/*
* Select an allocation group to look for a free inode in, based on the parent
* inode and then mode. Return the allocation group buffer.
*/
STATIC xfs_buf_t * /* allocation group buffer */
xfs_ialloc_ag_select(
xfs_trans_t *tp, /* transaction pointer */
xfs_ino_t parent, /* parent directory inode number */
mode_t mode, /* bits set to indicate file type */
int okalloc) /* ok to allocate more space */
{
xfs_buf_t *agbp; /* allocation group header buffer */
xfs_agnumber_t agcount; /* number of ag's in the filesystem */
xfs_agnumber_t agno; /* current ag number */
int flags; /* alloc buffer locking flags */
xfs_extlen_t ineed; /* blocks needed for inode allocation */
xfs_extlen_t longest = 0; /* longest extent available */
xfs_mount_t *mp; /* mount point structure */
int needspace; /* file mode implies space allocated */
xfs_perag_t *pag; /* per allocation group data */
xfs_agnumber_t pagno; /* parent (starting) ag number */
/*
* Files of these types need at least one block if length > 0
* (and they won't fit in the inode, but that's hard to figure out).
*/
needspace = S_ISDIR(mode) || S_ISREG(mode) || S_ISLNK(mode);
mp = tp->t_mountp;
agcount = mp->m_maxagi;
if (S_ISDIR(mode))
pagno = xfs_ialloc_next_ag(mp);
else {
pagno = XFS_INO_TO_AGNO(mp, parent);
if (pagno >= agcount)
pagno = 0;
}
ASSERT(pagno < agcount);
/*
* Loop through allocation groups, looking for one with a little
* free space in it. Note we don't look for free inodes, exactly.
* Instead, we include whether there is a need to allocate inodes
* to mean that blocks must be allocated for them,
* if none are currently free.
*/
agno = pagno;
flags = XFS_ALLOC_FLAG_TRYLOCK;
down_read(&mp->m_peraglock);
for (;;) {
pag = &mp->m_perag[agno];
if (!pag->pagi_init) {
if (xfs_ialloc_read_agi(mp, tp, agno, &agbp)) {
agbp = NULL;
goto nextag;
}
} else
agbp = NULL;
if (!pag->pagi_inodeok) {
xfs_ialloc_next_ag(mp);
goto unlock_nextag;
}
/*
* Is there enough free space for the file plus a block
* of inodes (if we need to allocate some)?
*/
ineed = pag->pagi_freecount ? 0 : XFS_IALLOC_BLOCKS(mp);
if (ineed && !pag->pagf_init) {
if (agbp == NULL &&
xfs_ialloc_read_agi(mp, tp, agno, &agbp)) {
agbp = NULL;
goto nextag;
}
(void)xfs_alloc_pagf_init(mp, tp, agno, flags);
}
if (!ineed || pag->pagf_init) {
if (ineed && !(longest = pag->pagf_longest))
longest = pag->pagf_flcount > 0;
if (!ineed ||
(pag->pagf_freeblks >= needspace + ineed &&
longest >= ineed &&
okalloc)) {
if (agbp == NULL &&
xfs_ialloc_read_agi(mp, tp, agno, &agbp)) {
agbp = NULL;
goto nextag;
}
up_read(&mp->m_peraglock);
return agbp;
}
}
unlock_nextag:
if (agbp)
xfs_trans_brelse(tp, agbp);
nextag:
/*
* No point in iterating over the rest, if we're shutting
* down.
*/
if (XFS_FORCED_SHUTDOWN(mp)) {
up_read(&mp->m_peraglock);
return NULL;
}
agno++;
if (agno >= agcount)
agno = 0;
if (agno == pagno) {
if (flags == 0) {
up_read(&mp->m_peraglock);
return NULL;
}
flags = 0;
}
}
}
/*
* Allocate an inode on disk.
* Mode is used to tell whether the new inode will need space, and whether
* it is a directory.
*
* The arguments IO_agbp and alloc_done are defined to work within
* the constraint of one allocation per transaction.
* xfs_dialloc() is designed to be called twice if it has to do an
* allocation to make more free inodes. On the first call,
* IO_agbp should be set to NULL. If an inode is available,
* i.e., xfs_dialloc() did not need to do an allocation, an inode
* number is returned. In this case, IO_agbp would be set to the
* current ag_buf and alloc_done set to false.
* If an allocation needed to be done, xfs_dialloc would return
* the current ag_buf in IO_agbp and set alloc_done to true.
* The caller should then commit the current transaction, allocate a new
* transaction, and call xfs_dialloc() again, passing in the previous
* value of IO_agbp. IO_agbp should be held across the transactions.
* Since the agbp is locked across the two calls, the second call is
* guaranteed to have a free inode available.
*
* Once we successfully pick an inode its number is returned and the
* on-disk data structures are updated. The inode itself is not read
* in, since doing so would break ordering constraints with xfs_reclaim.
*/
int
xfs_dialloc(
xfs_trans_t *tp, /* transaction pointer */
xfs_ino_t parent, /* parent inode (directory) */
mode_t mode, /* mode bits for new inode */
int okalloc, /* ok to allocate more space */
xfs_buf_t **IO_agbp, /* in/out ag header's buffer */
boolean_t *alloc_done, /* true if we needed to replenish
inode freelist */
xfs_ino_t *inop) /* inode number allocated */
{
xfs_agnumber_t agcount; /* number of allocation groups */
xfs_buf_t *agbp; /* allocation group header's buffer */
xfs_agnumber_t agno; /* allocation group number */
xfs_agi_t *agi; /* allocation group header structure */
xfs_btree_cur_t *cur; /* inode allocation btree cursor */
int error; /* error return value */
int i; /* result code */
int ialloced; /* inode allocation status */
int noroom = 0; /* no space for inode blk allocation */
xfs_ino_t ino; /* fs-relative inode to be returned */
/* REFERENCED */
int j; /* result code */
xfs_mount_t *mp; /* file system mount structure */
int offset; /* index of inode in chunk */
xfs_agino_t pagino; /* parent's a.g. relative inode # */
xfs_agnumber_t pagno; /* parent's allocation group number */
xfs_inobt_rec_incore_t rec; /* inode allocation record */
xfs_agnumber_t tagno; /* testing allocation group number */
xfs_btree_cur_t *tcur; /* temp cursor */
xfs_inobt_rec_incore_t trec; /* temp inode allocation record */
if (*IO_agbp == NULL) {
/*
* We do not have an agbp, so select an initial allocation
* group for inode allocation.
