/*
* This returns the number of iovecs needed to log the given inode item.
*
* We need one iovec for the inode log format structure, one for the
* inode core, and possibly one for the inode data/extents/b-tree root
* and one for the inode attribute data/extents/b-tree root.
*/
STATIC uint
xfs_inode_item_size(
xfs_inode_log_item_t *iip)
{
uint nvecs;
xfs_inode_t *ip;
ip = iip->ili_inode;
nvecs = 2;
/*
* Only log the data/extents/b-tree root if there is something
* left to log.
*/
iip->ili_format.ilf_fields |= XFS_ILOG_CORE;
switch (ip->i_d.di_format) {
case XFS_DINODE_FMT_EXTENTS:
iip->ili_format.ilf_fields &=
~(XFS_ILOG_DDATA | XFS_ILOG_DBROOT |
XFS_ILOG_DEV | XFS_ILOG_UUID);
if ((iip->ili_format.ilf_fields & XFS_ILOG_DEXT) &&
(ip->i_d.di_nextents > 0) &&
(ip->i_df.if_bytes > 0)) {
ASSERT(ip->i_df.if_u1.if_extents != NULL);
nvecs++;
} else {
iip->ili_format.ilf_fields &= ~XFS_ILOG_DEXT;
}
break;
case XFS_DINODE_FMT_BTREE:
ASSERT(ip->i_df.if_ext_max ==
XFS_IFORK_DSIZE(ip) / (uint)sizeof(xfs_bmbt_rec_t));
iip->ili_format.ilf_fields &=
~(XFS_ILOG_DDATA | XFS_ILOG_DEXT |
XFS_ILOG_DEV | XFS_ILOG_UUID);
if ((iip->ili_format.ilf_fields & XFS_ILOG_DBROOT) &&
(ip->i_df.if_broot_bytes > 0)) {
ASSERT(ip->i_df.if_broot != NULL);
nvecs++;
} else {
ASSERT(!(iip->ili_format.ilf_fields &
XFS_ILOG_DBROOT));
#ifdef XFS_TRANS_DEBUG
if (iip->ili_root_size > 0) {
ASSERT(iip->ili_root_size ==
ip->i_df.if_broot_bytes);
ASSERT(memcmp(iip->ili_orig_root,
ip->i_df.if_broot,
iip->ili_root_size) == 0);
} else {
ASSERT(ip->i_df.if_broot_bytes == 0);
}
#endif
iip->ili_format.ilf_fields &= ~XFS_ILOG_DBROOT;
}
break;
case XFS_DINODE_FMT_LOCAL:
iip->ili_format.ilf_fields &=
~(XFS_ILOG_DEXT | XFS_ILOG_DBROOT |
XFS_ILOG_DEV | XFS_ILOG_UUID);
if ((iip->ili_format.ilf_fields & XFS_ILOG_DDATA) &&
(ip->i_df.if_bytes > 0)) {
ASSERT(ip->i_df.if_u1.if_data != NULL);
ASSERT(ip->i_d.di_size > 0);
nvecs++;
} else {
iip->ili_format.ilf_fields &= ~XFS_ILOG_DDATA;
}
break;
case XFS_DINODE_FMT_DEV:
iip->ili_format.ilf_fields &=
~(XFS_ILOG_DDATA | XFS_ILOG_DBROOT |
XFS_ILOG_DEXT | XFS_ILOG_UUID);
break;
case XFS_DINODE_FMT_UUID:
iip->ili_format.ilf_fields &=
~(XFS_ILOG_DDATA | XFS_ILOG_DBROOT |
XFS_ILOG_DEXT | XFS_ILOG_DEV);
break;
default:
ASSERT(0);
break;
}
/*
* If there are no attributes associated with this file,
* then there cannot be anything more to log.
* Clear all attribute-related log flags.
*/
if (!XFS_IFORK_Q(ip)) {
iip->ili_format.ilf_fields &=
~(XFS_ILOG_ADATA | XFS_ILOG_ABROOT | XFS_ILOG_AEXT);
return nvecs;
}
/*
* Log any necessary attribute data.
*/
switch (ip->i_d.di_aformat) {
case XFS_DINODE_FMT_EXTENTS:
iip->ili_format.ilf_fields &=
~(XFS_ILOG_ADATA | XFS_ILOG_ABROOT);
if ((iip->ili_format.ilf_fields & XFS_ILOG_AEXT) &&
(ip->i_d.di_anextents > 0) &&
(ip->i_afp->if_bytes > 0)) {
ASSERT(ip->i_afp->if_u1.if_extents != NULL);
nvecs++;
} else {
iip->ili_format.ilf_fields &= ~XFS_ILOG_AEXT;
}
break;
case XFS_DINODE_FMT_BTREE:
iip->ili_format.ilf_fields &=
~(XFS_ILOG_ADATA | XFS_ILOG_AEXT);
if ((iip->ili_format.ilf_fields & XFS_ILOG_ABROOT) &&
(ip->i_afp->if_broot_bytes > 0)) {
ASSERT(ip->i_afp->if_broot != NULL);
nvecs++;
} else {
iip->ili_format.ilf_fields &= ~XFS_ILOG_ABROOT;
}
break;
case XFS_DINODE_FMT_LOCAL:
iip->ili_format.ilf_fields &=
~(XFS_ILOG_AEXT | XFS_ILOG_ABROOT);
if ((iip->ili_format.ilf_fields & XFS_ILOG_ADATA) &&
(ip->i_afp->if_bytes > 0)) {
ASSERT(ip->i_afp->if_u1.if_data != NULL);
nvecs++;
} else {
iip->ili_format.ilf_fields &= ~XFS_ILOG_ADATA;
}
break;
default:
ASSERT(0);
break;
}
return nvecs;
}
/*
* Remove an entry from a block format directory.
* If that makes the block small enough to fit in shortform, transform it.
*/
int /* error */
xfs_dir2_block_removename(
xfs_da_args_t *args) /* directory operation args */
{
xfs_dir2_data_hdr_t *hdr; /* block header */
xfs_dir2_leaf_entry_t *blp; /* block leaf pointer */
struct xfs_buf *bp; /* block buffer */
xfs_dir2_block_tail_t *btp; /* block tail */
xfs_dir2_data_entry_t *dep; /* block data entry */
xfs_inode_t *dp; /* incore inode */
int ent; /* block leaf entry index */
int error; /* error return value */
xfs_mount_t *mp; /* filesystem mount point */
int needlog; /* need to log block header */
int needscan; /* need to fixup bestfree */
xfs_dir2_sf_hdr_t sfh; /* shortform header */
int size; /* shortform size */
xfs_trans_t *tp; /* transaction pointer */
trace_xfs_dir2_block_removename(args);
/*
* Look up the entry in the block. Gets the buffer and entry index.
* It will always be there, the vnodeops level does a lookup first.
*/
if ((error = xfs_dir2_block_lookup_int(args, &bp, &ent))) {
return error;
}
dp = args->dp;
tp = args->trans;
mp = dp->i_mount;
hdr = bp->b_addr;
btp = xfs_dir2_block_tail_p(mp, hdr);
blp = xfs_dir2_block_leaf_p(btp);
/*
* Point to the data entry using the leaf entry.
*/
dep = (xfs_dir2_data_entry_t *)
((char *)hdr + xfs_dir2_dataptr_to_off(mp, be32_to_cpu(blp[ent].address)));
/*
* Mark the data entry's space free.
*/
needlog = needscan = 0;
xfs_dir2_data_make_free(tp, dp, bp,
(xfs_dir2_data_aoff_t)((char *)dep - (char *)hdr),
dp->d_ops->data_entsize(dep->namelen), &needlog, &needscan);
/*
* Fix up the block tail.
*/
be32_add_cpu(&btp->stale, 1);
xfs_dir2_block_log_tail(tp, bp);
/*
* Remove the leaf entry by marking it stale.
*/
blp[ent].address = cpu_to_be32(XFS_DIR2_NULL_DATAPTR);
xfs_dir2_block_log_leaf(tp, bp, ent, ent);
/*
* Fix up bestfree, log the header if necessary.
*/
if (needscan)
xfs_dir2_data_freescan(dp, hdr, &needlog);
if (needlog)
xfs_dir2_data_log_header(tp, dp, bp);
xfs_dir3_data_check(dp, bp);
/*
* See if the size as a shortform is good enough.
