int
xfs_dir3_data_readahead(
struct xfs_inode *dp,
xfs_dablk_t bno,
xfs_daddr_t mapped_bno)
{
return xfs_da_reada_buf(dp, bno, mapped_bno,
XFS_DATA_FORK, &xfs_dir3_data_reada_buf_ops);
}
STATIC int
xfs_dir_open(
struct inode *inode,
struct file *file)
{
struct xfs_inode *ip = XFS_I(inode);
int mode;
int error;
error = xfs_file_open(inode, file);
if (error)
return error;
mode = xfs_ilock_map_shared(ip);
if (ip->i_d.di_nextents > 0)
xfs_da_reada_buf(NULL, ip, 0, XFS_DATA_FORK);
xfs_iunlock(ip, mode);
return 0;
}
STATIC int
xfs_dir_open(
struct inode *inode,
struct file *file)
{
struct xfs_inode *ip = XFS_I(inode);
int mode;
int error;
error = xfs_file_open(inode, file);
if (error)
return error;
/*
* If there are any blocks, read-ahead block 0 as we're almost
* certain to have the next operation be a read there.
*/
mode = xfs_ilock_map_shared(ip);
if (ip->i_d.di_nextents > 0)
xfs_da_reada_buf(NULL, ip, 0, XFS_DATA_FORK);
xfs_iunlock(ip, mode);
return 0;
}
STATIC int
xfs_dir_node_getdents(xfs_trans_t *trans, xfs_inode_t *dp, uio_t *uio,
int *eofp, dirent_t *dbp, xfs_dir_put_t put)
{
xfs_da_intnode_t *node;
xfs_da_node_entry_t *btree;
xfs_dir_leafblock_t *leaf;
xfs_dablk_t bno, nextbno;
xfs_dahash_t cookhash;
xfs_mount_t *mp;
int error, eob, i;
xfs_dabuf_t *bp;
xfs_daddr_t nextda;
/*
* Pick up our context.
*/
mp = dp->i_mount;
bp = NULL;
bno = XFS_DA_COOKIE_BNO(mp, uio->uio_offset);
cookhash = XFS_DA_COOKIE_HASH(mp, uio->uio_offset);
xfs_dir_trace_g_du("node: start", dp, uio);
/*
* Re-find our place, even if we're confused about what our place is.
*
* First we check the block number from the magic cookie, it is a
* cache of where we ended last time. If we find a leaf block, and
* the starting hashval in that block is less than our desired
* hashval, then we run with it.
*/
if (bno > 0) {
error = xfs_da_read_buf(trans, dp, bno, -1, &bp, XFS_DATA_FORK);
if ((error != 0) && (error != EFSCORRUPTED))
return(error);
if (bp)
leaf = bp->data;
if (bp && INT_GET(leaf->hdr.info.magic, ARCH_CONVERT) != XFS_DIR_LEAF_MAGIC) {
xfs_dir_trace_g_dub("node: block not a leaf",
dp, uio, bno);
xfs_da_brelse(trans, bp);
bp = NULL;
}
if (bp && INT_GET(leaf->entries[0].hashval, ARCH_CONVERT) > cookhash) {
xfs_dir_trace_g_dub("node: leaf hash too large",
dp, uio, bno);
xfs_da_brelse(trans, bp);
bp = NULL;
}
if (bp &&
cookhash > INT_GET(leaf->entries[INT_GET(leaf->hdr.count, ARCH_CONVERT) - 1].hashval, ARCH_CONVERT)) {
xfs_dir_trace_g_dub("node: leaf hash too small",
dp, uio, bno);
xfs_da_brelse(trans, bp);
bp = NULL;
}
}
/*
* If we did not find a leaf block from the blockno in the cookie,
* or we there was no blockno in the cookie (eg: first time thru),
* the we start at the top of the Btree and re-find our hashval.
