int
xfs_rmap_insert(
struct xfs_btree_cur *rcur,
xfs_agblock_t agbno,
xfs_extlen_t len,
uint64_t owner,
uint64_t offset,
unsigned int flags)
{
int i;
int error;
trace_xfs_rmap_insert(rcur->bc_mp, rcur->bc_private.a.agno, agbno,
len, owner, offset, flags);
error = xfs_rmap_lookup_eq(rcur, agbno, len, owner, offset, flags, &i);
if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(rcur->bc_mp, i == 0, done);
rcur->bc_rec.r.rm_startblock = agbno;
rcur->bc_rec.r.rm_blockcount = len;
rcur->bc_rec.r.rm_owner = owner;
rcur->bc_rec.r.rm_offset = offset;
rcur->bc_rec.r.rm_flags = flags;
error = xfs_btree_insert(rcur, &i);
if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(rcur->bc_mp, i == 1, done);
done:
if (error)
trace_xfs_rmap_insert_error(rcur->bc_mp,
rcur->bc_private.a.agno, error, _RET_IP_);
return error;
}
/*
* Allocate new inodes in the allocation group specified by agbp.
* Return 0 for success, else error code.
*/
STATIC int /* error code or 0 */
xfs_ialloc_ag_alloc(
xfs_trans_t *tp, /* transaction pointer */
xfs_buf_t *agbp, /* alloc group buffer */
int *alloc)
{
xfs_agi_t *agi; /* allocation group header */
xfs_alloc_arg_t args; /* allocation argument structure */
xfs_btree_cur_t *cur; /* inode btree cursor */
xfs_agnumber_t agno;
int error;
int i;
xfs_agino_t newino; /* new first inode's number */
xfs_agino_t newlen; /* new number of inodes */
xfs_agino_t thisino; /* current inode number, for loop */
int isaligned = 0; /* inode allocation at stripe unit */
/* boundary */
struct xfs_perag *pag;
args.tp = tp;
args.mp = tp->t_mountp;
/*
* Locking will ensure that we don't have two callers in here
* at one time.
*/
newlen = XFS_IALLOC_INODES(args.mp);
if (args.mp->m_maxicount &&
args.mp->m_sb.sb_icount + newlen > args.mp->m_maxicount)
return XFS_ERROR(ENOSPC);
args.minlen = args.maxlen = XFS_IALLOC_BLOCKS(args.mp);
/*
* First try to allocate inodes contiguous with the last-allocated
* chunk of inodes. If the filesystem is striped, this will fill
* an entire stripe unit with inodes.
*/
agi = XFS_BUF_TO_AGI(agbp);
newino = be32_to_cpu(agi->agi_newino);
agno = be32_to_cpu(agi->agi_seqno);
args.agbno = XFS_AGINO_TO_AGBNO(args.mp, newino) +
XFS_IALLOC_BLOCKS(args.mp);
if (likely(newino != NULLAGINO &&
(args.agbno < be32_to_cpu(agi->agi_length)))) {
args.fsbno = XFS_AGB_TO_FSB(args.mp, agno, args.agbno);
args.type = XFS_ALLOCTYPE_THIS_BNO;
args.mod = args.total = args.wasdel = args.isfl =
args.userdata = args.minalignslop = 0;
args.prod = 1;
/*
* We need to take into account alignment here to ensure that
* we don't modify the free list if we fail to have an exact
* block. If we don't have an exact match, and every oher
* attempt allocation attempt fails, we'll end up cancelling
* a dirty transaction and shutting down.
*
* For an exact allocation, alignment must be 1,
* however we need to take cluster alignment into account when
* fixing up the freelist. Use the minalignslop field to
* indicate that extra blocks might be required for alignment,
* but not to use them in the actual exact allocation.
*/
args.alignment = 1;
args.minalignslop = xfs_ialloc_cluster_alignment(&args) - 1;
/* Allow space for the inode btree to split. */
args.minleft = args.mp->m_in_maxlevels - 1;
if ((error = xfs_alloc_vextent(&args)))
return error;
} else
args.fsbno = NULLFSBLOCK;
if (unlikely(args.fsbno == NULLFSBLOCK)) {
/*
* Set the alignment for the allocation.
* If stripe alignment is turned on then align at stripe unit
* boundary.
* If the cluster size is smaller than a filesystem block
* then we're doing I/O for inodes in filesystem block size
* pieces, so don't need alignment anyway.
*/
isaligned = 0;
if (args.mp->m_sinoalign) {
ASSERT(!(args.mp->m_flags & XFS_MOUNT_NOALIGN));
args.alignment = args.mp->m_dalign;
isaligned = 1;
} else
args.alignment = xfs_ialloc_cluster_alignment(&args);
/*
* Need to figure out where to allocate the inode blocks.
* Ideally they should be spaced out through the a.g.
* For now, just allocate blocks up front.
*/
args.agbno = be32_to_cpu(agi->agi_root);
args.fsbno = XFS_AGB_TO_FSB(args.mp, agno, args.agbno);
/*
* Allocate a fixed-size extent of inodes.
*/
args.type = XFS_ALLOCTYPE_NEAR_BNO;
args.mod = args.total = args.wasdel = args.isfl =
args.userdata = args.minalignslop = 0;
args.prod = 1;
/*
* Allow space for the inode btree to split.
*/
args.minleft = args.mp->m_in_maxlevels - 1;
if ((error = xfs_alloc_vextent(&args)))
return error;
}
/*
* If stripe alignment is turned on, then try again with cluster
* alignment.