*/
size = xfs_dir2_block_sfsize(dp, hdr, &sfh);
if (size > XFS_IFORK_DSIZE(dp))
return 0;
/*
* If it works, do the conversion.
*/
return xfs_dir2_block_to_sf(args, bp, size, &sfh);
}
/*
* Add a name to a shortform directory.
* There are two algorithms, "easy" and "hard" which we decide on
* before changing anything.
* Convert to block form if necessary, if the new entry won't fit.
*/
int /* error */
xfs_dir2_sf_addname(
xfs_da_args_t *args) /* operation arguments */
{
int add_entsize; /* size of the new entry */
xfs_inode_t *dp; /* incore directory inode */
int error; /* error return value */
int incr_isize; /* total change in size */
int new_isize; /* di_size after adding name */
int objchange; /* changing to 8-byte inodes */
xfs_dir2_data_aoff_t offset = 0; /* offset for new entry */
int old_isize; /* di_size before adding name */
int pick; /* which algorithm to use */
xfs_dir2_sf_t *sfp; /* shortform structure */
xfs_dir2_sf_entry_t *sfep = NULL; /* shortform entry */
xfs_dir2_trace_args("sf_addname", args);
ASSERT(xfs_dir2_sf_lookup(args) == ENOENT);
dp = args->dp;
ASSERT(dp->i_df.if_flags & XFS_IFINLINE);
/*
* Make sure the shortform value has some of its header.
*/
if (dp->i_d.di_size < offsetof(xfs_dir2_sf_hdr_t, parent)) {
ASSERT(XFS_FORCED_SHUTDOWN(dp->i_mount));
return XFS_ERROR(EIO);
}
ASSERT(dp->i_df.if_bytes == dp->i_d.di_size);
ASSERT(dp->i_df.if_u1.if_data != NULL);
sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
ASSERT(dp->i_d.di_size >= XFS_DIR2_SF_HDR_SIZE(sfp->hdr.i8count));
/*
* Compute entry (and change in) size.
*/
add_entsize = XFS_DIR2_SF_ENTSIZE_BYNAME(sfp, args->namelen);
incr_isize = add_entsize;
objchange = 0;
#if XFS_BIG_INUMS
/*
* Do we have to change to 8 byte inodes?
*/
if (args->inumber > XFS_DIR2_MAX_SHORT_INUM && sfp->hdr.i8count == 0) {
/*
* Yes, adjust the entry size and the total size.
*/
add_entsize +=
(uint)sizeof(xfs_dir2_ino8_t) -
(uint)sizeof(xfs_dir2_ino4_t);
incr_isize +=
(sfp->hdr.count + 2) *
((uint)sizeof(xfs_dir2_ino8_t) -
(uint)sizeof(xfs_dir2_ino4_t));
objchange = 1;
}
#endif
old_isize = (int)dp->i_d.di_size;
new_isize = old_isize + incr_isize;
/*
* Won't fit as shortform any more (due to size),
* or the pick routine says it won't (due to offset values).
*/
if (new_isize > XFS_IFORK_DSIZE(dp) ||
(pick =
xfs_dir2_sf_addname_pick(args, objchange, &sfep, &offset)) == 0) {
/*
* Just checking or no space reservation, it doesn't fit.
*/
if (args->justcheck || args->total == 0)
return XFS_ERROR(ENOSPC);
/*
* Convert to block form then add the name.
*/
error = xfs_dir2_sf_to_block(args);
if (error)
return error;
return xfs_dir2_block_addname(args);
}
/*
* Just checking, it fits.
*/
if (args->justcheck)
return 0;
/*
* Do it the easy way - just add it at the end.
*/
if (pick == 1)
xfs_dir2_sf_addname_easy(args, sfep, offset, new_isize);
/*
* Do it the hard way - look for a place to insert the new entry.
* Convert to 8 byte inode numbers first if necessary.
*/
else {
ASSERT(pick == 2);
#if XFS_BIG_INUMS
if (objchange)
xfs_dir2_sf_toino8(args);
#endif
xfs_dir2_sf_addname_hard(args, objchange, new_isize);
}
xfs_trans_log_inode(args->trans, dp, XFS_ILOG_CORE | XFS_ILOG_DDATA);
return 0;
}
/*
* Given a block directory (dp/block), calculate its size as a shortform (sf)
* directory and a header for the sf directory, if it will fit it the
* space currently present in the inode. If it won't fit, the output
* size is too big (but not accurate).
*/
int /* size for sf form */
xfs_dir2_block_sfsize(
xfs_inode_t *dp, /* incore inode pointer */
xfs_dir2_block_t *block, /* block directory data */
xfs_dir2_sf_hdr_t *sfhp) /* output: header for sf form */
{
xfs_dir2_dataptr_t addr; /* data entry address */
xfs_dir2_leaf_entry_t *blp; /* leaf area of the block */
xfs_dir2_block_tail_t *btp; /* tail area of the block */
int count; /* shortform entry count */
xfs_dir2_data_entry_t *dep; /* data entry in the block */
int i; /* block entry index */
int i8count; /* count of big-inode entries */
int isdot; /* entry is "." */
int isdotdot; /* entry is ".." */
xfs_mount_t *mp; /* mount structure pointer */
int namelen; /* total name bytes */
xfs_ino_t parent = 0; /* parent inode number */
int size=0; /* total computed size */
mp = dp->i_mount;
count = i8count = namelen = 0;
btp = XFS_DIR2_BLOCK_TAIL_P(mp, block);
blp = XFS_DIR2_BLOCK_LEAF_P(btp);
/*
* Iterate over the block's data entries by using the leaf pointers.
*/
for (i = 0; i < INT_GET(btp->count, ARCH_CONVERT); i++) {
if ((addr = INT_GET(blp[i].address, ARCH_CONVERT)) == XFS_DIR2_NULL_DATAPTR)
continue;
/*
* Calculate the pointer to the entry at hand.
*/
dep = (xfs_dir2_data_entry_t *)
((char *)block + XFS_DIR2_DATAPTR_TO_OFF(mp, addr));
/*
* Detect . and .., so we can special-case them.
* . is not included in sf directories.
* .. is included by just the parent inode number.
*/
isdot = dep->namelen == 1 && dep->name[0] == '.';
isdotdot =
dep->namelen == 2 &&
dep->name[0] == '.' && dep->name[1] == '.';
#if XFS_BIG_INUMS
if (!isdot)
i8count += INT_GET(dep->inumber, ARCH_CONVERT) > XFS_DIR2_MAX_SHORT_INUM;
#endif
if (!isdot && !isdotdot) {
count++;
namelen += dep->namelen;
} else if (isdotdot)
parent = INT_GET(dep->inumber, ARCH_CONVERT);
/*
* Calculate the new size, see if we should give up yet.
*/
/* Add by Jerry 2007/10/25
Fixed bug that XFS file system corrupted under Winthrax test.
This is the "special" structure alignment on the ARM.
Port a workaround solution from a patch on
http://lists.arm.linux.org.uk/pipermail/linux-arm-kernel/2004-March/020287.html
and http://www.nas-central.org/index.php/Buffalo_ARM9_Kernel_Port#XFS_Arm_Issues
This patch was never accepted by XFS because the code is incorrect for all other platforms.
*/
#if 0
size = XFS_DIR2_SF_HDR_SIZE(i8count) + /* header */
count + /* namelen */
count * (uint)sizeof(xfs_dir2_sf_off_t) + /* offset */
namelen + /* name */
(i8count ? /* inumber */
(uint)sizeof(xfs_dir2_ino8_t) * count :
(uint)sizeof(xfs_dir2_ino4_t) * count);
#else
size = XFS_DIR2_SF_HDR_SIZE(i8count) /* header */
+ namelen
+ (count * (sizeof(xfs_dir2_sf_entry_t) - 1))
- (count * (((i8count == 0) ? 1 : 0) *
(sizeof(xfs_dir2_ino8_t) - sizeof(xfs_dir2_ino4_t))));
#endif
if (size > XFS_IFORK_DSIZE(dp))
return size; /* size value is a failure */
}
/*
* Create the output header, if it worked.
