/*
* 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;
ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
ASSERT(ip->i_d.di_version == 1);
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));
mp = tp->t_mountp;
if (!xfs_sb_version_hasnlink(&mp->m_sb)) {
spin_lock(&mp->m_sb_lock);
if (!xfs_sb_version_hasnlink(&mp->m_sb)) {
xfs_sb_version_addnlink(&mp->m_sb);
spin_unlock(&mp->m_sb_lock);
xfs_mod_sb(tp, XFS_SB_VERSIONNUM);
} else {
spin_unlock(&mp->m_sb_lock);
}
}
/* Caller must log the inode */
}
/*
* 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(
struct xfs_log_item *lip,
struct xfs_log_iovec *vecp)
{
struct xfs_inode_log_item *iip = INODE_ITEM(lip);
struct xfs_inode *ip = iip->ili_inode;
uint nvecs;
size_t data_bytes;
xfs_mount_t *mp;
vecp->i_addr = &iip->ili_format;
vecp->i_len = sizeof(xfs_inode_log_format_t);
vecp->i_type = 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();
}
/*
* Make sure to get the latest timestamps from the Linux inode.
*/
xfs_synchronize_times(ip);
vecp->i_addr = &ip->i_d;
vecp->i_len = sizeof(struct xfs_icdinode);
vecp->i_type = 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);
ASSERT((ip->i_df.if_bytes /
(uint)sizeof(xfs_bmbt_rec_t)) > 0);
#ifdef XFS_NATIVE_HOST
if (ip->i_d.di_nextents == ip->i_df.if_bytes /
(uint)sizeof(xfs_bmbt_rec_t)) {
/*
* There are no delayed allocation
* extents, so just point to the
* real extents array.
*/
vecp->i_addr = ip->i_df.if_u1.if_extents;
vecp->i_len = ip->i_df.if_bytes;
vecp->i_type = XLOG_REG_TYPE_IEXT;
} else
#endif
{
xfs_inode_item_format_extents(ip, vecp,
XFS_DATA_FORK, 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 = ip->i_df.if_broot;
vecp->i_len = ip->i_df.if_broot_bytes;
vecp->i_type = 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 = 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->i_type = 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 == lip->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) {
#ifdef DEBUG
int nrecs = ip->i_afp->if_bytes /
(uint)sizeof(xfs_bmbt_rec_t);
ASSERT(nrecs > 0);
ASSERT(nrecs == ip->i_d.di_anextents);
ASSERT(ip->i_afp->if_bytes > 0);
ASSERT(ip->i_afp->if_u1.if_extents != NULL);
ASSERT(ip->i_d.di_anextents > 0);
#endif
#ifdef XFS_NATIVE_HOST
/*
* There are not delayed allocation extents
* for attributes, so just point at the array.
*/
vecp->i_addr = ip->i_afp->if_u1.if_extents;
vecp->i_len = ip->i_afp->if_bytes;
vecp->i_type = XLOG_REG_TYPE_IATTR_EXT;
#else
ASSERT(iip->ili_aextents_buf == NULL);
xfs_inode_item_format_extents(ip, vecp,
XFS_ATTR_FORK, XLOG_REG_TYPE_IATTR_EXT);
#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 = ip->i_afp->if_broot;
vecp->i_len = ip->i_afp->if_broot_bytes;
vecp->i_type = 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 = 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->i_type = XLOG_REG_TYPE_IATTR_LOCAL;
vecp++;
nvecs++;
iip->ili_format.ilf_asize = (unsigned)data_bytes;
}
break;
default:
ASSERT(0);
break;
}
ASSERT(nvecs == lip->li_desc->lid_size);
iip->ili_format.ilf_size = nvecs;
}
void
update_sb_version(xfs_mount_t *mp)
{
xfs_sb_t *sb;
__uint16_t vn;
sb = &mp->m_sb;
if (fs_attributes && !xfs_sb_version_hasattr(sb)) {
ASSERT(fs_attributes_allowed);
xfs_sb_version_addattr(sb);
}
if (fs_attributes2 && !xfs_sb_version_hasattr2(sb)) {
ASSERT(fs_attributes2_allowed);
xfs_sb_version_addattr2(sb);
}
if (fs_inode_nlink && !xfs_sb_version_hasnlink(sb)) {
ASSERT(fs_inode_nlink_allowed);
xfs_sb_version_addnlink(sb);
}
/*
* fix up the superblock version number and feature bits,
* turn off quota bits and flags if the filesystem doesn't
* have quotas.