*/
if (isaligned && args.fsbno == NULLFSBLOCK) {
args.type = XFS_ALLOCTYPE_NEAR_BNO;
args.agbno = be32_to_cpu(agi->agi_root);
args.fsbno = XFS_AGB_TO_FSB(args.mp, agno, args.agbno);
args.alignment = xfs_ialloc_cluster_alignment(&args);
if ((error = xfs_alloc_vextent(&args)))
return error;
}
if (args.fsbno == NULLFSBLOCK) {
*alloc = 0;
return 0;
}
ASSERT(args.len == args.minlen);
/*
* Stamp and write the inode buffers.
*
* Seed the new inode cluster with a random generation number. This
* prevents short-term reuse of generation numbers if a chunk is
* freed and then immediately reallocated. We use random numbers
* rather than a linear progression to prevent the next generation
* number from being easily guessable.
*/
error = xfs_ialloc_inode_init(args.mp, tp, agno, args.agbno,
args.len, prandom_u32());
if (error)
return error;
/*
* Convert the results.
*/
newino = XFS_OFFBNO_TO_AGINO(args.mp, args.agbno, 0);
be32_add_cpu(&agi->agi_count, newlen);
be32_add_cpu(&agi->agi_freecount, newlen);
pag = xfs_perag_get(args.mp, agno);
pag->pagi_freecount += newlen;
xfs_perag_put(pag);
agi->agi_newino = cpu_to_be32(newino);
/*
* Insert records describing the new inode chunk into the btree.
*/
cur = xfs_inobt_init_cursor(args.mp, tp, agbp, agno);
for (thisino = newino;
thisino < newino + newlen;
thisino += XFS_INODES_PER_CHUNK) {
cur->bc_rec.i.ir_startino = thisino;
cur->bc_rec.i.ir_freecount = XFS_INODES_PER_CHUNK;
cur->bc_rec.i.ir_free = XFS_INOBT_ALL_FREE;
error = xfs_btree_lookup(cur, XFS_LOOKUP_EQ, &i);
if (error) {
xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
return error;
}
ASSERT(i == 0);
error = xfs_btree_insert(cur, &i);
if (error) {
xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
return error;
}
ASSERT(i == 1);
}
xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR);
/*
* Log allocation group header fields
*/
xfs_ialloc_log_agi(tp, agbp,
XFS_AGI_COUNT | XFS_AGI_FREECOUNT | XFS_AGI_NEWINO);
/*
* Modify/log superblock values for inode count and inode free count.
*/
xfs_trans_mod_sb(tp, XFS_TRANS_SB_ICOUNT, (long)newlen);
xfs_trans_mod_sb(tp, XFS_TRANS_SB_IFREE, (long)newlen);
*alloc = 1;
return 0;
}
/*
* Convert an unwritten extent to a real extent or vice versa.
* Does not handle overlapping extents.
*/
STATIC int
xfs_rmap_convert(
struct xfs_btree_cur *cur,
xfs_agblock_t bno,
xfs_extlen_t len,
bool unwritten,
struct xfs_owner_info *oinfo)
{
struct xfs_mount *mp = cur->bc_mp;
struct xfs_rmap_irec r[4]; /* neighbor extent entries */
/* left is 0, right is 1, prev is 2 */
/* new is 3 */
uint64_t owner;
uint64_t offset;
uint64_t new_endoff;
unsigned int oldext;
unsigned int newext;
unsigned int flags = 0;
int i;
int state = 0;
int error;
xfs_owner_info_unpack(oinfo, &owner, &offset, &flags);
ASSERT(!(XFS_RMAP_NON_INODE_OWNER(owner) ||
(flags & (XFS_RMAP_ATTR_FORK | XFS_RMAP_BMBT_BLOCK))));
oldext = unwritten ? XFS_RMAP_UNWRITTEN : 0;
new_endoff = offset + len;
trace_xfs_rmap_convert(mp, cur->bc_private.a.agno, bno, len,
unwritten, oinfo);
/*
* For the initial lookup, look for an exact match or the left-adjacent
* record for our insertion point. This will also give us the record for
* start block contiguity tests.
*/
error = xfs_rmap_lookup_le(cur, bno, len, owner, offset, oldext, &i);
if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
error = xfs_rmap_get_rec(cur, &PREV, &i);
if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
trace_xfs_rmap_lookup_le_range_result(cur->bc_mp,
cur->bc_private.a.agno, PREV.rm_startblock,
PREV.rm_blockcount, PREV.rm_owner,
PREV.rm_offset, PREV.rm_flags);
ASSERT(PREV.rm_offset <= offset);
ASSERT(PREV.rm_offset + PREV.rm_blockcount >= new_endoff);
ASSERT((PREV.rm_flags & XFS_RMAP_UNWRITTEN) == oldext);
newext = ~oldext & XFS_RMAP_UNWRITTEN;
/*
* Set flags determining what part of the previous oldext allocation
* extent is being replaced by a newext allocation.
*/
if (PREV.rm_offset == offset)
state |= RMAP_LEFT_FILLING;
if (PREV.rm_offset + PREV.rm_blockcount == new_endoff)
state |= RMAP_RIGHT_FILLING;
/*
* Decrement the cursor to see if we have a left-adjacent record to our
* insertion point. This will give us the record for end block
* contiguity tests.
*/
error = xfs_btree_decrement(cur, 0, &i);
if (error)
goto done;
if (i) {
state |= RMAP_LEFT_VALID;
error = xfs_rmap_get_rec(cur, &LEFT, &i);
if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
XFS_WANT_CORRUPTED_GOTO(mp,
LEFT.rm_startblock + LEFT.rm_blockcount <= bno,
done);
trace_xfs_rmap_find_left_neighbor_result(cur->bc_mp,
cur->bc_private.a.agno, LEFT.rm_startblock,
LEFT.rm_blockcount, LEFT.rm_owner,
LEFT.rm_offset, LEFT.rm_flags);
if (LEFT.rm_startblock + LEFT.rm_blockcount == bno &&
LEFT.rm_offset + LEFT.rm_blockcount == offset &&
xfs_rmap_is_mergeable(&LEFT, owner, newext))
state |= RMAP_LEFT_CONTIG;
}
/*
* Increment the cursor to see if we have a right-adjacent record to our
* insertion point. This will give us the record for end block
* contiguity tests.