*/
sfhp->count = count;
sfhp->i8count = i8count;
XFS_DIR2_SF_PUT_INUMBER((xfs_dir2_sf_t *)sfhp, &parent, &sfhp->parent);
return size;
}
int /* */
xfs_dir2_sf_addname(
xfs_da_args_t *args) /* */
{
int add_entsize; /* */
xfs_inode_t *dp; /* */
int error; /* */
int incr_isize; /* */
int new_isize; /* */
int objchange; /* */
xfs_dir2_data_aoff_t offset = 0; /* */
int old_isize; /* */
int pick; /* */
xfs_dir2_sf_hdr_t *sfp; /* */
xfs_dir2_sf_entry_t *sfep = NULL; /* */
trace_xfs_dir2_sf_addname(args);
ASSERT(xfs_dir2_sf_lookup(args) == ENOENT);
dp = args->dp;
ASSERT(dp->i_df.if_flags & XFS_IFINLINE);
/*
*/
if (dp->i_d.di_size < offsetof(xfs_dir2_sf_hdr_t, parent)) {
ASSERT(XFS_FORCED_SHUTDOWN(dp->i_mount));
return XFS_ERROR(EIO);
}
ASSERT(dp->i_df.if_bytes == dp->i_d.di_size);
ASSERT(dp->i_df.if_u1.if_data != NULL);
sfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
ASSERT(dp->i_d.di_size >= xfs_dir2_sf_hdr_size(sfp->i8count));
/*
*/
add_entsize = xfs_dir2_sf_entsize(sfp, args->namelen);
incr_isize = add_entsize;
objchange = 0;
#if XFS_BIG_INUMS
/*
*/
if (args->inumber > XFS_DIR2_MAX_SHORT_INUM && sfp->i8count == 0) {
/*
*/
add_entsize +=
(uint)sizeof(xfs_dir2_ino8_t) -
(uint)sizeof(xfs_dir2_ino4_t);
incr_isize +=
(sfp->count + 2) *
((uint)sizeof(xfs_dir2_ino8_t) -
(uint)sizeof(xfs_dir2_ino4_t));
objchange = 1;
}
#endif
old_isize = (int)dp->i_d.di_size;
new_isize = old_isize + incr_isize;
/*
*/
if (new_isize > XFS_IFORK_DSIZE(dp) ||
(pick =
xfs_dir2_sf_addname_pick(args, objchange, &sfep, &offset)) == 0) {
/*
*/
if ((args->op_flags & XFS_DA_OP_JUSTCHECK) || args->total == 0)
return XFS_ERROR(ENOSPC);
/*
*/
error = xfs_dir2_sf_to_block(args);
if (error)
return error;
return xfs_dir2_block_addname(args);
}
/*
*/
if (args->op_flags & XFS_DA_OP_JUSTCHECK)
return 0;
/*
*/
if (pick == 1)
xfs_dir2_sf_addname_easy(args, sfep, offset, new_isize);
/*
*/
else {
ASSERT(pick == 2);
#if XFS_BIG_INUMS
if (objchange)
xfs_dir2_sf_toino8(args);
#endif
xfs_dir2_sf_addname_hard(args, objchange, new_isize);
}
xfs_trans_log_inode(args->trans, dp, XFS_ILOG_CORE | XFS_ILOG_DDATA);
return 0;
}
/*
* Replace the inode number of an entry in a shortform directory.
*/
int /* error */
xfs_dir2_sf_replace(
xfs_da_args_t *args) /* operation arguments */
{
xfs_inode_t *dp; /* incore directory inode */
int i; /* entry index */
#if XFS_BIG_INUMS || defined(DEBUG)
xfs_ino_t ino=0; /* entry old inode number */
#endif
#if XFS_BIG_INUMS
int i8elevated; /* sf_toino8 set i8count=1 */
#endif
xfs_dir2_sf_entry_t *sfep; /* shortform directory entry */
xfs_dir2_sf_t *sfp; /* shortform structure */
xfs_dir2_trace_args("sf_replace", args);
dp = args->dp;
ASSERT(dp->i_df.if_flags & XFS_IFINLINE);
/*
* Bail out if the shortform directory is way too small.
*/
if (dp->i_d.di_size < offsetof(xfs_dir2_sf_hdr_t, parent)) {
ASSERT(XFS_FORCED_SHUTDOWN(dp->i_mount));
return XFS_ERROR(EIO);
}
ASSERT(dp->i_df.if_bytes == dp->i_d.di_size);
ASSERT(dp->i_df.if_u1.if_data != NULL);
sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
ASSERT(dp->i_d.di_size >= XFS_DIR2_SF_HDR_SIZE(sfp->hdr.i8count));
#if XFS_BIG_INUMS
/*
* New inode number is large, and need to convert to 8-byte inodes.
*/
if (args->inumber > XFS_DIR2_MAX_SHORT_INUM && sfp->hdr.i8count == 0) {
int error; /* error return value */
int newsize; /* new inode size */
newsize =
dp->i_df.if_bytes +
(sfp->hdr.count + 1) *
((uint)sizeof(xfs_dir2_ino8_t) -
(uint)sizeof(xfs_dir2_ino4_t));
/*
* Won't fit as shortform, convert to block then do replace.
*/
if (newsize > XFS_IFORK_DSIZE(dp)) {
error = xfs_dir2_sf_to_block(args);
if (error) {
return error;
}
return xfs_dir2_block_replace(args);
}
/*
* Still fits, convert to 8-byte now.
*/
xfs_dir2_sf_toino8(args);
i8elevated = 1;
sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
} else
i8elevated = 0;
#endif
ASSERT(args->namelen != 1 || args->name[0] != '.');
/*
* Replace ..'s entry.
*/
if (args->namelen == 2 &&
args->name[0] == '.' && args->name[1] == '.') {
#if XFS_BIG_INUMS || defined(DEBUG)
ino = XFS_DIR2_SF_GET_INUMBER(sfp, &sfp->hdr.parent);
ASSERT(args->inumber != ino);
#endif
XFS_DIR2_SF_PUT_INUMBER(sfp, &args->inumber, &sfp->hdr.parent);
}
/*
* Normal entry, look for the name.
*/
else {
for (i = 0, sfep = XFS_DIR2_SF_FIRSTENTRY(sfp);
i < sfp->hdr.count;
i++, sfep = XFS_DIR2_SF_NEXTENTRY(sfp, sfep)) {
if (sfep->namelen == args->namelen &&
sfep->name[0] == args->name[0] &&
memcmp(args->name, sfep->name, args->namelen) == 0) {
#if XFS_BIG_INUMS || defined(DEBUG)
ino = XFS_DIR2_SF_GET_INUMBER(sfp,
XFS_DIR2_SF_INUMBERP(sfep));
ASSERT(args->inumber != ino);
#endif
XFS_DIR2_SF_PUT_INUMBER(sfp, &args->inumber,
XFS_DIR2_SF_INUMBERP(sfep));
break;
}
}
/*
* Didn't find it.
*/
if (i == sfp->hdr.count) {
ASSERT(args->oknoent);
#if XFS_BIG_INUMS
if (i8elevated)
xfs_dir2_sf_toino4(args);
#endif
return XFS_ERROR(ENOENT);
}
}
#if XFS_BIG_INUMS
/*
* See if the old number was large, the new number is small.
*/
if (ino > XFS_DIR2_MAX_SHORT_INUM &&
args->inumber <= XFS_DIR2_MAX_SHORT_INUM) {
/*
* And the old count was one, so need to convert to small.
*/
if (sfp->hdr.i8count == 1)
xfs_dir2_sf_toino4(args);
else
sfp->hdr.i8count--;
}
/*
* See if the old number was small, the new number is large.
