extlen_t len, /* length of extent */
167 int *stat) /* success/failure */
169 cur->bc_rec.a.ar_startblock = bno;
170 cur->bc_rec.a.ar_blockcount = len;
171 return xfs_btree_lookup(cur, XFS_LOOKUP_GE, stat);
175 * Lookup the first record less than or equal to [bno, len]
176 * in the btree given by cur.
180 struct xfs_btree_cur *cur, /* btree cursor */
181 xfs_agblock_t bno, /* starting block of extent */
182 xfs_extlen_t len, /* length of extent */
183 int *stat) /* success/failure */
185 cur->bc_rec.a.ar_startblock = bno;
186 cur->bc_rec.a.ar_blockcount = len;
187 return xfs_btree_lookup(cur, XFS_LOOKUP_LE, stat);
191 * Update the record referred to by cur to the value given
193 * This either works (return 0) or gets an EFSCORRUPTED error.
195 STATIC int /* error */
197 struct xfs_btree_cur *cur, /* btree cursor */
198 xfs_agblock_t bno, /* starting block of extent */
199 xfs_extlen_t len) /* length of extent */
201 union xfs_btree_rec rec;
203 rec.alloc.ar_startblock = cpu_to_be32(bno);
204 rec.alloc.ar_blockcount = cpu_to_be32(len);
205 return xfs_btree_update(cur, &rec);
209 * Get the data from the pointed-to record.
213 struct xfs_btree_cur *cur, /* btree cursor */
214 xfs_agblock_t *bno, /* output: starting block of extent */
215 xfs_extlen_t *len, /* output: length of extent */
216 int *stat) /* output: success/failure */
218 struct xfs_mount *mp = cur->bc_mp;
219 xfs_agnumber_t agno = cur->bc_private.a.agno;
220 union xfs_btree_rec *rec;
223 error = xfs_btree_get_rec(cur, &rec, stat);
224 if (error || !(*stat))
226 if (rec->alloc.ar_blockcount == 0)
229 *bno = be32_to_cpu(rec->alloc.ar_startblock);
230 *len = be32_to_cpu(rec->alloc.ar_blockcount);
232 /* check for valid extent range, including overflow */
233 if (!xfs_verify_agbno(mp, agno, *bno))
235 if (*bno > *bno + *len)
237 if (!xfs_verify_agbno(mp, agno, *bno + *len - 1))
244 "%s Freespace BTree record corruption in AG %d detected!",
245 cur->bc_btnum == XFS_BTNUM_BNO ? "Block" : "Size", agno);
247 "start block 0x%x block count 0x%x", *bno, *len);
248 return -EFSCORRUPTED;
252 * Compute aligned version of the found extent.
253 * Takes alignment and min length into account.
256 xfs_alloc_compute_aligned(
257 xfs_alloc_arg_t *args, /* allocation argument structure */
258 xfs_agblock_t foundbno, /* starting block in found extent */
259 xfs_extlen_t foundlen, /* length in found extent */
260 xfs_agblock_t *resbno, /* result block number */
261 xfs_extlen_t *reslen, /* result length */
264 xfs_agblock_t bno = foundbno;
265 xfs_extlen_t len = foundlen;
269 /* Trim busy sections out of found extent */
270 busy = xfs_extent_busy_trim(args, &bno, &len, busy_gen);
273 * If we have a largish extent that happens to start before min_agbno,
274 * see if we can shift it into range...
276 if (bno < args->min_agbno && bno + len > args->min_agbno) {
277 diff = args->min_agbno - bno;
284 if (args->alignment > 1 && len >= args->minlen) {
285 xfs_agblock_t aligned_bno = roundup(bno, args->alignment);
287 diff = aligned_bno - bno;
289 *resbno = aligned_bno;
290 *reslen = diff >= len ? 0 : len - diff;
300 * Compute best start block and diff for "near" allocations.
301 * freelen >= wantlen already checked by caller.
303 STATIC xfs_extlen_t /* difference value (absolute) */
304 xfs_alloc_compute_diff(
305 xfs_agblock_t wantbno, /* target starting block */
306 xfs_extlen_t wantlen, /* target length */
307 xfs_extlen_t alignment, /* target alignment */
308 int datatype, /* are we allocating data? */
309 xfs_agblock_t freebno, /* freespace's starting block */
310 xfs_extlen_t freelen, /* freespace's length */
311 xfs_agblock_t *newbnop) /* result: best start block from free */
313 xfs_agblock_t freeend; /* end of freespace extent */
314 xfs_agblock_t newbno1; /* return block number */
315 xfs_agblock_t newbno2; /* other new block number */
316 xfs_extlen_t newlen1=0; /* length with newbno1 */
317 xfs_extlen_t newlen2=0; /* length with newbno2 */
318 xfs_agblock_t wantend; /* end of target extent */
319 bool userdata = xfs_alloc_is_userdata(datatype);
321 ASSERT(freelen >= wantlen);
322 freeend = freebno + freelen;
323 wantend = wantbno + wantlen;
325 * We want to allocate from the start of a free extent if it is past
326 * the desired block or if we are allocating user data and the free
327 * extent is before desired block. The second case is there to allow
328 * for contiguous allocation from the remaining free space if the file
329 * grows in the short term.
331 if (freebno >= wantbno || (userdata && freeend < wantend)) {
332 if ((newbno1 = roundup(freebno, alignment)) >= freeend)
333 newbno1 = NULLAGBLOCK;
334 } else if (freeend >= wantend && alignment > 1) {
335 newbno1 = roundup(wantbno, alignment);
336 newbno2 = newbno1 - alignment;
337 if (newbno1 >= freeend)
338 newbno1 = NULLAGBLOCK;
340 newlen1 = XFS_EXTLEN_MIN(wantlen, freeend - newbno1);
341 if (newbno2 < freebno)
342 newbno2 = NULLAGBLOCK;
344 newlen2 = XFS_EXTLEN_MIN(wantlen, freeend - newbno2);
345 if (newbno1 != NULLAGBLOCK && newbno2 != NULLAGBLOCK) {
346 if (newlen1 < newlen2 ||
347 (newlen1 == newlen2 &&
348 XFS_ABSDIFF(newbno1, wantbno) >
349 XFS_ABSDIFF(newbno2, wantbno)))
351 } else if (newbno2 != NULLAGBLOCK)
353 } else if (freeend >= wantend) {
355 } else if (alignment > 1) {
356 newbno1 = roundup(freeend - wantlen, alignment);
357 if (newbno1 > freeend - wantlen &&
358 newbno1 - alignment >= freebno)
359 newbno1 -= alignment;
360 else if (newbno1 >= freeend)
361 newbno1 = NULLAGBLOCK;
363 newbno1 = freeend - wantlen;
365 return newbno1 == NULLAGBLOCK ? 0 : XFS_ABSDIFF(newbno1, wantbno);
369 * Fix up the length, based on mod and prod.
370 * len should be k * prod + mod for some k.
371 * If len is too small it is returned unchanged.
372 * If len hits maxlen it is left alone.
376 xfs_alloc_arg_t *args) /* allocation argument structure */
381 ASSERT(args->mod < args->prod);
383 ASSERT(rlen >= args->minlen);
384 ASSERT(rlen <= args->maxlen);
385 if (args->prod <= 1 || rlen < args->mod || rlen == args->maxlen ||
386 (args->mod == 0 && rlen < args->prod))
388 k = rlen % args->prod;
392 rlen = rlen - (k - args->mod);
394 rlen = rlen - args->prod + (args->mod - k);
395 /* casts to (int) catch length underflows */
396 if ((int)rlen < (int)args->minlen)
398 ASSERT(rlen >= args->minlen && rlen <= args->maxlen);
399 ASSERT(rlen % args->prod == args->mod);
400 ASSERT(args->pag->pagf_freeblks + args->pag->pagf_flcount >=
401 rlen + args->minleft);
406 * Update the two btrees, logically removing from freespace the extent
407 * starting at rbno, rlen blocks. The extent is contained within the
408 * actual (current) free extent fbno for flen blocks.
409 * Flags are passed in indicating whether the cursors are set to the
412 STATIC int /* error code */
413 xfs_alloc_fixup_trees(
414 xfs_btree_cur_t *cnt_cur, /* cursor for by-size btree */
415 xfs_btree_cur_t *bno_cur, /* cursor for by-block btree */
416 xfs_agblock_t fbno, /* starting block of free extent */
417 xfs_extlen_t flen, /* length of free extent */
418 xfs_agblock_t rbno, /* starting block of returned extent */
419 xfs_extlen_t rlen, /* length of returned extent */
420 int flags) /* flags, XFSA_FIXUP_... */
422 int error; /* error code */
423 int i; /* operation results */
424 xfs_agblock_t nfbno1; /* first new free startblock */
425 xfs_agblock_t nfbno2; /* second new free startblock */
426 xfs_extlen_t nflen1=0; /* first new free length */
427 xfs_extlen_t nflen2=0; /* second new free length */
428 struct xfs_mount *mp;
433 * Look up the record in the by-size tree if necessary.
435 if (flags & XFSA_FIXUP_CNT_OK) {
437 if ((error = xfs_alloc_get_rec(cnt_cur, &nfbno1, &nflen1, &i)))
439 XFS_WANT_CORRUPTED_RETURN(mp,
440 i == 1 && nfbno1 == fbno && nflen1 == flen);
443 if ((error = xfs_alloc_lookup_eq(cnt_cur, fbno, flen, &i)))
445 XFS_WANT_CORRUPTED_RETURN(mp, i == 1);
448 * Look up the record in the by-block tree if necessary.
450 if (flags & XFSA_FIXUP_BNO_OK) {
452 if ((error = xfs_alloc_get_rec(bno_cur, &nfbno1, &nflen1, &i)))
454 XFS_WANT_CORRUPTED_RETURN(mp,
455 i == 1 && nfbno1 == fbno && nflen1 == flen);
458 if ((error = xfs_alloc_lookup_eq(bno_cur, fbno, flen, &i)))
460 XFS_WANT_CORRUPTED_RETURN(mp, i == 1);
464 if (bno_cur->bc_nlevels == 1 && cnt_cur->bc_nlevels == 1) {
465 struct xfs_btree_block *bnoblock;
466 struct xfs_btree_block *cntblock;
468 bnoblock = XFS_BUF_TO_BLOCK(bno_cur->bc_bufs[0]);
469 cntblock = XFS_BUF_TO_BLOCK(cnt_cur->bc_bufs[0]);
471 XFS_WANT_CORRUPTED_RETURN(mp,
472 bnoblock->bb_numrecs == cntblock->bb_numrecs);
477 * Deal with all four cases: the allocated record is contained
478 * within the freespace record, so we can have new freespace
479 * at either (or both) end, or no freespace remaining.
481 if (rbno == fbno && rlen == flen)
482 nfbno1 = nfbno2 = NULLAGBLOCK;
483 else if (rbno == fbno) {
484 nfbno1 = rbno + rlen;
485 nflen1 = flen - rlen;
486 nfbno2 = NULLAGBLOCK;
487 } else if (rbno + rlen == fbno + flen) {
489 nflen1 = flen - rlen;
490 nfbno2 = NULLAGBLOCK;
493 nflen1 = rbno - fbno;
494 nfbno2 = rbno + rlen;
495 nflen2 = (fbno + flen) - nfbno2;
498 * Delete the entry from the by-size btree.
500 if ((error = xfs_btree_delete(cnt_cur, &i)))
502 XFS_WANT_CORRUPTED_RETURN(mp, i == 1);
504 * Add new by-size btree entry(s).
506 if (nfbno1 != NULLAGBLOCK) {
507 if ((error = xfs_alloc_lookup_eq(cnt_cur, nfbno1, nflen1, &i)))
509 XFS_WANT_CORRUPTED_RETURN(mp, i == 0);
510 if ((error = xfs_btree_insert(cnt_cur, &i)))
512 XFS_WANT_CORRUPTED_RETURN(mp, i == 1);
514 if (nfbno2 != NULLAGBLOCK) {
515 if ((error = xfs_alloc_lookup_eq(cnt_cur, nfbno2, nflen2, &i)))
517 XFS_WANT_CORRUPTED_RETURN(mp, i == 0);
518 if ((error = xfs_btree_insert(cnt_cur, &i)))
520 XFS_WANT_CORRUPTED_RETURN(mp, i == 1);
523 * Fix up the by-block btree entry(s).
525 if (nfbno1 == NULLAGBLOCK) {
527 * No remaining freespace, just delete the by-block tree entry.