*/
error = xfs_btree_increment(cur, 0, &i);
if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
error = xfs_btree_increment(cur, 0, &i);
if (error)
goto done;
if (i) {
state |= RMAP_RIGHT_VALID;
error = xfs_rmap_get_rec(cur, &RIGHT, &i);
if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
XFS_WANT_CORRUPTED_GOTO(mp, bno + len <= RIGHT.rm_startblock,
done);
trace_xfs_rmap_find_right_neighbor_result(cur->bc_mp,
cur->bc_private.a.agno, RIGHT.rm_startblock,
RIGHT.rm_blockcount, RIGHT.rm_owner,
RIGHT.rm_offset, RIGHT.rm_flags);
if (bno + len == RIGHT.rm_startblock &&
offset + len == RIGHT.rm_offset &&
xfs_rmap_is_mergeable(&RIGHT, owner, newext))
state |= RMAP_RIGHT_CONTIG;
}
/* check that left + prev + right is not too long */
if ((state & (RMAP_LEFT_FILLING | RMAP_LEFT_CONTIG |
RMAP_RIGHT_FILLING | RMAP_RIGHT_CONTIG)) ==
(RMAP_LEFT_FILLING | RMAP_LEFT_CONTIG |
RMAP_RIGHT_FILLING | RMAP_RIGHT_CONTIG) &&
(unsigned long)LEFT.rm_blockcount + len +
RIGHT.rm_blockcount > XFS_RMAP_LEN_MAX)
state &= ~RMAP_RIGHT_CONTIG;
trace_xfs_rmap_convert_state(mp, cur->bc_private.a.agno, state,
_RET_IP_);
/* reset the cursor back to PREV */
error = xfs_rmap_lookup_le(cur, bno, len, owner, offset, oldext, &i);
if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
/*
* Switch out based on the FILLING and CONTIG state bits.
*/
switch (state & (RMAP_LEFT_FILLING | RMAP_LEFT_CONTIG |
RMAP_RIGHT_FILLING | RMAP_RIGHT_CONTIG)) {
case RMAP_LEFT_FILLING | RMAP_LEFT_CONTIG |
RMAP_RIGHT_FILLING | RMAP_RIGHT_CONTIG:
/*
* Setting all of a previous oldext extent to newext.
* The left and right neighbors are both contiguous with new.
*/
error = xfs_btree_increment(cur, 0, &i);
if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
trace_xfs_rmap_delete(mp, cur->bc_private.a.agno,
RIGHT.rm_startblock, RIGHT.rm_blockcount,
RIGHT.rm_owner, RIGHT.rm_offset,
RIGHT.rm_flags);
error = xfs_btree_delete(cur, &i);
if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
error = xfs_btree_decrement(cur, 0, &i);
if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
trace_xfs_rmap_delete(mp, cur->bc_private.a.agno,
PREV.rm_startblock, PREV.rm_blockcount,
PREV.rm_owner, PREV.rm_offset,
PREV.rm_flags);
error = xfs_btree_delete(cur, &i);
if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
error = xfs_btree_decrement(cur, 0, &i);
if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
NEW = LEFT;
NEW.rm_blockcount += PREV.rm_blockcount + RIGHT.rm_blockcount;
error = xfs_rmap_update(cur, &NEW);
if (error)
goto done;
break;
case RMAP_LEFT_FILLING | RMAP_RIGHT_FILLING | RMAP_LEFT_CONTIG:
/*
* Setting all of a previous oldext extent to newext.
* The left neighbor is contiguous, the right is not.
*/
trace_xfs_rmap_delete(mp, cur->bc_private.a.agno,
PREV.rm_startblock, PREV.rm_blockcount,
PREV.rm_owner, PREV.rm_offset,
PREV.rm_flags);
error = xfs_btree_delete(cur, &i);
if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
error = xfs_btree_decrement(cur, 0, &i);
if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
NEW = LEFT;
NEW.rm_blockcount += PREV.rm_blockcount;
error = xfs_rmap_update(cur, &NEW);
if (error)
goto done;
break;
case RMAP_LEFT_FILLING | RMAP_RIGHT_FILLING | RMAP_RIGHT_CONTIG:
/*
* Setting all of a previous oldext extent to newext.
* The right neighbor is contiguous, the left is not.
*/
error = xfs_btree_increment(cur, 0, &i);
if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
trace_xfs_rmap_delete(mp, cur->bc_private.a.agno,
RIGHT.rm_startblock, RIGHT.rm_blockcount,
RIGHT.rm_owner, RIGHT.rm_offset,
RIGHT.rm_flags);
error = xfs_btree_delete(cur, &i);
if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
error = xfs_btree_decrement(cur, 0, &i);
if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
NEW = PREV;
NEW.rm_blockcount = len + RIGHT.rm_blockcount;
NEW.rm_flags = newext;
error = xfs_rmap_update(cur, &NEW);
if (error)
goto done;
break;
case RMAP_LEFT_FILLING | RMAP_RIGHT_FILLING:
/*
* Setting all of a previous oldext extent to newext.
* Neither the left nor right neighbors are contiguous with
* the new one.
*/
NEW = PREV;
NEW.rm_flags = newext;
error = xfs_rmap_update(cur, &NEW);
if (error)
goto done;
break;
case RMAP_LEFT_FILLING | RMAP_LEFT_CONTIG:
/*
* Setting the first part of a previous oldext extent to newext.
* The left neighbor is contiguous.
