/*
* This gets called when the inode's version needs to be changed from 1 to 2.
* Currently this happens when the nlink field overflows the old 16-bit value
* or when chproj is called to change the project for the first time.
* As a side effect the superblock version will also get rev'd
* to contain the NLINK bit.
*/
void
xfs_bump_ino_vers2(
xfs_trans_t *tp,
xfs_inode_t *ip)
{
xfs_mount_t *mp;
unsigned long s;
ASSERT(ismrlocked (&ip->i_lock, MR_UPDATE));
ASSERT(ip->i_d.di_version == XFS_DINODE_VERSION_1);
ip->i_d.di_version = XFS_DINODE_VERSION_2;
ip->i_d.di_onlink = 0;
memset(&(ip->i_d.di_pad[0]), 0, sizeof(ip->i_d.di_pad));
mp = tp->t_mountp;
if (!XFS_SB_VERSION_HASNLINK(&mp->m_sb)) {
s = XFS_SB_LOCK(mp);
if (!XFS_SB_VERSION_HASNLINK(&mp->m_sb)) {
XFS_SB_VERSION_ADDNLINK(&mp->m_sb);
XFS_SB_UNLOCK(mp, s);
xfs_mod_sb(tp, XFS_SB_VERSIONNUM);
} else {
XFS_SB_UNLOCK(mp, s);
}
}
/* Caller must log the inode */
}
int
xfs_sb_version_hasnlink(xfs_sb_t *sbp)
{
return XFS_SB_VERSION_HASNLINK(sbp);
}
/*
* This is called to fill in the vector of log iovecs for the
* given inode log item. It fills the first item with an inode
* log format structure, the second with the on-disk inode structure,
* and a possible third and/or fourth with the inode data/extents/b-tree
* root and inode attributes data/extents/b-tree root.
*/
STATIC void
xfs_inode_item_format(
xfs_inode_log_item_t *iip,
xfs_log_iovec_t *log_vector)
{
uint nvecs;
xfs_log_iovec_t *vecp;
xfs_inode_t *ip;
size_t data_bytes;
xfs_bmbt_rec_t *ext_buffer;
int nrecs;
xfs_mount_t *mp;
ip = iip->ili_inode;
vecp = log_vector;
vecp->i_addr = (xfs_caddr_t)&iip->ili_format;
vecp->i_len = sizeof(xfs_inode_log_format_t);
vecp++;
nvecs = 1;
/*
* Clear i_update_core if the timestamps (or any other
* non-transactional modification) need flushing/logging
* and we're about to log them with the rest of the core.
*
* This is the same logic as xfs_iflush() but this code can't
* run at the same time as xfs_iflush because we're in commit
* processing here and so we have the inode lock held in
* exclusive mode. Although it doesn't really matter
* for the timestamps if both routines were to grab the
* timestamps or not. That would be ok.
*
* We clear i_update_core before copying out the data.
* This is for coordination with our timestamp updates
* that don't hold the inode lock. They will always
* update the timestamps BEFORE setting i_update_core,
* so if we clear i_update_core after they set it we
* are guaranteed to see their updates to the timestamps
* either here. Likewise, if they set it after we clear it
* here, we'll see it either on the next commit of this
* inode or the next time the inode gets flushed via
* xfs_iflush(). This depends on strongly ordered memory
* semantics, but we have that. We use the SYNCHRONIZE
* macro to make sure that the compiler does not reorder
* the i_update_core access below the data copy below.
*/
if (ip->i_update_core) {
ip->i_update_core = 0;
SYNCHRONIZE();
}
/*
* We don't have to worry about re-ordering here because
* the update_size field is protected by the inode lock
* and we have that held in exclusive mode.
*/
if (ip->i_update_size)
ip->i_update_size = 0;
vecp->i_addr = (xfs_caddr_t)&ip->i_d;
vecp->i_len = sizeof(xfs_dinode_core_t);
vecp++;
nvecs++;
iip->ili_format.ilf_fields |= XFS_ILOG_CORE;
/*
* If this is really an old format inode, then we need to
* log it as such. This means that we have to copy the link
* count from the new field to the old. We don't have to worry
* about the new fields, because nothing trusts them as long as
* the old inode version number is there. If the superblock already
* has a new version number, then we don't bother converting back.