*/
if (bp == NULL) {
xfs_dir_trace_g_du("node: start at root" , dp, uio);
bno = 0;
for (;;) {
error = xfs_da_read_buf(trans, dp, bno, -1, &bp,
XFS_DATA_FORK);
if (error)
return(error);
if (bp == NULL)
return(XFS_ERROR(EFSCORRUPTED));
node = bp->data;
if (INT_GET(node->hdr.info.magic, ARCH_CONVERT) != XFS_DA_NODE_MAGIC)
break;
btree = &node->btree[0];
xfs_dir_trace_g_dun("node: node detail", dp, uio, node);
for (i = 0; i < INT_GET(node->hdr.count, ARCH_CONVERT); btree++, i++) {
if (INT_GET(btree->hashval, ARCH_CONVERT) >= cookhash) {
bno = INT_GET(btree->before, ARCH_CONVERT);
break;
}
}
if (i == INT_GET(node->hdr.count, ARCH_CONVERT)) {
xfs_da_brelse(trans, bp);
xfs_dir_trace_g_du("node: hash beyond EOF",
dp, uio);
uio->uio_offset = XFS_DA_MAKE_COOKIE(mp, 0, 0,
XFS_DA_MAXHASH);
*eofp = 1;
return(0);
}
xfs_dir_trace_g_dub("node: going to block",
dp, uio, bno);
xfs_da_brelse(trans, bp);
}
}
ASSERT(cookhash != XFS_DA_MAXHASH);
/*
* We've dropped down to the (first) leaf block that contains the
* hashval we are interested in. Continue rolling upward thru the
* leaf blocks until we fill up our buffer.
*/
for (;;) {
leaf = bp->data;
if (INT_GET(leaf->hdr.info.magic, ARCH_CONVERT) != XFS_DIR_LEAF_MAGIC) {
xfs_dir_trace_g_dul("node: not a leaf", dp, uio, leaf);
xfs_da_brelse(trans, bp);
return XFS_ERROR(EFSCORRUPTED);
}
xfs_dir_trace_g_dul("node: leaf detail", dp, uio, leaf);
if (nextbno = INT_GET(leaf->hdr.info.forw, ARCH_CONVERT)) {
nextda = xfs_da_reada_buf(trans, dp, nextbno,
XFS_DATA_FORK);
} else
nextda = -1;
error = xfs_dir_leaf_getdents_int(bp, dp, bno, uio, &eob, dbp,
put, nextda);
xfs_da_brelse(trans, bp);
bno = nextbno;
if (eob) {
xfs_dir_trace_g_dub("node: E-O-B", dp, uio, bno);
*eofp = 0;
return(error);
}
if (bno == 0)
break;
error = xfs_da_read_buf(trans, dp, bno, nextda, &bp,
XFS_DATA_FORK);
if (error)
return(error);
if (bp == NULL)
return(XFS_ERROR(EFSCORRUPTED));
}
*eofp = 1;
xfs_dir_trace_g_du("node: E-O-F", dp, uio);
return(0);
}
/*
* Getdents (readdir) for leaf and node directories.
* This reads the data blocks only, so is the same for both forms.
*/
int /* error */
xfs_dir2_leaf_getdents(
xfs_inode_t *dp, /* incore directory inode */
void *dirent,
size_t bufsize,
xfs_off_t *offset,
filldir_t filldir)
{
xfs_dabuf_t *bp; /* data block buffer */
int byteoff; /* offset in current block */
xfs_dir2_db_t curdb; /* db for current block */
xfs_dir2_off_t curoff; /* current overall offset */
xfs_dir2_data_t *data; /* data block structure */
xfs_dir2_data_entry_t *dep; /* data entry */
xfs_dir2_data_unused_t *dup; /* unused entry */
int error = 0; /* error return value */
int i; /* temporary loop index */
int j; /* temporary loop index */
int length; /* temporary length value */
xfs_bmbt_irec_t *map; /* map vector for blocks */
xfs_extlen_t map_blocks; /* number of fsbs in map */
xfs_dablk_t map_off; /* last mapped file offset */
int map_size; /* total entries in *map */
int map_valid; /* valid entries in *map */
xfs_mount_t *mp; /* filesystem mount point */
xfs_dir2_off_t newoff; /* new curoff after new blk */
int nmap; /* mappings to ask xfs_bmapi */
char *ptr = NULL; /* pointer to current data */
int ra_current; /* number of read-ahead blks */
int ra_index; /* *map index for read-ahead */
int ra_offset; /* map entry offset for ra */
int ra_want; /* readahead count wanted */
/*
* If the offset is at or past the largest allowed value,
* give up right away.
*/
if (*offset >= XFS_DIR2_MAX_DATAPTR)
return 0;
mp = dp->i_mount;
/*
* Set up to bmap a number of blocks based on the caller's
* buffer size, the directory block size, and the filesystem
* block size.
*/
map_size = howmany(bufsize + mp->m_dirblksize, mp->m_sb.sb_blocksize);
map = kmem_alloc(map_size * sizeof(*map), KM_SLEEP);
map_valid = ra_index = ra_offset = ra_current = map_blocks = 0;
bp = NULL;
/*
* Inside the loop we keep the main offset value as a byte offset
* in the directory file.