*/
NEW = PREV;
NEW.rm_offset += len;
NEW.rm_startblock += len;
NEW.rm_blockcount -= len;
error = xfs_rmap_update(cur, &NEW);
if (error)
goto done;
error = xfs_btree_decrement(cur, 0, &i);
if (error)
goto done;
NEW = LEFT;
NEW.rm_blockcount += len;
error = xfs_rmap_update(cur, &NEW);
if (error)
goto done;
break;
case RMAP_LEFT_FILLING:
/*
* Setting the first part of a previous oldext extent to newext.
* The left neighbor is not contiguous.
*/
NEW = PREV;
NEW.rm_startblock += len;
NEW.rm_offset += len;
NEW.rm_blockcount -= len;
error = xfs_rmap_update(cur, &NEW);
if (error)
goto done;
NEW.rm_startblock = bno;
NEW.rm_owner = owner;
NEW.rm_offset = offset;
NEW.rm_blockcount = len;
NEW.rm_flags = newext;
cur->bc_rec.r = NEW;
trace_xfs_rmap_insert(mp, cur->bc_private.a.agno, bno,
len, owner, offset, newext);
error = xfs_btree_insert(cur, &i);
if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
break;
case RMAP_RIGHT_FILLING | RMAP_RIGHT_CONTIG:
/*
* Setting the last part of a previous oldext extent to newext.
* The right neighbor is contiguous with the new allocation.
*/
NEW = PREV;
NEW.rm_blockcount -= len;
error = xfs_rmap_update(cur, &NEW);
if (error)
goto done;
error = xfs_btree_increment(cur, 0, &i);
if (error)
goto done;
NEW = RIGHT;
NEW.rm_offset = offset;
NEW.rm_startblock = bno;
NEW.rm_blockcount += len;
error = xfs_rmap_update(cur, &NEW);
if (error)
goto done;
break;
case RMAP_RIGHT_FILLING:
/*
* Setting the last part of a previous oldext extent to newext.
* The right neighbor is not contiguous.
*/
NEW = PREV;
NEW.rm_blockcount -= len;
error = xfs_rmap_update(cur, &NEW);
if (error)
goto done;
error = xfs_rmap_lookup_eq(cur, bno, len, owner, offset,
oldext, &i);
if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(mp, i == 0, done);
NEW.rm_startblock = bno;
NEW.rm_owner = owner;
NEW.rm_offset = offset;
NEW.rm_blockcount = len;
NEW.rm_flags = newext;
cur->bc_rec.r = NEW;
trace_xfs_rmap_insert(mp, cur->bc_private.a.agno, bno,
len, owner, offset, newext);
error = xfs_btree_insert(cur, &i);
if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
break;
case 0:
/*
* Setting the middle part of a previous oldext extent to
* newext. Contiguity is impossible here.
* One extent becomes three extents.
*/
/* new right extent - oldext */
NEW.rm_startblock = bno + len;
NEW.rm_owner = owner;
NEW.rm_offset = new_endoff;
NEW.rm_blockcount = PREV.rm_offset + PREV.rm_blockcount -
new_endoff;
NEW.rm_flags = PREV.rm_flags;
error = xfs_rmap_update(cur, &NEW);
if (error)
goto done;
/* new left extent - oldext */
NEW = PREV;
NEW.rm_blockcount = offset - PREV.rm_offset;
cur->bc_rec.r = NEW;
trace_xfs_rmap_insert(mp, cur->bc_private.a.agno,
NEW.rm_startblock, NEW.rm_blockcount,
NEW.rm_owner, NEW.rm_offset,
NEW.rm_flags);
error = xfs_btree_insert(cur, &i);
if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
/*
* Reset the cursor to the position of the new extent
* we are about to insert as we can't trust it after
* the previous insert.
*/
error = xfs_rmap_lookup_eq(cur, bno, len, owner, offset,
oldext, &i);
if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(mp, i == 0, done);
/* new middle extent - newext */
cur->bc_rec.r.rm_flags &= ~XFS_RMAP_UNWRITTEN;
cur->bc_rec.r.rm_flags |= newext;
trace_xfs_rmap_insert(mp, cur->bc_private.a.agno, bno, len,
owner, offset, newext);
error = xfs_btree_insert(cur, &i);
if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, done);
break;
case RMAP_LEFT_FILLING | RMAP_LEFT_CONTIG | RMAP_RIGHT_CONTIG:
case RMAP_RIGHT_FILLING | RMAP_LEFT_CONTIG | RMAP_RIGHT_CONTIG:
case RMAP_LEFT_FILLING | RMAP_RIGHT_CONTIG:
case RMAP_RIGHT_FILLING | RMAP_LEFT_CONTIG:
case RMAP_LEFT_CONTIG | RMAP_RIGHT_CONTIG:
case RMAP_LEFT_CONTIG:
case RMAP_RIGHT_CONTIG:
/*
* These cases are all impossible.
*/
ASSERT(0);
}
trace_xfs_rmap_convert_done(mp, cur->bc_private.a.agno, bno, len,
unwritten, oinfo);
done:
if (error)
trace_xfs_rmap_convert_error(cur->bc_mp,
cur->bc_private.a.agno, error, _RET_IP_);
return error;
}
/*
* When we allocate a new block, the first thing we do is add a reference to
* the extent in the rmap btree. This takes the form of a [agbno, length,
* owner, offset] record. Flags are encoded in the high bits of the offset
* field.