*/
mp = ip->i_mount;
ASSERT(ip->i_d.di_version == XFS_DINODE_VERSION_1 ||
XFS_SB_VERSION_HASNLINK(&mp->m_sb));
if (ip->i_d.di_version == XFS_DINODE_VERSION_1) {
if (!XFS_SB_VERSION_HASNLINK(&mp->m_sb)) {
/*
* Convert it back.
*/
ASSERT(ip->i_d.di_nlink <= XFS_MAXLINK_1);
ip->i_d.di_onlink = ip->i_d.di_nlink;
} else {
/*
* The superblock version has already been bumped,
* so just make the conversion to the new inode
* format permanent.
*/
ip->i_d.di_version = XFS_DINODE_VERSION_2;
ip->i_d.di_onlink = 0;
memset(&(ip->i_d.di_pad[0]), 0, sizeof(ip->i_d.di_pad));
}
}
switch (ip->i_d.di_format) {
case XFS_DINODE_FMT_EXTENTS:
ASSERT(!(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) {
ASSERT(ip->i_df.if_bytes > 0);
ASSERT(ip->i_df.if_u1.if_extents != NULL);
ASSERT(ip->i_d.di_nextents > 0);
ASSERT(iip->ili_extents_buf == NULL);
nrecs = ip->i_df.if_bytes /
(uint)sizeof(xfs_bmbt_rec_t);
ASSERT(nrecs > 0);
#if __BYTE_ORDER == __BIG_ENDIAN
if (nrecs == ip->i_d.di_nextents) {
/*
* There are no delayed allocation
* extents, so just point to the
* real extents array.
*/
vecp->i_addr =
(char *)(ip->i_df.if_u1.if_extents);
vecp->i_len = ip->i_df.if_bytes;
} else
#endif
{
/*
* There are delayed allocation extents
* in the inode, or we need to convert
* the extents to on disk format.
* Use xfs_iextents_copy()
* to copy only the real extents into
* a separate buffer. We'll free the
* buffer in the unlock routine.
*/
ext_buffer = kmem_alloc(ip->i_df.if_bytes,
KM_SLEEP);
iip->ili_extents_buf = ext_buffer;
vecp->i_addr = (xfs_caddr_t)ext_buffer;
vecp->i_len = xfs_iextents_copy(ip, ext_buffer,
XFS_DATA_FORK);
}
ASSERT(vecp->i_len <= ip->i_df.if_bytes);
iip->ili_format.ilf_dsize = vecp->i_len;
vecp++;
nvecs++;
}
break;
case XFS_DINODE_FMT_BTREE:
ASSERT(!(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) {
ASSERT(ip->i_df.if_broot_bytes > 0);
ASSERT(ip->i_df.if_broot != NULL);
vecp->i_addr = (xfs_caddr_t)ip->i_df.if_broot;
vecp->i_len = ip->i_df.if_broot_bytes;
vecp++;
nvecs++;
iip->ili_format.ilf_dsize = ip->i_df.if_broot_bytes;
}
break;
case XFS_DINODE_FMT_LOCAL:
ASSERT(!(iip->ili_format.ilf_fields &
(XFS_ILOG_DBROOT | XFS_ILOG_DEXT |
XFS_ILOG_DEV | XFS_ILOG_UUID)));
if (iip->ili_format.ilf_fields & XFS_ILOG_DDATA) {
ASSERT(ip->i_df.if_bytes > 0);
ASSERT(ip->i_df.if_u1.if_data != NULL);
ASSERT(ip->i_d.di_size > 0);
vecp->i_addr = (xfs_caddr_t)ip->i_df.if_u1.if_data;
/*
* Round i_bytes up to a word boundary.
* The underlying memory is guaranteed to
* to be there by xfs_idata_realloc().
*/
data_bytes = roundup(ip->i_df.if_bytes, 4);
ASSERT((ip->i_df.if_real_bytes == 0) ||
(ip->i_df.if_real_bytes == data_bytes));
vecp->i_len = (int)data_bytes;
vecp++;
nvecs++;
iip->ili_format.ilf_dsize = (unsigned)data_bytes;
}
break;
case XFS_DINODE_FMT_DEV:
ASSERT(!(iip->ili_format.ilf_fields &
(XFS_ILOG_DBROOT | XFS_ILOG_DEXT |
XFS_ILOG_DDATA | XFS_ILOG_UUID)));
if (iip->ili_format.ilf_fields & XFS_ILOG_DEV) {
iip->ili_format.ilf_u.ilfu_rdev =
ip->i_df.if_u2.if_rdev;
}
break;
case XFS_DINODE_FMT_UUID:
ASSERT(!(iip->ili_format.ilf_fields &
(XFS_ILOG_DBROOT | XFS_ILOG_DEXT |
XFS_ILOG_DDATA | XFS_ILOG_DEV)));
if (iip->ili_format.ilf_fields & XFS_ILOG_UUID) {
iip->ili_format.ilf_u.ilfu_uuid =
ip->i_df.if_u2.if_uuid;
}
break;
default:
ASSERT(0);
break;
}
/*
* If there are no attributes associated with the file,
* then we're done.
