static void
xfs_dquot_buf_verify(
struct xfs_buf *bp)
{
struct xfs_mount *mp = bp->b_target->bt_mount;
struct xfs_dqblk *d = (struct xfs_dqblk *)bp->b_addr;
struct xfs_disk_dquot *ddq;
xfs_dqid_t id = 0;
int i;
/*
* On the first read of the buffer, verify that each dquot is valid.
* We don't know what the id of the dquot is supposed to be, just that
* they should be increasing monotonically within the buffer. If the
* first id is corrupt, then it will fail on the second dquot in the
* buffer so corruptions could point to the wrong dquot in this case.
*/
for (i = 0; i < mp->m_quotainfo->qi_dqperchunk; i++) {
int error;
ddq = &d[i].dd_diskdq;
if (i == 0)
id = be32_to_cpu(ddq->d_id);
error = xfs_qm_dqcheck(mp, ddq, id + i, 0, XFS_QMOPT_DOWARN,
"xfs_dquot_read_verify");
if (error) {
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, d);
xfs_buf_ioerror(bp, EFSCORRUPTED);
break;
}
}
}
static void
xfs_allocbt_verify(
struct xfs_buf *bp)
{
struct xfs_mount *mp = bp->b_target->bt_mount;
struct xfs_btree_block *block = XFS_BUF_TO_BLOCK(bp);
struct xfs_perag *pag = bp->b_pag;
unsigned int level;
int sblock_ok; /* block passes checks */
/*
* magic number and level verification
*
* During growfs operations, we can't verify the exact level as the
* perag is not fully initialised and hence not attached to the buffer.
* In this case, check against the maximum tree depth.
*/
level = be16_to_cpu(block->bb_level);
switch (block->bb_magic) {
case cpu_to_be32(XFS_ABTB_MAGIC):
if (pag)
sblock_ok = level < pag->pagf_levels[XFS_BTNUM_BNOi];
else
sblock_ok = level < mp->m_ag_maxlevels;
break;
case cpu_to_be32(XFS_ABTC_MAGIC):
if (pag)
sblock_ok = level < pag->pagf_levels[XFS_BTNUM_CNTi];
else
sblock_ok = level < mp->m_ag_maxlevels;
break;
default:
sblock_ok = 0;
break;
}
/* numrecs verification */
sblock_ok = sblock_ok &&
be16_to_cpu(block->bb_numrecs) <= mp->m_alloc_mxr[level != 0];
/* sibling pointer verification */
sblock_ok = sblock_ok &&
(block->bb_u.s.bb_leftsib == cpu_to_be32(NULLAGBLOCK) ||
be32_to_cpu(block->bb_u.s.bb_leftsib) < mp->m_sb.sb_agblocks) &&
block->bb_u.s.bb_leftsib &&
(block->bb_u.s.bb_rightsib == cpu_to_be32(NULLAGBLOCK) ||
be32_to_cpu(block->bb_u.s.bb_rightsib) < mp->m_sb.sb_agblocks) &&
block->bb_u.s.bb_rightsib;
if (!sblock_ok) {
trace_xfs_btree_corrupt(bp, _RET_IP_);
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, block);
xfs_buf_ioerror(bp, EFSCORRUPTED);
}
}
static void
xfs_dquot_buf_read_verify(
struct xfs_buf *bp)
{
struct xfs_mount *mp = bp->b_target->bt_mount;
if (!xfs_dquot_buf_verify_crc(mp, bp) || !xfs_dquot_buf_verify(mp, bp)) {
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
xfs_buf_ioerror(bp, EFSCORRUPTED);
}
}
static void
xfs_dir3_block_read_verify(
struct xfs_buf *bp)
{
struct xfs_mount *mp = bp->b_target->bt_mount;
if ((xfs_sb_version_hascrc(&mp->m_sb) &&
!xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
XFS_DIR3_DATA_CRC_OFF)) ||
!xfs_dir3_block_verify(bp)) {
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
xfs_buf_ioerror(bp, EFSCORRUPTED);
}
}
static void
xfs_dir2_block_verify(
struct xfs_buf *bp)
{
struct xfs_mount *mp = bp->b_target->bt_mount;
struct xfs_dir2_data_hdr *hdr = bp->b_addr;
int block_ok = 0;
block_ok = hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC);