*/
STATIC int
xfs_rmap_map(
struct xfs_btree_cur *cur,
xfs_agblock_t bno,
xfs_extlen_t len,
bool unwritten,
struct xfs_owner_info *oinfo)
{
struct xfs_mount *mp = cur->bc_mp;
struct xfs_rmap_irec ltrec;
struct xfs_rmap_irec gtrec;
int have_gt;
int have_lt;
int error = 0;
int i;
uint64_t owner;
uint64_t offset;
unsigned int flags = 0;
bool ignore_off;
xfs_owner_info_unpack(oinfo, &owner, &offset, &flags);
ASSERT(owner != 0);
ignore_off = XFS_RMAP_NON_INODE_OWNER(owner) ||
(flags & XFS_RMAP_BMBT_BLOCK);
if (unwritten)
flags |= XFS_RMAP_UNWRITTEN;
trace_xfs_rmap_map(mp, cur->bc_private.a.agno, bno, len,
unwritten, oinfo);
/*
* For the initial lookup, look for an exact match or the left-adjacent
* record for our insertion point. This will also give us the record for
* start block contiguity tests.
*/
error = xfs_rmap_lookup_le(cur, bno, len, owner, offset, flags,
&have_lt);
if (error)
goto out_error;
XFS_WANT_CORRUPTED_GOTO(mp, have_lt == 1, out_error);
error = xfs_rmap_get_rec(cur, <rec, &have_lt);
if (error)
goto out_error;
XFS_WANT_CORRUPTED_GOTO(mp, have_lt == 1, out_error);
trace_xfs_rmap_lookup_le_range_result(cur->bc_mp,
cur->bc_private.a.agno, ltrec.rm_startblock,
ltrec.rm_blockcount, ltrec.rm_owner,
ltrec.rm_offset, ltrec.rm_flags);
if (!xfs_rmap_is_mergeable(<rec, owner, flags))
have_lt = 0;
XFS_WANT_CORRUPTED_GOTO(mp,
have_lt == 0 ||
ltrec.rm_startblock + ltrec.rm_blockcount <= bno, out_error);
/*
* Increment the cursor to see if we have a right-adjacent record to our
* insertion point. This will give us the record for end block
* contiguity tests.
*/
error = xfs_btree_increment(cur, 0, &have_gt);
if (error)
goto out_error;
if (have_gt) {
error = xfs_rmap_get_rec(cur, >rec, &have_gt);
if (error)
goto out_error;
XFS_WANT_CORRUPTED_GOTO(mp, have_gt == 1, out_error);
XFS_WANT_CORRUPTED_GOTO(mp, bno + len <= gtrec.rm_startblock,
out_error);
trace_xfs_rmap_find_right_neighbor_result(cur->bc_mp,
cur->bc_private.a.agno, gtrec.rm_startblock,
gtrec.rm_blockcount, gtrec.rm_owner,
gtrec.rm_offset, gtrec.rm_flags);
if (!xfs_rmap_is_mergeable(>rec, owner, flags))
have_gt = 0;
}
/*
* Note: cursor currently points one record to the right of ltrec, even
* if there is no record in the tree to the right.
*/
if (have_lt &&
ltrec.rm_startblock + ltrec.rm_blockcount == bno &&
(ignore_off || ltrec.rm_offset + ltrec.rm_blockcount == offset)) {
/*
* left edge contiguous, merge into left record.
*
* ltbno ltlen
* orig: |ooooooooo|
* adding: |aaaaaaaaa|
* result: |rrrrrrrrrrrrrrrrrrr|
* bno len
*/
ltrec.rm_blockcount += len;
if (have_gt &&
bno + len == gtrec.rm_startblock &&
(ignore_off || offset + len == gtrec.rm_offset) &&
(unsigned long)ltrec.rm_blockcount + len +
gtrec.rm_blockcount <= XFS_RMAP_LEN_MAX) {
/*
* right edge also contiguous, delete right record
* and merge into left record.
*
* ltbno ltlen gtbno gtlen
* orig: |ooooooooo| |ooooooooo|
* adding: |aaaaaaaaa|
* result: |rrrrrrrrrrrrrrrrrrrrrrrrrrrrr|
*/
ltrec.rm_blockcount += gtrec.rm_blockcount;
trace_xfs_rmap_delete(mp, cur->bc_private.a.agno,
gtrec.rm_startblock,
gtrec.rm_blockcount,
gtrec.rm_owner,
gtrec.rm_offset,
gtrec.rm_flags);
error = xfs_btree_delete(cur, &i);
if (error)
goto out_error;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, out_error);
}
/* point the cursor back to the left record and update */
error = xfs_btree_decrement(cur, 0, &have_gt);
if (error)
goto out_error;
error = xfs_rmap_update(cur, <rec);
if (error)
goto out_error;
} else if (have_gt &&
bno + len == gtrec.rm_startblock &&
(ignore_off || offset + len == gtrec.rm_offset)) {
/*
* right edge contiguous, merge into right record.
*
* gtbno gtlen
* Orig: |ooooooooo|
* adding: |aaaaaaaaa|
* Result: |rrrrrrrrrrrrrrrrrrr|
* bno len
*/
gtrec.rm_startblock = bno;
gtrec.rm_blockcount += len;
if (!ignore_off)
gtrec.rm_offset = offset;
error = xfs_rmap_update(cur, >rec);
if (error)
goto out_error;
} else {
/*
* no contiguous edge with identical owner, insert
* new record at current cursor position.
*/
cur->bc_rec.r.rm_startblock = bno;
cur->bc_rec.r.rm_blockcount = len;
cur->bc_rec.r.rm_owner = owner;
cur->bc_rec.r.rm_offset = offset;
cur->bc_rec.r.rm_flags = flags;
trace_xfs_rmap_insert(mp, cur->bc_private.a.agno, bno, len,
owner, offset, flags);
error = xfs_btree_insert(cur, &i);
if (error)
goto out_error;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, out_error);
}
trace_xfs_rmap_map_done(mp, cur->bc_private.a.agno, bno, len,
unwritten, oinfo);
out_error:
if (error)
trace_xfs_rmap_map_error(mp, cur->bc_private.a.agno,
error, _RET_IP_);
return error;
}
/*
* Find the extent in the rmap btree and remove it.