*/
agbp = xfs_ialloc_ag_select(tp, parent, mode, okalloc);
/*
* Couldn't find an allocation group satisfying the
* criteria, give up.
*/
if (!agbp) {
*inop = NULLFSINO;
return 0;
}
agi = XFS_BUF_TO_AGI(agbp);
ASSERT(be32_to_cpu(agi->agi_magicnum) == XFS_AGI_MAGIC);
} else {
/*
* Continue where we left off before. In this case, we
* know that the allocation group has free inodes.
*/
agbp = *IO_agbp;
agi = XFS_BUF_TO_AGI(agbp);
ASSERT(be32_to_cpu(agi->agi_magicnum) == XFS_AGI_MAGIC);
ASSERT(be32_to_cpu(agi->agi_freecount) > 0);
}
mp = tp->t_mountp;
agcount = mp->m_sb.sb_agcount;
agno = be32_to_cpu(agi->agi_seqno);
tagno = agno;
pagno = XFS_INO_TO_AGNO(mp, parent);
pagino = XFS_INO_TO_AGINO(mp, parent);
/*
* If we have already hit the ceiling of inode blocks then clear
* okalloc so we scan all available agi structures for a free
* inode.
*/
if (mp->m_maxicount &&
mp->m_sb.sb_icount + XFS_IALLOC_INODES(mp) > mp->m_maxicount) {
noroom = 1;
okalloc = 0;
}
/*
* Loop until we find an allocation group that either has free inodes
* or in which we can allocate some inodes. Iterate through the
* allocation groups upward, wrapping at the end.
*/
*alloc_done = B_FALSE;
while (!agi->agi_freecount) {
/*
* Don't do anything if we're not supposed to allocate
* any blocks, just go on to the next ag.
*/
if (okalloc) {
/*
* Try to allocate some new inodes in the allocation
* group.
*/
if ((error = xfs_ialloc_ag_alloc(tp, agbp, &ialloced))) {
xfs_trans_brelse(tp, agbp);
if (error == ENOSPC) {
*inop = NULLFSINO;
return 0;
} else
return error;
}
if (ialloced) {
/*
* We successfully allocated some inodes, return
* the current context to the caller so that it
* can commit the current transaction and call
* us again where we left off.
*/
ASSERT(be32_to_cpu(agi->agi_freecount) > 0);
*alloc_done = B_TRUE;
*IO_agbp = agbp;
*inop = NULLFSINO;
return 0;
}
}
/*
* If it failed, give up on this ag.
*/
xfs_trans_brelse(tp, agbp);
/*
* Go on to the next ag: get its ag header.
*/
nextag:
if (++tagno == agcount)
tagno = 0;
if (tagno == agno) {
*inop = NULLFSINO;
return noroom ? ENOSPC : 0;
}
down_read(&mp->m_peraglock);
if (mp->m_perag[tagno].pagi_inodeok == 0) {
up_read(&mp->m_peraglock);
goto nextag;
}
error = xfs_ialloc_read_agi(mp, tp, tagno, &agbp);
up_read(&mp->m_peraglock);
if (error)
goto nextag;
agi = XFS_BUF_TO_AGI(agbp);
ASSERT(be32_to_cpu(agi->agi_magicnum) == XFS_AGI_MAGIC);
}
/*
* Here with an allocation group that has a free inode.
* Reset agno since we may have chosen a new ag in the
* loop above.
*/
agno = tagno;
*IO_agbp = NULL;
cur = xfs_btree_init_cursor(mp, tp, agbp, be32_to_cpu(agi->agi_seqno),
XFS_BTNUM_INO, (xfs_inode_t *)0, 0);
/*
* If pagino is 0 (this is the root inode allocation) use newino.
* This must work because we've just allocated some.
*/
if (!pagino)
pagino = be32_to_cpu(agi->agi_newino);
#ifdef DEBUG
if (cur->bc_nlevels == 1) {
int freecount = 0;
if ((error = xfs_inobt_lookup_ge(cur, 0, 0, 0, &i)))
goto error0;
XFS_WANT_CORRUPTED_GOTO(i == 1, error0);
do {
if ((error = xfs_inobt_get_rec(cur, &rec.ir_startino,
&rec.ir_freecount, &rec.ir_free, &i)))
goto error0;
XFS_WANT_CORRUPTED_GOTO(i == 1, error0);
freecount += rec.ir_freecount;
if ((error = xfs_inobt_increment(cur, 0, &i)))
goto error0;
} while (i == 1);
ASSERT(freecount == be32_to_cpu(agi->agi_freecount) ||
XFS_FORCED_SHUTDOWN(mp));
}
#endif
/*
* If in the same a.g. as the parent, try to get near the parent.
*/
if (pagno == agno) {
if ((error = xfs_inobt_lookup_le(cur, pagino, 0, 0, &i)))
goto error0;
if (i != 0 &&
(error = xfs_inobt_get_rec(cur, &rec.ir_startino,
&rec.ir_freecount, &rec.ir_free, &j)) == 0 &&
j == 1 &&
rec.ir_freecount > 0) {
/*
* Found a free inode in the same chunk
* as parent, done.
*/
}
/*
* In the same a.g. as parent, but parent's chunk is full.
*/
else {
int doneleft; /* done, to the left */
int doneright; /* done, to the right */
if (error)
goto error0;
ASSERT(i == 1);
ASSERT(j == 1);
/*
* Duplicate the cursor, search left & right
* simultaneously.
*/
if ((error = xfs_btree_dup_cursor(cur, &tcur)))
goto error0;
/*
* Search left with tcur, back up 1 record.