*/
if (fs_quotas) {
if (!xfs_sb_version_hasquota(sb)) {
ASSERT(fs_quotas_allowed);
xfs_sb_version_addquota(sb);
}
/*
* protect against stray bits in the quota flag field
*/
if (sb->sb_qflags & ~XFS_MOUNT_QUOTA_ALL) {
/*
* update the incore superblock, if we're in
* no_modify mode, it'll never get flushed out
* so this is ok.
*/
do_warn(_("bogus quota flags 0x%x set in superblock"),
sb->sb_qflags & ~XFS_MOUNT_QUOTA_ALL);
sb->sb_qflags &= XFS_MOUNT_QUOTA_ALL;
if (!no_modify)
do_warn(_(", bogus flags will be cleared\n"));
else
do_warn(_(", bogus flags would be cleared\n"));
}
} else {
sb->sb_qflags = 0;
if (xfs_sb_version_hasquota(sb)) {
lost_quotas = 1;
vn = sb->sb_versionnum;
vn &= ~XFS_SB_VERSION_QUOTABIT;
if (!(vn & XFS_SB_VERSION_ALLFBITS))
vn = xfs_sb_version_toold(vn);
ASSERT(vn != 0);
sb->sb_versionnum = vn;
}
}
if (!fs_aligned_inodes && xfs_sb_version_hasalign(sb))
sb->sb_versionnum &= ~XFS_SB_VERSION_ALIGNBIT;
}
/*
* 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 */
xfs_agnumber_t agno;
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 */
unsigned int gen;
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;
/*
* We need to take into account alignment here to ensure that
* we don't modify the free list if we fail to have an exact
* block. If we don't have an exact match, and every oher
* attempt allocation attempt fails, we'll end up cancelling
* a dirty transaction and shutting down.
*
* For an exact allocation, alignment must be 1,
* however we need to take cluster alignment into account when
* fixing up the freelist. Use the minalignslop field to
* indicate that extra blocks might be required for alignment,
* but not to use them in the actual exact allocation.
*/
args.alignment = 1;
args.minalignslop = xfs_ialloc_cluster_alignment(&args) - 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
args.alignment = xfs_ialloc_cluster_alignment(&args);
/*
* 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);
args.alignment = xfs_ialloc_cluster_alignment(&args);
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;
/*
* Seed the new inode cluster with a random generation number. This
* prevents short-term reuse of generation numbers if a chunk is
* freed and then immediately reallocated. We use random numbers
* rather than a linear progression to prevent the next generation
* number from being easily guessable.
*/
gen = random32();
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);
free->di_core.di_magic = cpu_to_be16(XFS_DINODE_MAGIC);
free->di_core.di_version = version;
free->di_core.di_gen = cpu_to_be32(gen);
free->di_next_unlinked = cpu_to_be32(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_cpu(&agi->agi_count, newlen);
be32_add_cpu(&agi->agi_freecount, newlen);
agno = be32_to_cpu(agi->agi_seqno);
down_read(&args.mp->m_peraglock);
args.mp->m_perag[agno].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, agno,
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;
}
/*
* returns 0 if things are fine, 1 if we don't understand
* this superblock version. Sets superblock geometry-dependent
* global variables.
*/
int
parse_sb_version(xfs_sb_t *sb)
{
int issue_warning;
fs_attributes = 0;
fs_attributes2 = 0;
fs_inode_nlink = 0;
fs_quotas = 0;
fs_aligned_inodes = 0;
fs_sb_feature_bits = 0;
fs_ino_alignment = 0;
fs_has_extflgbit = 0;
have_uquotino = 0;
have_gquotino = 0;
have_pquotino = 0;
issue_warning = 0;
/*
* ok, check to make sure that the sb isn't newer
* than we are
*/
if (xfs_sb_version_hasextflgbit(sb)) {
fs_has_extflgbit = 1;
if (!fs_has_extflgbit_allowed) {
issue_warning = 1;
do_warn(
_("This filesystem has uninitialized extent flags.\n"));
}
}
if (xfs_sb_version_hasshared(sb)) {
fs_shared = 1;
if (!fs_shared_allowed) {
issue_warning = 1;
do_warn(_("This filesystem is marked shared.\n"));
}
}
if (issue_warning) {
do_warn(
_("This filesystem uses feature(s) not yet supported in this release.\n"
"Please run a more recent version of xfs_repair.\n"));