*/
if (ino <= XFS_DIR2_MAX_SHORT_INUM &&
args->inumber > XFS_DIR2_MAX_SHORT_INUM) {
/*
* add to the i8count unless we just converted to 8-byte
* inodes (which does an implied i8count = 1)
*/
ASSERT(sfp->hdr.i8count != 0);
if (!i8elevated)
sfp->hdr.i8count++;
}
#endif
xfs_dir2_sf_check(args);
xfs_trans_log_inode(args->trans, dp, XFS_ILOG_DDATA);
return 0;
}
int /* */
xfs_dir2_sf_replace(
xfs_da_args_t *args) /* */
{
xfs_inode_t *dp; /* */
int i; /* */
#if XFS_BIG_INUMS || defined(DEBUG)
xfs_ino_t ino=0; /* */
#endif
#if XFS_BIG_INUMS
int i8elevated; /* */
#endif
xfs_dir2_sf_entry_t *sfep; /* */
xfs_dir2_sf_hdr_t *sfp; /* */
trace_xfs_dir2_sf_replace(args);
dp = args->dp;
ASSERT(dp->i_df.if_flags & XFS_IFINLINE);
/*
*/
if (dp->i_d.di_size < offsetof(xfs_dir2_sf_hdr_t, parent)) {
ASSERT(XFS_FORCED_SHUTDOWN(dp->i_mount));
return XFS_ERROR(EIO);
}
ASSERT(dp->i_df.if_bytes == dp->i_d.di_size);
ASSERT(dp->i_df.if_u1.if_data != NULL);
sfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
ASSERT(dp->i_d.di_size >= xfs_dir2_sf_hdr_size(sfp->i8count));
#if XFS_BIG_INUMS
/*
*/
if (args->inumber > XFS_DIR2_MAX_SHORT_INUM && sfp->i8count == 0) {
int error; /* */
int newsize; /* */
newsize =
dp->i_df.if_bytes +
(sfp->count + 1) *
((uint)sizeof(xfs_dir2_ino8_t) -
(uint)sizeof(xfs_dir2_ino4_t));
/*
*/
if (newsize > XFS_IFORK_DSIZE(dp)) {
error = xfs_dir2_sf_to_block(args);
if (error) {
return error;
}
return xfs_dir2_block_replace(args);
}
/*
*/
xfs_dir2_sf_toino8(args);
i8elevated = 1;
sfp = (xfs_dir2_sf_hdr_t *)dp->i_df.if_u1.if_data;
} else
i8elevated = 0;
#endif
ASSERT(args->namelen != 1 || args->name[0] != '.');
/*
*/
if (args->namelen == 2 &&
args->name[0] == '.' && args->name[1] == '.') {
#if XFS_BIG_INUMS || defined(DEBUG)
ino = xfs_dir2_sf_get_parent_ino(sfp);
ASSERT(args->inumber != ino);
#endif
xfs_dir2_sf_put_parent_ino(sfp, args->inumber);
}
/*
*/
else {
for (i = 0, sfep = xfs_dir2_sf_firstentry(sfp);
i < sfp->count;
i++, sfep = xfs_dir2_sf_nextentry(sfp, sfep)) {
if (xfs_da_compname(args, sfep->name, sfep->namelen) ==
XFS_CMP_EXACT) {
#if XFS_BIG_INUMS || defined(DEBUG)
ino = xfs_dir2_sfe_get_ino(sfp, sfep);
ASSERT(args->inumber != ino);
#endif
xfs_dir2_sfe_put_ino(sfp, sfep, args->inumber);
break;
}
}
/*
*/
if (i == sfp->count) {
ASSERT(args->op_flags & XFS_DA_OP_OKNOENT);
#if XFS_BIG_INUMS
if (i8elevated)
xfs_dir2_sf_toino4(args);
#endif
return XFS_ERROR(ENOENT);
}
}
#if XFS_BIG_INUMS
/*
*/
if (ino > XFS_DIR2_MAX_SHORT_INUM &&
args->inumber <= XFS_DIR2_MAX_SHORT_INUM) {
/*
*/
if (sfp->i8count == 1)
xfs_dir2_sf_toino4(args);
else
sfp->i8count--;
}
/*
*/
if (ino <= XFS_DIR2_MAX_SHORT_INUM &&
args->inumber > XFS_DIR2_MAX_SHORT_INUM) {
/*
*/
ASSERT(sfp->i8count != 0);
if (!i8elevated)
sfp->i8count++;
}
#endif
xfs_dir2_sf_check(args);
xfs_trans_log_inode(args->trans, dp, XFS_ILOG_DDATA);
return 0;
}
int /* */
xfs_dir2_block_sfsize(
xfs_inode_t *dp, /* */
xfs_dir2_data_hdr_t *hdr, /* */
xfs_dir2_sf_hdr_t *sfhp) /* */
{
xfs_dir2_dataptr_t addr; /* */
xfs_dir2_leaf_entry_t *blp; /* */
xfs_dir2_block_tail_t *btp; /* */
int count; /* */
xfs_dir2_data_entry_t *dep; /* */
int i; /* */
int i8count; /* */
int isdot; /* */
int isdotdot; /* */
xfs_mount_t *mp; /* */
int namelen; /* */
xfs_ino_t parent = 0; /* */
int size=0; /* */
mp = dp->i_mount;
count = i8count = namelen = 0;
btp = xfs_dir2_block_tail_p(mp, hdr);
blp = xfs_dir2_block_leaf_p(btp);
/*
*/
for (i = 0; i < be32_to_cpu(btp->count); i++) {
if ((addr = be32_to_cpu(blp[i].address)) == XFS_DIR2_NULL_DATAPTR)
continue;
/*
*/
dep = (xfs_dir2_data_entry_t *)
((char *)hdr + xfs_dir2_dataptr_to_off(mp, addr));
/*
*/
isdot = dep->namelen == 1 && dep->name[0] == '.';
isdotdot =
dep->namelen == 2 &&
dep->name[0] == '.' && dep->name[1] == '.';
#if XFS_BIG_INUMS
if (!isdot)
i8count += be64_to_cpu(dep->inumber) > XFS_DIR2_MAX_SHORT_INUM;
#endif
if (!isdot && !isdotdot) {
count++;
namelen += dep->namelen;
} else if (isdotdot)
parent = be64_to_cpu(dep->inumber);
/*
*/
size = xfs_dir2_sf_hdr_size(i8count) + /* */
count + /* */
count * (uint)sizeof(xfs_dir2_sf_off_t) + /* */
namelen + /* */
(i8count ? /* */
(uint)sizeof(xfs_dir2_ino8_t) * count :
(uint)sizeof(xfs_dir2_ino4_t) * count);
if (size > XFS_IFORK_DSIZE(dp))
return size; /* */
}
/*
*/
sfhp->count = count;
sfhp->i8count = i8count;
xfs_dir2_sf_put_parent_ino(sfhp, parent);
return size;
}
/*
* Look up an inode by number in the given file system.
* The inode is looked up in the cache held in each AG.
* If the inode is found in the cache, initialise the vfs inode
* if necessary.
*
* If it is not in core, read it in from the file system's device,
* add it to the cache and initialise the vfs inode.
*
* The inode is locked according to the value of the lock_flags parameter.
* This flag parameter indicates how and if the inode's IO lock and inode lock
* should be taken.
*
* mp -- the mount point structure for the current file system. It points
* to the inode hash table.
* tp -- a pointer to the current transaction if there is one. This is
* simply passed through to the xfs_iread() call.
* ino -- the number of the inode desired. This is the unique identifier
* within the file system for the inode being requested.
* lock_flags -- flags indicating how to lock the inode. See the comment
* for xfs_ilock() for a list of valid values.
*/
int
xfs_iget(
xfs_mount_t *mp,
xfs_trans_t *tp,
xfs_ino_t ino,
uint flags,
uint lock_flags,
xfs_inode_t **ipp)
{
xfs_inode_t *ip;
int error;
xfs_perag_t *pag;
xfs_agino_t agino;
/* reject inode numbers outside existing AGs */
if (!ino || XFS_INO_TO_AGNO(mp, ino) >= mp->m_sb.sb_agcount)
return EINVAL;
/* get the perag structure and ensure that it's inode capable */
pag = xfs_perag_get(mp, XFS_INO_TO_AGNO(mp, ino));
agino = XFS_INO_TO_AGINO(mp, ino);
again:
error = 0;
rcu_read_lock();
ip = radix_tree_lookup(&pag->pag_ici_root, agino);
if (ip) {
error = xfs_iget_cache_hit(pag, ip, ino, flags, lock_flags);
if (error)
goto out_error_or_again;
} else {
rcu_read_unlock();
XFS_STATS_INC(xs_ig_missed);
error = xfs_iget_cache_miss(mp, pag, tp, ino, &ip,
flags, lock_flags);
if (error)
goto out_error_or_again;
}
xfs_perag_put(pag);
*ipp = ip;
ASSERT(ip->i_df.if_ext_max ==
XFS_IFORK_DSIZE(ip) / sizeof(xfs_bmbt_rec_t));
/*
* If we have a real type for an on-disk inode, we can set ops(&unlock)
* now. If it's a new inode being created, xfs_ialloc will handle it.
*/
if (xfs_iflags_test(ip, XFS_INEW) && ip->i_d.di_mode != 0)
xfs_setup_inode(ip);
return 0;
out_error_or_again:
if (error == EAGAIN) {
delay(1);
goto again;
}
xfs_perag_put(pag);
return error;
}
/*
* Convert a V2 leaf directory to a V2 block directory if possible.