* Assert that no attribute-related log flags are set.
*/
if (!XFS_IFORK_Q(ip)) {
ASSERT(nvecs == iip->ili_item.li_desc->lid_size);
iip->ili_format.ilf_size = nvecs;
ASSERT(!(iip->ili_format.ilf_fields &
(XFS_ILOG_ADATA | XFS_ILOG_ABROOT | XFS_ILOG_AEXT)));
return;
}
switch (ip->i_d.di_aformat) {
case XFS_DINODE_FMT_EXTENTS:
ASSERT(!(iip->ili_format.ilf_fields &
(XFS_ILOG_ADATA | XFS_ILOG_ABROOT)));
if (iip->ili_format.ilf_fields & XFS_ILOG_AEXT) {
ASSERT(ip->i_afp->if_bytes > 0);
ASSERT(ip->i_afp->if_u1.if_extents != NULL);
ASSERT(ip->i_d.di_anextents > 0);
#ifdef DEBUG
nrecs = ip->i_afp->if_bytes /
(uint)sizeof(xfs_bmbt_rec_t);
#endif
ASSERT(nrecs > 0);
ASSERT(nrecs == ip->i_d.di_anextents);
#if __BYTE_ORDER == __BIG_ENDIAN
/*
* There are not delayed allocation extents
* for attributes, so just point at the array.
*/
vecp->i_addr = (char *)(ip->i_afp->if_u1.if_extents);
vecp->i_len = ip->i_afp->if_bytes;
#else
ASSERT(iip->ili_aextents_buf == NULL);
/*
* Need to endian flip before logging
*/
ext_buffer = kmem_alloc(ip->i_afp->if_bytes,
KM_SLEEP);
iip->ili_aextents_buf = ext_buffer;
vecp->i_addr = (xfs_caddr_t)ext_buffer;
vecp->i_len = xfs_iextents_copy(ip, ext_buffer,
XFS_ATTR_FORK);
#endif
iip->ili_format.ilf_asize = vecp->i_len;
vecp++;
nvecs++;
}
break;
case XFS_DINODE_FMT_BTREE:
ASSERT(!(iip->ili_format.ilf_fields &
(XFS_ILOG_ADATA | XFS_ILOG_AEXT)));
if (iip->ili_format.ilf_fields & XFS_ILOG_ABROOT) {
ASSERT(ip->i_afp->if_broot_bytes > 0);
ASSERT(ip->i_afp->if_broot != NULL);
vecp->i_addr = (xfs_caddr_t)ip->i_afp->if_broot;
vecp->i_len = ip->i_afp->if_broot_bytes;
vecp++;
nvecs++;
iip->ili_format.ilf_asize = ip->i_afp->if_broot_bytes;
}
break;
case XFS_DINODE_FMT_LOCAL:
ASSERT(!(iip->ili_format.ilf_fields &
(XFS_ILOG_ABROOT | XFS_ILOG_AEXT)));
if (iip->ili_format.ilf_fields & XFS_ILOG_ADATA) {
ASSERT(ip->i_afp->if_bytes > 0);
ASSERT(ip->i_afp->if_u1.if_data != NULL);
vecp->i_addr = (xfs_caddr_t)ip->i_afp->if_u1.if_data;
/*
* Round i_bytes up to a word boundary.
* The underlying memory is guaranteed to
* to be there by xfs_idata_realloc().
*/
data_bytes = roundup(ip->i_afp->if_bytes, 4);
ASSERT((ip->i_afp->if_real_bytes == 0) ||
(ip->i_afp->if_real_bytes == data_bytes));
vecp->i_len = (int)data_bytes;
vecp++;
nvecs++;
iip->ili_format.ilf_asize = (unsigned)data_bytes;
}
break;
default:
ASSERT(0);
break;
}
ASSERT(nvecs == iip->ili_item.li_desc->lid_size);
iip->ili_format.ilf_size = nvecs;
}
/*
* Allocate new inodes in the allocation group specified by agbp.