block_ok = block_ok && __xfs_dir2_data_check(NULL, bp) == 0;
if (!block_ok) {
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, hdr);
xfs_buf_ioerror(bp, EFSCORRUPTED);
}
}
static void
xfs_attr3_rmt_write_verify(
struct xfs_buf *bp)
{
struct xfs_mount *mp = bp->b_target->bt_mount;
struct xfs_buf_log_item *bip = bp->b_fspriv;
char *ptr;
int len;
xfs_daddr_t bno;
/* no verification of non-crc buffers */
if (!xfs_sb_version_hascrc(&mp->m_sb))
return;
ptr = bp->b_addr;
bno = bp->b_bn;
len = BBTOB(bp->b_length);
ASSERT(len >= XFS_LBSIZE(mp));
while (len > 0) {
if (!xfs_attr3_rmt_verify(mp, ptr, XFS_LBSIZE(mp), bno)) {
XFS_CORRUPTION_ERROR(__func__,
XFS_ERRLEVEL_LOW, mp, bp->b_addr);
xfs_buf_ioerror(bp, EFSCORRUPTED);
return;
}
if (bip) {
struct xfs_attr3_rmt_hdr *rmt;
rmt = (struct xfs_attr3_rmt_hdr *)ptr;
rmt->rm_lsn = cpu_to_be64(bip->bli_item.li_lsn);
}
xfs_update_cksum(ptr, XFS_LBSIZE(mp), XFS_ATTR3_RMT_CRC_OFF);
len -= XFS_LBSIZE(mp);
ptr += XFS_LBSIZE(mp);
bno += mp->m_bsize;
}
ASSERT(len == 0);
}
static void
xfs_attr3_rmt_read_verify(
struct xfs_buf *bp)
{
struct xfs_mount *mp = bp->b_target->bt_mount;
char *ptr;
int len;
bool corrupt = false;
xfs_daddr_t bno;
/* no verification of non-crc buffers */
if (!xfs_sb_version_hascrc(&mp->m_sb))
return;
ptr = bp->b_addr;
bno = bp->b_bn;
len = BBTOB(bp->b_length);
ASSERT(len >= XFS_LBSIZE(mp));
while (len > 0) {
if (!xfs_verify_cksum(ptr, XFS_LBSIZE(mp),
XFS_ATTR3_RMT_CRC_OFF)) {
corrupt = true;
break;
}
if (!xfs_attr3_rmt_verify(mp, ptr, XFS_LBSIZE(mp), bno)) {
corrupt = true;
break;
}
len -= XFS_LBSIZE(mp);
ptr += XFS_LBSIZE(mp);
bno += mp->m_bsize;
}
if (corrupt) {
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
xfs_buf_ioerror(bp, EFSCORRUPTED);
} else
ASSERT(len == 0);
}
static void
xfs_dir3_block_write_verify(
struct xfs_buf *bp)
{
struct xfs_mount *mp = bp->b_target->bt_mount;
struct xfs_buf_log_item *bip = bp->b_fspriv;
struct xfs_dir3_blk_hdr *hdr3 = bp->b_addr;
if (!xfs_dir3_block_verify(bp)) {
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
xfs_buf_ioerror(bp, EFSCORRUPTED);
return;
}
if (!xfs_sb_version_hascrc(&mp->m_sb))
return;
if (bip)
hdr3->lsn = cpu_to_be64(bip->bli_item.li_lsn);
xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length), XFS_DIR3_DATA_CRC_OFF);
}
/*
* Copy out entries of shortform attribute lists for attr_list().
* Shortform attribute lists are not stored in hashval sorted order.
* If the output buffer is not large enough to hold them all, then we
* we have to calculate each entries' hashvalue and sort them before
* we can begin returning them to the user.
*/
static int
xfs_attr_shortform_list(xfs_attr_list_context_t *context)
{
attrlist_cursor_kern_t *cursor;
xfs_attr_sf_sort_t *sbuf, *sbp;
xfs_attr_shortform_t *sf;
xfs_attr_sf_entry_t *sfe;
xfs_inode_t *dp;
int sbsize, nsbuf, count, i;
int error;
ASSERT(context != NULL);
dp = context->dp;
ASSERT(dp != NULL);
ASSERT(dp->i_afp != NULL);
sf = (xfs_attr_shortform_t *)dp->i_afp->if_u1.if_data;
ASSERT(sf != NULL);
if (!sf->hdr.count)
return 0;
cursor = context->cursor;
ASSERT(cursor != NULL);
trace_xfs_attr_list_sf(context);
/*
* If the buffer is large enough and the cursor is at the start,
* do not bother with sorting since we will return everything in
* one buffer and another call using the cursor won't need to be
* made.