return(1);
}
if (!xfs_sb_good_version(sb)) {
do_warn(_("WARNING: unknown superblock version %d\n"),
XFS_SB_VERSION_NUM(sb));
do_warn(
_("This filesystem contains features not understood by this program.\n"));
return(1);
}
if (XFS_SB_VERSION_NUM(sb) >= XFS_SB_VERSION_4) {
if (!fs_sb_feature_bits_allowed) {
if (!no_modify) {
do_warn(
_("WARNING: you have disallowed superblock-feature-bits-allowed\n"
"\tbut this superblock has feature bits. The superblock\n"
"\twill be downgraded. This may cause loss of filesystem meta-data\n"));
} else {
do_warn(
_("WARNING: you have disallowed superblock-feature-bits-allowed\n"
"\tbut this superblock has feature bits. The superblock\n"
"\twould be downgraded. This might cause loss of filesystem\n"
"\tmeta-data.\n"));
}
} else {
fs_sb_feature_bits = 1;
}
}
if (xfs_sb_version_hasattr(sb)) {
if (!fs_attributes_allowed) {
if (!no_modify) {
do_warn(
_("WARNING: you have disallowed attributes but this filesystem\n"
"\thas attributes. The filesystem will be downgraded and\n"
"\tall attributes will be removed.\n"));
} else {
do_warn(
_("WARNING: you have disallowed attributes but this filesystem\n"
"\thas attributes. The filesystem would be downgraded and\n"
"\tall attributes would be removed.\n"));
}
} else {
fs_attributes = 1;
}
}
if (xfs_sb_version_hasattr2(sb)) {
if (!fs_attributes2_allowed) {
if (!no_modify) {
do_warn(
_("WARNING: you have disallowed attr2 attributes but this filesystem\n"
"\thas attributes. The filesystem will be downgraded and\n"
"\tall attr2 attributes will be removed.\n"));
} else {
do_warn(
_("WARNING: you have disallowed attr2 attributes but this filesystem\n"
"\thas attributes. The filesystem would be downgraded and\n"
"\tall attr2 attributes would be removed.\n"));
}
} else {
fs_attributes2 = 1;
}
}
if (xfs_sb_version_hasnlink(sb)) {
if (!fs_inode_nlink_allowed) {
if (!no_modify) {
do_warn(
_("WARNING: you have disallowed version 2 inodes but this filesystem\n"
"\thas version 2 inodes. The filesystem will be downgraded and\n"
"\tall version 2 inodes will be converted to version 1 inodes.\n"
"\tThis may cause some hard links to files to be destroyed\n"));
} else {
do_warn(
_("WARNING: you have disallowed version 2 inodes but this filesystem\n"
"\thas version 2 inodes. The filesystem would be downgraded and\n"
"\tall version 2 inodes would be converted to version 1 inodes.\n"
"\tThis might cause some hard links to files to be destroyed\n"));
}
} else {
fs_inode_nlink = 1;
}
}
if (xfs_sb_version_hasquota(sb)) {
if (!fs_quotas_allowed) {
if (!no_modify) {
do_warn(
_("WARNING: you have disallowed quotas but this filesystem\n"
"\thas quotas. The filesystem will be downgraded and\n"
"\tall quota information will be removed.\n"));
} else {
do_warn(
_("WARNING: you have disallowed quotas but this filesystem\n"
"\thas quotas. The filesystem would be downgraded and\n"
"\tall quota information would be removed.\n"));
}
} else {
fs_quotas = 1;
if (sb->sb_uquotino != 0 &&
sb->sb_uquotino != NULLFSINO)
have_uquotino = 1;
if (sb->sb_gquotino != 0 &&
sb->sb_gquotino != NULLFSINO)
have_gquotino = 1;
if (sb->sb_pquotino != 0 &&
sb->sb_pquotino != NULLFSINO)
have_pquotino = 1;
}
}
if (xfs_sb_version_hasalign(sb)) {
if (fs_aligned_inodes_allowed) {
fs_aligned_inodes = 1;
fs_ino_alignment = sb->sb_inoalignmt;
} else {
if (!no_modify) {
do_warn(
_("WARNING: you have disallowed aligned inodes but this filesystem\n"
"\thas aligned inodes. The filesystem will be downgraded.\n"
"\tThis will permanently degrade the performance of this filesystem.\n"));
} else {
do_warn(
_("WARNING: you have disallowed aligned inodes but this filesystem\n"
"\thas aligned inodes. The filesystem would be downgraded.\n"
"\tThis would permanently degrade the performance of this filesystem.\n"));
}
}
}
/*
* calculate maximum file offset for this geometry
*/
fs_max_file_offset = 0x7fffffffffffffffLL >> sb->sb_blocklog;
return(0);
}
/*
* Initialise a new set of inodes.