* Return 0 for success, else error code.
*/
STATIC int /* error code or 0 */
xfs_ialloc_ag_alloc(
xfs_trans_t *tp, /* transaction pointer */
xfs_buf_t *agbp, /* alloc group buffer */
int *alloc)
{
xfs_agi_t *agi; /* allocation group header */
xfs_alloc_arg_t args; /* allocation argument structure */
int blks_per_cluster; /* fs blocks per inode cluster */
xfs_btree_cur_t *cur; /* inode btree cursor */
xfs_daddr_t d; /* disk addr of buffer */
int error;
xfs_buf_t *fbuf; /* new free inodes' buffer */
xfs_dinode_t *free; /* new free inode structure */
int i; /* inode counter */
int j; /* block counter */
int nbufs; /* num bufs of new inodes */
xfs_agino_t newino; /* new first inode's number */
xfs_agino_t newlen; /* new number of inodes */
int ninodes; /* num inodes per buf */
xfs_agino_t thisino; /* current inode number, for loop */
int version; /* inode version number to use */
int isaligned = 0; /* inode allocation at stripe unit */
/* boundary */
args.tp = tp;
args.mp = tp->t_mountp;
/*
* Locking will ensure that we don't have two callers in here
* at one time.
*/
newlen = XFS_IALLOC_INODES(args.mp);
if (args.mp->m_maxicount &&
args.mp->m_sb.sb_icount + newlen > args.mp->m_maxicount)
return XFS_ERROR(ENOSPC);
args.minlen = args.maxlen = XFS_IALLOC_BLOCKS(args.mp);
/*
* First try to allocate inodes contiguous with the last-allocated
* chunk of inodes. If the filesystem is striped, this will fill
* an entire stripe unit with inodes.
*/
agi = XFS_BUF_TO_AGI(agbp);
newino = be32_to_cpu(agi->agi_newino);
args.agbno = XFS_AGINO_TO_AGBNO(args.mp, newino) +
XFS_IALLOC_BLOCKS(args.mp);
if (likely(newino != NULLAGINO &&
(args.agbno < be32_to_cpu(agi->agi_length)))) {
args.fsbno = XFS_AGB_TO_FSB(args.mp,
be32_to_cpu(agi->agi_seqno), args.agbno);
args.type = XFS_ALLOCTYPE_THIS_BNO;
args.mod = args.total = args.wasdel = args.isfl =
args.userdata = args.minalignslop = 0;
args.prod = 1;
args.alignment = 1;
/*
* Allow space for the inode btree to split.
*/
args.minleft = XFS_IN_MAXLEVELS(args.mp) - 1;
if ((error = xfs_alloc_vextent(&args)))
return error;
} else
args.fsbno = NULLFSBLOCK;
if (unlikely(args.fsbno == NULLFSBLOCK)) {
/*
* Set the alignment for the allocation.
* If stripe alignment is turned on then align at stripe unit
* boundary.
* If the cluster size is smaller than a filesystem block
* then we're doing I/O for inodes in filesystem block size
* pieces, so don't need alignment anyway.
*/
isaligned = 0;
if (args.mp->m_sinoalign) {
ASSERT(!(args.mp->m_flags & XFS_MOUNT_NOALIGN));
args.alignment = args.mp->m_dalign;
isaligned = 1;
} else if (XFS_SB_VERSION_HASALIGN(&args.mp->m_sb) &&
args.mp->m_sb.sb_inoalignmt >=
XFS_B_TO_FSBT(args.mp,
XFS_INODE_CLUSTER_SIZE(args.mp)))
args.alignment = args.mp->m_sb.sb_inoalignmt;
else
args.alignment = 1;
/*
* Need to figure out where to allocate the inode blocks.
* Ideally they should be spaced out through the a.g.
* For now, just allocate blocks up front.
*/
args.agbno = be32_to_cpu(agi->agi_root);
args.fsbno = XFS_AGB_TO_FSB(args.mp,
be32_to_cpu(agi->agi_seqno), args.agbno);
/*
* Allocate a fixed-size extent of inodes.
*/
args.type = XFS_ALLOCTYPE_NEAR_BNO;
args.mod = args.total = args.wasdel = args.isfl =
args.userdata = args.minalignslop = 0;
args.prod = 1;
/*
* Allow space for the inode btree to split.