*/
if ((error = xfs_inobt_decrement(tcur, 0, &i)))
goto error1;
doneleft = !i;
if (!doneleft) {
if ((error = xfs_inobt_get_rec(tcur,
&trec.ir_startino,
&trec.ir_freecount,
&trec.ir_free, &i)))
goto error1;
XFS_WANT_CORRUPTED_GOTO(i == 1, error1);
}
/*
* Search right with cur, go forward 1 record.
*/
if ((error = xfs_inobt_increment(cur, 0, &i)))
goto error1;
doneright = !i;
if (!doneright) {
if ((error = xfs_inobt_get_rec(cur,
&rec.ir_startino,
&rec.ir_freecount,
&rec.ir_free, &i)))
goto error1;
XFS_WANT_CORRUPTED_GOTO(i == 1, error1);
}
/*
* Loop until we find the closest inode chunk
* with a free one.
*/
while (!doneleft || !doneright) {
int useleft; /* using left inode
chunk this time */
/*
* Figure out which block is closer,
* if both are valid.
*/
if (!doneleft && !doneright)
useleft =
pagino -
(trec.ir_startino +
XFS_INODES_PER_CHUNK - 1) <
rec.ir_startino - pagino;
else
useleft = !doneleft;
/*
* If checking the left, does it have
* free inodes?
*/
if (useleft && trec.ir_freecount) {
/*
* Yes, set it up as the chunk to use.
*/
rec = trec;
xfs_btree_del_cursor(cur,
XFS_BTREE_NOERROR);
cur = tcur;
break;
}
/*
* If checking the right, does it have
* free inodes?
*/
if (!useleft && rec.ir_freecount) {
/*
* Yes, it's already set up.
*/
xfs_btree_del_cursor(tcur,
XFS_BTREE_NOERROR);
break;
}
/*
* If used the left, get another one
* further left.
*/
if (useleft) {
if ((error = xfs_inobt_decrement(tcur, 0,
&i)))
goto error1;
doneleft = !i;
if (!doneleft) {
if ((error = xfs_inobt_get_rec(
tcur,
&trec.ir_startino,
&trec.ir_freecount,
&trec.ir_free, &i)))
goto error1;
XFS_WANT_CORRUPTED_GOTO(i == 1,
error1);
}
}
/*
* If used the right, get another one
* further right.
*/
else {
if ((error = xfs_inobt_increment(cur, 0,
&i)))
goto error1;
doneright = !i;
if (!doneright) {
if ((error = xfs_inobt_get_rec(
cur,
&rec.ir_startino,
&rec.ir_freecount,
&rec.ir_free, &i)))
goto error1;
XFS_WANT_CORRUPTED_GOTO(i == 1,
error1);
}
}
}
ASSERT(!doneleft || !doneright);
}
}
/*
* In a different a.g. from the parent.
* See if the most recently allocated block has any free.
*/
else if (be32_to_cpu(agi->agi_newino) != NULLAGINO) {
if ((error = xfs_inobt_lookup_eq(cur,
be32_to_cpu(agi->agi_newino), 0, 0, &i)))
goto error0;
if (i == 1 &&
(error = xfs_inobt_get_rec(cur, &rec.ir_startino,
&rec.ir_freecount, &rec.ir_free, &j)) == 0 &&
j == 1 &&
rec.ir_freecount > 0) {
/*
* The last chunk allocated in the group still has
* a free inode.
*/
}
/*
* None left in the last group, search the whole a.g.
*/
else {
if (error)
goto error0;
if ((error = xfs_inobt_lookup_ge(cur, 0, 0, 0, &i)))
goto error0;
ASSERT(i == 1);
for (;;) {
if ((error = xfs_inobt_get_rec(cur,
&rec.ir_startino,
&rec.ir_freecount, &rec.ir_free,
&i)))
goto error0;
XFS_WANT_CORRUPTED_GOTO(i == 1, error0);
if (rec.ir_freecount > 0)
break;
if ((error = xfs_inobt_increment(cur, 0, &i)))
goto error0;
XFS_WANT_CORRUPTED_GOTO(i == 1, error0);
}
}
}
offset = XFS_IALLOC_FIND_FREE(&rec.ir_free);
ASSERT(offset >= 0);
ASSERT(offset < XFS_INODES_PER_CHUNK);
ASSERT((XFS_AGINO_TO_OFFSET(mp, rec.ir_startino) %
XFS_INODES_PER_CHUNK) == 0);
ino = XFS_AGINO_TO_INO(mp, agno, rec.ir_startino + offset);
XFS_INOBT_CLR_FREE(&rec, offset);
rec.ir_freecount--;
if ((error = xfs_inobt_update(cur, rec.ir_startino, rec.ir_freecount,
rec.ir_free)))
goto error0;
be32_add_cpu(&agi->agi_freecount, -1);
xfs_ialloc_log_agi(tp, agbp, XFS_AGI_FREECOUNT);
down_read(&mp->m_peraglock);
mp->m_perag[tagno].pagi_freecount--;
up_read(&mp->m_peraglock);
#ifdef DEBUG
if (cur->bc_nlevels == 1) {
int freecount = 0;
if ((error = xfs_inobt_lookup_ge(cur, 0, 0, 0, &i)))
goto error0;
do {
if ((error = xfs_inobt_get_rec(cur, &rec.ir_startino,
&rec.ir_freecount, &rec.ir_free, &i)))
goto error0;
XFS_WANT_CORRUPTED_GOTO(i == 1, error0);
freecount += rec.ir_freecount;
if ((error = xfs_inobt_increment(cur, 0, &i)))
goto error0;
} while (i == 1);
ASSERT(freecount == be32_to_cpu(agi->agi_freecount) ||
XFS_FORCED_SHUTDOWN(mp));
}
#endif
xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR);
xfs_trans_mod_sb(tp, XFS_TRANS_SB_IFREE, -1);
*inop = ino;
return 0;
error1:
xfs_btree_del_cursor(tcur, XFS_BTREE_ERROR);
error0:
xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
return error;
}
/*
* Allocate an inode on disk.
* Mode is used to tell whether the new inode will need space, and whether
* it is a directory.
*
* The arguments IO_agbp and alloc_done are defined to work within
* the constraint of one allocation per transaction.