529 if ((error = xfs_btree_delete(bno_cur, &i)))
531 XFS_WANT_CORRUPTED_RETURN(mp, i == 1);
534 * Update the by-block entry to start later|be shorter.
536 if ((error = xfs_alloc_update(bno_cur, nfbno1, nflen1)))
539 if (nfbno2 != NULLAGBLOCK) {
541 * 2 resulting free entries, need to add one.
543 if ((error = xfs_alloc_lookup_eq(bno_cur, nfbno2, nflen2, &i)))
545 XFS_WANT_CORRUPTED_RETURN(mp, i == 0);
546 if ((error = xfs_btree_insert(bno_cur, &i)))
548 XFS_WANT_CORRUPTED_RETURN(mp, i == 1);
553 static xfs_failaddr_t
557 struct xfs_mount *mp = bp->b_target->bt_mount;
558 struct xfs_agfl *agfl = XFS_BUF_TO_AGFL(bp);
562 * There is no verification of non-crc AGFLs because mkfs does not
563 * initialise the AGFL to zero or NULL. Hence the only valid part of the
564 * AGFL is what the AGF says is active. We can't get to the AGF, so we
565 * can't verify just those entries are valid.
567 if (!xfs_sb_version_hascrc(&mp->m_sb))
570 if (!uuid_equal(&agfl->agfl_uuid, &mp->m_sb.sb_meta_uuid))
571 return __this_address;
572 if (be32_to_cpu(agfl->agfl_magicnum) != XFS_AGFL_MAGIC)
573 return __this_address;
575 * during growfs operations, the perag is not fully initialised,
576 * so we can't use it for any useful checking. growfs ensures we can't
577 * use it by using uncached buffers that don't have the perag attached
578 * so we can detect and avoid this problem.
580 if (bp->b_pag && be32_to_cpu(agfl->agfl_seqno) != bp->b_pag->pag_agno)
581 return __this_address;
583 for (i = 0; i < xfs_agfl_size(mp); i++) {
584 if (be32_to_cpu(agfl->agfl_bno[i]) != NULLAGBLOCK &&
585 be32_to_cpu(agfl->agfl_bno[i]) >= mp->m_sb.sb_agblocks)
586 return __this_address;
589 if (!xfs_log_check_lsn(mp, be64_to_cpu(XFS_BUF_TO_AGFL(bp)->agfl_lsn)))
590 return __this_address;
595 xfs_agfl_read_verify(
598 struct xfs_mount *mp = bp->b_target->bt_mount;
602 * There is no verification of non-crc AGFLs because mkfs does not
603 * initialise the AGFL to zero or NULL. Hence the only valid part of the
604 * AGFL is what the AGF says is active. We can't get to the AGF, so we
605 * can't verify just those entries are valid.
607 if (!xfs_sb_version_hascrc(&mp->m_sb))
610 if (!xfs_buf_verify_cksum(bp, XFS_AGFL_CRC_OFF))
611 xfs_verifier_error(bp, -EFSBADCRC, __this_address);
613 fa = xfs_agfl_verify(bp);
615 xfs_verifier_error(bp, -EFSCORRUPTED, fa);
620 xfs_agfl_write_verify(
623 struct xfs_mount *mp = bp->b_target->bt_mount;
624 struct xfs_buf_log_item *bip = bp->b_log_item;
627 /* no verification of non-crc AGFLs */
628 if (!xfs_sb_version_hascrc(&mp->m_sb))
631 fa = xfs_agfl_verify(bp);
633 xfs_verifier_error(bp, -EFSCORRUPTED, fa);
638 XFS_BUF_TO_AGFL(bp)->agfl_lsn = cpu_to_be64(bip->bli_item.li_lsn);
640 xfs_buf_update_cksum(bp, XFS_AGFL_CRC_OFF);
643 const struct xfs_buf_ops xfs_agfl_buf_ops = {
645 .verify_read = xfs_agfl_read_verify,
646 .verify_write = xfs_agfl_write_verify,
647 .verify_struct = xfs_agfl_verify,
651 * Read in the allocation group free block array.
655 xfs_mount_t *mp, /* mount point structure */
656 xfs_trans_t *tp, /* transaction pointer */
657 xfs_agnumber_t agno, /* allocation group number */
658 xfs_buf_t **bpp) /* buffer for the ag free block array */
660 xfs_buf_t *bp; /* return value */
663 ASSERT(agno != NULLAGNUMBER);
664 error = xfs_trans_read_buf(
665 mp, tp, mp->m_ddev_targp,
666 XFS_AG_DADDR(mp, agno, XFS_AGFL_DADDR(mp)),
667 XFS_FSS_TO_BB(mp, 1), 0, &bp, &xfs_agfl_buf_ops);
670 xfs_buf_set_ref(bp, XFS_AGFL_REF);
676 xfs_alloc_update_counters(
677 struct xfs_trans *tp,
678 struct xfs_perag *pag,
679 struct xfs_buf *agbp,
682 struct xfs_agf *agf = XFS_BUF_TO_AGF(agbp);
684 pag->pagf_freeblks += len;
685 be32_add_cpu(&agf->agf_freeblks, len);
687 xfs_trans_agblocks_delta(tp, len);
688 if (unlikely(be32_to_cpu(agf->agf_freeblks) >
689 be32_to_cpu(agf->agf_length)))
690 return -EFSCORRUPTED;
692 xfs_alloc_log_agf(tp, agbp, XFS_AGF_FREEBLKS);
697 * Allocation group level functions.
701 * Allocate a variable extent in the allocation group agno.
702 * Type and bno are used to determine where in the allocation group the
704 * Extent's length (returned in *len) will be between minlen and maxlen,
705 * and of the form k * prod + mod unless there's nothing that large.
706 * Return the starting a.g. block, or NULLAGBLOCK if we can't do it.
708 STATIC int /* error */
709 xfs_alloc_ag_vextent(
710 xfs_alloc_arg_t *args) /* argument structure for allocation */
714 ASSERT(args->minlen > 0);
715 ASSERT(args->maxlen > 0);
716 ASSERT(args->minlen <= args->maxlen);
717 ASSERT(args->mod < args->prod);
718 ASSERT(args->alignment > 0);
721 * Branch to correct routine based on the type.
724 switch (args->type) {
725 case XFS_ALLOCTYPE_THIS_AG:
726 error = xfs_alloc_ag_vextent_size(args);
728 case XFS_ALLOCTYPE_NEAR_BNO:
729 error = xfs_alloc_ag_vextent_near(args);
731 case XFS_ALLOCTYPE_THIS_BNO:
732 error = xfs_alloc_ag_vextent_exact(args);
739 if (error || args->agbno == NULLAGBLOCK)
742 ASSERT(args->len >= args->minlen);
743 ASSERT(args->len <= args->maxlen);
744 ASSERT(!args->wasfromfl || args->resv != XFS_AG_RESV_AGFL);
745 ASSERT(args->agbno % args->alignment == 0);
747 /* if not file data, insert new block into the reverse map btree */
748 if (!xfs_rmap_should_skip_owner_update(&args->oinfo)) {
749 error = xfs_rmap_alloc(args->tp, args->agbp, args->agno,
750 args->agbno, args->len, &args->oinfo);
755 if (!args->wasfromfl) {
756 error = xfs_alloc_update_counters(args->tp, args->pag,
758 -((long)(args->len)));
762 ASSERT(!xfs_extent_busy_search(args->mp, args->agno,
763 args->agbno, args->len));
766 xfs_ag_resv_alloc_extent(args->pag, args->resv, args);
768 XFS_STATS_INC(args->mp, xs_allocx);
769 XFS_STATS_ADD(args->mp, xs_allocb, args->len);
774 * Allocate a variable extent at exactly agno/bno.
775 * Extent's length (returned in *len) will be between minlen and maxlen,
776 * and of the form k * prod + mod unless there's nothing that large.
777 * Return the starting a.g. block (bno), or NULLAGBLOCK if we can't do it.
779 STATIC int /* error */
780 xfs_alloc_ag_vextent_exact(
781 xfs_alloc_arg_t *args) /* allocation argument structure */
783 xfs_btree_cur_t *bno_cur;/* by block-number btree cursor */
784 xfs_btree_cur_t *cnt_cur;/* by count btree cursor */
786 xfs_agblock_t fbno; /* start block of found extent */
787 xfs_extlen_t flen; /* length of found extent */
788 xfs_agblock_t tbno; /* start block of busy extent */
789 xfs_extlen_t tlen; /* length of busy extent */
790 xfs_agblock_t tend; /* end block of busy extent */
791 int i; /* success/failure of operation */
794 ASSERT(args->alignment == 1);
797 * Allocate/initialize a cursor for the by-number freespace btree.
799 bno_cur = xfs_allocbt_init_cursor(args->mp, args->tp, args->agbp,
800 args->agno, XFS_BTNUM_BNO);
803 * Lookup bno and minlen in the btree (minlen is irrelevant, really).
804 * Look for the closest free block <= bno, it must contain bno
805 * if any free block does.
807 error = xfs_alloc_lookup_le(bno_cur, args->agbno, args->minlen, &i);
814 * Grab the freespace record.
816 error = xfs_alloc_get_rec(bno_cur, &fbno, &flen, &i);
819 XFS_WANT_CORRUPTED_GOTO(args->mp, i == 1, error0);
820 ASSERT(fbno <= args->agbno);
823 * Check for overlapping busy extents.
827 xfs_extent_busy_trim(args, &tbno, &tlen, &busy_gen);
830 * Give up if the start of the extent is busy, or the freespace isn't
831 * long enough for the minimum request.
833 if (tbno > args->agbno)
835 if (tlen < args->minlen)
838 if (tend < args->agbno + args->minlen)
842 * End of extent will be smaller of the freespace end and the
843 * maximal requested end.
845 * Fix the length according to mod and prod if given.
847 args->len = XFS_AGBLOCK_MIN(tend, args->agbno + args->maxlen)
849 xfs_alloc_fix_len(args);
850 ASSERT(args->agbno + args->len <= tend);
853 * We are allocating agbno for args->len
854 * Allocate/initialize a cursor for the by-size btree.
856 cnt_cur = xfs_allocbt_init_cursor(args->mp, args->tp, args->agbp,
857 args->agno, XFS_BTNUM_CNT);
858 ASSERT(args->agbno + args->len <=
859 be32_to_cpu(XFS_BUF_TO_AGF(args->agbp)->agf_length));
860 error = xfs_alloc_fixup_trees(cnt_cur, bno_cur, fbno, flen, args->agbno,
861 args->len, XFSA_FIXUP_BNO_OK);
863 xfs_btree_del_cursor(cnt_cur, XFS_BTREE_ERROR);
867 xfs_btree_del_cursor(bno_cur, XFS_BTREE_NOERROR);
868 xfs_btree_del_cursor(cnt_cur, XFS_BTREE_NOERROR);
871 trace_xfs_alloc_exact_done(args);
875 /* Didn't find it, return null. */
876 xfs_btree_del_cursor(bno_cur, XFS_BTREE_NOERROR);
877 args->agbno = NULLAGBLOCK;
878 trace_xfs_alloc_exact_notfound(args);
882 xfs_btree_del_cursor(bno_cur, XFS_BTREE_ERROR);
883 trace_xfs_alloc_exact_error(args);
888 * Search the btree in a given direction via the search cursor and compare
889 * the records found against the good extent we've already found.
892 xfs_alloc_find_best_extent(
893 struct xfs_alloc_arg *args, /* allocation argument structure */
894 struct xfs_btree_cur **gcur, /* good cursor */
895 struct xfs_btree_cur **scur, /* searching cursor */
896 xfs_agblock_t gdiff, /* difference for search comparison */
897 xfs_agblock_t *sbno, /* extent found by search */
898 xfs_extlen_t *slen, /* extent length */
899 xfs_agblock_t *sbnoa, /* aligned extent found by search */
900 xfs_extlen_t *slena, /* aligned extent length */
901 int dir) /* 0 = search right, 1 = search left */
909 /* The good extent is perfect, no need to search. */
914 * Look until we find a better one, run out of space or run off the end.
917 error = xfs_alloc_get_rec(*scur, sbno, slen, &i);
920 XFS_WANT_CORRUPTED_GOTO(args->mp, i == 1, error0);
921 xfs_alloc_compute_aligned(args, *sbno, *slen,
922 sbnoa, slena, &busy_gen);
925 * The good extent is closer than this one.