*/
STATIC int
xfs_ialloc_inode_init(
struct xfs_mount *mp,
struct xfs_trans *tp,
xfs_agnumber_t agno,
xfs_agblock_t agbno,
xfs_agblock_t length,
unsigned int gen)
{
struct xfs_buf *fbuf;
struct xfs_dinode *free;
int blks_per_cluster, nbufs, ninodes;
int version;
int i, j;
xfs_daddr_t d;
/*
* 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 (mp->m_sb.sb_blocksize >= XFS_INODE_CLUSTER_SIZE(mp)) {
blks_per_cluster = 1;
nbufs = length;
ninodes = mp->m_sb.sb_inopblock;
} else {
blks_per_cluster = XFS_INODE_CLUSTER_SIZE(mp) /
mp->m_sb.sb_blocksize;
nbufs = length / blks_per_cluster;
ninodes = blks_per_cluster * 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(&mp->m_sb))
version = 2;
else
version = 1;
for (j = 0; j < nbufs; j++) {
/*
* Get the block.
*/
d = XFS_AGB_TO_DADDR(mp, agno, agbno + (j * blks_per_cluster));
fbuf = xfs_trans_get_buf(tp, mp->m_ddev_targp, d,
mp->m_bsize * blks_per_cluster,
XBF_LOCK);
if (!fbuf)
return ENOMEM;
/*
* Initialize all inodes in this buffer and then log them.
*
* XXX: It would be much better if we had just one transaction
* to log a whole cluster of inodes instead of all the
* individual transactions causing a lot of log traffic.
*/
xfs_buf_zero(fbuf, 0, ninodes << mp->m_sb.sb_inodelog);
for (i = 0; i < ninodes; i++) {
int ioffset = i << mp->m_sb.sb_inodelog;
uint isize = sizeof(struct xfs_dinode);
free = xfs_make_iptr(mp, fbuf, i);
free->di_magic = cpu_to_be16(XFS_DINODE_MAGIC);
free->di_version = version;
free->di_gen = cpu_to_be32(gen);
free->di_next_unlinked = cpu_to_be32(NULLAGINO);
xfs_trans_log_buf(tp, fbuf, ioffset, ioffset + isize - 1);
}
xfs_trans_inode_alloc_buf(tp, fbuf);
}
return 0;
}
/*
* Initialise a new set of inodes. When called without a transaction context
* (e.g. from recovery) we initiate a delayed write of the inode buffers rather
* than logging them (which in a transaction context puts them into the AIL
* for writeback rather than the xfsbufd queue).
*/
int
xfs_ialloc_inode_init(
struct xfs_mount *mp,
struct xfs_trans *tp,
struct list_head *buffer_list,
xfs_agnumber_t agno,
xfs_agblock_t agbno,
xfs_agblock_t length,
unsigned int gen)
{
struct xfs_buf *fbuf;
struct xfs_dinode *free;
int blks_per_cluster, nbufs, ninodes;
int version;
int i, j;
xfs_daddr_t d;
xfs_ino_t ino = 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 (mp->m_sb.sb_blocksize >= XFS_INODE_CLUSTER_SIZE(mp)) {
blks_per_cluster = 1;
nbufs = length;
ninodes = mp->m_sb.sb_inopblock;
} else {
blks_per_cluster = XFS_INODE_CLUSTER_SIZE(mp) /
mp->m_sb.sb_blocksize;
nbufs = length / blks_per_cluster;
ninodes = blks_per_cluster * 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.
*
* For v3 inodes, we also need to write the inode number into the inode,
* so calculate the first inode number of the chunk here as
* XFS_OFFBNO_TO_AGINO() only works within a filesystem block, not
* across multiple filesystem blocks (such as a cluster) and so cannot
* be used in the cluster buffer loop below.
*
* Further, because we are writing the inode directly into the buffer
* and calculating a CRC on the entire inode, we have ot log the entire
* inode so that the entire range the CRC covers is present in the log.
* That means for v3 inode we log the entire buffer rather than just the
* inode cores.
*/
if (xfs_sb_version_hascrc(&mp->m_sb)) {
version = 3;
ino = XFS_AGINO_TO_INO(mp, agno,
XFS_OFFBNO_TO_AGINO(mp, agbno, 0));
/*
* log the initialisation that is about to take place as an
* logical operation. This means the transaction does not
* need to log the physical changes to the inode buffers as log
* recovery will know what initialisation is actually needed.
* Hence we only need to log the buffers as "ordered" buffers so
* they track in the AIL as if they were physically logged.
*/
if (tp)
xfs_icreate_log(tp, agno, agbno, XFS_IALLOC_INODES(mp),
mp->m_sb.sb_inodesize, length, gen);
} else if (xfs_sb_version_hasnlink(&mp->m_sb))
version = 2;
else
version = 1;
for (j = 0; j < nbufs; j++) {
/*
* Get the block.