*/
curoff = xfs_dir2_dataptr_to_byte(mp, *offset);
/*
* Force this conversion through db so we truncate the offset
* down to get the start of the data block.
*/
map_off = xfs_dir2_db_to_da(mp, xfs_dir2_byte_to_db(mp, curoff));
/*
* Loop over directory entries until we reach the end offset.
* Get more blocks and readahead as necessary.
*/
while (curoff < XFS_DIR2_LEAF_OFFSET) {
/*
* If we have no buffer, or we're off the end of the
* current buffer, need to get another one.
*/
if (!bp || ptr >= (char *)bp->data + mp->m_dirblksize) {
/*
* If we have a buffer, we need to release it and
* take it out of the mapping.
*/
if (bp) {
xfs_da_brelse(NULL, bp);
bp = NULL;
map_blocks -= mp->m_dirblkfsbs;
/*
* Loop to get rid of the extents for the
* directory block.
*/
for (i = mp->m_dirblkfsbs; i > 0; ) {
j = MIN((int)map->br_blockcount, i);
map->br_blockcount -= j;
map->br_startblock += j;
map->br_startoff += j;
/*
* If mapping is done, pitch it from
* the table.
*/
if (!map->br_blockcount && --map_valid)
memmove(&map[0], &map[1],
sizeof(map[0]) *
map_valid);
i -= j;
}
}
/*
* Recalculate the readahead blocks wanted.
*/
ra_want = howmany(bufsize + mp->m_dirblksize,
mp->m_sb.sb_blocksize) - 1;
ASSERT(ra_want >= 0);
/*
* If we don't have as many as we want, and we haven't
* run out of data blocks, get some more mappings.
*/
if (1 + ra_want > map_blocks &&
map_off <
xfs_dir2_byte_to_da(mp, XFS_DIR2_LEAF_OFFSET)) {
/*
* Get more bmaps, fill in after the ones
* we already have in the table.
*/
nmap = map_size - map_valid;
error = xfs_bmapi(NULL, dp,
map_off,
xfs_dir2_byte_to_da(mp,
XFS_DIR2_LEAF_OFFSET) - map_off,
XFS_BMAPI_METADATA, NULL, 0,
&map[map_valid], &nmap, NULL, NULL);
/*
* Don't know if we should ignore this or
* try to return an error.
* The trouble with returning errors
* is that readdir will just stop without
* actually passing the error through.
*/
if (error)
break; /* XXX */
/*
* If we got all the mappings we asked for,
* set the final map offset based on the
* last bmap value received.
* Otherwise, we've reached the end.
*/
if (nmap == map_size - map_valid)
map_off =
map[map_valid + nmap - 1].br_startoff +
map[map_valid + nmap - 1].br_blockcount;
else
map_off =
xfs_dir2_byte_to_da(mp,
XFS_DIR2_LEAF_OFFSET);
/*
* Look for holes in the mapping, and
* eliminate them. Count up the valid blocks.
*/
for (i = map_valid; i < map_valid + nmap; ) {
if (map[i].br_startblock ==
HOLESTARTBLOCK) {
nmap--;
length = map_valid + nmap - i;
if (length)
memmove(&map[i],
&map[i + 1],
sizeof(map[i]) *
length);
} else {
map_blocks +=
map[i].br_blockcount;
i++;
}
}
map_valid += nmap;
}
/*
* No valid mappings, so no more data blocks.
*/
if (!map_valid) {
curoff = xfs_dir2_da_to_byte(mp, map_off);
break;
}
/*
* Read the directory block starting at the first
* mapping.
*/
curdb = xfs_dir2_da_to_db(mp, map->br_startoff);
error = xfs_da_read_buf(NULL, dp, map->br_startoff,
map->br_blockcount >= mp->m_dirblkfsbs ?
XFS_FSB_TO_DADDR(mp, map->br_startblock) :
-1,
&bp, XFS_DATA_FORK);
/*
* Should just skip over the data block instead
* of giving up.
*/
if (error)
break; /* XXX */
/*
* Adjust the current amount of read-ahead: we just
* read a block that was previously ra.
*/
if (ra_current)
ra_current -= mp->m_dirblkfsbs;
/*
* Do we need more readahead?
*/
for (ra_index = ra_offset = i = 0;
ra_want > ra_current && i < map_blocks;
i += mp->m_dirblkfsbs) {
ASSERT(ra_index < map_valid);
/*
* Read-ahead a contiguous directory block.
