int
xfs_iomap_write_unwritten(
xfs_inode_t *ip,
xfs_off_t offset,
size_t count)
{
xfs_mount_t *mp = ip->i_mount;
xfs_iocore_t *io = &ip->i_iocore;
xfs_trans_t *tp;
xfs_fileoff_t offset_fsb;
xfs_filblks_t count_fsb;
xfs_filblks_t numblks_fsb;
xfs_bmbt_irec_t imap;
int committed;
int error;
int nres;
int nimaps;
xfs_fsblock_t firstfsb;
xfs_bmap_free_t free_list;
xfs_iomap_enter_trace(XFS_IOMAP_UNWRITTEN,
&ip->i_iocore, offset, count);
offset_fsb = XFS_B_TO_FSBT(mp, offset);
count_fsb = XFS_B_TO_FSB(mp, (xfs_ufsize_t)offset + count);
count_fsb = (xfs_filblks_t)(count_fsb - offset_fsb);
do {
nres = XFS_DIOSTRAT_SPACE_RES(mp, 0);
/*
* set up a transaction to convert the range of extents
* from unwritten to real. Do allocations in a loop until
* we have covered the range passed in.
*/
tp = xfs_trans_alloc(mp, XFS_TRANS_STRAT_WRITE);
error = xfs_trans_reserve(tp, nres,
XFS_WRITE_LOG_RES(mp), 0,
XFS_TRANS_PERM_LOG_RES,
XFS_WRITE_LOG_COUNT);
if (error) {
xfs_trans_cancel(tp, 0);
goto error0;
}
xfs_ilock(ip, XFS_ILOCK_EXCL);
xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
xfs_trans_ihold(tp, ip);
/*
* Modify the unwritten extent state of the buffer.
*/
XFS_BMAP_INIT(&free_list, &firstfsb);
nimaps = 1;
error = xfs_bmapi(tp, ip, offset_fsb, count_fsb,
XFS_BMAPI_WRITE, &firstfsb,
1, &imap, &nimaps, &free_list);
if (error)
goto error_on_bmapi_transaction;
error = xfs_bmap_finish(&(tp), &(free_list),
firstfsb, &committed);
if (error)
goto error_on_bmapi_transaction;
error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES, NULL);
xfs_iunlock(ip, XFS_ILOCK_EXCL);
if (error)
goto error0;
if ( !(io->io_flags & XFS_IOCORE_RT) && !imap.br_startblock) {
cmn_err(CE_PANIC,"Access to block zero: fs <%s> "
"inode: %lld start_block : %llx start_off : "
"%llx blkcnt : %llx extent-state : %x \n",
(ip->i_mount)->m_fsname,
(long long)ip->i_ino,
imap.br_startblock,imap.br_startoff,
imap.br_blockcount,imap.br_state);
}
if ((numblks_fsb = imap.br_blockcount) == 0) {
/*
* The numblks_fsb value should always get
* smaller, otherwise the loop is stuck.
*/
ASSERT(imap.br_blockcount);
break;
}
offset_fsb += numblks_fsb;
count_fsb -= numblks_fsb;
} while (count_fsb > 0);
return 0;
error_on_bmapi_transaction:
xfs_bmap_cancel(&free_list);
xfs_trans_cancel(tp, (XFS_TRANS_RELEASE_LOG_RES | XFS_TRANS_ABORT));
xfs_iunlock(ip, XFS_ILOCK_EXCL);
error0:
return XFS_ERROR(error);
}
/*
* Try to truncate the given file to 0 length. Currently called
* only out of xfs_remove when it has to truncate a file to free
* up space for the remove to proceed.
*/
int
xfs_truncate_file(
xfs_mount_t *mp,
xfs_inode_t *ip)
{
xfs_trans_t *tp;
int error;
#ifdef QUOTADEBUG
/*
* This is called to truncate the quotainodes too.
*/
if (XFS_IS_UQUOTA_ON(mp)) {
if (ip->i_ino != mp->m_sb.sb_uquotino)
ASSERT(ip->i_udquot);
}
if (XFS_IS_GQUOTA_ON(mp)) {
if (ip->i_ino != mp->m_sb.sb_gquotino)
ASSERT(ip->i_gdquot);
}
#endif
/*
* Make the call to xfs_itruncate_start before starting the
* transaction, because we cannot make the call while we're
* in a transaction.