* Note the generous fudge factor of 16 overhead bytes per entry.
* If bufsize is zero then put_listent must be a search function
* and can just scan through what we have.
*/
if (context->bufsize == 0 ||
(XFS_ISRESET_CURSOR(cursor) &&
(dp->i_afp->if_bytes + sf->hdr.count * 16) < context->bufsize)) {
for (i = 0, sfe = &sf->list[0]; i < sf->hdr.count; i++) {
error = context->put_listent(context,
sfe->flags,
sfe->nameval,
(int)sfe->namelen,
(int)sfe->valuelen);
if (error)
return error;
/*
* Either search callback finished early or
* didn't fit it all in the buffer after all.
*/
if (context->seen_enough)
break;
sfe = XFS_ATTR_SF_NEXTENTRY(sfe);
}
trace_xfs_attr_list_sf_all(context);
return 0;
}
/* do no more for a search callback */
if (context->bufsize == 0)
return 0;
/*
* It didn't all fit, so we have to sort everything on hashval.
*/
sbsize = sf->hdr.count * sizeof(*sbuf);
sbp = sbuf = kmem_alloc(sbsize, KM_SLEEP | KM_NOFS);
/*
* Scan the attribute list for the rest of the entries, storing
* the relevant info from only those that match into a buffer.
*/
nsbuf = 0;
for (i = 0, sfe = &sf->list[0]; i < sf->hdr.count; i++) {
if (unlikely(
((char *)sfe < (char *)sf) ||
((char *)sfe >= ((char *)sf + dp->i_afp->if_bytes)))) {
XFS_CORRUPTION_ERROR("xfs_attr_shortform_list",
XFS_ERRLEVEL_LOW,
context->dp->i_mount, sfe);
kmem_free(sbuf);
return -EFSCORRUPTED;
}
sbp->entno = i;
sbp->hash = xfs_da_hashname(sfe->nameval, sfe->namelen);
sbp->name = sfe->nameval;
sbp->namelen = sfe->namelen;
/* These are bytes, and both on-disk, don't endian-flip */
sbp->valuelen = sfe->valuelen;
sbp->flags = sfe->flags;
sfe = XFS_ATTR_SF_NEXTENTRY(sfe);
sbp++;
nsbuf++;
}
/*
* Sort the entries on hash then entno.
*/
xfs_sort(sbuf, nsbuf, sizeof(*sbuf), xfs_attr_shortform_compare);
/*
* Re-find our place IN THE SORTED LIST.
*/
count = 0;
cursor->initted = 1;
cursor->blkno = 0;
for (sbp = sbuf, i = 0; i < nsbuf; i++, sbp++) {
if (sbp->hash == cursor->hashval) {
if (cursor->offset == count) {
break;
}
count++;
} else if (sbp->hash > cursor->hashval) {
break;
}
}
if (i == nsbuf) {
kmem_free(sbuf);
return 0;
}
/*
* Loop putting entries into the user buffer.
*/
for ( ; i < nsbuf; i++, sbp++) {
if (cursor->hashval != sbp->hash) {
cursor->hashval = sbp->hash;
cursor->offset = 0;
}
error = context->put_listent(context,
sbp->flags,
sbp->name,
sbp->namelen,
sbp->valuelen);
if (error) {
kmem_free(sbuf);
return error;
}
if (context->seen_enough)
break;
cursor->offset++;
}
kmem_free(sbuf);
return 0;
}
STATIC int
xfs_attr_node_list(xfs_attr_list_context_t *context)
{
attrlist_cursor_kern_t *cursor;
xfs_attr_leafblock_t *leaf;
xfs_da_intnode_t *node;
struct xfs_attr3_icleaf_hdr leafhdr;
struct xfs_da3_icnode_hdr nodehdr;
struct xfs_da_node_entry *btree;
int error, i;
struct xfs_buf *bp;
struct xfs_inode *dp = context->dp;
struct xfs_mount *mp = dp->i_mount;
trace_xfs_attr_node_list(context);
cursor = context->cursor;
cursor->initted = 1;
/*
* Do all sorts of validation on the passed-in cursor structure.