*/
d = XFS_AGB_TO_DADDR(mp, agno, agbno + (j * blks_per_cluster));
fbuf = xfs_trans_get_buf(tp, mp->m_ddev_targp, d,
mp->m_bsize * blks_per_cluster,
XBF_UNMAPPED);
if (!fbuf)
return ENOMEM;
/* Initialize the inode buffers and log them appropriately. */
fbuf->b_ops = &xfs_inode_buf_ops;
xfs_buf_zero(fbuf, 0, BBTOB(fbuf->b_length));
for (i = 0; i < ninodes; i++) {
int ioffset = i << mp->m_sb.sb_inodelog;
uint isize = xfs_dinode_size(version);
free = xfs_make_iptr(mp, fbuf, i);
free->di_magic = cpu_to_be16(XFS_DINODE_MAGIC);
free->di_version = version;
free->di_gen = cpu_to_be32(gen);
free->di_next_unlinked = cpu_to_be32(NULLAGINO);
if (version == 3) {
free->di_ino = cpu_to_be64(ino);
ino++;
uuid_copy(&free->di_uuid, &mp->m_sb.sb_uuid);
xfs_dinode_calc_crc(mp, free);
} else if (tp) {
/* just log the inode core */
xfs_trans_log_buf(tp, fbuf, ioffset,
ioffset + isize - 1);
}
}
if (tp) {
/*
* Mark the buffer as an inode allocation buffer so it
* sticks in AIL at the point of this allocation
* transaction. This ensures the they are on disk before
* the tail of the log can be moved past this
* transaction (i.e. by preventing relogging from moving
* it forward in the log).
*/
xfs_trans_inode_alloc_buf(tp, fbuf);
if (version == 3) {
/*
* Mark the buffer as ordered so that they are
* not physically logged in the transaction but
* still tracked in the AIL as part of the
* transaction and pin the log appropriately.
*/
xfs_trans_ordered_buf(tp, fbuf);
xfs_trans_log_buf(tp, fbuf, 0,
BBTOB(fbuf->b_length) - 1);
}
} else {
fbuf->b_flags |= XBF_DONE;
xfs_buf_delwri_queue(fbuf, buffer_list);
xfs_buf_relse(fbuf);
}
}
return 0;
}
/*
* Initialise a new set of inodes.
*/
STATIC int
xfs_ialloc_inode_init(
struct xfs_mount *mp,
struct xfs_trans *tp,
xfs_agnumber_t agno,
xfs_agblock_t agbno,
xfs_agblock_t length,
unsigned int gen)
{
struct xfs_buf *fbuf;
struct xfs_dinode *free;
int blks_per_cluster, nbufs, ninodes;
int version;
int i, j;
xfs_daddr_t d;
xfs_ino_t ino = 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 (mp->m_sb.sb_blocksize >= XFS_INODE_CLUSTER_SIZE(mp)) {
blks_per_cluster = 1;
nbufs = length;
ninodes = mp->m_sb.sb_inopblock;
} else {
blks_per_cluster = XFS_INODE_CLUSTER_SIZE(mp) /
mp->m_sb.sb_blocksize;
nbufs = length / blks_per_cluster;
ninodes = blks_per_cluster * 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.
*
* For v3 inodes, we also need to write the inode number into the inode,
* so calculate the first inode number of the chunk here as
* XFS_OFFBNO_TO_AGINO() only works within a filesystem block, not
* across multiple filesystem blocks (such as a cluster) and so cannot
* be used in the cluster buffer loop below.
*
* Further, because we are writing the inode directly into the buffer
* and calculating a CRC on the entire inode, we have ot log the entire
* inode so that the entire range the CRC covers is present in the log.
* That means for v3 inode we log the entire buffer rather than just the
* inode cores.
*/
if (xfs_sb_version_hascrc(&mp->m_sb)) {
version = 3;
ino = XFS_AGINO_TO_INO(mp, agno,
XFS_OFFBNO_TO_AGINO(mp, agbno, 0));
} else if (xfs_sb_version_hasnlink(&mp->m_sb))
version = 2;
else
version = 1;
for (j = 0; j < nbufs; j++) {
/*
* Get the block.
*/
d = XFS_AGB_TO_DADDR(mp, agno, agbno + (j * blks_per_cluster));
fbuf = xfs_trans_get_buf(tp, mp->m_ddev_targp, d,
mp->m_bsize * blks_per_cluster,
XBF_UNMAPPED);
if (!fbuf)
return ENOMEM;
/*
* Initialize all inodes in this buffer and then log them.