* xfs_dialloc() is designed to be called twice if it has to do an
* allocation to make more free inodes. On the first call,
* IO_agbp should be set to NULL. If an inode is available,
* i.e., xfs_dialloc() did not need to do an allocation, an inode
* number is returned. In this case, IO_agbp would be set to the
* current ag_buf and alloc_done set to false.
* If an allocation needed to be done, xfs_dialloc would return
* the current ag_buf in IO_agbp and set alloc_done to true.
* The caller should then commit the current transaction, allocate a new
* transaction, and call xfs_dialloc() again, passing in the previous
* value of IO_agbp. IO_agbp should be held across the transactions.
* Since the agbp is locked across the two calls, the second call is
* guaranteed to have a free inode available.
*
* Once we successfully pick an inode its number is returned and the
* on-disk data structures are updated. The inode itself is not read
* in, since doing so would break ordering constraints with xfs_reclaim.
*/
int
xfs_dialloc(
xfs_trans_t *tp, /* transaction pointer */
xfs_ino_t parent, /* parent inode (directory) */
umode_t mode, /* mode bits for new inode */
int okalloc, /* ok to allocate more space */
xfs_buf_t **IO_agbp, /* in/out ag header's buffer */
boolean_t *alloc_done, /* true if we needed to replenish
inode freelist */
xfs_ino_t *inop) /* inode number allocated */
{
xfs_agnumber_t agcount; /* number of allocation groups */
xfs_buf_t *agbp; /* allocation group header's buffer */
xfs_agnumber_t agno; /* allocation group number */
xfs_agi_t *agi; /* allocation group header structure */
xfs_btree_cur_t *cur; /* inode allocation btree cursor */
int error; /* error return value */
int i; /* result code */
int ialloced; /* inode allocation status */
int noroom = 0; /* no space for inode blk allocation */
xfs_ino_t ino; /* fs-relative inode to be returned */
/* REFERENCED */
int j; /* result code */
xfs_mount_t *mp; /* file system mount structure */
int offset; /* index of inode in chunk */
xfs_agino_t pagino; /* parent's AG relative inode # */
xfs_agnumber_t pagno; /* parent's AG number */
xfs_inobt_rec_incore_t rec; /* inode allocation record */
xfs_agnumber_t tagno; /* testing allocation group number */
xfs_btree_cur_t *tcur; /* temp cursor */
xfs_inobt_rec_incore_t trec; /* temp inode allocation record */
struct xfs_perag *pag;
if (*IO_agbp == NULL) {
/*
* We do not have an agbp, so select an initial allocation
* group for inode allocation.
*/
agbp = xfs_ialloc_ag_select(tp, parent, mode, okalloc);
/*
* Couldn't find an allocation group satisfying the
* criteria, give up.
*/
if (!agbp) {
*inop = NULLFSINO;
return 0;
}
agi = XFS_BUF_TO_AGI(agbp);
ASSERT(agi->agi_magicnum == cpu_to_be32(XFS_AGI_MAGIC));
} else {
/*
* Continue where we left off before. In this case, we
* know that the allocation group has free inodes.
*/
agbp = *IO_agbp;
agi = XFS_BUF_TO_AGI(agbp);
ASSERT(agi->agi_magicnum == cpu_to_be32(XFS_AGI_MAGIC));
ASSERT(be32_to_cpu(agi->agi_freecount) > 0);
}
mp = tp->t_mountp;
agcount = mp->m_sb.sb_agcount;
agno = be32_to_cpu(agi->agi_seqno);
tagno = agno;
pagno = XFS_INO_TO_AGNO(mp, parent);
pagino = XFS_INO_TO_AGINO(mp, parent);
/*
* If we have already hit the ceiling of inode blocks then clear
* okalloc so we scan all available agi structures for a free
* inode.
*/
if (mp->m_maxicount &&
mp->m_sb.sb_icount + XFS_IALLOC_INODES(mp) > mp->m_maxicount) {
noroom = 1;
okalloc = 0;
}
/*
* Loop until we find an allocation group that either has free inodes
* or in which we can allocate some inodes. Iterate through the
* allocation groups upward, wrapping at the end.
*/
*alloc_done = B_FALSE;
while (!agi->agi_freecount) {
/*
* Don't do anything if we're not supposed to allocate
* any blocks, just go on to the next ag.
*/
if (okalloc) {
/*
* Try to allocate some new inodes in the allocation
* group.
*/
if ((error = xfs_ialloc_ag_alloc(tp, agbp, &ialloced))) {
xfs_trans_brelse(tp, agbp);
if (error == ENOSPC) {
*inop = NULLFSINO;
return 0;
} else
return error;
}
if (ialloced) {
/*
* We successfully allocated some inodes, return
* the current context to the caller so that it
* can commit the current transaction and call
* us again where we left off.
*/
ASSERT(be32_to_cpu(agi->agi_freecount) > 0);
*alloc_done = B_TRUE;
*IO_agbp = agbp;
*inop = NULLFSINO;
return 0;
}
}
/*
* If it failed, give up on this ag.
*/
xfs_trans_brelse(tp, agbp);
/*
* Go on to the next ag: get its ag header.
*/
nextag:
if (++tagno == agcount)
tagno = 0;
if (tagno == agno) {
*inop = NULLFSINO;
return noroom ? ENOSPC : 0;
}
pag = xfs_perag_get(mp, tagno);
if (pag->pagi_inodeok == 0) {
xfs_perag_put(pag);
goto nextag;
}
error = xfs_ialloc_read_agi(mp, tp, tagno, &agbp);
xfs_perag_put(pag);
if (error)
goto nextag;
agi = XFS_BUF_TO_AGI(agbp);
ASSERT(agi->agi_magicnum == cpu_to_be32(XFS_AGI_MAGIC));
}
/*
* Here with an allocation group that has a free inode.
* Reset agno since we may have chosen a new ag in the
* loop above.
*/
agno = tagno;
*IO_agbp = NULL;
pag = xfs_perag_get(mp, agno);
restart_pagno:
cur = xfs_inobt_init_cursor(mp, tp, agbp, be32_to_cpu(agi->agi_seqno));
/*
* If pagino is 0 (this is the root inode allocation) use newino.