928 if (*sbnoa > args->max_agbno)
930 if (*sbnoa >= args->agbno + gdiff)
933 if (*sbnoa < args->min_agbno)
935 if (*sbnoa <= args->agbno - gdiff)
940 * Same distance, compare length and pick the best.
942 if (*slena >= args->minlen) {
943 args->len = XFS_EXTLEN_MIN(*slena, args->maxlen);
944 xfs_alloc_fix_len(args);
946 sdiff = xfs_alloc_compute_diff(args->agbno, args->len,
948 args->datatype, *sbnoa,
952 * Choose closer size and invalidate other cursor.
960 error = xfs_btree_increment(*scur, 0, &i);
962 error = xfs_btree_decrement(*scur, 0, &i);
968 xfs_btree_del_cursor(*scur, XFS_BTREE_NOERROR);
973 xfs_btree_del_cursor(*gcur, XFS_BTREE_NOERROR);
978 /* caller invalidates cursors */
983 * Allocate a variable extent near bno in the allocation group agno.
984 * Extent's length (returned in len) will be between minlen and maxlen,
985 * and of the form k * prod + mod unless there's nothing that large.
986 * Return the starting a.g. block, or NULLAGBLOCK if we can't do it.
988 STATIC int /* error */
989 xfs_alloc_ag_vextent_near(
990 xfs_alloc_arg_t *args) /* allocation argument structure */
992 xfs_btree_cur_t *bno_cur_gt; /* cursor for bno btree, right side */
993 xfs_btree_cur_t *bno_cur_lt; /* cursor for bno btree, left side */
994 xfs_btree_cur_t *cnt_cur; /* cursor for count btree */
995 xfs_agblock_t gtbno; /* start bno of right side entry */
996 xfs_agblock_t gtbnoa; /* aligned ... */
997 xfs_extlen_t gtdiff; /* difference to right side entry */
998 xfs_extlen_t gtlen; /* length of right side entry */
999 xfs_extlen_t gtlena; /* aligned ... */
1000 xfs_agblock_t gtnew; /* useful start bno of right side */
1001 int error; /* error code */
1002 int i; /* result code, temporary */
1003 int j; /* result code, temporary */
1004 xfs_agblock_t ltbno; /* start bno of left side entry */
1005 xfs_agblock_t ltbnoa; /* aligned ... */
1006 xfs_extlen_t ltdiff; /* difference to left side entry */
1007 xfs_extlen_t ltlen; /* length of left side entry */
1008 xfs_extlen_t ltlena; /* aligned ... */
1009 xfs_agblock_t ltnew; /* useful start bno of left side */
1010 xfs_extlen_t rlen; /* length of returned extent */
1015 * Randomly don't execute the first algorithm.
1017 int dofirst; /* set to do first algorithm */
1019 dofirst = prandom_u32() & 1;
1022 /* handle unitialized agbno range so caller doesn't have to */
1023 if (!args->min_agbno && !args->max_agbno)
1024 args->max_agbno = args->mp->m_sb.sb_agblocks - 1;
1025 ASSERT(args->min_agbno <= args->max_agbno);
1027 /* clamp agbno to the range if it's outside */
1028 if (args->agbno < args->min_agbno)
1029 args->agbno = args->min_agbno;
1030 if (args->agbno > args->max_agbno)
1031 args->agbno = args->max_agbno;
1042 * Get a cursor for the by-size btree.
1044 cnt_cur = xfs_allocbt_init_cursor(args->mp, args->tp, args->agbp,
1045 args->agno, XFS_BTNUM_CNT);
1048 * See if there are any free extents as big as maxlen.
1050 if ((error = xfs_alloc_lookup_ge(cnt_cur, 0, args->maxlen, &i)))
1053 * If none, then pick up the last entry in the tree unless the
1057 if ((error = xfs_alloc_ag_vextent_small(args, cnt_cur, <bno,
1060 if (i == 0 || ltlen == 0) {
1061 xfs_btree_del_cursor(cnt_cur, XFS_BTREE_NOERROR);
1062 trace_xfs_alloc_near_noentry(args);
1067 args->wasfromfl = 0;
1071 * If the requested extent is large wrt the freespaces available
1072 * in this a.g., then the cursor will be pointing to a btree entry
1073 * near the right edge of the tree. If it's in the last btree leaf
1074 * block, then we just examine all the entries in that block
1075 * that are big enough, and pick the best one.
1076 * This is written as a while loop so we can break out of it,
1077 * but we never loop back to the top.
1079 while (xfs_btree_islastblock(cnt_cur, 0)) {
1082 xfs_extlen_t blen=0;
1083 xfs_agblock_t bnew=0;
1090 * Start from the entry that lookup found, sequence through
1091 * all larger free blocks. If we're actually pointing at a
1092 * record smaller than maxlen, go to the start of this block,
1093 * and skip all those smaller than minlen.
1095 if (ltlen || args->alignment > 1) {
1096 cnt_cur->bc_ptrs[0] = 1;
1098 if ((error = xfs_alloc_get_rec(cnt_cur, <bno,
1101 XFS_WANT_CORRUPTED_GOTO(args->mp, i == 1, error0);
1102 if (ltlen >= args->minlen)
1104 if ((error = xfs_btree_increment(cnt_cur, 0, &i)))
1107 ASSERT(ltlen >= args->minlen);
1111 i = cnt_cur->bc_ptrs[0];
1112 for (j = 1, blen = 0, bdiff = 0;
1113 !error && j && (blen < args->maxlen || bdiff > 0);
1114 error = xfs_btree_increment(cnt_cur, 0, &j)) {
1116 * For each entry, decide if it's better than
1117 * the previous best entry.
1119 if ((error = xfs_alloc_get_rec(cnt_cur, <bno, <len, &i)))
1121 XFS_WANT_CORRUPTED_GOTO(args->mp, i == 1, error0);
1122 busy = xfs_alloc_compute_aligned(args, ltbno, ltlen,
1123 <bnoa, <lena, &busy_gen);
1124 if (ltlena < args->minlen)
1126 if (ltbnoa < args->min_agbno || ltbnoa > args->max_agbno)
1128 args->len = XFS_EXTLEN_MIN(ltlena, args->maxlen);
1129 xfs_alloc_fix_len(args);
1130 ASSERT(args->len >= args->minlen);
1131 if (args->len < blen)
1133 ltdiff = xfs_alloc_compute_diff(args->agbno, args->len,
1134 args->alignment, args->datatype, ltbnoa,
1136 if (ltnew != NULLAGBLOCK &&
1137 (args->len > blen || ltdiff < bdiff)) {
1141 besti = cnt_cur->bc_ptrs[0];
1145 * It didn't work. We COULD be in a case where
1146 * there's a good record somewhere, so try again.
1151 * Point at the best entry, and retrieve it again.
1153 cnt_cur->bc_ptrs[0] = besti;
1154 if ((error = xfs_alloc_get_rec(cnt_cur, <bno, <len, &i)))
1156 XFS_WANT_CORRUPTED_GOTO(args->mp, i == 1, error0);
1157 ASSERT(ltbno + ltlen <= be32_to_cpu(XFS_BUF_TO_AGF(args->agbp)->agf_length));
1161 * We are allocating starting at bnew for blen blocks.
1164 ASSERT(bnew >= ltbno);
1165 ASSERT(bnew + blen <= ltbno + ltlen);
1167 * Set up a cursor for the by-bno tree.
1169 bno_cur_lt = xfs_allocbt_init_cursor(args->mp, args->tp,
1170 args->agbp, args->agno, XFS_BTNUM_BNO);
1172 * Fix up the btree entries.
1174 if ((error = xfs_alloc_fixup_trees(cnt_cur, bno_cur_lt, ltbno,
1175 ltlen, bnew, blen, XFSA_FIXUP_CNT_OK)))
1177 xfs_btree_del_cursor(cnt_cur, XFS_BTREE_NOERROR);
1178 xfs_btree_del_cursor(bno_cur_lt, XFS_BTREE_NOERROR);
1180 trace_xfs_alloc_near_first(args);
1185 * Search in the by-bno tree to the left and to the right
1186 * simultaneously, until in each case we find a space big enough,
1187 * or run into the edge of the tree. When we run into the edge,
1188 * we deallocate that cursor.
1189 * If both searches succeed, we compare the two spaces and pick
1191 * With alignment, it's possible for both to fail; the upper
1192 * level algorithm that picks allocation groups for allocations
1193 * is not supposed to do this.
1196 * Allocate and initialize the cursor for the leftward search.
1198 bno_cur_lt = xfs_allocbt_init_cursor(args->mp, args->tp, args->agbp,
1199 args->agno, XFS_BTNUM_BNO);
1201 * Lookup <= bno to find the leftward search's starting point.
1203 if ((error = xfs_alloc_lookup_le(bno_cur_lt, args->agbno, args->maxlen, &i)))
1207 * Didn't find anything; use this cursor for the rightward
1210 bno_cur_gt = bno_cur_lt;
1214 * Found something. Duplicate the cursor for the rightward search.
1216 else if ((error = xfs_btree_dup_cursor(bno_cur_lt, &bno_cur_gt)))
1219 * Increment the cursor, so we will point at the entry just right
1220 * of the leftward entry if any, or to the leftmost entry.
1222 if ((error = xfs_btree_increment(bno_cur_gt, 0, &i)))
1226 * It failed, there are no rightward entries.
1228 xfs_btree_del_cursor(bno_cur_gt, XFS_BTREE_NOERROR);
1232 * Loop going left with the leftward cursor, right with the
1233 * rightward cursor, until either both directions give up or
1234 * we find an entry at least as big as minlen.
1238 if ((error = xfs_alloc_get_rec(bno_cur_lt, <bno, <len, &i)))
1240 XFS_WANT_CORRUPTED_GOTO(args->mp, i == 1, error0);
1241 busy |= xfs_alloc_compute_aligned(args, ltbno, ltlen,
1242 <bnoa, <lena, &busy_gen);
1243 if (ltlena >= args->minlen && ltbnoa >= args->min_agbno)
1245 if ((error = xfs_btree_decrement(bno_cur_lt, 0, &i)))
1247 if (!i || ltbnoa < args->min_agbno) {
1248 xfs_btree_del_cursor(bno_cur_lt,
1254 if ((error = xfs_alloc_get_rec(bno_cur_gt, >bno, >len, &i)))
1256 XFS_WANT_CORRUPTED_GOTO(args->mp, i == 1, error0);
1257 busy |= xfs_alloc_compute_aligned(args, gtbno, gtlen,
1258 >bnoa, >lena, &busy_gen);
1259 if (gtlena >= args->minlen && gtbnoa <= args->max_agbno)
1261 if ((error = xfs_btree_increment(bno_cur_gt, 0, &i)))
1263 if (!i || gtbnoa > args->max_agbno) {
1264 xfs_btree_del_cursor(bno_cur_gt,
1269 } while (bno_cur_lt || bno_cur_gt);
1272 * Got both cursors still active, need to find better entry.
1274 if (bno_cur_lt && bno_cur_gt) {
1275 if (ltlena >= args->minlen) {
1277 * Left side is good, look for a right side entry.
1279 args->len = XFS_EXTLEN_MIN(ltlena, args->maxlen);
1280 xfs_alloc_fix_len(args);
1281 ltdiff = xfs_alloc_compute_diff(args->agbno, args->len,
1282 args->alignment, args->datatype, ltbnoa,
1285 error = xfs_alloc_find_best_extent(args,
1286 &bno_cur_lt, &bno_cur_gt,
1287 ltdiff, >bno, >len,
1289 0 /* search right */);
1291 ASSERT(gtlena >= args->minlen);
1294 * Right side is good, look for a left side entry.
1296 args->len = XFS_EXTLEN_MIN(gtlena, args->maxlen);
1297 xfs_alloc_fix_len(args);
1298 gtdiff = xfs_alloc_compute_diff(args->agbno, args->len,
1299 args->alignment, args->datatype, gtbnoa,
1302 error = xfs_alloc_find_best_extent(args,
1303 &bno_cur_gt, &bno_cur_lt,
1304 gtdiff, <bno, <len,
1306 1 /* search left */);
1314 * If we couldn't get anything, give up.
1316 if (bno_cur_lt == NULL && bno_cur_gt == NULL) {
1317 xfs_btree_del_cursor(cnt_cur, XFS_BTREE_NOERROR);
1320 trace_xfs_alloc_near_busy(args);
1321 xfs_extent_busy_flush(args->mp, args->pag, busy_gen);
1324 trace_xfs_alloc_size_neither(args);
1325 args->agbno = NULLAGBLOCK;
1330 * At this point we have selected a freespace entry, either to the
1331 * left or to the right. If it's on the right, copy all the
1332 * useful variables to the "left" set so we only have one
1333 * copy of this code.