*
* XXX: It would be much better if we had just one transaction
* to log a whole cluster of inodes instead of all the
* individual transactions causing a lot of log traffic.
*/
fbuf->b_ops = &xfs_inode_buf_ops;
xfs_buf_zero(fbuf, 0, BBTOB(fbuf->b_length));
for (i = 0; i < ninodes; i++) {
int ioffset = i << mp->m_sb.sb_inodelog;
uint isize = xfs_dinode_size(version);
free = xfs_make_iptr(mp, fbuf, i);
free->di_magic = cpu_to_be16(XFS_DINODE_MAGIC);
free->di_version = version;
free->di_gen = cpu_to_be32(gen);
free->di_next_unlinked = cpu_to_be32(NULLAGINO);
if (version == 3) {
free->di_ino = cpu_to_be64(ino);
ino++;
uuid_copy(&free->di_uuid, &mp->m_sb.sb_uuid);
xfs_dinode_calc_crc(mp, free);
} else {
/* just log the inode core */
xfs_trans_log_buf(tp, fbuf, ioffset,
ioffset + isize - 1);
}
}
if (version == 3) {
/* need to log the entire buffer */
xfs_trans_log_buf(tp, fbuf, 0,
BBTOB(fbuf->b_length) - 1);
}
xfs_trans_inode_alloc_buf(tp, fbuf);
}
return 0;
}
STATIC void
xfs_inode_item_format(
struct xfs_log_item *lip,
struct xfs_log_iovec *vecp)
{
struct xfs_inode_log_item *iip = INODE_ITEM(lip);
struct xfs_inode *ip = iip->ili_inode;
uint nvecs;
size_t data_bytes;
xfs_mount_t *mp;
vecp->i_addr = &iip->ili_format;
vecp->i_len = sizeof(xfs_inode_log_format_t);
vecp->i_type = XLOG_REG_TYPE_IFORMAT;
vecp++;
nvecs = 1;
vecp->i_addr = &ip->i_d;
vecp->i_len = sizeof(struct xfs_icdinode);
vecp->i_type = XLOG_REG_TYPE_ICORE;
vecp++;
nvecs++;
/*
*/
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)) {
/*
*/
ASSERT(ip->i_d.di_nlink <= XFS_MAXLINK_1);
ip->i_d.di_onlink = ip->i_d.di_nlink;
} else {
/*
*/
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:
iip->ili_fields &=
~(XFS_ILOG_DDATA | XFS_ILOG_DBROOT |
XFS_ILOG_DEV | XFS_ILOG_UUID);
if ((iip->ili_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);
ASSERT(ip->i_df.if_bytes / sizeof(xfs_bmbt_rec_t) > 0);
ASSERT(iip->ili_extents_buf == NULL);
#ifdef XFS_NATIVE_HOST
if (ip->i_d.di_nextents == ip->i_df.if_bytes /
(uint)sizeof(xfs_bmbt_rec_t)) {
/*
*/
vecp->i_addr = ip->i_df.if_u1.if_extents;
vecp->i_len = ip->i_df.if_bytes;
vecp->i_type = XLOG_REG_TYPE_IEXT;
} else
#endif
{
xfs_inode_item_format_extents(ip, vecp,
XFS_DATA_FORK, XLOG_REG_TYPE_IEXT);
}
ASSERT(vecp->i_len <= ip->i_df.if_bytes);
iip->ili_format.ilf_dsize = vecp->i_len;
vecp++;
nvecs++;
} else {
iip->ili_fields &= ~XFS_ILOG_DEXT;
}
break;
case XFS_DINODE_FMT_BTREE:
iip->ili_fields &=
~(XFS_ILOG_DDATA | XFS_ILOG_DEXT |
XFS_ILOG_DEV | XFS_ILOG_UUID);
if ((iip->ili_fields & XFS_ILOG_DBROOT) &&
ip->i_df.if_broot_bytes > 0) {
ASSERT(ip->i_df.if_broot != NULL);
vecp->i_addr = ip->i_df.if_broot;
vecp->i_len = ip->i_df.if_broot_bytes;
vecp->i_type = XLOG_REG_TYPE_IBROOT;
vecp++;
nvecs++;
iip->ili_format.ilf_dsize = ip->i_df.if_broot_bytes;
} else {
ASSERT(!(iip->ili_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_fields &= ~XFS_ILOG_DBROOT;
}
break;
case XFS_DINODE_FMT_LOCAL:
iip->ili_fields &=
~(XFS_ILOG_DEXT | XFS_ILOG_DBROOT |
XFS_ILOG_DEV | XFS_ILOG_UUID);
if ((iip->ili_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);