* This must work because we've just allocated some.
*/
if (!pagino)
pagino = be32_to_cpu(agi->agi_newino);
error = xfs_check_agi_freecount(cur, agi);
if (error)
goto error0;
/*
* If in the same AG as the parent, try to get near the parent.
*/
if (pagno == agno) {
int doneleft; /* done, to the left */
int doneright; /* done, to the right */
int searchdistance = 10;
error = xfs_inobt_lookup(cur, pagino, XFS_LOOKUP_LE, &i);
if (error)
goto error0;
XFS_WANT_CORRUPTED_GOTO(i == 1, error0);
error = xfs_inobt_get_rec(cur, &rec, &j);
if (error)
goto error0;
XFS_WANT_CORRUPTED_GOTO(i == 1, error0);
if (rec.ir_freecount > 0) {
/*
* Found a free inode in the same chunk
* as the parent, done.
*/
goto alloc_inode;
}
/*
* In the same AG as parent, but parent's chunk is full.
*/
/* duplicate the cursor, search left & right simultaneously */
error = xfs_btree_dup_cursor(cur, &tcur);
if (error)
goto error0;
/*
* Skip to last blocks looked up if same parent inode.
*/
if (pagino != NULLAGINO &&
pag->pagl_pagino == pagino &&
pag->pagl_leftrec != NULLAGINO &&
pag->pagl_rightrec != NULLAGINO) {
error = xfs_ialloc_get_rec(tcur, pag->pagl_leftrec,
&trec, &doneleft, 1);
if (error)
goto error1;
error = xfs_ialloc_get_rec(cur, pag->pagl_rightrec,
&rec, &doneright, 0);
if (error)
goto error1;
} else {
/* search left with tcur, back up 1 record */
error = xfs_ialloc_next_rec(tcur, &trec, &doneleft, 1);
if (error)
goto error1;
/* search right with cur, go forward 1 record. */
error = xfs_ialloc_next_rec(cur, &rec, &doneright, 0);
if (error)
goto error1;
}
/*
* Loop until we find an inode chunk with a free inode.
*/
while (!doneleft || !doneright) {
int useleft; /* using left inode chunk this time */
if (!--searchdistance) {
/*
* Not in range - save last search
* location and allocate a new inode
*/
xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR);
pag->pagl_leftrec = trec.ir_startino;
pag->pagl_rightrec = rec.ir_startino;
pag->pagl_pagino = pagino;
goto newino;
}
/* figure out the closer block if both are valid. */
if (!doneleft && !doneright) {
useleft = pagino -
(trec.ir_startino + XFS_INODES_PER_CHUNK - 1) <
rec.ir_startino - pagino;
} else {
useleft = !doneleft;
}
/* free inodes to the left? */
if (useleft && trec.ir_freecount) {
rec = trec;
xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR);
cur = tcur;
pag->pagl_leftrec = trec.ir_startino;
pag->pagl_rightrec = rec.ir_startino;
pag->pagl_pagino = pagino;
goto alloc_inode;
}
/* free inodes to the right? */
if (!useleft && rec.ir_freecount) {
xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR);
pag->pagl_leftrec = trec.ir_startino;
pag->pagl_rightrec = rec.ir_startino;
pag->pagl_pagino = pagino;
goto alloc_inode;
}
/* get next record to check */
if (useleft) {
error = xfs_ialloc_next_rec(tcur, &trec,
&doneleft, 1);
} else {
error = xfs_ialloc_next_rec(cur, &rec,
&doneright, 0);
}
if (error)
goto error1;
}
/*
* We've reached the end of the btree. because
* we are only searching a small chunk of the
* btree each search, there is obviously free
* inodes closer to the parent inode than we
* are now. restart the search again.
*/
pag->pagl_pagino = NULLAGINO;
pag->pagl_leftrec = NULLAGINO;
pag->pagl_rightrec = NULLAGINO;
xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR);
xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR);
goto restart_pagno;
}
/*
* In a different AG from the parent.
* See if the most recently allocated block has any free.
*/
newino:
if (agi->agi_newino != cpu_to_be32(NULLAGINO)) {
error = xfs_inobt_lookup(cur, be32_to_cpu(agi->agi_newino),
XFS_LOOKUP_EQ, &i);
if (error)
goto error0;
if (i == 1) {
error = xfs_inobt_get_rec(cur, &rec, &j);
if (error)
goto error0;
if (j == 1 && rec.ir_freecount > 0) {
/*
* The last chunk allocated in the group
* still has a free inode.
*/
goto alloc_inode;
}
}
}
/*
* None left in the last group, search the whole AG
*/
error = xfs_inobt_lookup(cur, 0, XFS_LOOKUP_GE, &i);
if (error)
goto error0;
XFS_WANT_CORRUPTED_GOTO(i == 1, error0);
for (;;) {
error = xfs_inobt_get_rec(cur, &rec, &i);
if (error)
goto error0;
XFS_WANT_CORRUPTED_GOTO(i == 1, error0);
if (rec.ir_freecount > 0)
break;
error = xfs_btree_increment(cur, 0, &i);
if (error)
goto error0;
XFS_WANT_CORRUPTED_GOTO(i == 1, error0);
}
alloc_inode:
offset = xfs_ialloc_find_free(&rec.ir_free);
ASSERT(offset >= 0);
ASSERT(offset < XFS_INODES_PER_CHUNK);
ASSERT((XFS_AGINO_TO_OFFSET(mp, rec.ir_startino) %
XFS_INODES_PER_CHUNK) == 0);
ino = XFS_AGINO_TO_INO(mp, agno, rec.ir_startino + offset);
rec.ir_free &= ~XFS_INOBT_MASK(offset);
rec.ir_freecount--;
error = xfs_inobt_update(cur, &rec);
if (error)
goto error0;
be32_add_cpu(&agi->agi_freecount, -1);
xfs_ialloc_log_agi(tp, agbp, XFS_AGI_FREECOUNT);
pag->pagi_freecount--;
error = xfs_check_agi_freecount(cur, agi);
if (error)
goto error0;
xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR);
xfs_trans_mod_sb(tp, XFS_TRANS_SB_IFREE, -1);
xfs_perag_put(pag);
*inop = ino;
return 0;
error1:
xfs_btree_del_cursor(tcur, XFS_BTREE_ERROR);
error0:
xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
xfs_perag_put(pag);
return error;
}
/*
* Return inode number table for the filesystem.