*
* The record we find should always be an exact match for the extent that we're
* looking for, since we insert them into the btree without modification.
*
* Special Case #1: when growing the filesystem, we "free" an extent when
* growing the last AG. This extent is new space and so it is not tracked as
* used space in the btree. The growfs code will pass in an owner of
* XFS_RMAP_OWN_NULL to indicate that it expected that there is no owner of this
* extent. We verify that - the extent lookup result in a record that does not
* overlap.
*
* Special Case #2: EFIs do not record the owner of the extent, so when
* recovering EFIs from the log we pass in XFS_RMAP_OWN_UNKNOWN to tell the rmap
* btree to ignore the owner (i.e. wildcard match) so we don't trigger
* corruption checks during log recovery.
*/
STATIC int
xfs_rmap_unmap(
struct xfs_btree_cur *cur,
xfs_agblock_t bno,
xfs_extlen_t len,
bool unwritten,
struct xfs_owner_info *oinfo)
{
struct xfs_mount *mp = cur->bc_mp;
struct xfs_rmap_irec ltrec;
uint64_t ltoff;
int error = 0;
int i;
uint64_t owner;
uint64_t offset;
unsigned int flags;
bool ignore_off;
xfs_owner_info_unpack(oinfo, &owner, &offset, &flags);
ignore_off = XFS_RMAP_NON_INODE_OWNER(owner) ||
(flags & XFS_RMAP_BMBT_BLOCK);
if (unwritten)
flags |= XFS_RMAP_UNWRITTEN;
trace_xfs_rmap_unmap(mp, cur->bc_private.a.agno, bno, len,
unwritten, oinfo);
/*
* We should always have a left record because there's a static record
* for the AG headers at rm_startblock == 0 created by mkfs/growfs that
* will not ever be removed from the tree.
*/
error = xfs_rmap_lookup_le(cur, bno, len, owner, offset, flags, &i);
if (error)
goto out_error;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, out_error);
error = xfs_rmap_get_rec(cur, <rec, &i);
if (error)
goto out_error;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, out_error);
trace_xfs_rmap_lookup_le_range_result(cur->bc_mp,
cur->bc_private.a.agno, ltrec.rm_startblock,
ltrec.rm_blockcount, ltrec.rm_owner,
ltrec.rm_offset, ltrec.rm_flags);
ltoff = ltrec.rm_offset;
/*
* For growfs, the incoming extent must be beyond the left record we
* just found as it is new space and won't be used by anyone. This is
* just a corruption check as we don't actually do anything with this
* extent. Note that we need to use >= instead of > because it might
* be the case that the "left" extent goes all the way to EOFS.
*/
if (owner == XFS_RMAP_OWN_NULL) {
XFS_WANT_CORRUPTED_GOTO(mp, bno >= ltrec.rm_startblock +
ltrec.rm_blockcount, out_error);
goto out_done;
}
/* Make sure the unwritten flag matches. */
XFS_WANT_CORRUPTED_GOTO(mp, (flags & XFS_RMAP_UNWRITTEN) ==
(ltrec.rm_flags & XFS_RMAP_UNWRITTEN), out_error);
/* Make sure the extent we found covers the entire freeing range. */
XFS_WANT_CORRUPTED_GOTO(mp, ltrec.rm_startblock <= bno &&
ltrec.rm_startblock + ltrec.rm_blockcount >=
bno + len, out_error);
/* Make sure the owner matches what we expect to find in the tree. */
XFS_WANT_CORRUPTED_GOTO(mp, owner == ltrec.rm_owner ||
XFS_RMAP_NON_INODE_OWNER(owner), out_error);
/* Check the offset, if necessary. */
if (!XFS_RMAP_NON_INODE_OWNER(owner)) {
if (flags & XFS_RMAP_BMBT_BLOCK) {
XFS_WANT_CORRUPTED_GOTO(mp,
ltrec.rm_flags & XFS_RMAP_BMBT_BLOCK,
out_error);
} else {
XFS_WANT_CORRUPTED_GOTO(mp,
ltrec.rm_offset <= offset, out_error);
XFS_WANT_CORRUPTED_GOTO(mp,
ltoff + ltrec.rm_blockcount >= offset + len,
out_error);
}
}
if (ltrec.rm_startblock == bno && ltrec.rm_blockcount == len) {
/* exact match, simply remove the record from rmap tree */
trace_xfs_rmap_delete(mp, cur->bc_private.a.agno,
ltrec.rm_startblock, ltrec.rm_blockcount,
ltrec.rm_owner, ltrec.rm_offset,
ltrec.rm_flags);
error = xfs_btree_delete(cur, &i);
if (error)
goto out_error;
XFS_WANT_CORRUPTED_GOTO(mp, i == 1, out_error);
} else if (ltrec.rm_startblock == bno) {
/*
* overlap left hand side of extent: move the start, trim the
* length and update the current record.
*
* ltbno ltlen
* Orig: |oooooooooooooooooooo|
* Freeing: |fffffffff|
* Result: |rrrrrrrrrr|
* bno len
*/
ltrec.rm_startblock += len;
ltrec.rm_blockcount -= len;
if (!ignore_off)
ltrec.rm_offset += len;
error = xfs_rmap_update(cur, <rec);
if (error)
goto out_error;
} else if (ltrec.rm_startblock + ltrec.rm_blockcount == bno + len) {
/*
* overlap right hand side of extent: trim the length and update
* the current record.
*
* ltbno ltlen
* Orig: |oooooooooooooooooooo|
* Freeing: |fffffffff|
* Result: |rrrrrrrrrr|
* bno len
*/
ltrec.rm_blockcount -= len;
error = xfs_rmap_update(cur, <rec);
if (error)
goto out_error;
} else {
/*
* overlap middle of extent: trim the length of the existing
* record to the length of the new left-extent size, increment
* the insertion position so we can insert a new record
* containing the remaining right-extent space.