*/
int /* error */
xfs_dir2_leaf_to_block(
xfs_da_args_t *args, /* operation arguments */
struct xfs_buf *lbp, /* leaf buffer */
struct xfs_buf *dbp) /* data buffer */
{
__be16 *bestsp; /* leaf bests table */
xfs_dir2_data_hdr_t *hdr; /* block header */
xfs_dir2_block_tail_t *btp; /* block tail */
xfs_inode_t *dp; /* incore directory inode */
xfs_dir2_data_unused_t *dup; /* unused data entry */
int error; /* error return value */
int from; /* leaf from index */
xfs_dir2_leaf_t *leaf; /* leaf structure */
xfs_dir2_leaf_entry_t *lep; /* leaf entry */
xfs_dir2_leaf_tail_t *ltp; /* leaf tail structure */
xfs_mount_t *mp; /* file system mount point */
int needlog; /* need to log data header */
int needscan; /* need to scan for bestfree */
xfs_dir2_sf_hdr_t sfh; /* shortform header */
int size; /* bytes used */
__be16 *tagp; /* end of entry (tag) */
int to; /* block/leaf to index */
xfs_trans_t *tp; /* transaction pointer */
struct xfs_dir2_leaf_entry *ents;
struct xfs_dir3_icleaf_hdr leafhdr;
trace_xfs_dir2_leaf_to_block(args);
dp = args->dp;
tp = args->trans;
mp = dp->i_mount;
leaf = lbp->b_addr;
dp->d_ops->leaf_hdr_from_disk(&leafhdr, leaf);
ents = dp->d_ops->leaf_ents_p(leaf);
ltp = xfs_dir2_leaf_tail_p(args->geo, leaf);
ASSERT(leafhdr.magic == XFS_DIR2_LEAF1_MAGIC ||
leafhdr.magic == XFS_DIR3_LEAF1_MAGIC);
/*
* If there are data blocks other than the first one, take this
* opportunity to remove trailing empty data blocks that may have
* been left behind during no-space-reservation operations.
* These will show up in the leaf bests table.
*/
while (dp->i_d.di_size > args->geo->blksize) {
int hdrsz;
hdrsz = dp->d_ops->data_entry_offset;
bestsp = xfs_dir2_leaf_bests_p(ltp);
if (be16_to_cpu(bestsp[be32_to_cpu(ltp->bestcount) - 1]) ==
args->geo->blksize - hdrsz) {
if ((error =
xfs_dir2_leaf_trim_data(args, lbp,
(xfs_dir2_db_t)(be32_to_cpu(ltp->bestcount) - 1))))
return error;
} else
return 0;
}
/*
* Read the data block if we don't already have it, give up if it fails.
*/
if (!dbp) {
error = xfs_dir3_data_read(tp, dp, args->geo->datablk, -1, &dbp);
if (error)
return error;
}
hdr = dbp->b_addr;
ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
hdr->magic == cpu_to_be32(XFS_DIR3_DATA_MAGIC));
/*
* Size of the "leaf" area in the block.
*/
size = (uint)sizeof(xfs_dir2_block_tail_t) +
(uint)sizeof(*lep) * (leafhdr.count - leafhdr.stale);
/*
* Look at the last data entry.
*/
tagp = (__be16 *)((char *)hdr + args->geo->blksize) - 1;
dup = (xfs_dir2_data_unused_t *)((char *)hdr + be16_to_cpu(*tagp));
/*
* If it's not free or is too short we can't do it.
*/
if (be16_to_cpu(dup->freetag) != XFS_DIR2_DATA_FREE_TAG ||
be16_to_cpu(dup->length) < size)
return 0;
/*
* Start converting it to block form.
*/
xfs_dir3_block_init(mp, tp, dbp, dp);
needlog = 1;
needscan = 0;
/*
* Use up the space at the end of the block (blp/btp).
*/
xfs_dir2_data_use_free(args, dbp, dup, args->geo->blksize - size, size,
&needlog, &needscan);
/*
* Initialize the block tail.
*/
btp = xfs_dir2_block_tail_p(args->geo, hdr);
btp->count = cpu_to_be32(leafhdr.count - leafhdr.stale);
btp->stale = 0;
xfs_dir2_block_log_tail(tp, dbp);
/*
* Initialize the block leaf area. We compact out stale entries.
*/
lep = xfs_dir2_block_leaf_p(btp);
for (from = to = 0; from < leafhdr.count; from++) {
if (ents[from].address == cpu_to_be32(XFS_DIR2_NULL_DATAPTR))
continue;
lep[to++] = ents[from];
}
ASSERT(to == be32_to_cpu(btp->count));
xfs_dir2_block_log_leaf(tp, dbp, 0, be32_to_cpu(btp->count) - 1);
/*
* Scan the bestfree if we need it and log the data block header.
*/
if (needscan)
xfs_dir2_data_freescan(dp, hdr, &needlog);
if (needlog)
xfs_dir2_data_log_header(args, dbp);
/*
* Pitch the old leaf block.
*/
error = xfs_da_shrink_inode(args, args->geo->leafblk, lbp);
if (error)
return error;
/*
* Now see if the resulting block can be shrunken to shortform.
*/
size = xfs_dir2_block_sfsize(dp, hdr, &sfh);
if (size > XFS_IFORK_DSIZE(dp))
return 0;
return xfs_dir2_block_to_sf(args, dbp, size, &sfh);
}
int
xfs_symlink(
struct xfs_inode *dp,
struct xfs_name *link_name,
const char *target_path,
umode_t mode,
struct xfs_inode **ipp)
{
struct xfs_mount *mp = dp->i_mount;
struct xfs_trans *tp = NULL;
struct xfs_inode *ip = NULL;
int error = 0;
int pathlen;
struct xfs_bmap_free free_list;
xfs_fsblock_t first_block;
bool unlock_dp_on_error = false;
uint cancel_flags;
int committed;
xfs_fileoff_t first_fsb;
xfs_filblks_t fs_blocks;
int nmaps;
struct xfs_bmbt_irec mval[XFS_SYMLINK_MAPS];
xfs_daddr_t d;
const char *cur_chunk;
int byte_cnt;
int n;
xfs_buf_t *bp;
prid_t prid;
struct xfs_dquot *udqp = NULL;
struct xfs_dquot *gdqp = NULL;
struct xfs_dquot *pdqp = NULL;
uint resblks;
*ipp = NULL;
trace_xfs_symlink(dp, link_name);
if (XFS_FORCED_SHUTDOWN(mp))
return XFS_ERROR(EIO);
/*
* Check component lengths of the target path name.
*/
pathlen = strlen(target_path);
if (pathlen >= MAXPATHLEN) /* total string too long */
return XFS_ERROR(ENAMETOOLONG);
udqp = gdqp = NULL;
if (dp->i_d.di_flags & XFS_DIFLAG_PROJINHERIT)
prid = xfs_get_projid(dp);
else
prid = XFS_PROJID_DEFAULT;
/*
* Make sure that we have allocated dquot(s) on disk.
*/
error = xfs_qm_vop_dqalloc(dp,
xfs_kuid_to_uid(current_fsuid()),
xfs_kgid_to_gid(current_fsgid()), prid,
XFS_QMOPT_QUOTALL | XFS_QMOPT_INHERIT,
&udqp, &gdqp, &pdqp);
if (error)
goto std_return;
tp = xfs_trans_alloc(mp, XFS_TRANS_SYMLINK);
cancel_flags = XFS_TRANS_RELEASE_LOG_RES;
/*
* The symlink will fit into the inode data fork?
* There can't be any attributes so we get the whole variable part.
*/
if (pathlen <= XFS_LITINO(mp, dp->i_d.di_version))
fs_blocks = 0;
else
fs_blocks = xfs_symlink_blocks(mp, pathlen);
resblks = XFS_SYMLINK_SPACE_RES(mp, link_name->len, fs_blocks);
error = xfs_trans_reserve(tp, &M_RES(mp)->tr_symlink, resblks, 0);
if (error == ENOSPC && fs_blocks == 0) {
resblks = 0;
error = xfs_trans_reserve(tp, &M_RES(mp)->tr_symlink, 0, 0);
}
if (error) {
cancel_flags = 0;
goto error_return;
}
xfs_ilock(dp, XFS_ILOCK_EXCL | XFS_ILOCK_PARENT);
unlock_dp_on_error = true;
/*
* Check whether the directory allows new symlinks or not.