vecp->i_addr = ip->i_df.if_u1.if_data;
/*
*/
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->i_type = XLOG_REG_TYPE_ILOCAL;
vecp++;
nvecs++;
iip->ili_format.ilf_dsize = (unsigned)data_bytes;
} else {
iip->ili_fields &= ~XFS_ILOG_DDATA;
}
break;
case XFS_DINODE_FMT_DEV:
iip->ili_fields &=
~(XFS_ILOG_DDATA | XFS_ILOG_DBROOT |
XFS_ILOG_DEXT | XFS_ILOG_UUID);
if (iip->ili_fields & XFS_ILOG_DEV) {
iip->ili_format.ilf_u.ilfu_rdev =
ip->i_df.if_u2.if_rdev;
}
break;
case XFS_DINODE_FMT_UUID:
iip->ili_fields &=
~(XFS_ILOG_DDATA | XFS_ILOG_DBROOT |
XFS_ILOG_DEXT | XFS_ILOG_DEV);
if (iip->ili_fields & XFS_ILOG_UUID) {
iip->ili_format.ilf_u.ilfu_uuid =
ip->i_df.if_u2.if_uuid;
}
break;
default:
ASSERT(0);
break;
}
/*
*/
if (!XFS_IFORK_Q(ip)) {
iip->ili_fields &=
~(XFS_ILOG_ADATA | XFS_ILOG_ABROOT | XFS_ILOG_AEXT);
goto out;
}
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) {
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);
#ifdef XFS_NATIVE_HOST
/*
*/
vecp->i_addr = ip->i_afp->if_u1.if_extents;
vecp->i_len = ip->i_afp->if_bytes;
vecp->i_type = XLOG_REG_TYPE_IATTR_EXT;
#else
ASSERT(iip->ili_aextents_buf == NULL);
xfs_inode_item_format_extents(ip, vecp,
XFS_ATTR_FORK, XLOG_REG_TYPE_IATTR_EXT);
#endif
iip->ili_format.ilf_asize = vecp->i_len;
vecp++;
nvecs++;
} 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);
vecp->i_addr = ip->i_afp->if_broot;
vecp->i_len = ip->i_afp->if_broot_bytes;
vecp->i_type = XLOG_REG_TYPE_IATTR_BROOT;
vecp++;
nvecs++;
iip->ili_format.ilf_asize = ip->i_afp->if_broot_bytes;
} 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) {
ASSERT(ip->i_afp->if_u1.if_data != NULL);
vecp->i_addr = ip->i_afp->if_u1.if_data;
/*
*/
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->i_type = XLOG_REG_TYPE_IATTR_LOCAL;
vecp++;
nvecs++;
iip->ili_format.ilf_asize = (unsigned)data_bytes;
} else {
iip->ili_fields &= ~XFS_ILOG_ADATA;
}
break;
default:
ASSERT(0);
break;
}
out:
/*
*/
iip->ili_format.ilf_fields = XFS_ILOG_CORE |
(iip->ili_fields & ~XFS_ILOG_TIMESTAMP);
iip->ili_format.ilf_size = nvecs;
}
/*
* 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(
struct xfs_log_item *lip,
struct xfs_log_iovec *vecp)
{
struct xfs_inode_log_item *iip = INODE_ITEM(lip);
struct xfs_inode *ip = iip->ili_inode;
uint nvecs;
size_t data_bytes;
xfs_mount_t *mp;
vecp->i_addr = &iip->ili_format;
vecp->i_len = sizeof(xfs_inode_log_format_t);
vecp->i_type = XLOG_REG_TYPE_IFORMAT;
vecp++;
nvecs = 1;
vecp->i_addr = &ip->i_d;
vecp->i_len = sizeof(struct xfs_icdinode);
vecp->i_type = XLOG_REG_TYPE_ICORE;
vecp++;
nvecs++;
/*
* 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:
iip->ili_fields &=
~(XFS_ILOG_DDATA | XFS_ILOG_DBROOT |
XFS_ILOG_DEV | XFS_ILOG_UUID);
if ((iip->ili_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);
ASSERT(ip->i_df.if_bytes / sizeof(xfs_bmbt_rec_t) > 0);
ASSERT(iip->ili_extents_buf == NULL);
#ifdef XFS_NATIVE_HOST
if (ip->i_d.di_nextents == ip->i_df.if_bytes /
(uint)sizeof(xfs_bmbt_rec_t)) {
/*
* There are no delayed allocation
* extents, so just point to the
* real extents array.