1336 bno_cur_lt = bno_cur_gt;
1347 * Fix up the length and compute the useful address.
1349 args->len = XFS_EXTLEN_MIN(ltlena, args->maxlen);
1350 xfs_alloc_fix_len(args);
1352 (void)xfs_alloc_compute_diff(args->agbno, rlen, args->alignment,
1353 args->datatype, ltbnoa, ltlena, <new);
1354 ASSERT(ltnew >= ltbno);
1355 ASSERT(ltnew + rlen <= ltbnoa + ltlena);
1356 ASSERT(ltnew + rlen <= be32_to_cpu(XFS_BUF_TO_AGF(args->agbp)->agf_length));
1357 ASSERT(ltnew >= args->min_agbno && ltnew <= args->max_agbno);
1358 args->agbno = ltnew;
1360 if ((error = xfs_alloc_fixup_trees(cnt_cur, bno_cur_lt, ltbno, ltlen,
1361 ltnew, rlen, XFSA_FIXUP_BNO_OK)))
1365 trace_xfs_alloc_near_greater(args);
1367 trace_xfs_alloc_near_lesser(args);
1369 xfs_btree_del_cursor(cnt_cur, XFS_BTREE_NOERROR);
1370 xfs_btree_del_cursor(bno_cur_lt, XFS_BTREE_NOERROR);
1374 trace_xfs_alloc_near_error(args);
1375 if (cnt_cur != NULL)
1376 xfs_btree_del_cursor(cnt_cur, XFS_BTREE_ERROR);
1377 if (bno_cur_lt != NULL)
1378 xfs_btree_del_cursor(bno_cur_lt, XFS_BTREE_ERROR);
1379 if (bno_cur_gt != NULL)
1380 xfs_btree_del_cursor(bno_cur_gt, XFS_BTREE_ERROR);
1385 * Allocate a variable extent anywhere in the allocation group agno.
1386 * Extent's length (returned in len) will be between minlen and maxlen,
1387 * and of the form k * prod + mod unless there's nothing that large.
1388 * Return the starting a.g. block, or NULLAGBLOCK if we can't do it.
1390 STATIC int /* error */
1391 xfs_alloc_ag_vextent_size(
1392 xfs_alloc_arg_t *args) /* allocation argument structure */
1394 xfs_btree_cur_t *bno_cur; /* cursor for bno btree */
1395 xfs_btree_cur_t *cnt_cur; /* cursor for cnt btree */
1396 int error; /* error result */
1397 xfs_agblock_t fbno; /* start of found freespace */
1398 xfs_extlen_t flen; /* length of found freespace */
1399 int i; /* temp status variable */
1400 xfs_agblock_t rbno; /* returned block number */
1401 xfs_extlen_t rlen; /* length of returned extent */
1407 * Allocate and initialize a cursor for the by-size btree.
1409 cnt_cur = xfs_allocbt_init_cursor(args->mp, args->tp, args->agbp,
1410 args->agno, XFS_BTNUM_CNT);
1415 * Look for an entry >= maxlen+alignment-1 blocks.
1417 if ((error = xfs_alloc_lookup_ge(cnt_cur, 0,
1418 args->maxlen + args->alignment - 1, &i)))
1422 * If none then we have to settle for a smaller extent. In the case that
1423 * there are no large extents, this will return the last entry in the
1424 * tree unless the tree is empty. In the case that there are only busy
1425 * large extents, this will return the largest small extent unless there
1426 * are no smaller extents available.
1429 error = xfs_alloc_ag_vextent_small(args, cnt_cur,
1433 if (i == 0 || flen == 0) {
1434 xfs_btree_del_cursor(cnt_cur, XFS_BTREE_NOERROR);
1435 trace_xfs_alloc_size_noentry(args);
1439 busy = xfs_alloc_compute_aligned(args, fbno, flen, &rbno,
1443 * Search for a non-busy extent that is large enough.
1446 error = xfs_alloc_get_rec(cnt_cur, &fbno, &flen, &i);
1449 XFS_WANT_CORRUPTED_GOTO(args->mp, i == 1, error0);
1451 busy = xfs_alloc_compute_aligned(args, fbno, flen,
1452 &rbno, &rlen, &busy_gen);
1454 if (rlen >= args->maxlen)
1457 error = xfs_btree_increment(cnt_cur, 0, &i);
1462 * Our only valid extents must have been busy.
1463 * Make it unbusy by forcing the log out and
1466 xfs_btree_del_cursor(cnt_cur,
1468 trace_xfs_alloc_size_busy(args);
1469 xfs_extent_busy_flush(args->mp,
1470 args->pag, busy_gen);
1477 * In the first case above, we got the last entry in the
1478 * by-size btree. Now we check to see if the space hits maxlen
1479 * once aligned; if not, we search left for something better.
1480 * This can't happen in the second case above.
1482 rlen = XFS_EXTLEN_MIN(args->maxlen, rlen);
1483 XFS_WANT_CORRUPTED_GOTO(args->mp, rlen == 0 ||
1484 (rlen <= flen && rbno + rlen <= fbno + flen), error0);
1485 if (rlen < args->maxlen) {
1486 xfs_agblock_t bestfbno;
1487 xfs_extlen_t bestflen;
1488 xfs_agblock_t bestrbno;
1489 xfs_extlen_t bestrlen;
1496 if ((error = xfs_btree_decrement(cnt_cur, 0, &i)))
1500 if ((error = xfs_alloc_get_rec(cnt_cur, &fbno, &flen,
1503 XFS_WANT_CORRUPTED_GOTO(args->mp, i == 1, error0);
1504 if (flen < bestrlen)
1506 busy = xfs_alloc_compute_aligned(args, fbno, flen,
1507 &rbno, &rlen, &busy_gen);
1508 rlen = XFS_EXTLEN_MIN(args->maxlen, rlen);
1509 XFS_WANT_CORRUPTED_GOTO(args->mp, rlen == 0 ||
1510 (rlen <= flen && rbno + rlen <= fbno + flen),
1512 if (rlen > bestrlen) {
1517 if (rlen == args->maxlen)
1521 if ((error = xfs_alloc_lookup_eq(cnt_cur, bestfbno, bestflen,
1524 XFS_WANT_CORRUPTED_GOTO(args->mp, i == 1, error0);
1530 args->wasfromfl = 0;
1532 * Fix up the length.
1535 if (rlen < args->minlen) {
1537 xfs_btree_del_cursor(cnt_cur, XFS_BTREE_NOERROR);
1538 trace_xfs_alloc_size_busy(args);
1539 xfs_extent_busy_flush(args->mp, args->pag, busy_gen);
1544 xfs_alloc_fix_len(args);
1547 XFS_WANT_CORRUPTED_GOTO(args->mp, rlen <= flen, error0);
1549 * Allocate and initialize a cursor for the by-block tree.
1551 bno_cur = xfs_allocbt_init_cursor(args->mp, args->tp, args->agbp,
1552 args->agno, XFS_BTNUM_BNO);
1553 if ((error = xfs_alloc_fixup_trees(cnt_cur, bno_cur, fbno, flen,
1554 rbno, rlen, XFSA_FIXUP_CNT_OK)))
1556 xfs_btree_del_cursor(cnt_cur, XFS_BTREE_NOERROR);
1557 xfs_btree_del_cursor(bno_cur, XFS_BTREE_NOERROR);
1558 cnt_cur = bno_cur = NULL;
1561 XFS_WANT_CORRUPTED_GOTO(args->mp,
1562 args->agbno + args->len <=
1563 be32_to_cpu(XFS_BUF_TO_AGF(args->agbp)->agf_length),
1565 trace_xfs_alloc_size_done(args);
1569 trace_xfs_alloc_size_error(args);
1571 xfs_btree_del_cursor(cnt_cur, XFS_BTREE_ERROR);
1573 xfs_btree_del_cursor(bno_cur, XFS_BTREE_ERROR);
1577 xfs_btree_del_cursor(cnt_cur, XFS_BTREE_NOERROR);
1578 trace_xfs_alloc_size_nominleft(args);
1579 args->agbno = NULLAGBLOCK;
1584 * Deal with the case where only small freespaces remain.
1585 * Either return the contents of the last freespace record,
1586 * or allocate space from the freelist if there is nothing in the tree.
1588 STATIC int /* error */
1589 xfs_alloc_ag_vextent_small(
1590 xfs_alloc_arg_t *args, /* allocation argument structure */
1591 xfs_btree_cur_t *ccur, /* by-size cursor */
1592 xfs_agblock_t *fbnop, /* result block number */
1593 xfs_extlen_t *flenp, /* result length */
1594 int *stat) /* status: 0-freelist, 1-normal/none */
1596 struct xfs_owner_info oinfo;
1602 if ((error = xfs_btree_decrement(ccur, 0, &i)))
1605 if ((error = xfs_alloc_get_rec(ccur, &fbno, &flen, &i)))
1607 XFS_WANT_CORRUPTED_GOTO(args->mp, i == 1, error0);
1610 * Nothing in the btree, try the freelist. Make sure
1611 * to respect minleft even when pulling from the
1614 else if (args->minlen == 1 && args->alignment == 1 &&
1615 args->resv != XFS_AG_RESV_AGFL &&
1616 (be32_to_cpu(XFS_BUF_TO_AGF(args->agbp)->agf_flcount)
1618 error = xfs_alloc_get_freelist(args->tp, args->agbp, &fbno, 0);
1621 if (fbno != NULLAGBLOCK) {
1622 xfs_extent_busy_reuse(args->mp, args->agno, fbno, 1,
1623 xfs_alloc_allow_busy_reuse(args->datatype));
1625 if (xfs_alloc_is_userdata(args->datatype)) {
1628 bp = xfs_btree_get_bufs(args->mp, args->tp,
1629 args->agno, fbno, 0);
1631 error = -EFSCORRUPTED;
1634 xfs_trans_binval(args->tp, bp);
1638 XFS_WANT_CORRUPTED_GOTO(args->mp,
1639 args->agbno + args->len <=
1640 be32_to_cpu(XFS_BUF_TO_AGF(args->agbp)->agf_length),
1642 args->wasfromfl = 1;
1643 trace_xfs_alloc_small_freelist(args);
1646 * If we're feeding an AGFL block to something that
1647 * doesn't live in the free space, we need to clear
1648 * out the OWN_AG rmap.
1650 xfs_rmap_ag_owner(&oinfo, XFS_RMAP_OWN_AG);
1651 error = xfs_rmap_free(args->tp, args->agbp, args->agno,
1660 * Nothing in the freelist.
1666 * Can't allocate from the freelist for some reason.
1673 * Can't do the allocation, give up.
1675 if (flen < args->minlen) {
1676 args->agbno = NULLAGBLOCK;
1677 trace_xfs_alloc_small_notenough(args);
1683 trace_xfs_alloc_small_done(args);
1687 trace_xfs_alloc_small_error(args);
1692 * Free the extent starting at agno/bno for length.
1698 xfs_agnumber_t agno,
1701 struct xfs_owner_info *oinfo,
1702 enum xfs_ag_resv_type type)
1704 xfs_btree_cur_t *bno_cur; /* cursor for by-block btree */
1705 xfs_btree_cur_t *cnt_cur; /* cursor for by-size btree */
1706 int error; /* error return value */
1707 xfs_agblock_t gtbno; /* start of right neighbor block */
1708 xfs_extlen_t gtlen; /* length of right neighbor block */
1709 int haveleft; /* have a left neighbor block */
1710 int haveright; /* have a right neighbor block */
1711 int i; /* temp, result code */
1712 xfs_agblock_t ltbno; /* start of left neighbor block */
1713 xfs_extlen_t ltlen; /* length of left neighbor block */
1714 xfs_mount_t *mp; /* mount point struct for filesystem */
1715 xfs_agblock_t nbno; /* new starting block of freespace */
1716 xfs_extlen_t nlen; /* new length of freespace */
1717 xfs_perag_t *pag; /* per allocation group data */
1719 bno_cur = cnt_cur = NULL;
1722 if (!xfs_rmap_should_skip_owner_update(oinfo)) {
1723 error = xfs_rmap_free(tp, agbp, agno, bno, len, oinfo);
1729 * Allocate and initialize a cursor for the by-block btree.