*/
if (i > ra_current &&
map[ra_index].br_blockcount >=
mp->m_dirblkfsbs) {
//TODO: make sure this is right, should readahead be
//replaced by readbuf?
//xfs_buf_readahead(mp->m_ddev_targp,
// XFS_FSB_TO_DADDR(mp,
// map[ra_index].br_startblock +
// ra_offset),
// (int)BTOBB(mp->m_dirblksize));
//TODO: figure out flags, flags =0
//libxfs_readbuf(mp->m_dev,
// XFS_FSB_TO_DADDR(mp,
// map[ra_index].br_startblock +
// ra_offset),
// (int)BTOBB(mp->m_dirblksize), 0);
ra_current = i;
}
/*
* Read-ahead a non-contiguous directory block.
* This doesn't use our mapping, but this
* is a very rare case.
*/
else if (i > ra_current) {
(void)xfs_da_reada_buf(NULL, dp,
map[ra_index].br_startoff +
ra_offset, XFS_DATA_FORK);
ra_current = i;
}
/*
* Advance offset through the mapping table.
*/
for (j = 0; j < mp->m_dirblkfsbs; j++) {
/*
* The rest of this extent but not
* more than a dir block.
*/
length = MIN(mp->m_dirblkfsbs,
(int)(map[ra_index].br_blockcount -
ra_offset));
j += length;
ra_offset += length;
/*
* Advance to the next mapping if
* this one is used up.
*/
if (ra_offset ==
map[ra_index].br_blockcount) {
ra_offset = 0;
ra_index++;
}
}
}
/*
* Having done a read, we need to set a new offset.
*/
newoff = xfs_dir2_db_off_to_byte(mp, curdb, 0);
/*
* Start of the current block.
*/
if (curoff < newoff)
curoff = newoff;
/*
* Make sure we're in the right block.
*/
else if (curoff > newoff)
ASSERT(xfs_dir2_byte_to_db(mp, curoff) ==
curdb);
data = bp->data;
xfs_dir2_data_check(dp, bp);
/*
* Find our position in the block.
*/
ptr = (char *)&data->u;
byteoff = xfs_dir2_byte_to_off(mp, curoff);
/*
* Skip past the header.
*/
if (byteoff == 0)
curoff += (uint)sizeof(data->hdr);
/*
* Skip past entries until we reach our offset.
*/
else {
while ((char *)ptr - (char *)data < byteoff) {
dup = (xfs_dir2_data_unused_t *)ptr;
if (be16_to_cpu(dup->freetag)
== XFS_DIR2_DATA_FREE_TAG) {
length = be16_to_cpu(dup->length);
ptr += length;
continue;
}
dep = (xfs_dir2_data_entry_t *)ptr;
length =
xfs_dir2_data_entsize(dep->namelen);
ptr += length;
}
/*
* Now set our real offset.
*/
curoff =
xfs_dir2_db_off_to_byte(mp,
xfs_dir2_byte_to_db(mp, curoff),
(char *)ptr - (char *)data);
if (ptr >= (char *)data + mp->m_dirblksize) {
continue;
}
}
}
/*
* We have a pointer to an entry.
* Is it a live one?
*/
dup = (xfs_dir2_data_unused_t *)ptr;
/*
* No, it's unused, skip over it.
*/
if (be16_to_cpu(dup->freetag) == XFS_DIR2_DATA_FREE_TAG) {
length = be16_to_cpu(dup->length);
ptr += length;
curoff += length;
continue;
}
dep = (xfs_dir2_data_entry_t *)ptr;
length = xfs_dir2_data_entsize(dep->namelen);
if (filldir(dirent, (char *)dep->name, dep->namelen,
xfs_dir2_byte_to_dataptr(mp, curoff) & 0x7fffffff,
be64_to_cpu(dep->inumber), DT_UNKNOWN))
break;
/*
* Advance to next entry in the block.
*/
ptr += length;
curoff += length;
/* bufsize may have just been a guess; don't go negative */
bufsize = bufsize > length ? bufsize - length : 0;
}
/*
* All done. Set output offset value to current offset.
*/
if (curoff > xfs_dir2_dataptr_to_byte(mp, XFS_DIR2_MAX_DATAPTR))
*offset = XFS_DIR2_MAX_DATAPTR & 0x7fffffff;
else
*offset = xfs_dir2_byte_to_dataptr(mp, curoff) & 0x7fffffff;
kmem_free(map);
if (bp)
xfs_da_brelse(NULL, bp);
return error;
}