*/
vecp->i_addr = ip->i_df.if_u1.if_extents;
vecp->i_len = ip->i_df.if_bytes;
vecp->i_type = XLOG_REG_TYPE_IEXT;
} else
#endif
{
xfs_inode_item_format_extents(ip, vecp,
XFS_DATA_FORK, XLOG_REG_TYPE_IEXT);
}
ASSERT(vecp->i_len <= ip->i_df.if_bytes);
iip->ili_format.ilf_dsize = vecp->i_len;
vecp++;
nvecs++;
} else {
iip->ili_fields &= ~XFS_ILOG_DEXT;
}
break;
case XFS_DINODE_FMT_BTREE:
iip->ili_fields &=
~(XFS_ILOG_DDATA | XFS_ILOG_DEXT |
XFS_ILOG_DEV | XFS_ILOG_UUID);
if ((iip->ili_fields & XFS_ILOG_DBROOT) &&
ip->i_df.if_broot_bytes > 0) {
ASSERT(ip->i_df.if_broot != NULL);
vecp->i_addr = ip->i_df.if_broot;
vecp->i_len = ip->i_df.if_broot_bytes;
vecp->i_type = XLOG_REG_TYPE_IBROOT;
vecp++;
nvecs++;
iip->ili_format.ilf_dsize = ip->i_df.if_broot_bytes;
} else {
ASSERT(!(iip->ili_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_fields &= ~XFS_ILOG_DBROOT;
}
break;
case XFS_DINODE_FMT_LOCAL:
iip->ili_fields &=
~(XFS_ILOG_DEXT | XFS_ILOG_DBROOT |
XFS_ILOG_DEV | XFS_ILOG_UUID);
if ((iip->ili_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);
vecp->i_addr = 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->i_type = XLOG_REG_TYPE_ILOCAL;
vecp++;
nvecs++;
iip->ili_format.ilf_dsize = (unsigned)data_bytes;
} else {
iip->ili_fields &= ~XFS_ILOG_DDATA;
}
break;
case XFS_DINODE_FMT_DEV:
iip->ili_fields &=
~(XFS_ILOG_DDATA | XFS_ILOG_DBROOT |
XFS_ILOG_DEXT | XFS_ILOG_UUID);
if (iip->ili_fields & XFS_ILOG_DEV) {
iip->ili_format.ilf_u.ilfu_rdev =
ip->i_df.if_u2.if_rdev;
}
break;
case XFS_DINODE_FMT_UUID:
iip->ili_fields &=
~(XFS_ILOG_DDATA | XFS_ILOG_DBROOT |
XFS_ILOG_DEXT | XFS_ILOG_DEV);
if (iip->ili_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.
*/
if (!XFS_IFORK_Q(ip)) {
iip->ili_fields &=
~(XFS_ILOG_ADATA | XFS_ILOG_ABROOT | XFS_ILOG_AEXT);
goto out;
}
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) {
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);
#ifdef XFS_NATIVE_HOST
/*
* There are not delayed allocation extents
* for attributes, so just point at the array.
*/
vecp->i_addr = ip->i_afp->if_u1.if_extents;
vecp->i_len = ip->i_afp->if_bytes;
vecp->i_type = XLOG_REG_TYPE_IATTR_EXT;
#else
ASSERT(iip->ili_aextents_buf == NULL);
xfs_inode_item_format_extents(ip, vecp,
XFS_ATTR_FORK, XLOG_REG_TYPE_IATTR_EXT);
#endif
iip->ili_format.ilf_asize = vecp->i_len;
vecp++;
nvecs++;
} 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);
vecp->i_addr = ip->i_afp->if_broot;
vecp->i_len = ip->i_afp->if_broot_bytes;
vecp->i_type = XLOG_REG_TYPE_IATTR_BROOT;
vecp++;
nvecs++;
iip->ili_format.ilf_asize = ip->i_afp->if_broot_bytes;
} 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) {
ASSERT(ip->i_afp->if_u1.if_data != NULL);
vecp->i_addr = 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->i_type = XLOG_REG_TYPE_IATTR_LOCAL;
vecp++;
nvecs++;
iip->ili_format.ilf_asize = (unsigned)data_bytes;
} else {
iip->ili_fields &= ~XFS_ILOG_ADATA;
}
break;
default:
ASSERT(0);
break;
}
out:
/*
* Now update the log format that goes out to disk from the in-core
* values. We always write the inode core to make the arithmetic
* games in recovery easier, which isn't a big deal as just about any
* transaction would dirty it anyway.
*/
iip->ili_format.ilf_fields = XFS_ILOG_CORE |
(iip->ili_fields & ~XFS_ILOG_TIMESTAMP);
iip->ili_format.ilf_size = nvecs;
}