1731 bno_cur = xfs_allocbt_init_cursor(mp, tp, agbp, agno, XFS_BTNUM_BNO);
1733 * Look for a neighboring block on the left (lower block numbers)
1734 * that is contiguous with this space.
1736 if ((error = xfs_alloc_lookup_le(bno_cur, bno, len, &haveleft)))
1740 * There is a block to our left.
1742 if ((error = xfs_alloc_get_rec(bno_cur, <bno, <len, &i)))
1744 XFS_WANT_CORRUPTED_GOTO(mp, i == 1, error0);
1746 * It's not contiguous, though.
1748 if (ltbno + ltlen < bno)
1752 * If this failure happens the request to free this
1753 * space was invalid, it's (partly) already free.
1756 XFS_WANT_CORRUPTED_GOTO(mp,
1757 ltbno + ltlen <= bno, error0);
1761 * Look for a neighboring block on the right (higher block numbers)
1762 * that is contiguous with this space.
1764 if ((error = xfs_btree_increment(bno_cur, 0, &haveright)))
1768 * There is a block to our right.
1770 if ((error = xfs_alloc_get_rec(bno_cur, >bno, >len, &i)))
1772 XFS_WANT_CORRUPTED_GOTO(mp, i == 1, error0);
1774 * It's not contiguous, though.
1776 if (bno + len < gtbno)
1780 * If this failure happens the request to free this
1781 * space was invalid, it's (partly) already free.
1784 XFS_WANT_CORRUPTED_GOTO(mp, gtbno >= bno + len, error0);
1788 * Now allocate and initialize a cursor for the by-size tree.
1790 cnt_cur = xfs_allocbt_init_cursor(mp, tp, agbp, agno, XFS_BTNUM_CNT);
1792 * Have both left and right contiguous neighbors.
1793 * Merge all three into a single free block.
1795 if (haveleft && haveright) {
1797 * Delete the old by-size entry on the left.
1799 if ((error = xfs_alloc_lookup_eq(cnt_cur, ltbno, ltlen, &i)))
1801 XFS_WANT_CORRUPTED_GOTO(mp, i == 1, error0);
1802 if ((error = xfs_btree_delete(cnt_cur, &i)))
1804 XFS_WANT_CORRUPTED_GOTO(mp, i == 1, error0);
1806 * Delete the old by-size entry on the right.
1808 if ((error = xfs_alloc_lookup_eq(cnt_cur, gtbno, gtlen, &i)))
1810 XFS_WANT_CORRUPTED_GOTO(mp, i == 1, error0);
1811 if ((error = xfs_btree_delete(cnt_cur, &i)))
1813 XFS_WANT_CORRUPTED_GOTO(mp, i == 1, error0);
1815 * Delete the old by-block entry for the right block.
1817 if ((error = xfs_btree_delete(bno_cur, &i)))
1819 XFS_WANT_CORRUPTED_GOTO(mp, i == 1, error0);
1821 * Move the by-block cursor back to the left neighbor.
1823 if ((error = xfs_btree_decrement(bno_cur, 0, &i)))
1825 XFS_WANT_CORRUPTED_GOTO(mp, i == 1, error0);
1828 * Check that this is the right record: delete didn't
1829 * mangle the cursor.
1832 xfs_agblock_t xxbno;
1835 if ((error = xfs_alloc_get_rec(bno_cur, &xxbno, &xxlen,
1838 XFS_WANT_CORRUPTED_GOTO(mp,
1839 i == 1 && xxbno == ltbno && xxlen == ltlen,
1844 * Update remaining by-block entry to the new, joined block.
1847 nlen = len + ltlen + gtlen;
1848 if ((error = xfs_alloc_update(bno_cur, nbno, nlen)))
1852 * Have only a left contiguous neighbor.
1853 * Merge it together with the new freespace.
1855 else if (haveleft) {
1857 * Delete the old by-size entry on the left.
1859 if ((error = xfs_alloc_lookup_eq(cnt_cur, ltbno, ltlen, &i)))
1861 XFS_WANT_CORRUPTED_GOTO(mp, i == 1, error0);
1862 if ((error = xfs_btree_delete(cnt_cur, &i)))
1864 XFS_WANT_CORRUPTED_GOTO(mp, i == 1, error0);
1866 * Back up the by-block cursor to the left neighbor, and
1867 * update its length.
1869 if ((error = xfs_btree_decrement(bno_cur, 0, &i)))
1871 XFS_WANT_CORRUPTED_GOTO(mp, i == 1, error0);
1874 if ((error = xfs_alloc_update(bno_cur, nbno, nlen)))
1878 * Have only a right contiguous neighbor.
1879 * Merge it together with the new freespace.
1881 else if (haveright) {
1883 * Delete the old by-size entry on the right.
1885 if ((error = xfs_alloc_lookup_eq(cnt_cur, gtbno, gtlen, &i)))
1887 XFS_WANT_CORRUPTED_GOTO(mp, i == 1, error0);
1888 if ((error = xfs_btree_delete(cnt_cur, &i)))
1890 XFS_WANT_CORRUPTED_GOTO(mp, i == 1, error0);
1892 * Update the starting block and length of the right
1893 * neighbor in the by-block tree.
1897 if ((error = xfs_alloc_update(bno_cur, nbno, nlen)))
1901 * No contiguous neighbors.
1902 * Insert the new freespace into the by-block tree.
1907 if ((error = xfs_btree_insert(bno_cur, &i)))
1909 XFS_WANT_CORRUPTED_GOTO(mp, i == 1, error0);
1911 xfs_btree_del_cursor(bno_cur, XFS_BTREE_NOERROR);
1914 * In all cases we need to insert the new freespace in the by-size tree.
1916 if ((error = xfs_alloc_lookup_eq(cnt_cur, nbno, nlen, &i)))
1918 XFS_WANT_CORRUPTED_GOTO(mp, i == 0, error0);
1919 if ((error = xfs_btree_insert(cnt_cur, &i)))
1921 XFS_WANT_CORRUPTED_GOTO(mp, i == 1, error0);
1922 xfs_btree_del_cursor(cnt_cur, XFS_BTREE_NOERROR);
1926 * Update the freespace totals in the ag and superblock.
1928 pag = xfs_perag_get(mp, agno);
1929 error = xfs_alloc_update_counters(tp, pag, agbp, len);
1930 xfs_ag_resv_free_extent(pag, type, tp, len);
1935 XFS_STATS_INC(mp, xs_freex);
1936 XFS_STATS_ADD(mp, xs_freeb, len);
1938 trace_xfs_free_extent(mp, agno, bno, len, type, haveleft, haveright);
1943 trace_xfs_free_extent(mp, agno, bno, len, type, -1, -1);
1945 xfs_btree_del_cursor(bno_cur, XFS_BTREE_ERROR);
1947 xfs_btree_del_cursor(cnt_cur, XFS_BTREE_ERROR);
1952 * Visible (exported) allocation/free functions.
1953 * Some of these are used just by xfs_alloc_btree.c and this file.
1957 * Compute and fill in value of m_ag_maxlevels.
1960 xfs_alloc_compute_maxlevels(
1961 xfs_mount_t *mp) /* file system mount structure */
1963 mp->m_ag_maxlevels = xfs_btree_compute_maxlevels(mp->m_alloc_mnr,
1964 (mp->m_sb.sb_agblocks + 1) / 2);
1968 * Find the length of the longest extent in an AG. The 'need' parameter
1969 * specifies how much space we're going to need for the AGFL and the
1970 * 'reserved' parameter tells us how many blocks in this AG are reserved for
1974 xfs_alloc_longest_free_extent(
1975 struct xfs_perag *pag,
1977 xfs_extlen_t reserved)
1979 xfs_extlen_t delta = 0;
1982 * If the AGFL needs a recharge, we'll have to subtract that from the
1985 if (need > pag->pagf_flcount)
1986 delta = need - pag->pagf_flcount;
1989 * If we cannot maintain others' reservations with space from the
1990 * not-longest freesp extents, we'll have to subtract /that/ from
1991 * the longest extent too.
1993 if (pag->pagf_freeblks - pag->pagf_longest < reserved)
1994 delta += reserved - (pag->pagf_freeblks - pag->pagf_longest);
1997 * If the longest extent is long enough to satisfy all the
1998 * reservations and AGFL rules in place, we can return this extent.
2000 if (pag->pagf_longest > delta)
2001 return pag->pagf_longest - delta;
2003 /* Otherwise, let the caller try for 1 block if there's space. */
2004 return pag->pagf_flcount > 0 || pag->pagf_longest > 0;
2008 xfs_alloc_min_freelist(
2009 struct xfs_mount *mp,
2010 struct xfs_perag *pag)
2012 unsigned int min_free;
2014 /* space needed by-bno freespace btree */
2015 min_free = min_t(unsigned int, pag->pagf_levels[XFS_BTNUM_BNOi] + 1,
2016 mp->m_ag_maxlevels);
2017 /* space needed by-size freespace btree */
2018 min_free += min_t(unsigned int, pag->pagf_levels[XFS_BTNUM_CNTi] + 1,
2019 mp->m_ag_maxlevels);
2020 /* space needed reverse mapping used space btree */
2021 if (xfs_sb_version_hasrmapbt(&mp->m_sb))
2022 min_free += min_t(unsigned int,
2023 pag->pagf_levels[XFS_BTNUM_RMAPi] + 1,
2024 mp->m_rmap_maxlevels);
2030 * Check if the operation we are fixing up the freelist for should go ahead or
2031 * not. If we are freeing blocks, we always allow it, otherwise the allocation
2032 * is dependent on whether the size and shape of free space available will
2033 * permit the requested allocation to take place.
2036 xfs_alloc_space_available(
2037 struct xfs_alloc_arg *args,
2038 xfs_extlen_t min_free,
2041 struct xfs_perag *pag = args->pag;
2042 xfs_extlen_t alloc_len, longest;
2043 xfs_extlen_t reservation; /* blocks that are still reserved */
2046 if (flags & XFS_ALLOC_FLAG_FREEING)
2049 reservation = xfs_ag_resv_needed(pag, args->resv);
2051 /* do we have enough contiguous free space for the allocation? */
2052 alloc_len = args->minlen + (args->alignment - 1) + args->minalignslop;
2053 longest = xfs_alloc_longest_free_extent(pag, min_free, reservation);
2054 if (longest < alloc_len)
2057 /* do we have enough free space remaining for the allocation? */
2058 available = (int)(pag->pagf_freeblks + pag->pagf_flcount -
2059 reservation - min_free - args->minleft);
2060 if (available < (int)max(args->total, alloc_len))
2064 * Clamp maxlen to the amount of free space available for the actual
2065 * extent allocation.
2067 if (available < (int)args->maxlen && !(flags & XFS_ALLOC_FLAG_CHECK)) {
2068 args->maxlen = available;
2069 ASSERT(args->maxlen > 0);
2070 ASSERT(args->maxlen >= args->minlen);
2077 xfs_free_agfl_block(
2078 struct xfs_trans *tp,
2079 xfs_agnumber_t agno,
2080 xfs_agblock_t agbno,
2081 struct xfs_buf *agbp,
2082 struct xfs_owner_info *oinfo)
2087 error = xfs_free_ag_extent(tp, agbp, agno, agbno, 1, oinfo,
2092 bp = xfs_btree_get_bufs(tp->t_mountp, tp, agno, agbno, 0);
2094 return -EFSCORRUPTED;
2095 xfs_trans_binval(tp, bp);
2101 * Check the agfl fields of the agf for inconsistency or corruption. The purpose
2102 * is to detect an agfl header padding mismatch between current and early v5
2103 * kernels. This problem manifests as a 1-slot size difference between the
2104 * on-disk flcount and the active [first, last] range of a wrapped agfl. This
2105 * may also catch variants of agfl count corruption unrelated to padding. Either
2106 * way, we'll reset the agfl and warn the user.
2108 * Return true if a reset is required before the agfl can be used, false
2112 xfs_agfl_needs_reset(
2113 struct xfs_mount *mp,
2114 struct xfs_agf *agf)
2116 uint32_t f = be32_to_cpu(agf->agf_flfirst);
2117 uint32_t l = be32_to_cpu(agf->agf_fllast);
2118 uint32_t c = be32_to_cpu(agf->agf_flcount);
2119 int agfl_size = xfs_agfl_size(mp);
2122 /* no agfl header on v4 supers */
2123 if (!xfs_sb_version_hascrc(&mp->m_sb))
2127 * The agf read verifier catches severe corruption of these fields.