*/
if (dp->i_d.di_flags & XFS_DIFLAG_NOSYMLINKS) {
error = XFS_ERROR(EPERM);
goto error_return;
}
/*
* Reserve disk quota : blocks and inode.
*/
error = xfs_trans_reserve_quota(tp, mp, udqp, gdqp,
pdqp, resblks, 1, 0);
if (error)
goto error_return;
/*
* Check for ability to enter directory entry, if no space reserved.
*/
error = xfs_dir_canenter(tp, dp, link_name, resblks);
if (error)
goto error_return;
/*
* Initialize the bmap freelist prior to calling either
* bmapi or the directory create code.
*/
xfs_bmap_init(&free_list, &first_block);
/*
* Allocate an inode for the symlink.
*/
error = xfs_dir_ialloc(&tp, dp, S_IFLNK | (mode & ~S_IFMT), 1, 0,
prid, resblks > 0, &ip, NULL);
if (error) {
if (error == ENOSPC)
goto error_return;
goto error1;
}
/*
* An error after we've joined dp to the transaction will result in the
* transaction cancel unlocking dp so don't do it explicitly in the
* error path.
*/
xfs_trans_ijoin(tp, dp, XFS_ILOCK_EXCL);
unlock_dp_on_error = false;
/*
* Also attach the dquot(s) to it, if applicable.
*/
xfs_qm_vop_create_dqattach(tp, ip, udqp, gdqp, pdqp);
if (resblks)
resblks -= XFS_IALLOC_SPACE_RES(mp);
/*
* If the symlink will fit into the inode, write it inline.
*/
if (pathlen <= XFS_IFORK_DSIZE(ip)) {
xfs_idata_realloc(ip, pathlen, XFS_DATA_FORK);
memcpy(ip->i_df.if_u1.if_data, target_path, pathlen);
ip->i_d.di_size = pathlen;
/*
* The inode was initially created in extent format.
*/
ip->i_df.if_flags &= ~(XFS_IFEXTENTS | XFS_IFBROOT);
ip->i_df.if_flags |= XFS_IFINLINE;
ip->i_d.di_format = XFS_DINODE_FMT_LOCAL;
xfs_trans_log_inode(tp, ip, XFS_ILOG_DDATA | XFS_ILOG_CORE);
} else {
int offset;
first_fsb = 0;
nmaps = XFS_SYMLINK_MAPS;
error = xfs_bmapi_write(tp, ip, first_fsb, fs_blocks,
XFS_BMAPI_METADATA, &first_block, resblks,
mval, &nmaps, &free_list);
if (error)
goto error2;
if (resblks)
resblks -= fs_blocks;
ip->i_d.di_size = pathlen;
xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
cur_chunk = target_path;
offset = 0;
for (n = 0; n < nmaps; n++) {
char *buf;
d = XFS_FSB_TO_DADDR(mp, mval[n].br_startblock);
byte_cnt = XFS_FSB_TO_B(mp, mval[n].br_blockcount);
bp = xfs_trans_get_buf(tp, mp->m_ddev_targp, d,
BTOBB(byte_cnt), 0);
if (!bp) {
error = ENOMEM;
goto error2;
}
bp->b_ops = &xfs_symlink_buf_ops;
byte_cnt = XFS_SYMLINK_BUF_SPACE(mp, byte_cnt);
byte_cnt = min(byte_cnt, pathlen);
buf = bp->b_addr;
buf += xfs_symlink_hdr_set(mp, ip->i_ino, offset,
byte_cnt, bp);
memcpy(buf, cur_chunk, byte_cnt);
cur_chunk += byte_cnt;
pathlen -= byte_cnt;
offset += byte_cnt;
xfs_trans_log_buf(tp, bp, 0, (buf + byte_cnt - 1) -
(char *)bp->b_addr);
}
ASSERT(pathlen == 0);
}
/*
* Create the directory entry for the symlink.
*/
error = xfs_dir_createname(tp, dp, link_name, ip->i_ino,
&first_block, &free_list, resblks);
if (error)
goto error2;
xfs_trans_ichgtime(tp, dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
xfs_trans_log_inode(tp, dp, XFS_ILOG_CORE);
/*
* If this is a synchronous mount, make sure that the
* symlink transaction goes to disk before returning to
* the user.
*/
if (mp->m_flags & (XFS_MOUNT_WSYNC|XFS_MOUNT_DIRSYNC)) {
xfs_trans_set_sync(tp);
}
error = xfs_bmap_finish(&tp, &free_list, &committed);
if (error) {
goto error2;
}
error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
xfs_qm_dqrele(udqp);
xfs_qm_dqrele(gdqp);
xfs_qm_dqrele(pdqp);
*ipp = ip;
return 0;
error2:
IRELE(ip);
error1:
xfs_bmap_cancel(&free_list);
cancel_flags |= XFS_TRANS_ABORT;
error_return:
xfs_trans_cancel(tp, cancel_flags);
xfs_qm_dqrele(udqp);
xfs_qm_dqrele(gdqp);
xfs_qm_dqrele(pdqp);
if (unlock_dp_on_error)
xfs_iunlock(dp, XFS_ILOCK_EXCL);
std_return:
return error;
}
/*
* Convert a V2 leaf directory to a V2 block directory if possible.
*/
int /* error */
xfs_dir2_leaf_to_block(
xfs_da_args_t *args, /* operation arguments */
xfs_dabuf_t *lbp, /* leaf buffer */
xfs_dabuf_t *dbp) /* data buffer */
{
__be16 *bestsp; /* leaf bests table */
xfs_dir2_block_t *block; /* block structure */
xfs_dir2_block_tail_t *btp; /* block tail */
xfs_inode_t *dp; /* incore directory inode */
xfs_dir2_data_unused_t *dup; /* unused data entry */
int error; /* error return value */
int from; /* leaf from index */
xfs_dir2_leaf_t *leaf; /* leaf structure */
xfs_dir2_leaf_entry_t *lep; /* leaf entry */
xfs_dir2_leaf_tail_t *ltp; /* leaf tail structure */
xfs_mount_t *mp; /* file system mount point */
int needlog; /* need to log data header */
int needscan; /* need to scan for bestfree */
xfs_dir2_sf_hdr_t sfh; /* shortform header */
int size; /* bytes used */
__be16 *tagp; /* end of entry (tag) */
int to; /* block/leaf to index */
xfs_trans_t *tp; /* transaction pointer */
trace_xfs_dir2_leaf_to_block(args);
dp = args->dp;
tp = args->trans;
mp = dp->i_mount;
leaf = lbp->data;
ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_DIR2_LEAF1_MAGIC);
ltp = xfs_dir2_leaf_tail_p(mp, leaf);
/*
* If there are data blocks other than the first one, take this
* opportunity to remove trailing empty data blocks that may have
* been left behind during no-space-reservation operations.
* These will show up in the leaf bests table.
*/
while (dp->i_d.di_size > mp->m_dirblksize) {
bestsp = xfs_dir2_leaf_bests_p(ltp);
if (be16_to_cpu(bestsp[be32_to_cpu(ltp->bestcount) - 1]) ==
mp->m_dirblksize - (uint)sizeof(block->hdr)) {
if ((error =
xfs_dir2_leaf_trim_data(args, lbp,
(xfs_dir2_db_t)(be32_to_cpu(ltp->bestcount) - 1))))
goto out;
} else {
error = 0;
goto out;
}
}
/*
* Read the data block if we don't already have it, give up if it fails.
*/
if (dbp == NULL &&
(error = xfs_da_read_buf(tp, dp, mp->m_dirdatablk, -1, &dbp,
XFS_DATA_FORK))) {
goto out;
}
block = dbp->data;
ASSERT(be32_to_cpu(block->hdr.magic) == XFS_DIR2_DATA_MAGIC);
/*
* Size of the "leaf" area in the block.
*/
size = (uint)sizeof(block->tail) +
(uint)sizeof(*lep) * (be16_to_cpu(leaf->hdr.count) - be16_to_cpu(leaf->hdr.stale));
/*
* Look at the last data entry.
*/
tagp = (__be16 *)((char *)block + mp->m_dirblksize) - 1;
dup = (xfs_dir2_data_unused_t *)((char *)block + be16_to_cpu(*tagp));
/*
* If it's not free or is too short we can't do it.