*/
int /* error status */
xfs_inumbers(
struct xfs_mount *mp,/* mount point for filesystem */
xfs_ino_t *lastino,/* last inode returned */
int *count,/* size of buffer/count returned */
void __user *ubuffer,/* buffer with inode descriptions */
inumbers_fmt_pf formatter)
{
xfs_agnumber_t agno = XFS_INO_TO_AGNO(mp, *lastino);
xfs_agino_t agino = XFS_INO_TO_AGINO(mp, *lastino);
struct xfs_btree_cur *cur = NULL;
struct xfs_buf *agbp = NULL;
struct xfs_inogrp *buffer;
int bcount;
int left = *count;
int bufidx = 0;
int error = 0;
*count = 0;
if (agno >= mp->m_sb.sb_agcount ||
*lastino != XFS_AGINO_TO_INO(mp, agno, agino))
return error;
bcount = MIN(left, (int)(PAGE_SIZE / sizeof(*buffer)));
buffer = kmem_alloc(bcount * sizeof(*buffer), KM_SLEEP);
do {
struct xfs_inobt_rec_incore r;
int stat;
if (!agbp) {
error = xfs_ialloc_read_agi(mp, NULL, agno, &agbp);
if (error)
break;
cur = xfs_inobt_init_cursor(mp, NULL, agbp, agno,
XFS_BTNUM_INO);
error = xfs_inobt_lookup(cur, agino, XFS_LOOKUP_GE,
&stat);
if (error)
break;
if (!stat)
goto next_ag;
}
error = xfs_inobt_get_rec(cur, &r, &stat);
if (error)
break;
if (!stat)
goto next_ag;
agino = r.ir_startino + XFS_INODES_PER_CHUNK - 1;
buffer[bufidx].xi_startino =
XFS_AGINO_TO_INO(mp, agno, r.ir_startino);
buffer[bufidx].xi_alloccount = r.ir_count - r.ir_freecount;
buffer[bufidx].xi_allocmask = ~r.ir_free;
if (++bufidx == bcount) {
long written;
error = formatter(ubuffer, buffer, bufidx, &written);
if (error)
break;
ubuffer += written;
*count += bufidx;
bufidx = 0;
}
if (!--left)
break;
error = xfs_btree_increment(cur, 0, &stat);
if (error)
break;
if (stat)
continue;
next_ag:
xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
cur = NULL;
xfs_buf_relse(agbp);
agbp = NULL;
agino = 0;
agno++;
} while (agno < mp->m_sb.sb_agcount);
if (!error) {
if (bufidx) {
long written;
error = formatter(ubuffer, buffer, bufidx, &written);
if (!error)
*count += bufidx;
}
*lastino = XFS_AGINO_TO_INO(mp, agno, agino);
}
kmem_free(buffer);
if (cur)
xfs_btree_del_cursor(cur, (error ? XFS_BTREE_ERROR :
XFS_BTREE_NOERROR));
if (agbp)
xfs_buf_relse(agbp);
return error;
}
/*
* Return stat information in bulk (by-inode) for the filesystem.
*/
int /* error status */
xfs_bulkstat(
xfs_mount_t *mp, /* mount point for filesystem */
xfs_ino_t *lastinop, /* last inode returned */
int *ubcountp, /* size of buffer/count returned */
bulkstat_one_pf formatter, /* func that'd fill a single buf */
size_t statstruct_size, /* sizeof struct filling */
char __user *ubuffer, /* buffer with inode stats */
int *done) /* 1 if there are more stats to get */
{
xfs_buf_t *agbp; /* agi header buffer */
xfs_agino_t agino; /* inode # in allocation group */
xfs_agnumber_t agno; /* allocation group number */
xfs_btree_cur_t *cur; /* btree cursor for ialloc btree */
size_t irbsize; /* size of irec buffer in bytes */
xfs_inobt_rec_incore_t *irbuf; /* start of irec buffer */
int nirbuf; /* size of irbuf */
int ubcount; /* size of user's buffer */
struct xfs_bulkstat_agichunk ac;
int error = 0;
/*
* Get the last inode value, see if there's nothing to do.
*/
agno = XFS_INO_TO_AGNO(mp, *lastinop);
agino = XFS_INO_TO_AGINO(mp, *lastinop);
if (agno >= mp->m_sb.sb_agcount ||
*lastinop != XFS_AGINO_TO_INO(mp, agno, agino)) {
*done = 1;
*ubcountp = 0;
return 0;
}
ubcount = *ubcountp; /* statstruct's */
ac.ac_ubuffer = &ubuffer;
ac.ac_ubleft = ubcount * statstruct_size; /* bytes */;
ac.ac_ubelem = 0;
*ubcountp = 0;
*done = 0;
irbuf = kmem_zalloc_greedy(&irbsize, PAGE_SIZE, PAGE_SIZE * 4);
if (!irbuf)
return -ENOMEM;
nirbuf = irbsize / sizeof(*irbuf);
/*
* Loop over the allocation groups, starting from the last
* inode returned; 0 means start of the allocation group.
*/
while (agno < mp->m_sb.sb_agcount) {
struct xfs_inobt_rec_incore *irbp = irbuf;
struct xfs_inobt_rec_incore *irbufend = irbuf + nirbuf;
bool end_of_ag = false;
int icount = 0;
int stat;
error = xfs_ialloc_read_agi(mp, NULL, agno, &agbp);
if (error)
break;
/*
* Allocate and initialize a btree cursor for ialloc btree.
*/
cur = xfs_inobt_init_cursor(mp, NULL, agbp, agno,
XFS_BTNUM_INO);
if (agino > 0) {
/*
* In the middle of an allocation group, we need to get
* the remainder of the chunk we're in.