2128 * Repeat some sanity checks to cover a packed -> unpacked mismatch if
2129 * the verifier allows it.
2131 if (f >= agfl_size || l >= agfl_size)
2137 * Check consistency between the on-disk count and the active range. An
2138 * agfl padding mismatch manifests as an inconsistent flcount.
2143 active = agfl_size - f + l + 1;
2151 * Reset the agfl to an empty state. Ignore/drop any existing blocks since the
2152 * agfl content cannot be trusted. Warn the user that a repair is required to
2153 * recover leaked blocks.
2155 * The purpose of this mechanism is to handle filesystems affected by the agfl
2156 * header padding mismatch problem. A reset keeps the filesystem online with a
2157 * relatively minor free space accounting inconsistency rather than suffer the
2158 * inevitable crash from use of an invalid agfl block.
2162 struct xfs_trans *tp,
2163 struct xfs_buf *agbp,
2164 struct xfs_perag *pag)
2166 struct xfs_mount *mp = tp->t_mountp;
2167 struct xfs_agf *agf = XFS_BUF_TO_AGF(agbp);
2169 ASSERT(pag->pagf_agflreset);
2170 trace_xfs_agfl_reset(mp, agf, 0, _RET_IP_);
2173 "WARNING: Reset corrupted AGFL on AG %u. %d blocks leaked. "
2174 "Please unmount and run xfs_repair.",
2175 pag->pag_agno, pag->pagf_flcount);
2177 agf->agf_flfirst = 0;
2178 agf->agf_fllast = cpu_to_be32(xfs_agfl_size(mp) - 1);
2179 agf->agf_flcount = 0;
2180 xfs_alloc_log_agf(tp, agbp, XFS_AGF_FLFIRST | XFS_AGF_FLLAST |
2183 pag->pagf_flcount = 0;
2184 pag->pagf_agflreset = false;
2188 * Defer an AGFL block free. This is effectively equivalent to
2189 * xfs_bmap_add_free() with some special handling particular to AGFL blocks.
2191 * Deferring AGFL frees helps prevent log reservation overruns due to too many
2192 * allocation operations in a transaction. AGFL frees are prone to this problem
2193 * because for one they are always freed one at a time. Further, an immediate
2194 * AGFL block free can cause a btree join and require another block free before
2195 * the real allocation can proceed. Deferring the free disconnects freeing up
2196 * the AGFL slot from freeing the block.
2199 xfs_defer_agfl_block(
2200 struct xfs_mount *mp,
2201 struct xfs_defer_ops *dfops,
2202 xfs_agnumber_t agno,
2203 xfs_fsblock_t agbno,
2204 struct xfs_owner_info *oinfo)
2206 struct xfs_extent_free_item *new; /* new element */
2208 ASSERT(xfs_bmap_free_item_zone != NULL);
2209 ASSERT(oinfo != NULL);
2211 new = kmem_zone_alloc(xfs_bmap_free_item_zone, KM_SLEEP);
2212 new->xefi_startblock = XFS_AGB_TO_FSB(mp, agno, agbno);
2213 new->xefi_blockcount = 1;
2214 new->xefi_oinfo = *oinfo;
2216 trace_xfs_agfl_free_defer(mp, agno, 0, agbno, 1);
2218 xfs_defer_add(dfops, XFS_DEFER_OPS_TYPE_AGFL_FREE, &new->xefi_list);
2222 * Decide whether to use this allocation group for this allocation.
2223 * If so, fix up the btree freelist's size.
2226 xfs_alloc_fix_freelist(
2227 struct xfs_alloc_arg *args, /* allocation argument structure */
2228 int flags) /* XFS_ALLOC_FLAG_... */
2230 struct xfs_mount *mp = args->mp;
2231 struct xfs_perag *pag = args->pag;
2232 struct xfs_trans *tp = args->tp;
2233 struct xfs_buf *agbp = NULL;
2234 struct xfs_buf *agflbp = NULL;
2235 struct xfs_alloc_arg targs; /* local allocation arguments */
2236 xfs_agblock_t bno; /* freelist block */
2237 xfs_extlen_t need; /* total blocks needed in freelist */
2240 if (!pag->pagf_init) {
2241 error = xfs_alloc_read_agf(mp, tp, args->agno, flags, &agbp);
2244 if (!pag->pagf_init) {
2245 ASSERT(flags & XFS_ALLOC_FLAG_TRYLOCK);
2246 ASSERT(!(flags & XFS_ALLOC_FLAG_FREEING));
2247 goto out_agbp_relse;
2252 * If this is a metadata preferred pag and we are user data then try
2253 * somewhere else if we are not being asked to try harder at this
2256 if (pag->pagf_metadata && xfs_alloc_is_userdata(args->datatype) &&
2257 (flags & XFS_ALLOC_FLAG_TRYLOCK)) {
2258 ASSERT(!(flags & XFS_ALLOC_FLAG_FREEING));
2259 goto out_agbp_relse;
2262 need = xfs_alloc_min_freelist(mp, pag);
2263 if (!xfs_alloc_space_available(args, need, flags |
2264 XFS_ALLOC_FLAG_CHECK))
2265 goto out_agbp_relse;
2268 * Get the a.g. freespace buffer.
2269 * Can fail if we're not blocking on locks, and it's held.
2272 error = xfs_alloc_read_agf(mp, tp, args->agno, flags, &agbp);
2276 ASSERT(flags & XFS_ALLOC_FLAG_TRYLOCK);
2277 ASSERT(!(flags & XFS_ALLOC_FLAG_FREEING));
2282 /* reset a padding mismatched agfl before final free space check */
2283 if (pag->pagf_agflreset)
2284 xfs_agfl_reset(tp, agbp, pag);
2286 /* If there isn't enough total space or single-extent, reject it. */
2287 need = xfs_alloc_min_freelist(mp, pag);
2288 if (!xfs_alloc_space_available(args, need, flags))
2289 goto out_agbp_relse;
2292 * Make the freelist shorter if it's too long.
2294 * Note that from this point onwards, we will always release the agf and
2295 * agfl buffers on error. This handles the case where we error out and
2296 * the buffers are clean or may not have been joined to the transaction
2297 * and hence need to be released manually. If they have been joined to
2298 * the transaction, then xfs_trans_brelse() will handle them
2299 * appropriately based on the recursion count and dirty state of the
2302 * XXX (dgc): When we have lots of free space, does this buy us
2303 * anything other than extra overhead when we need to put more blocks
2304 * back on the free list? Maybe we should only do this when space is
2305 * getting low or the AGFL is more than half full?
2307 * The NOSHRINK flag prevents the AGFL from being shrunk if it's too
2308 * big; the NORMAP flag prevents AGFL expand/shrink operations from
2309 * updating the rmapbt. Both flags are used in xfs_repair while we're
2310 * rebuilding the rmapbt, and neither are used by the kernel. They're
2311 * both required to ensure that rmaps are correctly recorded for the
2312 * regenerated AGFL, bnobt, and cntbt. See repair/phase5.c and
2313 * repair/rmap.c in xfsprogs for details.
2315 memset(&targs, 0, sizeof(targs));
2316 if (flags & XFS_ALLOC_FLAG_NORMAP)
2317 xfs_rmap_skip_owner_update(&targs.oinfo);
2319 xfs_rmap_ag_owner(&targs.oinfo, XFS_RMAP_OWN_AG);
2320 while (!(flags & XFS_ALLOC_FLAG_NOSHRINK) && pag->pagf_flcount > need) {
2321 error = xfs_alloc_get_freelist(tp, agbp, &bno, 0);
2323 goto out_agbp_relse;
2325 /* defer agfl frees if dfops is provided */
2326 if (tp->t_agfl_dfops) {
2327 xfs_defer_agfl_block(mp, tp->t_agfl_dfops, args->agno,
2330 error = xfs_free_agfl_block(tp, args->agno, bno, agbp,
2333 goto out_agbp_relse;
2340 targs.agno = args->agno;
2341 targs.alignment = targs.minlen = targs.prod = 1;
2342 targs.type = XFS_ALLOCTYPE_THIS_AG;
2344 error = xfs_alloc_read_agfl(mp, tp, targs.agno, &agflbp);
2346 goto out_agbp_relse;
2348 /* Make the freelist longer if it's too short. */
2349 while (pag->pagf_flcount < need) {
2351 targs.maxlen = need - pag->pagf_flcount;
2352 targs.resv = XFS_AG_RESV_AGFL;
2354 /* Allocate as many blocks as possible at once. */
2355 error = xfs_alloc_ag_vextent(&targs);
2357 goto out_agflbp_relse;
2360 * Stop if we run out. Won't happen if callers are obeying
2361 * the restrictions correctly. Can happen for free calls
2362 * on a completely full ag.
2364 if (targs.agbno == NULLAGBLOCK) {
2365 if (flags & XFS_ALLOC_FLAG_FREEING)
2367 goto out_agflbp_relse;
2370 * Put each allocated block on the list.
2372 for (bno = targs.agbno; bno < targs.agbno + targs.len; bno++) {
2373 error = xfs_alloc_put_freelist(tp, agbp,
2376 goto out_agflbp_relse;
2379 xfs_trans_brelse(tp, agflbp);
2384 xfs_trans_brelse(tp, agflbp);
2387 xfs_trans_brelse(tp, agbp);
2394 * Get a block from the freelist.
2395 * Returns with the buffer for the block gotten.
2398 xfs_alloc_get_freelist(
2399 xfs_trans_t *tp, /* transaction pointer */
2400 xfs_buf_t *agbp, /* buffer containing the agf structure */
2401 xfs_agblock_t *bnop, /* block address retrieved from freelist */
2402 int btreeblk) /* destination is a AGF btree */
2404 xfs_agf_t *agf; /* a.g. freespace structure */
2405 xfs_buf_t *agflbp;/* buffer for a.g. freelist structure */
2406 xfs_agblock_t bno; /* block number returned */
2410 xfs_mount_t *mp = tp->t_mountp;
2411 xfs_perag_t *pag; /* per allocation group data */
2414 * Freelist is empty, give up.
2416 agf = XFS_BUF_TO_AGF(agbp);
2417 if (!agf->agf_flcount) {
2418 *bnop = NULLAGBLOCK;
2422 * Read the array of free blocks.
2424 error = xfs_alloc_read_agfl(mp, tp, be32_to_cpu(agf->agf_seqno),
2431 * Get the block number and update the data structures.
2433 agfl_bno = XFS_BUF_TO_AGFL_BNO(mp, agflbp);
2434 bno = be32_to_cpu(agfl_bno[be32_to_cpu(agf->agf_flfirst)]);
2435 be32_add_cpu(&agf->agf_flfirst, 1);
2436 xfs_trans_brelse(tp, agflbp);
2437 if (be32_to_cpu(agf->agf_flfirst) == xfs_agfl_size(mp))
2438 agf->agf_flfirst = 0;
2440 pag = xfs_perag_get(mp, be32_to_cpu(agf->agf_seqno));
2441 ASSERT(!pag->pagf_agflreset);
2442 be32_add_cpu(&agf->agf_flcount, -1);
2443 xfs_trans_agflist_delta(tp, -1);
2444 pag->pagf_flcount--;
2447 logflags = XFS_AGF_FLFIRST | XFS_AGF_FLCOUNT;
2449 be32_add_cpu(&agf->agf_btreeblks, 1);
2450 pag->pagf_btreeblks++;
2451 logflags |= XFS_AGF_BTREEBLKS;
2454 xfs_alloc_log_agf(tp, agbp, logflags);
2461 * Log the given fields from the agf structure.