*/
xfs_ilock(ip, XFS_IOLOCK_EXCL);
xfs_itruncate_start(ip, XFS_ITRUNC_DEFINITE, (xfs_fsize_t)0);
tp = xfs_trans_alloc(mp, XFS_TRANS_TRUNCATE_FILE);
if ((error = xfs_trans_reserve(tp, 0, XFS_ITRUNCATE_LOG_RES(mp), 0,
XFS_TRANS_PERM_LOG_RES,
XFS_ITRUNCATE_LOG_COUNT))) {
xfs_trans_cancel(tp, 0);
xfs_iunlock(ip, XFS_IOLOCK_EXCL);
return error;
}
/*
* Follow the normal truncate locking protocol. Since we
* hold the inode in the transaction, we know that it's number
* of references will stay constant.
*/
xfs_ilock(ip, XFS_ILOCK_EXCL);
xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL);
xfs_trans_ihold(tp, ip);
/*
* Signal a sync xaction. The only case where that isn't
* the case is if we're truncating an already unlinked file
* on a wsync fs. In that case, we know the blocks can't
* reappear in the file because the links to file are
* permanently toast. Currently, we're always going to
* want a sync transaction because this code is being
* called from places where nlink is guaranteed to be 1
* but I'm leaving the tests in to protect against future
* changes -- rcc.
*/
error = xfs_itruncate_finish(&tp, ip, (xfs_fsize_t)0,
XFS_DATA_FORK,
((ip->i_d.di_nlink != 0 ||
!(mp->m_flags & XFS_MOUNT_WSYNC))
? 1 : 0));
if (error) {
xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES |
XFS_TRANS_ABORT);
} else {
xfs_ichgtime(ip, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES,
NULL);
}
xfs_iunlock(ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL);
return error;
}
/*
* Truncate file. Must have write permission and not be a directory.
*/
int
xfs_setattr_size(
struct xfs_inode *ip,
struct iattr *iattr)
{
struct xfs_mount *mp = ip->i_mount;
struct inode *inode = VFS_I(ip);
xfs_off_t oldsize, newsize;
struct xfs_trans *tp;
int error;
uint lock_flags = 0;
uint commit_flags = 0;
trace_xfs_setattr(ip);
if (mp->m_flags & XFS_MOUNT_RDONLY)
return XFS_ERROR(EROFS);
if (XFS_FORCED_SHUTDOWN(mp))
return XFS_ERROR(EIO);
error = -inode_change_ok(inode, iattr);
if (error)
return XFS_ERROR(error);
ASSERT(xfs_isilocked(ip, XFS_IOLOCK_EXCL));
ASSERT(S_ISREG(ip->i_d.di_mode));
ASSERT((iattr->ia_valid & (ATTR_UID|ATTR_GID|ATTR_ATIME|ATTR_ATIME_SET|
ATTR_MTIME_SET|ATTR_KILL_PRIV|ATTR_TIMES_SET)) == 0);
oldsize = inode->i_size;
newsize = iattr->ia_size;
/*
* Short circuit the truncate case for zero length files.
*/
if (newsize == 0 && oldsize == 0 && ip->i_d.di_nextents == 0) {
if (!(iattr->ia_valid & (ATTR_CTIME|ATTR_MTIME)))
return 0;
/*
* Use the regular setattr path to update the timestamps.
*/
iattr->ia_valid &= ~ATTR_SIZE;
return xfs_setattr_nonsize(ip, iattr, 0);
}
/*
* Make sure that the dquots are attached to the inode.
*/
error = xfs_qm_dqattach(ip, 0);
if (error)
return error;
/*
* Now we can make the changes. Before we join the inode to the
* transaction, take care of the part of the truncation that must be
* done without the inode lock. This needs to be done before joining
* the inode to the transaction, because the inode cannot be unlocked
* once it is a part of the transaction.
*/
if (newsize > oldsize) {
/*
* Do the first part of growing a file: zero any data in the
* last block that is beyond the old EOF. We need to do this
* before the inode is joined to the transaction to modify
* i_size.
*/
error = xfs_zero_eof(ip, newsize, oldsize);
if (error)
return error;
}
/*
* We are going to log the inode size change in this transaction so
* any previous writes that are beyond the on disk EOF and the new
* EOF that have not been written out need to be written here. If we
* do not write the data out, we expose ourselves to the null files
* problem.
*
* Only flush from the on disk size to the smaller of the in memory
* file size or the new size as that's the range we really care about
* here and prevents waiting for other data not within the range we
* care about here.