* If anything is amiss, ignore the cursor and look up the hashval
* starting from the btree root.
*/
bp = NULL;
if (cursor->blkno > 0) {
error = xfs_da3_node_read(NULL, dp, cursor->blkno, -1,
&bp, XFS_ATTR_FORK);
if ((error != 0) && (error != -EFSCORRUPTED))
return error;
if (bp) {
struct xfs_attr_leaf_entry *entries;
node = bp->b_addr;
switch (be16_to_cpu(node->hdr.info.magic)) {
case XFS_DA_NODE_MAGIC:
case XFS_DA3_NODE_MAGIC:
trace_xfs_attr_list_wrong_blk(context);
xfs_trans_brelse(NULL, bp);
bp = NULL;
break;
case XFS_ATTR_LEAF_MAGIC:
case XFS_ATTR3_LEAF_MAGIC:
leaf = bp->b_addr;
xfs_attr3_leaf_hdr_from_disk(mp->m_attr_geo,
&leafhdr, leaf);
entries = xfs_attr3_leaf_entryp(leaf);
if (cursor->hashval > be32_to_cpu(
entries[leafhdr.count - 1].hashval)) {
trace_xfs_attr_list_wrong_blk(context);
xfs_trans_brelse(NULL, bp);
bp = NULL;
} else if (cursor->hashval <= be32_to_cpu(
entries[0].hashval)) {
trace_xfs_attr_list_wrong_blk(context);
xfs_trans_brelse(NULL, bp);
bp = NULL;
}
break;
default:
trace_xfs_attr_list_wrong_blk(context);
xfs_trans_brelse(NULL, bp);
bp = NULL;
}
}
}
/*
* We did not find what we expected given the cursor's contents,
* so we start from the top and work down based on the hash value.
* Note that start of node block is same as start of leaf block.
*/
if (bp == NULL) {
cursor->blkno = 0;
for (;;) {
__uint16_t magic;
error = xfs_da3_node_read(NULL, dp,
cursor->blkno, -1, &bp,
XFS_ATTR_FORK);
if (error)
return error;
node = bp->b_addr;
magic = be16_to_cpu(node->hdr.info.magic);
if (magic == XFS_ATTR_LEAF_MAGIC ||
magic == XFS_ATTR3_LEAF_MAGIC)
break;
if (magic != XFS_DA_NODE_MAGIC &&
magic != XFS_DA3_NODE_MAGIC) {
XFS_CORRUPTION_ERROR("xfs_attr_node_list(3)",
XFS_ERRLEVEL_LOW,
context->dp->i_mount,
node);
xfs_trans_brelse(NULL, bp);
return -EFSCORRUPTED;
}
dp->d_ops->node_hdr_from_disk(&nodehdr, node);
btree = dp->d_ops->node_tree_p(node);
for (i = 0; i < nodehdr.count; btree++, i++) {
if (cursor->hashval
<= be32_to_cpu(btree->hashval)) {
cursor->blkno = be32_to_cpu(btree->before);
trace_xfs_attr_list_node_descend(context,
btree);
break;
}
}
if (i == nodehdr.count) {
xfs_trans_brelse(NULL, bp);
return 0;
}
xfs_trans_brelse(NULL, bp);
}
}
ASSERT(bp != NULL);
/*
* Roll upward through the blocks, processing each leaf block in
* order. As long as there is space in the result buffer, keep
* adding the information.
*/
for (;;) {
leaf = bp->b_addr;
error = xfs_attr3_leaf_list_int(bp, context);
if (error) {
xfs_trans_brelse(NULL, bp);
return error;
}
xfs_attr3_leaf_hdr_from_disk(mp->m_attr_geo, &leafhdr, leaf);
if (context->seen_enough || leafhdr.forw == 0)
break;
cursor->blkno = leafhdr.forw;
xfs_trans_brelse(NULL, bp);
error = xfs_attr3_leaf_read(NULL, dp, cursor->blkno, -1, &bp);
if (error)
return error;
}
xfs_trans_brelse(NULL, bp);
return 0;
}
/*
* Read the disk inode attributes into the in-core inode structure.
*
* For version 5 superblocks, if we are initialising a new inode and we are not
* utilising the XFS_MOUNT_IKEEP inode cluster mode, we can simple build the new
* inode core with a random generation number. If we are keeping inodes around,
* we need to read the inode cluster to get the existing generation number off
* disk. Further, if we are using version 4 superblocks (i.e. v1/v2 inode
* format) then log recovery is dependent on the di_flushiter field being
* initialised from the current on-disk value and hence we must also read the
* inode off disk.