2465 xfs_trans_t *tp, /* transaction pointer */
2466 xfs_buf_t *bp, /* buffer for a.g. freelist header */
2467 int fields) /* mask of fields to be logged (XFS_AGF_...) */
2469 int first; /* first byte offset */
2470 int last; /* last byte offset */
2471 static const short offsets[] = {
2472 offsetof(xfs_agf_t, agf_magicnum),
2473 offsetof(xfs_agf_t, agf_versionnum),
2474 offsetof(xfs_agf_t, agf_seqno),
2475 offsetof(xfs_agf_t, agf_length),
2476 offsetof(xfs_agf_t, agf_roots[0]),
2477 offsetof(xfs_agf_t, agf_levels[0]),
2478 offsetof(xfs_agf_t, agf_flfirst),
2479 offsetof(xfs_agf_t, agf_fllast),
2480 offsetof(xfs_agf_t, agf_flcount),
2481 offsetof(xfs_agf_t, agf_freeblks),
2482 offsetof(xfs_agf_t, agf_longest),
2483 offsetof(xfs_agf_t, agf_btreeblks),
2484 offsetof(xfs_agf_t, agf_uuid),
2485 offsetof(xfs_agf_t, agf_rmap_blocks),
2486 offsetof(xfs_agf_t, agf_refcount_blocks),
2487 offsetof(xfs_agf_t, agf_refcount_root),
2488 offsetof(xfs_agf_t, agf_refcount_level),
2489 /* needed so that we don't log the whole rest of the structure: */
2490 offsetof(xfs_agf_t, agf_spare64),
2494 trace_xfs_agf(tp->t_mountp, XFS_BUF_TO_AGF(bp), fields, _RET_IP_);
2496 xfs_trans_buf_set_type(tp, bp, XFS_BLFT_AGF_BUF);
2498 xfs_btree_offsets(fields, offsets, XFS_AGF_NUM_BITS, &first, &last);
2499 xfs_trans_log_buf(tp, bp, (uint)first, (uint)last);
2503 * Interface for inode allocation to force the pag data to be initialized.
2506 xfs_alloc_pagf_init(
2507 xfs_mount_t *mp, /* file system mount structure */
2508 xfs_trans_t *tp, /* transaction pointer */
2509 xfs_agnumber_t agno, /* allocation group number */
2510 int flags) /* XFS_ALLOC_FLAGS_... */
2515 if ((error = xfs_alloc_read_agf(mp, tp, agno, flags, &bp)))
2518 xfs_trans_brelse(tp, bp);
2523 * Put the block on the freelist for the allocation group.
2526 xfs_alloc_put_freelist(
2527 xfs_trans_t *tp, /* transaction pointer */
2528 xfs_buf_t *agbp, /* buffer for a.g. freelist header */
2529 xfs_buf_t *agflbp,/* buffer for a.g. free block array */
2530 xfs_agblock_t bno, /* block being freed */
2531 int btreeblk) /* block came from a AGF btree */
2533 xfs_agf_t *agf; /* a.g. freespace structure */
2534 __be32 *blockp;/* pointer to array entry */
2537 xfs_mount_t *mp; /* mount structure */
2538 xfs_perag_t *pag; /* per allocation group data */
2542 agf = XFS_BUF_TO_AGF(agbp);
2545 if (!agflbp && (error = xfs_alloc_read_agfl(mp, tp,
2546 be32_to_cpu(agf->agf_seqno), &agflbp)))
2548 be32_add_cpu(&agf->agf_fllast, 1);
2549 if (be32_to_cpu(agf->agf_fllast) == xfs_agfl_size(mp))
2550 agf->agf_fllast = 0;
2552 pag = xfs_perag_get(mp, be32_to_cpu(agf->agf_seqno));
2553 ASSERT(!pag->pagf_agflreset);
2554 be32_add_cpu(&agf->agf_flcount, 1);
2555 xfs_trans_agflist_delta(tp, 1);
2556 pag->pagf_flcount++;
2558 logflags = XFS_AGF_FLLAST | XFS_AGF_FLCOUNT;
2560 be32_add_cpu(&agf->agf_btreeblks, -1);
2561 pag->pagf_btreeblks--;
2562 logflags |= XFS_AGF_BTREEBLKS;
2566 xfs_alloc_log_agf(tp, agbp, logflags);
2568 ASSERT(be32_to_cpu(agf->agf_flcount) <= xfs_agfl_size(mp));
2570 agfl_bno = XFS_BUF_TO_AGFL_BNO(mp, agflbp);
2571 blockp = &agfl_bno[be32_to_cpu(agf->agf_fllast)];
2572 *blockp = cpu_to_be32(bno);
2573 startoff = (char *)blockp - (char *)agflbp->b_addr;
2575 xfs_alloc_log_agf(tp, agbp, logflags);
2577 xfs_trans_buf_set_type(tp, agflbp, XFS_BLFT_AGFL_BUF);
2578 xfs_trans_log_buf(tp, agflbp, startoff,
2579 startoff + sizeof(xfs_agblock_t) - 1);
2583 static xfs_failaddr_t
2587 struct xfs_mount *mp = bp->b_target->bt_mount;
2588 struct xfs_agf *agf = XFS_BUF_TO_AGF(bp);
2590 if (xfs_sb_version_hascrc(&mp->m_sb)) {
2591 if (!uuid_equal(&agf->agf_uuid, &mp->m_sb.sb_meta_uuid))
2592 return __this_address;
2593 if (!xfs_log_check_lsn(mp,
2594 be64_to_cpu(XFS_BUF_TO_AGF(bp)->agf_lsn)))
2595 return __this_address;
2598 if (!(agf->agf_magicnum == cpu_to_be32(XFS_AGF_MAGIC) &&
2599 XFS_AGF_GOOD_VERSION(be32_to_cpu(agf->agf_versionnum)) &&
2600 be32_to_cpu(agf->agf_freeblks) <= be32_to_cpu(agf->agf_length) &&
2601 be32_to_cpu(agf->agf_flfirst) < xfs_agfl_size(mp) &&
2602 be32_to_cpu(agf->agf_fllast) < xfs_agfl_size(mp) &&
2603 be32_to_cpu(agf->agf_flcount) <= xfs_agfl_size(mp)))
2604 return __this_address;
2606 if (be32_to_cpu(agf->agf_levels[XFS_BTNUM_BNO]) < 1 ||
2607 be32_to_cpu(agf->agf_levels[XFS_BTNUM_CNT]) < 1 ||
2608 be32_to_cpu(agf->agf_levels[XFS_BTNUM_BNO]) > XFS_BTREE_MAXLEVELS ||
2609 be32_to_cpu(agf->agf_levels[XFS_BTNUM_CNT]) > XFS_BTREE_MAXLEVELS)
2610 return __this_address;
2612 if (xfs_sb_version_hasrmapbt(&mp->m_sb) &&
2613 (be32_to_cpu(agf->agf_levels[XFS_BTNUM_RMAP]) < 1 ||
2614 be32_to_cpu(agf->agf_levels[XFS_BTNUM_RMAP]) > XFS_BTREE_MAXLEVELS))
2615 return __this_address;
2618 * during growfs operations, the perag is not fully initialised,
2619 * so we can't use it for any useful checking. growfs ensures we can't
2620 * use it by using uncached buffers that don't have the perag attached
2621 * so we can detect and avoid this problem.
2623 if (bp->b_pag && be32_to_cpu(agf->agf_seqno) != bp->b_pag->pag_agno)
2624 return __this_address;
2626 if (xfs_sb_version_haslazysbcount(&mp->m_sb) &&
2627 be32_to_cpu(agf->agf_btreeblks) > be32_to_cpu(agf->agf_length))
2628 return __this_address;
2630 if (xfs_sb_version_hasreflink(&mp->m_sb) &&
2631 (be32_to_cpu(agf->agf_refcount_level) < 1 ||
2632 be32_to_cpu(agf->agf_refcount_level) > XFS_BTREE_MAXLEVELS))
2633 return __this_address;
2640 xfs_agf_read_verify(
2643 struct xfs_mount *mp = bp->b_target->bt_mount;
2646 if (xfs_sb_version_hascrc(&mp->m_sb) &&
2647 !xfs_buf_verify_cksum(bp, XFS_AGF_CRC_OFF))
2648 xfs_verifier_error(bp, -EFSBADCRC, __this_address);
2650 fa = xfs_agf_verify(bp);
2651 if (XFS_TEST_ERROR(fa, mp, XFS_ERRTAG_ALLOC_READ_AGF))
2652 xfs_verifier_error(bp, -EFSCORRUPTED, fa);
2657 xfs_agf_write_verify(
2660 struct xfs_mount *mp = bp->b_target->bt_mount;
2661 struct xfs_buf_log_item *bip = bp->b_log_item;
2664 fa = xfs_agf_verify(bp);
2666 xfs_verifier_error(bp, -EFSCORRUPTED, fa);
2670 if (!xfs_sb_version_hascrc(&mp->m_sb))
2674 XFS_BUF_TO_AGF(bp)->agf_lsn = cpu_to_be64(bip->bli_item.li_lsn);
2676 xfs_buf_update_cksum(bp, XFS_AGF_CRC_OFF);
2679 const struct xfs_buf_ops xfs_agf_buf_ops = {
2681 .verify_read = xfs_agf_read_verify,
2682 .verify_write = xfs_agf_write_verify,
2683 .verify_struct = xfs_agf_verify,
2687 * Read in the allocation group header (free/alloc section).
2691 struct xfs_mount *mp, /* mount point structure */
2692 struct xfs_trans *tp, /* transaction pointer */
2693 xfs_agnumber_t agno, /* allocation group number */
2694 int flags, /* XFS_BUF_ */
2695 struct xfs_buf **bpp) /* buffer for the ag freelist header */
2699 trace_xfs_read_agf(mp, agno);
2701 ASSERT(agno != NULLAGNUMBER);
2702 error = xfs_trans_read_buf(
2703 mp, tp, mp->m_ddev_targp,
2704 XFS_AG_DADDR(mp, agno, XFS_AGF_DADDR(mp)),
2705 XFS_FSS_TO_BB(mp, 1), flags, bpp, &xfs_agf_buf_ops);
2711 ASSERT(!(*bpp)->b_error);
2712 xfs_buf_set_ref(*bpp, XFS_AGF_REF);
2717 * Read in the allocation group header (free/alloc section).
2721 struct xfs_mount *mp, /* mount point structure */
2722 struct xfs_trans *tp, /* transaction pointer */
2723 xfs_agnumber_t agno, /* allocation group number */
2724 int flags, /* XFS_ALLOC_FLAG_... */
2725 struct xfs_buf **bpp) /* buffer for the ag freelist header */
2727 struct xfs_agf *agf; /* ag freelist header */
2728 struct xfs_perag *pag; /* per allocation group data */
2731 trace_xfs_alloc_read_agf(mp, agno);
2733 ASSERT(agno != NULLAGNUMBER);
2734 error = xfs_read_agf(mp, tp, agno,
2735 (flags & XFS_ALLOC_FLAG_TRYLOCK) ? XBF_TRYLOCK : 0,
2741 ASSERT(!(*bpp)->b_error);
2743 agf = XFS_BUF_TO_AGF(*bpp);
2744 pag = xfs_perag_get(mp, agno);
2745 if (!pag->pagf_init) {
2746 pag->pagf_freeblks = be32_to_cpu(agf->agf_freeblks);
2747 pag->pagf_btreeblks = be32_to_cpu(agf->agf_btreeblks);
2748 pag->pagf_flcount = be32_to_cpu(agf->agf_flcount);
2749 pag->pagf_longest = be32_to_cpu(agf->agf_longest);
2750 pag->pagf_levels[XFS_BTNUM_BNOi] =
2751 be32_to_cpu(agf->agf_levels[XFS_BTNUM_BNOi]);
2752 pag->pagf_levels[XFS_BTNUM_CNTi] =
2753 be32_to_cpu(agf->agf_levels[XFS_BTNUM_CNTi]);
2754 pag->pagf_levels[XFS_BTNUM_RMAPi] =
2755 be32_to_cpu(agf->agf_levels[XFS_BTNUM_RMAPi]);
2756 pag->pagf_refcount_level = be32_to_cpu(agf->agf_refcount_level);
2757 spin_lock_init(&pag->pagb_lock);
2758 pag->pagb_count = 0;
2759 pag->pagb_tree = RB_ROOT;
2761 pag->pagf_agflreset = xfs_agfl_needs_reset(mp, agf);
2764 else if (!XFS_FORCED_SHUTDOWN(mp)) {
2765 ASSERT(pag->pagf_freeblks == be32_to_cpu(agf->agf_freeblks));
2766 ASSERT(pag->pagf_btreeblks == be32_to_cpu(agf->agf_btreeblks));
2767 ASSERT(pag->pagf_flcount == be32_to_cpu(agf->agf_flcount));
2768 ASSERT(pag->pagf_longest == be32_to_cpu(agf->agf_longest));
2769 ASSERT(pag->pagf_levels[XFS_BTNUM_BNOi] ==
2770 be32_to_cpu(agf->agf_levels[XFS_BTNUM_BNOi]));
2771 ASSERT(pag->pagf_levels[XFS_BTNUM_CNTi] ==
2772 be32_to_cpu(agf->agf_levels[XFS_BTNUM_CNTi]));
2780 * Allocate an extent (variable-size).