*/
struct xfs_inobt_rec_incore r;
error = xfs_bulkstat_grab_ichunk(cur, agino, &icount, &r);
if (error)
goto del_cursor;
if (icount) {
irbp->ir_startino = r.ir_startino;
irbp->ir_holemask = r.ir_holemask;
irbp->ir_count = r.ir_count;
irbp->ir_freecount = r.ir_freecount;
irbp->ir_free = r.ir_free;
irbp++;
}
/* Increment to the next record */
error = xfs_btree_increment(cur, 0, &stat);
} else {
/* Start of ag. Lookup the first inode chunk */
error = xfs_inobt_lookup(cur, 0, XFS_LOOKUP_GE, &stat);
}
if (error || stat == 0) {
end_of_ag = true;
goto del_cursor;
}
/*
* Loop through inode btree records in this ag,
* until we run out of inodes or space in the buffer.
*/
while (irbp < irbufend && icount < ubcount) {
struct xfs_inobt_rec_incore r;
error = xfs_inobt_get_rec(cur, &r, &stat);
if (error || stat == 0) {
end_of_ag = true;
goto del_cursor;
}
/*
* If this chunk has any allocated inodes, save it.
* Also start read-ahead now for this chunk.
*/
if (r.ir_freecount < r.ir_count) {
xfs_bulkstat_ichunk_ra(mp, agno, &r);
irbp->ir_startino = r.ir_startino;
irbp->ir_holemask = r.ir_holemask;
irbp->ir_count = r.ir_count;
irbp->ir_freecount = r.ir_freecount;
irbp->ir_free = r.ir_free;
irbp++;
icount += r.ir_count - r.ir_freecount;
}
error = xfs_btree_increment(cur, 0, &stat);
if (error || stat == 0) {
end_of_ag = true;
goto del_cursor;
}
cond_resched();
}
/*
* Drop the btree buffers and the agi buffer as we can't hold any
* of the locks these represent when calling iget. If there is a
* pending error, then we are done.
*/
del_cursor:
xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR);
xfs_buf_relse(agbp);
if (error)
break;
/*
* Now format all the good inodes into the user's buffer. The
* call to xfs_bulkstat_ag_ichunk() sets up the agino pointer
* for the next loop iteration.
*/
irbufend = irbp;
for (irbp = irbuf;
irbp < irbufend && ac.ac_ubleft >= statstruct_size;
irbp++) {
error = xfs_bulkstat_ag_ichunk(mp, agno, irbp,
formatter, statstruct_size, &ac,
&agino);
if (error)
break;
cond_resched();
}
/*
* If we've run out of space or had a formatting error, we
* are now done
*/
if (ac.ac_ubleft < statstruct_size || error)
break;
if (end_of_ag) {
agno++;
agino = 0;
}
}
/*
* Done, we're either out of filesystem or space to put the data.
*/
kmem_free(irbuf);
*ubcountp = ac.ac_ubelem;
/*
* We found some inodes, so clear the error status and return them.
* The lastino pointer will point directly at the inode that triggered
* any error that occurred, so on the next call the error will be
* triggered again and propagated to userspace as there will be no
* formatted inodes in the buffer.
*/
if (ac.ac_ubelem)
error = 0;
/*
* If we ran out of filesystem, lastino will point off the end of
* the filesystem so the next call will return immediately.
*/
*lastinop = XFS_AGINO_TO_INO(mp, agno, agino);
if (agno >= mp->m_sb.sb_agcount)
*done = 1;
return error;
}
/*
* Allocate an inode on disk.
*
* Mode is used to tell whether the new inode will need space, and whether it
* is a directory.
*
* This function is designed to be called twice if it has to do an allocation
* to make more free inodes. On the first call, *IO_agbp should be set to NULL.
* If an inode is available without having to performn an allocation, an inode
* number is returned. In this case, *IO_agbp is set to NULL. If an allocation
* needs to be done, xfs_dialloc returns the current AGI buffer in *IO_agbp.
* The caller should then commit the current transaction, allocate a
* new transaction, and call xfs_dialloc() again, passing in the previous value
* of *IO_agbp. IO_agbp should be held across the transactions. Since the AGI
* buffer is locked across the two calls, the second call is guaranteed to have
* a free inode available.
*
* Once we successfully pick an inode its number is returned and the on-disk
* data structures are updated. The inode itself is not read in, since doing so
* would break ordering constraints with xfs_reclaim.
*/
int
xfs_dialloc(
struct xfs_trans *tp,
xfs_ino_t parent,
umode_t mode,
int okalloc,
struct xfs_buf **IO_agbp,
xfs_ino_t *inop)
{
struct xfs_mount *mp = tp->t_mountp;
struct xfs_buf *agbp;
xfs_agnumber_t agno;
int error;
int ialloced;
int noroom = 0;
xfs_agnumber_t start_agno;
struct xfs_perag *pag;
if (*IO_agbp) {
/*
* If the caller passes in a pointer to the AGI buffer,
* continue where we left off before. In this case, we
* know that the allocation group has free inodes.
*/
agbp = *IO_agbp;
goto out_alloc;
}
/*
* We do not have an agbp, so select an initial allocation
* group for inode allocation.
*/
start_agno = xfs_ialloc_ag_select(tp, parent, mode, okalloc);
if (start_agno == NULLAGNUMBER) {
*inop = NULLFSINO;
return 0;
}
/*
* If we have already hit the ceiling of inode blocks then clear
* okalloc so we scan all available agi structures for a free
* inode.
*/
if (mp->m_maxicount &&
mp->m_sb.sb_icount + XFS_IALLOC_INODES(mp) > mp->m_maxicount) {
noroom = 1;
okalloc = 0;
}
/*
* Loop until we find an allocation group that either has free inodes
* or in which we can allocate some inodes. Iterate through the
* allocation groups upward, wrapping at the end.
*/
agno = start_agno;
for (;;) {
pag = xfs_perag_get(mp, agno);
if (!pag->pagi_inodeok) {
xfs_ialloc_next_ag(mp);
goto nextag;
}
if (!pag->pagi_init) {
error = xfs_ialloc_pagi_init(mp, tp, agno);
if (error)
goto out_error;
}
/*
* Do a first racy fast path check if this AG is usable.