*/
int
xfs_iread(
xfs_mount_t *mp,
xfs_trans_t *tp,
xfs_inode_t *ip,
uint iget_flags)
{
xfs_buf_t *bp;
xfs_dinode_t *dip;
int error;
/*
* Fill in the location information in the in-core inode.
*/
error = xfs_imap(mp, tp, ip->i_ino, &ip->i_imap, iget_flags);
if (error)
return error;
/* shortcut IO on inode allocation if possible */
if ((iget_flags & XFS_IGET_CREATE) &&
xfs_sb_version_hascrc(&mp->m_sb) &&
!(mp->m_flags & XFS_MOUNT_IKEEP)) {
/* initialise the on-disk inode core */
memset(&ip->i_d, 0, sizeof(ip->i_d));
ip->i_d.di_magic = XFS_DINODE_MAGIC;
ip->i_d.di_gen = prandom_u32();
if (xfs_sb_version_hascrc(&mp->m_sb)) {
ip->i_d.di_version = 3;
ip->i_d.di_ino = ip->i_ino;
uuid_copy(&ip->i_d.di_uuid, &mp->m_sb.sb_uuid);
} else
ip->i_d.di_version = 2;
return 0;
}
/*
* Get pointers to the on-disk inode and the buffer containing it.
*/
error = xfs_imap_to_bp(mp, tp, &ip->i_imap, &dip, &bp, 0, iget_flags);
if (error)
return error;
/* even unallocated inodes are verified */
if (!xfs_dinode_verify(mp, ip, dip)) {
xfs_alert(mp, "%s: validation failed for inode %lld failed",
__func__, ip->i_ino);
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, dip);
error = -EFSCORRUPTED;
goto out_brelse;
}
/*
* If the on-disk inode is already linked to a directory
* entry, copy all of the inode into the in-core inode.
* xfs_iformat_fork() handles copying in the inode format
* specific information.
* Otherwise, just get the truly permanent information.
*/
if (dip->di_mode) {
xfs_dinode_from_disk(&ip->i_d, dip);
error = xfs_iformat_fork(ip, dip);
if (error) {
#ifdef DEBUG
xfs_alert(mp, "%s: xfs_iformat() returned error %d",
__func__, error);
#endif /* DEBUG */
goto out_brelse;
}
} else {
/*
* Partial initialisation of the in-core inode. Just the bits
* that xfs_ialloc won't overwrite or relies on being correct.
*/
ip->i_d.di_magic = be16_to_cpu(dip->di_magic);
ip->i_d.di_version = dip->di_version;
ip->i_d.di_gen = be32_to_cpu(dip->di_gen);
ip->i_d.di_flushiter = be16_to_cpu(dip->di_flushiter);
if (dip->di_version == 3) {
ip->i_d.di_ino = be64_to_cpu(dip->di_ino);
uuid_copy(&ip->i_d.di_uuid, &dip->di_uuid);
}
/*
* Make sure to pull in the mode here as well in
* case the inode is released without being used.
* This ensures that xfs_inactive() will see that
* the inode is already free and not try to mess
* with the uninitialized part of it.
*/
ip->i_d.di_mode = 0;
}
/*
* Automatically convert version 1 inode formats in memory to version 2
* inode format. If the inode is modified, it will get logged and
* rewritten as a version 2 inode. We can do this because we set the
* superblock feature bit for v2 inodes unconditionally during mount
* and it means the reast of the code can assume the inode version is 2
* or higher.
*/
if (ip->i_d.di_version == 1) {
ip->i_d.di_version = 2;
memset(&(ip->i_d.di_pad[0]), 0, sizeof(ip->i_d.di_pad));
ip->i_d.di_nlink = ip->i_d.di_onlink;
ip->i_d.di_onlink = 0;
xfs_set_projid(ip, 0);
}
ip->i_delayed_blks = 0;
/*
* Mark the buffer containing the inode as something to keep
* around for a while. This helps to keep recently accessed
* meta-data in-core longer.