*/
if (be16_to_cpu(dup->freetag) != XFS_DIR2_DATA_FREE_TAG ||
be16_to_cpu(dup->length) < size) {
error = 0;
goto out;
}
/*
* Start converting it to block form.
*/
block->hdr.magic = cpu_to_be32(XFS_DIR2_BLOCK_MAGIC);
needlog = 1;
needscan = 0;
/*
* Use up the space at the end of the block (blp/btp).
*/
xfs_dir2_data_use_free(tp, dbp, dup, mp->m_dirblksize - size, size,
&needlog, &needscan);
/*
* Initialize the block tail.
*/
btp = xfs_dir2_block_tail_p(mp, block);
btp->count = cpu_to_be32(be16_to_cpu(leaf->hdr.count) - be16_to_cpu(leaf->hdr.stale));
btp->stale = 0;
xfs_dir2_block_log_tail(tp, dbp);
/*
* Initialize the block leaf area. We compact out stale entries.
*/
lep = xfs_dir2_block_leaf_p(btp);
for (from = to = 0; from < be16_to_cpu(leaf->hdr.count); from++) {
if (be32_to_cpu(leaf->ents[from].address) == XFS_DIR2_NULL_DATAPTR)
continue;
lep[to++] = leaf->ents[from];
}
ASSERT(to == be32_to_cpu(btp->count));
xfs_dir2_block_log_leaf(tp, dbp, 0, be32_to_cpu(btp->count) - 1);
/*
* Scan the bestfree if we need it and log the data block header.
*/
if (needscan)
xfs_dir2_data_freescan(mp, (xfs_dir2_data_t *)block, &needlog);
if (needlog)
xfs_dir2_data_log_header(tp, dbp);
/*
* Pitch the old leaf block.
*/
error = xfs_da_shrink_inode(args, mp->m_dirleafblk, lbp);
lbp = NULL;
if (error) {
goto out;
}
/*
* Now see if the resulting block can be shrunken to shortform.
*/
if ((size = xfs_dir2_block_sfsize(dp, block, &sfh)) >
XFS_IFORK_DSIZE(dp)) {
error = 0;
goto out;
}
return xfs_dir2_block_to_sf(args, dbp, size, &sfh);
out:
if (lbp)
xfs_da_buf_done(lbp);
if (dbp)
xfs_da_buf_done(dbp);
return error;
}
/*
* Look up an inode by number in the given file system.
* The inode is looked up in the hash table for the file system
* represented by the mount point parameter mp. Each bucket of
* the hash table is guarded by an individual semaphore.
*
* If the inode is found in the hash table, its corresponding vnode
* is obtained with a call to vn_get(). This call takes care of
* coordination with the reclamation of the inode and vnode. Note
* that the vmap structure is filled in while holding the hash lock.
* This gives us the state of the inode/vnode when we found it and
* is used for coordination in vn_get().
*
* If it is not in core, read it in from the file system's device and
* add the inode into the hash table.
*
* The inode is locked according to the value of the lock_flags parameter.
* This flag parameter indicates how and if the inode's IO lock and inode lock
* should be taken.
*
* mp -- the mount point structure for the current file system. It points
* to the inode hash table.
* tp -- a pointer to the current transaction if there is one. This is
* simply passed through to the xfs_iread() call.
* ino -- the number of the inode desired. This is the unique identifier
* within the file system for the inode being requested.
* lock_flags -- flags indicating how to lock the inode. See the comment
* for xfs_ilock() for a list of valid values.
* bno -- the block number starting the buffer containing the inode,
* if known (as by bulkstat), else 0.
*/
int
xfs_iget(
xfs_mount_t *mp,
xfs_trans_t *tp,
xfs_ino_t ino,
uint flags,
uint lock_flags,
xfs_inode_t **ipp,
xfs_daddr_t bno)
{
xfs_ihash_t *ih;
xfs_inode_t *ip;
xfs_inode_t *iq;
xfs_vnode_t *vp;
ulong version;
int error;
/* REFERENCED */
int newnode;
xfs_chash_t *ch;
xfs_chashlist_t *chl, *chlnew;
vmap_t vmap;
SPLDECL(s);
XFS_STATS_INC(xs_ig_attempts);
ih = XFS_IHASH(mp, ino);
again:
read_lock(&ih->ih_lock);
for (ip = ih->ih_next; ip != NULL; ip = ip->i_next) {
if (ip->i_ino == ino) {
vp = XFS_ITOV(ip);
VMAP(vp, vmap);
/*
* Inode cache hit: if ip is not at the front of
* its hash chain, move it there now.
* Do this with the lock held for update, but
* do statistics after releasing the lock.
*/
if (ip->i_prevp != &ih->ih_next
&& rwlock_trypromote(&ih->ih_lock)) {
if ((iq = ip->i_next)) {
iq->i_prevp = ip->i_prevp;
}
*ip->i_prevp = iq;
iq = ih->ih_next;
iq->i_prevp = &ip->i_next;
ip->i_next = iq;
ip->i_prevp = &ih->ih_next;
ih->ih_next = ip;
write_unlock(&ih->ih_lock);
} else {
read_unlock(&ih->ih_lock);
}
XFS_STATS_INC(xs_ig_found);
/*
* Get a reference to the vnode/inode.
* vn_get() takes care of coordination with
* the file system inode release and reclaim
* functions. If it returns NULL, the inode
* has been reclaimed so just start the search
* over again. We probably won't find it,
* but we could be racing with another cpu
* looking for the same inode so we have to at
* least look.
*/
if (!(vp = vn_get(vp, &vmap))) {
XFS_STATS_INC(xs_ig_frecycle);
goto again;
}
if (lock_flags != 0) {
ip->i_flags &= ~XFS_IRECLAIM;
xfs_ilock(ip, lock_flags);
}
newnode = (ip->i_d.di_mode == 0);
if (newnode) {
xfs_iocore_inode_reinit(ip);
}
ip->i_flags &= ~XFS_ISTALE;
vn_trace_exit(vp, "xfs_iget.found",
(inst_t *)__return_address);
goto return_ip;
}
}
/*
* Inode cache miss: save the hash chain version stamp and unlock
* the chain, so we don't deadlock in vn_alloc.
*/
XFS_STATS_INC(xs_ig_missed);
version = ih->ih_version;
read_unlock(&ih->ih_lock);
/*
* Read the disk inode attributes into a new inode structure and get
* a new vnode for it. This should also initialize i_ino and i_mount.
*/
error = xfs_iread(mp, tp, ino, &ip, bno);
if (error) {
return error;
}
error = xfs_vn_allocate(mp, ip, &vp);
if (error) {
return error;
}
vn_trace_exit(vp, "xfs_iget.alloc", (inst_t *)__return_address);
xfs_inode_lock_init(ip, vp);
xfs_iocore_inode_init(ip);
if (lock_flags != 0) {
xfs_ilock(ip, lock_flags);
}
/*
* Put ip on its hash chain, unless someone else hashed a duplicate
* after we released the hash lock.
*/
write_lock(&ih->ih_lock);
if (ih->ih_version != version) {
for (iq = ih->ih_next; iq != NULL; iq = iq->i_next) {
if (iq->i_ino == ino) {
write_unlock(&ih->ih_lock);
xfs_idestroy(ip);
XFS_STATS_INC(xs_ig_dup);
goto again;
}
}
}
/*
* These values _must_ be set before releasing ihlock!
*/
ip->i_hash = ih;
if ((iq = ih->ih_next)) {
iq->i_prevp = &ip->i_next;
}
ip->i_next = iq;
ip->i_prevp = &ih->ih_next;
ih->ih_next = ip;
ip->i_udquot = ip->i_gdquot = NULL;
ih->ih_version++;
write_unlock(&ih->ih_lock);
/*
* put ip on its cluster's hash chain
*/
ASSERT(ip->i_chash == NULL && ip->i_cprev == NULL &&
ip->i_cnext == NULL);
chlnew = NULL;
ch = XFS_CHASH(mp, ip->i_blkno);
chlredo:
s = mutex_spinlock(&ch->ch_lock);
for (chl = ch->ch_list; chl != NULL; chl = chl->chl_next) {
if (chl->chl_blkno == ip->i_blkno) {
/* insert this inode into the doubly-linked list
* where chl points */
if ((iq = chl->chl_ip)) {
ip->i_cprev = iq->i_cprev;
iq->i_cprev->i_cnext = ip;
iq->i_cprev = ip;
ip->i_cnext = iq;
} else {
ip->i_cnext = ip;
ip->i_cprev = ip;
}
chl->chl_ip = ip;
ip->i_chash = chl;
break;
}
}
/* no hash list found for this block; add a new hash list */
if (chl == NULL) {
if (chlnew == NULL) {
mutex_spinunlock(&ch->ch_lock, s);
ASSERT(xfs_chashlist_zone != NULL);
chlnew = (xfs_chashlist_t *)
kmem_zone_alloc(xfs_chashlist_zone,
KM_SLEEP);
ASSERT(chlnew != NULL);
goto chlredo;
} else {
ip->i_cnext = ip;
ip->i_cprev = ip;
ip->i_chash = chlnew;
chlnew->chl_ip = ip;
chlnew->chl_blkno = ip->i_blkno;
chlnew->chl_next = ch->ch_list;
ch->ch_list = chlnew;
chlnew = NULL;
}
} else {
if (chlnew != NULL) {
kmem_zone_free(xfs_chashlist_zone, chlnew);
}
}
mutex_spinunlock(&ch->ch_lock, s);
/*
* Link ip to its mount and thread it on the mount's inode list.