*/
if (!pag->pagi_freecount && !okalloc)
goto nextag;
/*
* Then read in the AGI buffer and recheck with the AGI buffer
* lock held.
*/
error = xfs_ialloc_read_agi(mp, tp, agno, &agbp);
if (error)
goto out_error;
if (pag->pagi_freecount) {
xfs_perag_put(pag);
goto out_alloc;
}
if (!okalloc)
goto nextag_relse_buffer;
error = xfs_ialloc_ag_alloc(tp, agbp, &ialloced);
if (error) {
xfs_trans_brelse(tp, agbp);
if (error != ENOSPC)
goto out_error;
xfs_perag_put(pag);
*inop = NULLFSINO;
return 0;
}
if (ialloced) {
/*
* We successfully allocated some inodes, return
* the current context to the caller so that it
* can commit the current transaction and call
* us again where we left off.
*/
ASSERT(pag->pagi_freecount > 0);
xfs_perag_put(pag);
*IO_agbp = agbp;
*inop = NULLFSINO;
return 0;
}
nextag_relse_buffer:
xfs_trans_brelse(tp, agbp);
nextag:
xfs_perag_put(pag);
if (++agno == mp->m_sb.sb_agcount)
agno = 0;
if (agno == start_agno) {
*inop = NULLFSINO;
return noroom ? ENOSPC : 0;
}
}
out_alloc:
*IO_agbp = NULL;
return xfs_dialloc_ag(tp, agbp, parent, inop);
out_error:
xfs_perag_put(pag);
return XFS_ERROR(error);
}
/*
* Figure out how many blocks to reserve for an AG repair. We calculate the
* worst case estimate for the number of blocks we'd need to rebuild one of
* any type of per-AG btree.
*/
xfs_extlen_t
xrep_calc_ag_resblks(
struct xfs_scrub *sc)
{
struct xfs_mount *mp = sc->mp;
struct xfs_scrub_metadata *sm = sc->sm;
struct xfs_perag *pag;
struct xfs_buf *bp;
xfs_agino_t icount = NULLAGINO;
xfs_extlen_t aglen = NULLAGBLOCK;
xfs_extlen_t usedlen;
xfs_extlen_t freelen;
xfs_extlen_t bnobt_sz;
xfs_extlen_t inobt_sz;
xfs_extlen_t rmapbt_sz;
xfs_extlen_t refcbt_sz;
int error;
if (!(sm->sm_flags & XFS_SCRUB_IFLAG_REPAIR))
return 0;
pag = xfs_perag_get(mp, sm->sm_agno);
if (pag->pagi_init) {
/* Use in-core icount if possible. */
icount = pag->pagi_count;
} else {
/* Try to get the actual counters from disk. */
error = xfs_ialloc_read_agi(mp, NULL, sm->sm_agno, &bp);
if (!error) {
icount = pag->pagi_count;
xfs_buf_relse(bp);
}
}
/* Now grab the block counters from the AGF. */
error = xfs_alloc_read_agf(mp, NULL, sm->sm_agno, 0, &bp);
if (!error) {
aglen = be32_to_cpu(XFS_BUF_TO_AGF(bp)->agf_length);
freelen = be32_to_cpu(XFS_BUF_TO_AGF(bp)->agf_freeblks);
usedlen = aglen - freelen;
xfs_buf_relse(bp);
}
xfs_perag_put(pag);
/* If the icount is impossible, make some worst-case assumptions. */
if (icount == NULLAGINO ||
!xfs_verify_agino(mp, sm->sm_agno, icount)) {
xfs_agino_t first, last;
xfs_agino_range(mp, sm->sm_agno, &first, &last);
icount = last - first + 1;
}
/* If the block counts are impossible, make worst-case assumptions. */
if (aglen == NULLAGBLOCK ||
aglen != xfs_ag_block_count(mp, sm->sm_agno) ||
freelen >= aglen) {
aglen = xfs_ag_block_count(mp, sm->sm_agno);
freelen = aglen;
usedlen = aglen;
}
trace_xrep_calc_ag_resblks(mp, sm->sm_agno, icount, aglen,
freelen, usedlen);
/*
* Figure out how many blocks we'd need worst case to rebuild
* each type of btree. Note that we can only rebuild the
* bnobt/cntbt or inobt/finobt as pairs.
*/
bnobt_sz = 2 * xfs_allocbt_calc_size(mp, freelen);
if (xfs_sb_version_hassparseinodes(&mp->m_sb))
inobt_sz = xfs_iallocbt_calc_size(mp, icount /
XFS_INODES_PER_HOLEMASK_BIT);
else
inobt_sz = xfs_iallocbt_calc_size(mp, icount /
XFS_INODES_PER_CHUNK);
if (xfs_sb_version_hasfinobt(&mp->m_sb))
inobt_sz *= 2;
if (xfs_sb_version_hasreflink(&mp->m_sb))
refcbt_sz = xfs_refcountbt_calc_size(mp, usedlen);
else
refcbt_sz = 0;
if (xfs_sb_version_hasrmapbt(&mp->m_sb)) {
/*
* Guess how many blocks we need to rebuild the rmapbt.
* For non-reflink filesystems we can't have more records than
* used blocks. However, with reflink it's possible to have
* more than one rmap record per AG block. We don't know how
* many rmaps there could be in the AG, so we start off with
* what we hope is an generous over-estimation.
*/
if (xfs_sb_version_hasreflink(&mp->m_sb))
rmapbt_sz = xfs_rmapbt_calc_size(mp,
(unsigned long long)aglen * 2);
else
rmapbt_sz = xfs_rmapbt_calc_size(mp, usedlen);
} else {
rmapbt_sz = 0;
}
trace_xrep_calc_ag_resblks_btsize(mp, sm->sm_agno, bnobt_sz,
inobt_sz, rmapbt_sz, refcbt_sz);
return max(max(bnobt_sz, inobt_sz), max(rmapbt_sz, refcbt_sz));
}