*/
args.minleft = XFS_IN_MAXLEVELS(args.mp) - 1;
if ((error = xfs_alloc_vextent(&args)))
return error;
}
/*
* If stripe alignment is turned on, then try again with cluster
* alignment.
*/
if (isaligned && args.fsbno == NULLFSBLOCK) {
args.type = XFS_ALLOCTYPE_NEAR_BNO;
args.agbno = be32_to_cpu(agi->agi_root);
args.fsbno = XFS_AGB_TO_FSB(args.mp,
be32_to_cpu(agi->agi_seqno), args.agbno);
if (XFS_SB_VERSION_HASALIGN(&args.mp->m_sb) &&
args.mp->m_sb.sb_inoalignmt >=
XFS_B_TO_FSBT(args.mp, XFS_INODE_CLUSTER_SIZE(args.mp)))
args.alignment = args.mp->m_sb.sb_inoalignmt;
else
args.alignment = 1;
if ((error = xfs_alloc_vextent(&args)))
return error;
}
if (args.fsbno == NULLFSBLOCK) {
*alloc = 0;
return 0;
}
ASSERT(args.len == args.minlen);
/*
* Convert the results.
*/
newino = XFS_OFFBNO_TO_AGINO(args.mp, args.agbno, 0);
/*
* Loop over the new block(s), filling in the inodes.
* For small block sizes, manipulate the inodes in buffers
* which are multiples of the blocks size.
*/
if (args.mp->m_sb.sb_blocksize >= XFS_INODE_CLUSTER_SIZE(args.mp)) {
blks_per_cluster = 1;
nbufs = (int)args.len;
ninodes = args.mp->m_sb.sb_inopblock;
} else {
blks_per_cluster = XFS_INODE_CLUSTER_SIZE(args.mp) /
args.mp->m_sb.sb_blocksize;
nbufs = (int)args.len / blks_per_cluster;
ninodes = blks_per_cluster * args.mp->m_sb.sb_inopblock;
}
/*
* Figure out what version number to use in the inodes we create.
* If the superblock version has caught up to the one that supports
* the new inode format, then use the new inode version. Otherwise
* use the old version so that old kernels will continue to be
* able to use the file system.
*/
if (XFS_SB_VERSION_HASNLINK(&args.mp->m_sb))
version = XFS_DINODE_VERSION_2;
else
version = XFS_DINODE_VERSION_1;
for (j = 0; j < nbufs; j++) {
/*
* Get the block.
*/
d = XFS_AGB_TO_DADDR(args.mp, be32_to_cpu(agi->agi_seqno),
args.agbno + (j * blks_per_cluster));
fbuf = xfs_trans_get_buf(tp, args.mp->m_ddev_targp, d,
args.mp->m_bsize * blks_per_cluster,
XFS_BUF_LOCK);
ASSERT(fbuf);
ASSERT(!XFS_BUF_GETERROR(fbuf));
/*
* Set initial values for the inodes in this buffer.
*/
xfs_biozero(fbuf, 0, ninodes << args.mp->m_sb.sb_inodelog);
for (i = 0; i < ninodes; i++) {
free = XFS_MAKE_IPTR(args.mp, fbuf, i);
INT_SET(free->di_core.di_magic, ARCH_CONVERT, XFS_DINODE_MAGIC);
INT_SET(free->di_core.di_version, ARCH_CONVERT, version);
INT_SET(free->di_next_unlinked, ARCH_CONVERT, NULLAGINO);
xfs_ialloc_log_di(tp, fbuf, i,
XFS_DI_CORE_BITS | XFS_DI_NEXT_UNLINKED);
}
xfs_trans_inode_alloc_buf(tp, fbuf);
}
be32_add(&agi->agi_count, newlen);
be32_add(&agi->agi_freecount, newlen);
down_read(&args.mp->m_peraglock);
args.mp->m_perag[be32_to_cpu(agi->agi_seqno)].pagi_freecount += newlen;
up_read(&args.mp->m_peraglock);
agi->agi_newino = cpu_to_be32(newino);
/*
* Insert records describing the new inode chunk into the btree.