*
* ltbno ltlen
* Orig: |oooooooooooooooooooo|
* Freeing: |fffffffff|
* Result: |rrrrr| |rrrr|
* bno len
*/
xfs_extlen_t orig_len = ltrec.rm_blockcount;
ltrec.rm_blockcount = bno - ltrec.rm_startblock;
error = xfs_rmap_update(cur, <rec);
if (error)
goto out_error;
error = xfs_btree_increment(cur, 0, &i);
if (error)
goto out_error;
cur->bc_rec.r.rm_startblock = bno + len;
cur->bc_rec.r.rm_blockcount = orig_len - len -
ltrec.rm_blockcount;
cur->bc_rec.r.rm_owner = ltrec.rm_owner;
if (ignore_off)
cur->bc_rec.r.rm_offset = 0;
else
cur->bc_rec.r.rm_offset = offset + len;
cur->bc_rec.r.rm_flags = flags;
trace_xfs_rmap_insert(mp, cur->bc_private.a.agno,
cur->bc_rec.r.rm_startblock,
cur->bc_rec.r.rm_blockcount,
cur->bc_rec.r.rm_owner,
cur->bc_rec.r.rm_offset,
cur->bc_rec.r.rm_flags);
error = xfs_btree_insert(cur, &i);
if (error)
goto out_error;
}
out_done:
trace_xfs_rmap_unmap_done(mp, cur->bc_private.a.agno, bno, len,
unwritten, oinfo);
out_error:
if (error)
trace_xfs_rmap_unmap_error(mp, cur->bc_private.a.agno,
error, _RET_IP_);
return error;
}
/*
* Allocate new inodes in the allocation group specified by agbp.
* Return 0 for success, else error code.
*/
STATIC int /* error code or 0 */
xfs_ialloc_ag_alloc(
xfs_trans_t *tp, /* transaction pointer */
xfs_buf_t *agbp, /* alloc group buffer */
int *alloc)
{
xfs_agi_t *agi; /* allocation group header */
xfs_alloc_arg_t args; /* allocation argument structure */
int blks_per_cluster; /* fs blocks per inode cluster */
xfs_btree_cur_t *cur; /* inode btree cursor */
xfs_daddr_t d; /* disk addr of buffer */
xfs_agnumber_t agno;
int error;
xfs_buf_t *fbuf; /* new free inodes' buffer */
xfs_dinode_t *free; /* new free inode structure */
int i; /* inode counter */
int j; /* block counter */
int nbufs; /* num bufs of new inodes */
xfs_agino_t newino; /* new first inode's number */
xfs_agino_t newlen; /* new number of inodes */
int ninodes; /* num inodes per buf */
xfs_agino_t thisino; /* current inode number, for loop */
int version; /* inode version number to use */
int isaligned = 0; /* inode allocation at stripe unit */
/* boundary */
unsigned int gen;
args.tp = tp;
args.mp = tp->t_mountp;
/*
* Locking will ensure that we don't have two callers in here
* at one time.
*/
newlen = XFS_IALLOC_INODES(args.mp);
if (args.mp->m_maxicount &&
args.mp->m_sb.sb_icount + newlen > args.mp->m_maxicount)
return XFS_ERROR(ENOSPC);
args.minlen = args.maxlen = XFS_IALLOC_BLOCKS(args.mp);
/*
* First try to allocate inodes contiguous with the last-allocated
* chunk of inodes. If the filesystem is striped, this will fill
* an entire stripe unit with inodes.
*/
agi = XFS_BUF_TO_AGI(agbp);
newino = be32_to_cpu(agi->agi_newino);
args.agbno = XFS_AGINO_TO_AGBNO(args.mp, newino) +
XFS_IALLOC_BLOCKS(args.mp);
if (likely(newino != NULLAGINO &&
(args.agbno < be32_to_cpu(agi->agi_length)))) {
args.fsbno = XFS_AGB_TO_FSB(args.mp,
be32_to_cpu(agi->agi_seqno), args.agbno);
args.type = XFS_ALLOCTYPE_THIS_BNO;
args.mod = args.total = args.wasdel = args.isfl =
args.userdata = args.minalignslop = 0;
args.prod = 1;
/*
* We need to take into account alignment here to ensure that
* we don't modify the free list if we fail to have an exact
* block. If we don't have an exact match, and every oher
* attempt allocation attempt fails, we'll end up cancelling
* a dirty transaction and shutting down.
*
* For an exact allocation, alignment must be 1,
* however we need to take cluster alignment into account when
* fixing up the freelist. Use the minalignslop field to
* indicate that extra blocks might be required for alignment,
* but not to use them in the actual exact allocation.
*/
args.alignment = 1;
args.minalignslop = xfs_ialloc_cluster_alignment(&args) - 1;
/* Allow space for the inode btree to split. */
args.minleft = XFS_IN_MAXLEVELS(args.mp) - 1;
if ((error = xfs_alloc_vextent(&args)))
return error;
} else
args.fsbno = NULLFSBLOCK;
if (unlikely(args.fsbno == NULLFSBLOCK)) {
/*
* Set the alignment for the allocation.
* If stripe alignment is turned on then align at stripe unit
* boundary.
* If the cluster size is smaller than a filesystem block
* then we're doing I/O for inodes in filesystem block size
* pieces, so don't need alignment anyway.
*/
isaligned = 0;
if (args.mp->m_sinoalign) {
ASSERT(!(args.mp->m_flags & XFS_MOUNT_NOALIGN));
args.alignment = args.mp->m_dalign;
isaligned = 1;
} else
args.alignment = xfs_ialloc_cluster_alignment(&args);
/*
* Need to figure out where to allocate the inode blocks.
* Ideally they should be spaced out through the a.g.