*/
xfs_buf_set_ref(bp, XFS_INO_REF);
/*
* Use xfs_trans_brelse() to release the buffer containing the on-disk
* inode, because it was acquired with xfs_trans_read_buf() in
* xfs_imap_to_bp() above. If tp is NULL, this is just a normal
* brelse(). If we're within a transaction, then xfs_trans_brelse()
* will only release the buffer if it is not dirty within the
* transaction. It will be OK to release the buffer in this case,
* because inodes on disk are never destroyed and we will be locking the
* new in-core inode before putting it in the cache where other
* processes can find it. Thus we don't have to worry about the inode
* being changed just because we released the buffer.
*/
out_brelse:
xfs_trans_brelse(tp, bp);
return error;
}
/*
* Add an entry to a block directory.
*/
int /* error */
xfs_dir2_block_addname(
xfs_da_args_t *args) /* directory op arguments */
{
xfs_dir2_data_free_t *bf; /* bestfree table in block */
xfs_dir2_block_t *block; /* directory block structure */
xfs_dir2_leaf_entry_t *blp; /* block leaf entries */
xfs_dabuf_t *bp; /* buffer for block */
xfs_dir2_block_tail_t *btp; /* block tail */
int compact; /* need to compact leaf ents */
xfs_dir2_data_entry_t *dep; /* block data entry */
xfs_inode_t *dp; /* directory inode */
xfs_dir2_data_unused_t *dup; /* block unused entry */
int error; /* error return value */
xfs_dir2_data_unused_t *enddup=NULL; /* unused at end of data */
xfs_dahash_t hash; /* hash value of found entry */
int high; /* high index for binary srch */
int highstale; /* high stale index */
int lfloghigh=0; /* last final leaf to log */
int lfloglow=0; /* first final leaf to log */
int len; /* length of the new entry */
int low; /* low index for binary srch */
int lowstale; /* low stale index */
int mid=0; /* midpoint for binary srch */
xfs_mount_t *mp; /* filesystem mount point */
int needlog; /* need to log header */
int needscan; /* need to rescan freespace */
__be16 *tagp; /* pointer to tag value */
xfs_trans_t *tp; /* transaction structure */
trace_xfs_dir2_block_addname(args);
dp = args->dp;
tp = args->trans;
mp = dp->i_mount;
/*
* Read the (one and only) directory block into dabuf bp.
*/
if ((error =
xfs_da_read_buf(tp, dp, mp->m_dirdatablk, -1, &bp, XFS_DATA_FORK))) {
return error;
}
ASSERT(bp != NULL);
block = bp->data;
/*
* Check the magic number, corrupted if wrong.
*/
if (unlikely(be32_to_cpu(block->hdr.magic) != XFS_DIR2_BLOCK_MAGIC)) {
XFS_CORRUPTION_ERROR("xfs_dir2_block_addname",
XFS_ERRLEVEL_LOW, mp, block);
xfs_da_brelse(tp, bp);
return XFS_ERROR(EFSCORRUPTED);
}
len = xfs_dir2_data_entsize(args->namelen);
/*
* Set up pointers to parts of the block.
*/
bf = block->hdr.bestfree;
btp = xfs_dir2_block_tail_p(mp, block);
blp = xfs_dir2_block_leaf_p(btp);
/*
* No stale entries? Need space for entry and new leaf.
*/
if (!btp->stale) {
/*
* Tag just before the first leaf entry.
*/
tagp = (__be16 *)blp - 1;
/*
* Data object just before the first leaf entry.
*/
enddup = (xfs_dir2_data_unused_t *)((char *)block + be16_to_cpu(*tagp));
/*
* If it's not free then can't do this add without cleaning up:
* the space before the first leaf entry needs to be free so it
* can be expanded to hold the pointer to the new entry.
*/
if (be16_to_cpu(enddup->freetag) != XFS_DIR2_DATA_FREE_TAG)
dup = enddup = NULL;
/*
* Check out the biggest freespace and see if it's the same one.
*/
else {
dup = (xfs_dir2_data_unused_t *)
((char *)block + be16_to_cpu(bf[0].offset));
if (dup == enddup) {
/*
* It is the biggest freespace, is it too small
* to hold the new leaf too?
*/
if (be16_to_cpu(dup->length) < len + (uint)sizeof(*blp)) {
/*
* Yes, we use the second-largest
* entry instead if it works.
*/
if (be16_to_cpu(bf[1].length) >= len)
dup = (xfs_dir2_data_unused_t *)
((char *)block +
be16_to_cpu(bf[1].offset));
else
dup = NULL;
}
} else {
/*
* Not the same free entry,
* just check its length.