*/
cur = xfs_btree_init_cursor(args.mp, tp, agbp,
be32_to_cpu(agi->agi_seqno),
XFS_BTNUM_INO, (xfs_inode_t *)0, 0);
for (thisino = newino;
thisino < newino + newlen;
thisino += XFS_INODES_PER_CHUNK) {
if ((error = xfs_inobt_lookup_eq(cur, thisino,
XFS_INODES_PER_CHUNK, XFS_INOBT_ALL_FREE, &i))) {
xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
return error;
}
ASSERT(i == 0);
if ((error = xfs_inobt_insert(cur, &i))) {
xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
return error;
}
ASSERT(i == 1);
}
xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR);
/*
* Log allocation group header fields
*/
xfs_ialloc_log_agi(tp, agbp,
XFS_AGI_COUNT | XFS_AGI_FREECOUNT | XFS_AGI_NEWINO);
/*
* Modify/log superblock values for inode count and inode free count.
*/
xfs_trans_mod_sb(tp, XFS_TRANS_SB_ICOUNT, (long)newlen);
xfs_trans_mod_sb(tp, XFS_TRANS_SB_IFREE, (long)newlen);
*alloc = 1;
return 0;
}
/*
* Writes a modified inode's changes out to the inode's on disk home.
* Originally based on xfs_iflush_int() from xfs_inode.c in the kernel.
*/
int
libxfs_iflush_int(xfs_inode_t *ip, xfs_buf_t *bp)
{
xfs_inode_log_item_t *iip;
xfs_dinode_t *dip;
xfs_mount_t *mp;
ASSERT(XFS_BUF_FSPRIVATE(bp, void *) != NULL);
ASSERT(ip->i_d.di_format != XFS_DINODE_FMT_BTREE ||
ip->i_d.di_nextents > ip->i_df.if_ext_max);
iip = ip->i_itemp;
mp = ip->i_mount;
/* set *dip = inode's place in the buffer */
dip = (xfs_dinode_t *)xfs_buf_offset(bp, ip->i_boffset);
#ifdef DEBUG
ASSERT(ip->i_d.di_magic == XFS_DINODE_MAGIC);
if ((ip->i_d.di_mode & IFMT) == IFREG) {
ASSERT( (ip->i_d.di_format == XFS_DINODE_FMT_EXTENTS) ||
(ip->i_d.di_format == XFS_DINODE_FMT_BTREE) );
}
else if ((ip->i_d.di_mode & IFMT) == IFDIR) {
ASSERT( (ip->i_d.di_format == XFS_DINODE_FMT_EXTENTS) ||
(ip->i_d.di_format == XFS_DINODE_FMT_BTREE) ||
(ip->i_d.di_format == XFS_DINODE_FMT_LOCAL) );
}
ASSERT(ip->i_d.di_nextents+ip->i_d.di_anextents <= ip->i_d.di_nblocks);
ASSERT(ip->i_d.di_forkoff <= mp->m_sb.sb_inodesize);
#endif
/*
* Copy the dirty parts of the inode into the on-disk
* inode. We always copy out the core of the inode,
* because if the inode is dirty at all the core must
* be.
*/
xfs_xlate_dinode_core((xfs_caddr_t)&(dip->di_core), &(ip->i_d), -1,
ARCH_CONVERT);
/*
* If this is really an old format inode and the superblock version
* has not been updated to support only new format inodes, then
* convert back to the old inode format. If the superblock version
* has been updated, then make the conversion permanent.
*/
ASSERT(ip->i_d.di_version == XFS_DINODE_VERSION_1 ||
XFS_SB_VERSION_HASNLINK(&mp->m_sb));
if (ip->i_d.di_version == XFS_DINODE_VERSION_1) {
if (!XFS_SB_VERSION_HASNLINK(&mp->m_sb)) {
/*
* Convert it back.
*/
ASSERT(ip->i_d.di_nlink <= XFS_MAXLINK_1);
INT_SET(dip->di_core.di_onlink, ARCH_CONVERT,
ip->i_d.di_nlink);
} else {
/*
* The superblock version has already been bumped,
* so just make the conversion to the new inode
* format permanent.