2781 * Depending on the allocation type, we either look in a single allocation
2782 * group or loop over the allocation groups to find the result.
2786 xfs_alloc_arg_t *args) /* allocation argument structure */
2788 xfs_agblock_t agsize; /* allocation group size */
2790 int flags; /* XFS_ALLOC_FLAG_... locking flags */
2791 xfs_mount_t *mp; /* mount structure pointer */
2792 xfs_agnumber_t sagno; /* starting allocation group number */
2793 xfs_alloctype_t type; /* input allocation type */
2795 xfs_agnumber_t rotorstep = xfs_rotorstep; /* inode32 agf stepper */
2798 type = args->otype = args->type;
2799 args->agbno = NULLAGBLOCK;
2801 * Just fix this up, for the case where the last a.g. is shorter
2802 * (or there's only one a.g.) and the caller couldn't easily figure
2803 * that out (xfs_bmap_alloc).
2805 agsize = mp->m_sb.sb_agblocks;
2806 if (args->maxlen > agsize)
2807 args->maxlen = agsize;
2808 if (args->alignment == 0)
2809 args->alignment = 1;
2810 ASSERT(XFS_FSB_TO_AGNO(mp, args->fsbno) < mp->m_sb.sb_agcount);
2811 ASSERT(XFS_FSB_TO_AGBNO(mp, args->fsbno) < agsize);
2812 ASSERT(args->minlen <= args->maxlen);
2813 ASSERT(args->minlen <= agsize);
2814 ASSERT(args->mod < args->prod);
2815 if (XFS_FSB_TO_AGNO(mp, args->fsbno) >= mp->m_sb.sb_agcount ||
2816 XFS_FSB_TO_AGBNO(mp, args->fsbno) >= agsize ||
2817 args->minlen > args->maxlen || args->minlen > agsize ||
2818 args->mod >= args->prod) {
2819 args->fsbno = NULLFSBLOCK;
2820 trace_xfs_alloc_vextent_badargs(args);
2825 case XFS_ALLOCTYPE_THIS_AG:
2826 case XFS_ALLOCTYPE_NEAR_BNO:
2827 case XFS_ALLOCTYPE_THIS_BNO:
2829 * These three force us into a single a.g.
2831 args->agno = XFS_FSB_TO_AGNO(mp, args->fsbno);
2832 args->pag = xfs_perag_get(mp, args->agno);
2833 error = xfs_alloc_fix_freelist(args, 0);
2835 trace_xfs_alloc_vextent_nofix(args);
2839 trace_xfs_alloc_vextent_noagbp(args);
2842 args->agbno = XFS_FSB_TO_AGBNO(mp, args->fsbno);
2843 if ((error = xfs_alloc_ag_vextent(args)))
2846 case XFS_ALLOCTYPE_START_BNO:
2848 * Try near allocation first, then anywhere-in-ag after
2849 * the first a.g. fails.
2851 if ((args->datatype & XFS_ALLOC_INITIAL_USER_DATA) &&
2852 (mp->m_flags & XFS_MOUNT_32BITINODES)) {
2853 args->fsbno = XFS_AGB_TO_FSB(mp,
2854 ((mp->m_agfrotor / rotorstep) %
2855 mp->m_sb.sb_agcount), 0);
2858 args->agbno = XFS_FSB_TO_AGBNO(mp, args->fsbno);
2859 args->type = XFS_ALLOCTYPE_NEAR_BNO;
2861 case XFS_ALLOCTYPE_FIRST_AG:
2863 * Rotate through the allocation groups looking for a winner.
2865 if (type == XFS_ALLOCTYPE_FIRST_AG) {
2867 * Start with allocation group given by bno.
2869 args->agno = XFS_FSB_TO_AGNO(mp, args->fsbno);
2870 args->type = XFS_ALLOCTYPE_THIS_AG;
2875 * Start with the given allocation group.
2877 args->agno = sagno = XFS_FSB_TO_AGNO(mp, args->fsbno);
2878 flags = XFS_ALLOC_FLAG_TRYLOCK;
2881 * Loop over allocation groups twice; first time with
2882 * trylock set, second time without.
2885 args->pag = xfs_perag_get(mp, args->agno);
2886 error = xfs_alloc_fix_freelist(args, flags);
2888 trace_xfs_alloc_vextent_nofix(args);
2892 * If we get a buffer back then the allocation will fly.
2895 if ((error = xfs_alloc_ag_vextent(args)))
2900 trace_xfs_alloc_vextent_loopfailed(args);
2903 * Didn't work, figure out the next iteration.
2905 if (args->agno == sagno &&
2906 type == XFS_ALLOCTYPE_START_BNO)
2907 args->type = XFS_ALLOCTYPE_THIS_AG;
2909 * For the first allocation, we can try any AG to get
2910 * space. However, if we already have allocated a
2911 * block, we don't want to try AGs whose number is below
2912 * sagno. Otherwise, we may end up with out-of-order
2913 * locking of AGF, which might cause deadlock.
2915 if (++(args->agno) == mp->m_sb.sb_agcount) {
2916 if (args->firstblock != NULLFSBLOCK)
2922 * Reached the starting a.g., must either be done
2923 * or switch to non-trylock mode.
2925 if (args->agno == sagno) {
2927 args->agbno = NULLAGBLOCK;
2928 trace_xfs_alloc_vextent_allfailed(args);
2933 if (type == XFS_ALLOCTYPE_START_BNO) {
2934 args->agbno = XFS_FSB_TO_AGBNO(mp,
2936 args->type = XFS_ALLOCTYPE_NEAR_BNO;
2939 xfs_perag_put(args->pag);
2942 if (args->agno == sagno)
2943 mp->m_agfrotor = (mp->m_agfrotor + 1) %
2944 (mp->m_sb.sb_agcount * rotorstep);
2946 mp->m_agfrotor = (args->agno * rotorstep + 1) %
2947 (mp->m_sb.sb_agcount * rotorstep);
2954 if (args->agbno == NULLAGBLOCK)
2955 args->fsbno = NULLFSBLOCK;
2957 args->fsbno = XFS_AGB_TO_FSB(mp, args->agno, args->agbno);
2959 ASSERT(args->len >= args->minlen);
2960 ASSERT(args->len <= args->maxlen);
2961 ASSERT(args->agbno % args->alignment == 0);
2962 XFS_AG_CHECK_DADDR(mp, XFS_FSB_TO_DADDR(mp, args->fsbno),
2966 /* Zero the extent if we were asked to do so */
2967 if (args->datatype & XFS_ALLOC_USERDATA_ZERO) {
2968 error = xfs_zero_extent(args->ip, args->fsbno, args->len);
2974 xfs_perag_put(args->pag);
2977 xfs_perag_put(args->pag);
2981 /* Ensure that the freelist is at full capacity. */
2983 xfs_free_extent_fix_freelist(
2984 struct xfs_trans *tp,
2985 xfs_agnumber_t agno,
2986 struct xfs_buf **agbp)
2988 struct xfs_alloc_arg args;
2991 memset(&args, 0, sizeof(struct xfs_alloc_arg));
2993 args.mp = tp->t_mountp;
2997 * validate that the block number is legal - the enables us to detect
2998 * and handle a silent filesystem corruption rather than crashing.
3000 if (args.agno >= args.mp->m_sb.sb_agcount)
3001 return -EFSCORRUPTED;
3003 args.pag = xfs_perag_get(args.mp, args.agno);
3006 error = xfs_alloc_fix_freelist(&args, XFS_ALLOC_FLAG_FREEING);
3012 xfs_perag_put(args.pag);
3018 * Just break up the extent address and hand off to xfs_free_ag_extent
3019 * after fixing up the freelist.
3023 struct xfs_trans *tp, /* transaction pointer */
3024 xfs_fsblock_t bno, /* starting block number of extent */
3025 xfs_extlen_t len, /* length of extent */
3026 struct xfs_owner_info *oinfo, /* extent owner */
3027 enum xfs_ag_resv_type type, /* block reservation type */
3030 struct xfs_mount *mp = tp->t_mountp;
3031 struct xfs_buf *agbp;
3032 xfs_agnumber_t agno = XFS_FSB_TO_AGNO(mp, bno);
3033 xfs_agblock_t agbno = XFS_FSB_TO_AGBNO(mp, bno);
3035 unsigned int busy_flags = 0;
3038 ASSERT(type != XFS_AG_RESV_AGFL);
3040 if (XFS_TEST_ERROR(false, mp,
3041 XFS_ERRTAG_FREE_EXTENT))
3044 error = xfs_free_extent_fix_freelist(tp, agno, &agbp);
3048 XFS_WANT_CORRUPTED_GOTO(mp, agbno < mp->m_sb.sb_agblocks, err);
3050 /* validate the extent size is legal now we have the agf locked */
3051 XFS_WANT_CORRUPTED_GOTO(mp,
3052 agbno + len <= be32_to_cpu(XFS_BUF_TO_AGF(agbp)->agf_length),
3055 error = xfs_free_ag_extent(tp, agbp, agno, agbno, len, oinfo, type);
3060 busy_flags |= XFS_EXTENT_BUSY_SKIP_DISCARD;
3061 xfs_extent_busy_insert(tp, agno, agbno, len, busy_flags);
3065 xfs_trans_brelse(tp, agbp);
3069 struct xfs_alloc_query_range_info {
3070 xfs_alloc_query_range_fn fn;
3074 /* Format btree record and pass to our callback. */
3076 xfs_alloc_query_range_helper(
3077 struct xfs_btree_cur *cur,
3078 union xfs_btree_rec *rec,
3081 struct xfs_alloc_query_range_info *query = priv;
3082 struct xfs_alloc_rec_incore irec;
3084 irec.ar_startblock = be32_to_cpu(rec->alloc.ar_startblock);
3085 irec.ar_blockcount = be32_to_cpu(rec->alloc.ar_blockcount);
3086 return query->fn(cur, &irec, query->priv);
3089 /* Find all free space within a given range of blocks. */
3091 xfs_alloc_query_range(
3092 struct xfs_btree_cur *cur,
3093 struct xfs_alloc_rec_incore *low_rec,
3094 struct xfs_alloc_rec_incore *high_rec,
3095 xfs_alloc_query_range_fn fn,
3098 union xfs_btree_irec low_brec;
3099 union xfs_btree_irec high_brec;
3100 struct xfs_alloc_query_range_info query;
3102 ASSERT(cur->bc_btnum == XFS_BTNUM_BNO);
3103 low_brec.a = *low_rec;
3104 high_brec.a = *high_rec;
3107 return xfs_btree_query_range(cur, &low_brec, &high_brec,
3108 xfs_alloc_query_range_helper, &query);
3111 /* Find all free space records. */
3113 xfs_alloc_query_all(
3114 struct xfs_btree_cur *cur,
3115 xfs_alloc_query_range_fn fn,
3118 struct xfs_alloc_query_range_info query;
3120 ASSERT(cur->bc_btnum == XFS_BTNUM_BNO);
3123 return xfs_btree_query_all(cur, xfs_alloc_query_range_helper, &query);
3126 /* Is there a record covering a given extent? */
3128 xfs_alloc_has_record(
3129 struct xfs_btree_cur *cur,
3134 union xfs_btree_irec low;
3135 union xfs_btree_irec high;
3137 memset(&low, 0, sizeof(low));
3138 low.a.ar_startblock = bno;
3139 memset(&high, 0xFF, sizeof(high));
3140 high.a.ar_startblock = bno + len - 1;
3142 return xfs_btree_has_record(cur, &low, &high, exists);
3146 * Walk all the blocks in the AGFL. The @walk_fn can return any negative
3147 * error code or XFS_BTREE_QUERY_RANGE_ABORT.
3151 struct xfs_mount *mp,
3152 struct xfs_agf *agf,
3153 struct xfs_buf *agflbp,
3154 xfs_agfl_walk_fn walk_fn,
3161 agfl_bno = XFS_BUF_TO_AGFL_BNO(mp, agflbp);
3162 i = be32_to_cpu(agf->agf_flfirst);
3164 /* Nothing to walk in an empty AGFL. */
3165 if (agf->agf_flcount == cpu_to_be32(0))
3168 /* Otherwise, walk from first to last, wrapping as needed. */
3170 error = walk_fn(mp, be32_to_cpu(agfl_bno[i]), priv);
3173 if (i == be32_to_cpu(agf->agf_fllast))
3175 if (++i == xfs_agfl_size(mp))