*/
if (be16_to_cpu(dup->length) < len) {
dup = NULL;
}
}
}
compact = 0;
}
/*
* If there are stale entries we'll use one for the leaf.
* Is the biggest entry enough to avoid compaction?
*/
else if (be16_to_cpu(bf[0].length) >= len) {
dup = (xfs_dir2_data_unused_t *)
((char *)block + be16_to_cpu(bf[0].offset));
compact = 0;
}
/*
* Will need to compact to make this work.
*/
else {
/*
* Tag just before the first leaf entry.
*/
tagp = (__be16 *)blp - 1;
/*
* Data object just before the first leaf entry.
*/
dup = (xfs_dir2_data_unused_t *)((char *)block + be16_to_cpu(*tagp));
/*
* If it's not free then the data will go where the
* leaf data starts now, if it works at all.
*/
if (be16_to_cpu(dup->freetag) == XFS_DIR2_DATA_FREE_TAG) {
if (be16_to_cpu(dup->length) + (be32_to_cpu(btp->stale) - 1) *
(uint)sizeof(*blp) < len)
dup = NULL;
} else if ((be32_to_cpu(btp->stale) - 1) * (uint)sizeof(*blp) < len)
dup = NULL;
else
dup = (xfs_dir2_data_unused_t *)blp;
compact = 1;
}
/*
* If this isn't a real add, we're done with the buffer.
*/
if (args->op_flags & XFS_DA_OP_JUSTCHECK)
xfs_da_brelse(tp, bp);
/*
* If we don't have space for the new entry & leaf ...
*/
if (!dup) {
/*
* Not trying to actually do anything, or don't have
* a space reservation: return no-space.
*/
if ((args->op_flags & XFS_DA_OP_JUSTCHECK) || args->total == 0)
return XFS_ERROR(ENOSPC);
/*
* Convert to the next larger format.
* Then add the new entry in that format.
*/
error = xfs_dir2_block_to_leaf(args, bp);
xfs_da_buf_done(bp);
if (error)
return error;
return xfs_dir2_leaf_addname(args);
}
/*
* Just checking, and it would work, so say so.
*/
if (args->op_flags & XFS_DA_OP_JUSTCHECK)
return 0;
needlog = needscan = 0;
/*
* If need to compact the leaf entries, do it now.
* Leave the highest-numbered stale entry stale.
* XXX should be the one closest to mid but mid is not yet computed.
*/
if (compact) {
int fromidx; /* source leaf index */
int toidx; /* target leaf index */
for (fromidx = toidx = be32_to_cpu(btp->count) - 1,
highstale = lfloghigh = -1;
fromidx >= 0;
fromidx--) {
if (be32_to_cpu(blp[fromidx].address) == XFS_DIR2_NULL_DATAPTR) {
if (highstale == -1)
highstale = toidx;
else {
if (lfloghigh == -1)
lfloghigh = toidx;
continue;
}
}
if (fromidx < toidx)
blp[toidx] = blp[fromidx];
toidx--;
}
lfloglow = toidx + 1 - (be32_to_cpu(btp->stale) - 1);
lfloghigh -= be32_to_cpu(btp->stale) - 1;
be32_add_cpu(&btp->count, -(be32_to_cpu(btp->stale) - 1));
xfs_dir2_data_make_free(tp, bp,
(xfs_dir2_data_aoff_t)((char *)blp - (char *)block),
(xfs_dir2_data_aoff_t)((be32_to_cpu(btp->stale) - 1) * sizeof(*blp)),
&needlog, &needscan);
blp += be32_to_cpu(btp->stale) - 1;
btp->stale = cpu_to_be32(1);
/*
* If we now need to rebuild the bestfree map, do so.
* This needs to happen before the next call to use_free.
*/
if (needscan) {
xfs_dir2_data_freescan(mp, (xfs_dir2_data_t *)block, &needlog);
needscan = 0;
}
}
/*
* Set leaf logging boundaries to impossible state.
* For the no-stale case they're set explicitly.
*/
else if (btp->stale) {
lfloglow = be32_to_cpu(btp->count);
lfloghigh = -1;
}
/*
* Find the slot that's first lower than our hash value, -1 if none.