*/
ip->i_d.di_version = XFS_DINODE_VERSION_2;
INT_SET(dip->di_core.di_version, ARCH_CONVERT,
XFS_DINODE_VERSION_2);
ip->i_d.di_onlink = 0;
INT_ZERO(dip->di_core.di_onlink, ARCH_CONVERT);
bzero(&(ip->i_d.di_pad[0]), sizeof(ip->i_d.di_pad));
bzero(&(dip->di_core.di_pad[0]),
sizeof(dip->di_core.di_pad));
ASSERT(ip->i_d.di_projid == 0);
}
}
if (xfs_iflush_fork(ip, dip, iip, XFS_DATA_FORK, bp) == EFSCORRUPTED)
return EFSCORRUPTED;
if (XFS_IFORK_Q(ip)) {
/* The only error from xfs_iflush_fork is on the data fork. */
xfs_iflush_fork(ip, dip, iip, XFS_ATTR_FORK, bp);
}
return 0;
}
/*
* Allocate an inode on disk and return a copy of it's in-core version.
* Set mode, nlink, and rdev appropriately within the inode.
* The uid and gid for the inode are set according to the contents of
* the given cred structure.
*
* This was once shared with the kernel, but has diverged to the point
* where its no longer worth the hassle of maintaining common code.
*/
int
libxfs_ialloc(
xfs_trans_t *tp,
xfs_inode_t *pip,
mode_t mode,
nlink_t nlink,
dev_t rdev,
cred_t *cr,
xfs_prid_t prid,
int okalloc,
xfs_buf_t **ialloc_context,
boolean_t *call_again,
xfs_inode_t **ipp)
{
xfs_ino_t ino;
xfs_inode_t *ip;
uint flags;
int error;
/*
* Call the space management code to pick
* the on-disk inode to be allocated.
*/
error = xfs_dialloc(tp, pip ? pip->i_ino : 0, mode, okalloc,
ialloc_context, call_again, &ino);
if (error != 0)
return error;
if (*call_again || ino == NULLFSINO) {
*ipp = NULL;
return 0;
}
ASSERT(*ialloc_context == NULL);
error = xfs_trans_iget(tp->t_mountp, tp, ino, 0, &ip);
if (error != 0)
return error;
ASSERT(ip != NULL);
ip->i_d.di_mode = (__uint16_t)mode;
ip->i_d.di_onlink = 0;
ip->i_d.di_nlink = nlink;
ASSERT(ip->i_d.di_nlink == nlink);
ip->i_d.di_uid = cr->cr_uid;
ip->i_d.di_gid = cr->cr_gid;
ip->i_d.di_projid = prid;
bzero(&(ip->i_d.di_pad[0]), sizeof(ip->i_d.di_pad));
/*
* If the superblock version is up to where we support new format
* inodes and this is currently an old format inode, then change
* the inode version number now. This way we only do the conversion
* here rather than here and in the flush/logging code.
*/
if (XFS_SB_VERSION_HASNLINK(&tp->t_mountp->m_sb) &&
ip->i_d.di_version == XFS_DINODE_VERSION_1) {
ip->i_d.di_version = XFS_DINODE_VERSION_2;
/* old link count, projid field, pad field already zeroed */
}
ip->i_d.di_size = 0;
ip->i_d.di_nextents = 0;
ASSERT(ip->i_d.di_nblocks == 0);
xfs_ichgtime(ip, XFS_ICHGTIME_CHG|XFS_ICHGTIME_ACC|XFS_ICHGTIME_MOD);
/*
* di_gen will have been taken care of in xfs_iread.
*/
ip->i_d.di_extsize = 0;
ip->i_d.di_dmevmask = 0;
ip->i_d.di_dmstate = 0;
ip->i_d.di_flags = 0;
flags = XFS_ILOG_CORE;
switch (mode & IFMT) {
case IFIFO:
case IFCHR:
case IFBLK:
case IFSOCK:
ip->i_d.di_format = XFS_DINODE_FMT_DEV;
ip->i_df.if_u2.if_rdev = makedev(major(rdev), minor(rdev)); ip->i_df.if_flags = 0;
flags |= XFS_ILOG_DEV;
break;
case IFREG:
case IFDIR:
case IFLNK:
ip->i_d.di_format = XFS_DINODE_FMT_EXTENTS;
ip->i_df.if_flags = XFS_IFEXTENTS;
ip->i_df.if_bytes = ip->i_df.if_real_bytes = 0;
ip->i_df.if_u1.if_extents = NULL;
break;
default:
ASSERT(0);
}
/* Attribute fork settings for new inode. */
ip->i_d.di_aformat = XFS_DINODE_FMT_EXTENTS;
ip->i_d.di_anextents = 0;
/*
* Log the new values stuffed into the inode.
*/
xfs_trans_log_inode(tp, ip, flags);
*ipp = ip;
return 0;
}
