/*
* xfs_inode_item_format_extents - convert in-core extents to on-disk form
*
* For either the data or attr fork in extent format, we need to endian convert
* the in-core extent as we place them into the on-disk inode. In this case, we
* need to do this conversion before we write the extents into the log. Because
* we don't have the disk inode to write into here, we allocate a buffer and
* format the extents into it via xfs_iextents_copy(). We free the buffer in
* the unlock routine after the copy for the log has been made.
*
* In the case of the data fork, the in-core and on-disk fork sizes can be
* different due to delayed allocation extents. We only log on-disk extents
* here, so always use the physical fork size to determine the size of the
* buffer we need to allocate.
*/
STATIC void
xfs_inode_item_format_extents(
struct xfs_inode *ip,
struct xfs_log_iovec *vecp,
int whichfork,
int type)
{
xfs_bmbt_rec_t *ext_buffer;
ext_buffer = kmem_alloc(XFS_IFORK_SIZE(ip, whichfork), KM_SLEEP);
if (whichfork == XFS_DATA_FORK)
ip->i_itemp->ili_extents_buf = ext_buffer;
else
ip->i_itemp->ili_aextents_buf = ext_buffer;
vecp->i_addr = ext_buffer;
vecp->i_len = xfs_iextents_copy(ip, ext_buffer, whichfork);
vecp->i_type = type;
}
/*
* 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);
XLOG_VEC_SET_TYPE(vecp, XLOG_REG_TYPE_IFORMAT);
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;
/*
* Make sure to get the latest atime from the Linux inode.
*/
xfs_synchronize_atime(ip);
/*
* make sure the linux inode is dirty
*/
xfs_mark_inode_dirty_sync(ip);
vecp->i_addr = (xfs_caddr_t)&ip->i_d;
vecp->i_len = sizeof(struct xfs_icdinode);
XLOG_VEC_SET_TYPE(vecp, XLOG_REG_TYPE_ICORE);
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 == 1 || xfs_sb_version_hasnlink(&mp->m_sb));
if (ip->i_d.di_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 = 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);
#ifdef XFS_NATIVE_HOST
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;
XLOG_VEC_SET_TYPE(vecp, XLOG_REG_TYPE_IEXT);
} 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);
XLOG_VEC_SET_TYPE(vecp, XLOG_REG_TYPE_IEXT);
}
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;
XLOG_VEC_SET_TYPE(vecp, XLOG_REG_TYPE_IBROOT);
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;
XLOG_VEC_SET_TYPE(vecp, XLOG_REG_TYPE_ILOCAL);
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);
#ifdef XFS_NATIVE_HOST
/*
* 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
XLOG_VEC_SET_TYPE(vecp, XLOG_REG_TYPE_IATTR_EXT);
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;
XLOG_VEC_SET_TYPE(vecp, XLOG_REG_TYPE_IATTR_BROOT);
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;
XLOG_VEC_SET_TYPE(vecp, XLOG_REG_TYPE_IATTR_LOCAL);
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;
}
STATIC void
xfs_inode_item_format_attr_fork(
struct xfs_inode_log_item *iip,
struct xfs_inode_log_format *ilf,
struct xfs_log_vec *lv,
struct xfs_log_iovec **vecp)
{
struct xfs_inode *ip = iip->ili_inode;
size_t data_bytes;
switch (ip->i_d.di_aformat) {
case XFS_DINODE_FMT_EXTENTS:
iip->ili_fields &=
~(XFS_ILOG_ADATA | XFS_ILOG_ABROOT);
if ((iip->ili_fields & XFS_ILOG_AEXT) &&
ip->i_d.di_anextents > 0 &&
ip->i_afp->if_bytes > 0) {
struct xfs_bmbt_rec *p;
ASSERT(ip->i_afp->if_bytes / sizeof(xfs_bmbt_rec_t) ==
ip->i_d.di_anextents);
ASSERT(ip->i_afp->if_u1.if_extents != NULL);
p = xlog_prepare_iovec(lv, vecp, XLOG_REG_TYPE_IATTR_EXT);
data_bytes = xfs_iextents_copy(ip, p, XFS_ATTR_FORK);
xlog_finish_iovec(lv, *vecp, data_bytes);
ilf->ilf_asize = data_bytes;
ilf->ilf_size++;
} else {
iip->ili_fields &= ~XFS_ILOG_AEXT;
}
break;
case XFS_DINODE_FMT_BTREE:
iip->ili_fields &=
~(XFS_ILOG_ADATA | XFS_ILOG_AEXT);
if ((iip->ili_fields & XFS_ILOG_ABROOT) &&
ip->i_afp->if_broot_bytes > 0) {
ASSERT(ip->i_afp->if_broot != NULL);
xlog_copy_iovec(lv, vecp, XLOG_REG_TYPE_IATTR_BROOT,
ip->i_afp->if_broot,
ip->i_afp->if_broot_bytes);
ilf->ilf_asize = ip->i_afp->if_broot_bytes;
ilf->ilf_size++;
} else {
iip->ili_fields &= ~XFS_ILOG_ABROOT;
}
break;
case XFS_DINODE_FMT_LOCAL:
iip->ili_fields &=
~(XFS_ILOG_AEXT | XFS_ILOG_ABROOT);
if ((iip->ili_fields & XFS_ILOG_ADATA) &&
ip->i_afp->if_bytes > 0) {
/*
* 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);
ASSERT(ip->i_afp->if_u1.if_data != NULL);
xlog_copy_iovec(lv, vecp, XLOG_REG_TYPE_IATTR_LOCAL,
ip->i_afp->if_u1.if_data,
data_bytes);
ilf->ilf_asize = (unsigned)data_bytes;
ilf->ilf_size++;
} else {
iip->ili_fields &= ~XFS_ILOG_ADATA;
}
break;
default:
ASSERT(0);
break;
}
}