*/
for (low = 0, high = be32_to_cpu(btp->count) - 1; low <= high; ) {
mid = (low + high) >> 1;
if ((hash = be32_to_cpu(blp[mid].hashval)) == args->hashval)
break;
if (hash < args->hashval)
low = mid + 1;
else
high = mid - 1;
}
while (mid >= 0 && be32_to_cpu(blp[mid].hashval) >= args->hashval) {
mid--;
}
/*
* No stale entries, will use enddup space to hold new leaf.
*/
if (!btp->stale) {
/*
* Mark the space needed for the new leaf entry, now in use.
*/
xfs_dir2_data_use_free(tp, bp, enddup,
(xfs_dir2_data_aoff_t)
((char *)enddup - (char *)block + be16_to_cpu(enddup->length) -
sizeof(*blp)),
(xfs_dir2_data_aoff_t)sizeof(*blp),
&needlog, &needscan);
/*
* Update the tail (entry count).
*/
be32_add_cpu(&btp->count, 1);
/*
* If we now need to rebuild the bestfree map, do so.
* This needs to happen before the next call to use_free.
*/
if (needscan) {
xfs_dir2_data_freescan(mp, (xfs_dir2_data_t *)block,
&needlog);
needscan = 0;
}
/*
* Adjust pointer to the first leaf entry, we're about to move
* the table up one to open up space for the new leaf entry.
* Then adjust our index to match.
*/
blp--;
mid++;
if (mid)
memmove(blp, &blp[1], mid * sizeof(*blp));
lfloglow = 0;
lfloghigh = mid;
}
/*
* Use a stale leaf for our new entry.
*/
else {
for (lowstale = mid;
lowstale >= 0 &&
be32_to_cpu(blp[lowstale].address) != XFS_DIR2_NULL_DATAPTR;
lowstale--)
continue;
for (highstale = mid + 1;
highstale < be32_to_cpu(btp->count) &&
be32_to_cpu(blp[highstale].address) != XFS_DIR2_NULL_DATAPTR &&
(lowstale < 0 || mid - lowstale > highstale - mid);
highstale++)
continue;
/*
* Move entries toward the low-numbered stale entry.
*/
if (lowstale >= 0 &&
(highstale == be32_to_cpu(btp->count) ||
mid - lowstale <= highstale - mid)) {
if (mid - lowstale)
memmove(&blp[lowstale], &blp[lowstale + 1],
(mid - lowstale) * sizeof(*blp));
lfloglow = MIN(lowstale, lfloglow);
lfloghigh = MAX(mid, lfloghigh);
}
/*
* Move entries toward the high-numbered stale entry.
*/
else {
ASSERT(highstale < be32_to_cpu(btp->count));
mid++;
if (highstale - mid)
memmove(&blp[mid + 1], &blp[mid],
(highstale - mid) * sizeof(*blp));
lfloglow = MIN(mid, lfloglow);
lfloghigh = MAX(highstale, lfloghigh);
}
be32_add_cpu(&btp->stale, -1);
}
/*
* Point to the new data entry.
*/
dep = (xfs_dir2_data_entry_t *)dup;
/*
* Fill in the leaf entry.
*/
blp[mid].hashval = cpu_to_be32(args->hashval);
blp[mid].address = cpu_to_be32(xfs_dir2_byte_to_dataptr(mp,
(char *)dep - (char *)block));
xfs_dir2_block_log_leaf(tp, bp, lfloglow, lfloghigh);
/*
* Mark space for the data entry used.
*/
xfs_dir2_data_use_free(tp, bp, dup,
(xfs_dir2_data_aoff_t)((char *)dup - (char *)block),
(xfs_dir2_data_aoff_t)len, &needlog, &needscan);
/*
* Create the new data entry.
*/
dep->inumber = cpu_to_be64(args->inumber);
dep->namelen = args->namelen;
memcpy(dep->name, args->name, args->namelen);
tagp = xfs_dir2_data_entry_tag_p(dep);
*tagp = cpu_to_be16((char *)dep - (char *)block);
/*
* Clean up the bestfree array and log the header, tail, and entry.
*/
if (needscan)
xfs_dir2_data_freescan(mp, (xfs_dir2_data_t *)block, &needlog);
if (needlog)
xfs_dir2_data_log_header(tp, bp);
xfs_dir2_block_log_tail(tp, bp);
xfs_dir2_data_log_entry(tp, bp, dep);
xfs_dir2_data_check(dp, bp);
xfs_da_buf_done(bp);
return 0;
}