*/
XFS_MOUNT_ILOCK(mp);
if ((iq = mp->m_inodes)) {
ASSERT(iq->i_mprev->i_mnext == iq);
ip->i_mprev = iq->i_mprev;
iq->i_mprev->i_mnext = ip;
iq->i_mprev = ip;
ip->i_mnext = iq;
} else {
ip->i_mnext = ip;
ip->i_mprev = ip;
}
mp->m_inodes = ip;
XFS_MOUNT_IUNLOCK(mp);
newnode = 1;
return_ip:
ASSERT(ip->i_df.if_ext_max ==
XFS_IFORK_DSIZE(ip) / sizeof(xfs_bmbt_rec_t));
ASSERT(((ip->i_d.di_flags & XFS_DIFLAG_REALTIME) != 0) ==
((ip->i_iocore.io_flags & XFS_IOCORE_RT) != 0));
*ipp = ip;
/*
* If we have a real type for an on-disk inode, we can set ops(&unlock)
* now. If it's a new inode being created, xfs_ialloc will handle it.
*/
XVFS_INIT_VNODE(XFS_MTOVFS(mp), vp, XFS_ITOBHV(ip), 1);
return 0;
}
/*
* Look up an inode by number in the given file system.
* The inode is looked up in the cache held in each AG.
* If the inode is found in the cache, initialise the vfs inode
* if necessary.
*
* If it is not in core, read it in from the file system's device,
* add it to the cache and initialise the vfs inode.
*
* The inode is locked according to the value of the lock_flags parameter.
* This flag parameter indicates how and if the inode's IO lock and inode lock
* should be taken.
*
* mp -- the mount point structure for the current file system. It points
* to the inode hash table.
* tp -- a pointer to the current transaction if there is one. This is
* simply passed through to the xfs_iread() call.
* ino -- the number of the inode desired. This is the unique identifier
* within the file system for the inode being requested.
* lock_flags -- flags indicating how to lock the inode. See the comment
* for xfs_ilock() for a list of valid values.
*/
int
xfs_iget(
xfs_mount_t *mp,
xfs_trans_t *tp,
xfs_ino_t ino,
uint flags,
uint lock_flags,
xfs_inode_t **ipp)
{
xfs_inode_t *ip;
int error;
xfs_perag_t *pag;
xfs_agino_t agino;
/*
* xfs_reclaim_inode() uses the ILOCK to ensure an inode
* doesn't get freed while it's being referenced during a
* radix tree traversal here. It assumes this function
* aqcuires only the ILOCK (and therefore it has no need to
* involve the IOLOCK in this synchronization).
*/
ASSERT((lock_flags & (XFS_IOLOCK_EXCL | XFS_IOLOCK_SHARED)) == 0);
/* reject inode numbers outside existing AGs */
if (!ino || XFS_INO_TO_AGNO(mp, ino) >= mp->m_sb.sb_agcount)
return EINVAL;
/* get the perag structure and ensure that it's inode capable */
pag = xfs_perag_get(mp, XFS_INO_TO_AGNO(mp, ino));
agino = XFS_INO_TO_AGINO(mp, ino);
again:
error = 0;
rcu_read_lock();
ip = radix_tree_lookup(&pag->pag_ici_root, agino);
if (ip) {
error = xfs_iget_cache_hit(pag, ip, ino, flags, lock_flags);
if (error)
goto out_error_or_again;
} else {
rcu_read_unlock();
XFS_STATS_INC(xs_ig_missed);
error = xfs_iget_cache_miss(mp, pag, tp, ino, &ip,
flags, lock_flags);
if (error)
goto out_error_or_again;
}
xfs_perag_put(pag);
*ipp = ip;
ASSERT(ip->i_df.if_ext_max ==
XFS_IFORK_DSIZE(ip) / sizeof(xfs_bmbt_rec_t));
/*
* If we have a real type for an on-disk inode, we can set ops(&unlock)
* now. If it's a new inode being created, xfs_ialloc will handle it.
*/
if (xfs_iflags_test(ip, XFS_INEW) && ip->i_d.di_mode != 0)
xfs_setup_inode(ip);
return 0;
out_error_or_again:
if (error == EAGAIN) {
delay(1);
goto again;
}
xfs_perag_put(pag);
return error;
}
/*
* Given a block directory (dp/block), calculate its size as a shortform (sf)
* directory and a header for the sf directory, if it will fit it the
* space currently present in the inode. If it won't fit, the output
* size is too big (but not accurate).
*/
int /* size for sf form */
xfs_dir2_block_sfsize(
xfs_inode_t *dp, /* incore inode pointer */
xfs_dir2_block_t *block, /* block directory data */
xfs_dir2_sf_hdr_t *sfhp) /* output: header for sf form */
{
xfs_dir2_dataptr_t addr; /* data entry address */
xfs_dir2_leaf_entry_t *blp; /* leaf area of the block */
xfs_dir2_block_tail_t *btp; /* tail area of the block */
int count; /* shortform entry count */
xfs_dir2_data_entry_t *dep; /* data entry in the block */
int i; /* block entry index */
int i8count; /* count of big-inode entries */
int isdot; /* entry is "." */
int isdotdot; /* entry is ".." */
xfs_mount_t *mp; /* mount structure pointer */
int namelen; /* total name bytes */
xfs_ino_t parent = 0; /* parent inode number */
int size=0; /* total computed size */
mp = dp->i_mount;
count = i8count = namelen = 0;
btp = xfs_dir2_block_tail_p(mp, block);
blp = xfs_dir2_block_leaf_p(btp);
/*
* Iterate over the block's data entries by using the leaf pointers.
*/
for (i = 0; i < be32_to_cpu(btp->count); i++) {
if ((addr = be32_to_cpu(blp[i].address)) == XFS_DIR2_NULL_DATAPTR)
continue;
/*
* Calculate the pointer to the entry at hand.
*/
dep = (xfs_dir2_data_entry_t *)
((char *)block + xfs_dir2_dataptr_to_off(mp, addr));
/*
* Detect . and .., so we can special-case them.
* . is not included in sf directories.
* .. is included by just the parent inode number.
*/
isdot = dep->namelen == 1 && dep->name[0] == '.';
isdotdot =
dep->namelen == 2 &&
dep->name[0] == '.' && dep->name[1] == '.';
#if XFS_BIG_INUMS
if (!isdot)
i8count += be64_to_cpu(dep->inumber) > XFS_DIR2_MAX_SHORT_INUM;
#endif
if (!isdot && !isdotdot) {
count++;
namelen += dep->namelen;
} else if (isdotdot)
parent = be64_to_cpu(dep->inumber);
/*
* Calculate the new size, see if we should give up yet.
*/
size = xfs_dir2_sf_hdr_size(i8count) + /* header */
count + /* namelen */
count * (uint)sizeof(xfs_dir2_sf_off_t) + /* offset */
namelen + /* name */
(i8count ? /* inumber */
(uint)sizeof(xfs_dir2_ino8_t) * count :
(uint)sizeof(xfs_dir2_ino4_t) * count);
if (size > XFS_IFORK_DSIZE(dp))
return size; /* size value is a failure */
}
/*
* Create the output header, if it worked.
*/
sfhp->count = count;
sfhp->i8count = i8count;
xfs_dir2_sf_put_inumber((xfs_dir2_sf_t *)sfhp, &parent, &sfhp->parent);
return size;
}