* For now, just allocate blocks up front.
*/
args.agbno = be32_to_cpu(agi->agi_root);
args.fsbno = XFS_AGB_TO_FSB(args.mp,
be32_to_cpu(agi->agi_seqno), args.agbno);
/*
* Allocate a fixed-size extent of inodes.
*/
args.type = XFS_ALLOCTYPE_NEAR_BNO;
args.mod = args.total = args.wasdel = args.isfl =
args.userdata = args.minalignslop = 0;
args.prod = 1;
/*
* Allow space for the inode btree to split.
*/
args.minleft = XFS_IN_MAXLEVELS(args.mp) - 1;
if ((error = xfs_alloc_vextent(&args)))
return error;
}
/*
* If stripe alignment is turned on, then try again with cluster
* alignment.
*/
if (isaligned && args.fsbno == NULLFSBLOCK) {
args.type = XFS_ALLOCTYPE_NEAR_BNO;
args.agbno = be32_to_cpu(agi->agi_root);
args.fsbno = XFS_AGB_TO_FSB(args.mp,
be32_to_cpu(agi->agi_seqno), args.agbno);
args.alignment = xfs_ialloc_cluster_alignment(&args);
if ((error = xfs_alloc_vextent(&args)))
return error;
}
if (args.fsbno == NULLFSBLOCK) {
*alloc = 0;
return 0;
}
ASSERT(args.len == args.minlen);
/*
* Convert the results.
*/
newino = XFS_OFFBNO_TO_AGINO(args.mp, args.agbno, 0);
/*
* Loop over the new block(s), filling in the inodes.
* For small block sizes, manipulate the inodes in buffers
* which are multiples of the blocks size.
*/
if (args.mp->m_sb.sb_blocksize >= XFS_INODE_CLUSTER_SIZE(args.mp)) {
blks_per_cluster = 1;
nbufs = (int)args.len;
ninodes = args.mp->m_sb.sb_inopblock;
} else {
blks_per_cluster = XFS_INODE_CLUSTER_SIZE(args.mp) /
args.mp->m_sb.sb_blocksize;
nbufs = (int)args.len / blks_per_cluster;
ninodes = blks_per_cluster * args.mp->m_sb.sb_inopblock;
}
/*
* Figure out what version number to use in the inodes we create.
* If the superblock version has caught up to the one that supports
* the new inode format, then use the new inode version. Otherwise
* use the old version so that old kernels will continue to be
* able to use the file system.
*/
if (xfs_sb_version_hasnlink(&args.mp->m_sb))
version = 2;
else
version = 1;
/*
* Seed the new inode cluster with a random generation number. This
* prevents short-term reuse of generation numbers if a chunk is
* freed and then immediately reallocated. We use random numbers
* rather than a linear progression to prevent the next generation
* number from being easily guessable.
*/
gen = random32();
for (j = 0; j < nbufs; j++) {
/*
* Get the block.
*/
d = XFS_AGB_TO_DADDR(args.mp, be32_to_cpu(agi->agi_seqno),
args.agbno + (j * blks_per_cluster));
fbuf = xfs_trans_get_buf(tp, args.mp->m_ddev_targp, d,
args.mp->m_bsize * blks_per_cluster,
XFS_BUF_LOCK);
ASSERT(fbuf);
ASSERT(!XFS_BUF_GETERROR(fbuf));
/*
* Initialize all inodes in this buffer and then log them.
*
* XXX: It would be much better if we had just one transaction to
* log a whole cluster of inodes instead of all the indivdual
* transactions causing a lot of log traffic.
*/
xfs_biozero(fbuf, 0, ninodes << args.mp->m_sb.sb_inodelog);
for (i = 0; i < ninodes; i++) {
int ioffset = i << args.mp->m_sb.sb_inodelog;
uint isize = sizeof(struct xfs_dinode);
free = xfs_make_iptr(args.mp, fbuf, i);
free->di_magic = cpu_to_be16(XFS_DINODE_MAGIC);
free->di_version = version;
free->di_gen = cpu_to_be32(gen);
free->di_next_unlinked = cpu_to_be32(NULLAGINO);
xfs_trans_log_buf(tp, fbuf, ioffset, ioffset + isize - 1);
}
xfs_trans_inode_alloc_buf(tp, fbuf);
}
be32_add_cpu(&agi->agi_count, newlen);
be32_add_cpu(&agi->agi_freecount, newlen);
agno = be32_to_cpu(agi->agi_seqno);
down_read(&args.mp->m_peraglock);
args.mp->m_perag[agno].pagi_freecount += newlen;
up_read(&args.mp->m_peraglock);
agi->agi_newino = cpu_to_be32(newino);
/*
* Insert records describing the new inode chunk into the btree.
*/
cur = xfs_inobt_init_cursor(args.mp, tp, agbp, agno);
for (thisino = newino;
thisino < newino + newlen;
thisino += XFS_INODES_PER_CHUNK) {
if ((error = xfs_inobt_lookup_eq(cur, thisino,
XFS_INODES_PER_CHUNK, XFS_INOBT_ALL_FREE, &i))) {
xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
return error;
}
ASSERT(i == 0);
if ((error = xfs_btree_insert(cur, &i))) {
xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
return error;
}
ASSERT(i == 1);
}
xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR);
/*
* Log allocation group header fields
*/
xfs_ialloc_log_agi(tp, agbp,
XFS_AGI_COUNT | XFS_AGI_FREECOUNT | XFS_AGI_NEWINO);
/*
* Modify/log superblock values for inode count and inode free count.
*/
xfs_trans_mod_sb(tp, XFS_TRANS_SB_ICOUNT, (long)newlen);
xfs_trans_mod_sb(tp, XFS_TRANS_SB_IFREE, (long)newlen);
*alloc = 1;
return 0;
}