*/
if (oldsize != ip->i_d.di_size && newsize > ip->i_d.di_size) {
error = -filemap_write_and_wait_range(VFS_I(ip)->i_mapping,
ip->i_d.di_size, newsize);
if (error)
return error;
}
/*
* Wait for all direct I/O to complete.
*/
inode_dio_wait(inode);
error = -block_truncate_page(inode->i_mapping, newsize, xfs_get_blocks);
if (error)
return error;
tp = xfs_trans_alloc(mp, XFS_TRANS_SETATTR_SIZE);
error = xfs_trans_reserve(tp, &M_RES(mp)->tr_itruncate, 0, 0);
if (error)
goto out_trans_cancel;
truncate_setsize(inode, newsize);
commit_flags = XFS_TRANS_RELEASE_LOG_RES;
lock_flags |= XFS_ILOCK_EXCL;
xfs_ilock(ip, XFS_ILOCK_EXCL);
xfs_trans_ijoin(tp, ip, 0);
/*
* Only change the c/mtime if we are changing the size or we are
* explicitly asked to change it. This handles the semantic difference
* between truncate() and ftruncate() as implemented in the VFS.
*
* The regular truncate() case without ATTR_CTIME and ATTR_MTIME is a
* special case where we need to update the times despite not having
* these flags set. For all other operations the VFS set these flags
* explicitly if it wants a timestamp update.
*/
if (newsize != oldsize &&
!(iattr->ia_valid & (ATTR_CTIME | ATTR_MTIME))) {
iattr->ia_ctime = iattr->ia_mtime =
current_fs_time(inode->i_sb);
iattr->ia_valid |= ATTR_CTIME | ATTR_MTIME;
}
/*
* The first thing we do is set the size to new_size permanently on
* disk. This way we don't have to worry about anyone ever being able
* to look at the data being freed even in the face of a crash.
* What we're getting around here is the case where we free a block, it
* is allocated to another file, it is written to, and then we crash.
* If the new data gets written to the file but the log buffers
* containing the free and reallocation don't, then we'd end up with
* garbage in the blocks being freed. As long as we make the new size
* permanent before actually freeing any blocks it doesn't matter if
* they get written to.
*/
ip->i_d.di_size = newsize;
xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
if (newsize <= oldsize) {
error = xfs_itruncate_extents(&tp, ip, XFS_DATA_FORK, newsize);
if (error)
goto out_trans_abort;
/*
* Truncated "down", so we're removing references to old data
* here - if we delay flushing for a long time, we expose
* ourselves unduly to the notorious NULL files problem. So,
* we mark this inode and flush it when the file is closed,
* and do not wait the usual (long) time for writeout.
*/
xfs_iflags_set(ip, XFS_ITRUNCATED);
/* A truncate down always removes post-EOF blocks. */
xfs_inode_clear_eofblocks_tag(ip);
}
if (iattr->ia_valid & ATTR_MODE)
xfs_setattr_mode(ip, iattr);
if (iattr->ia_valid & (ATTR_ATIME|ATTR_CTIME|ATTR_MTIME))
xfs_setattr_time(ip, iattr);
xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
XFS_STATS_INC(xs_ig_attrchg);
if (mp->m_flags & XFS_MOUNT_WSYNC)
xfs_trans_set_sync(tp);
error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
out_unlock:
if (lock_flags)
xfs_iunlock(ip, lock_flags);
return error;
out_trans_abort:
commit_flags |= XFS_TRANS_ABORT;
out_trans_cancel:
xfs_trans_cancel(tp, commit_flags);
goto out_unlock;
}
/*
* Truncate file. Must have write permission and not be a directory.
*/
int
xfs_setattr_size(
struct xfs_inode *ip,
struct iattr *iattr,
int flags)
{
struct xfs_mount *mp = ip->i_mount;
struct inode *inode = VFS_I(ip);
int mask = iattr->ia_valid;
struct xfs_trans *tp;
int error;
uint lock_flags;
uint commit_flags = 0;
trace_xfs_setattr(ip);
if (mp->m_flags & XFS_MOUNT_RDONLY)
return XFS_ERROR(EROFS);
if (XFS_FORCED_SHUTDOWN(mp))
return XFS_ERROR(EIO);
error = -inode_change_ok(inode, iattr);
if (error)
return XFS_ERROR(error);
ASSERT(S_ISREG(ip->i_d.di_mode));
ASSERT((mask & (ATTR_MODE|ATTR_UID|ATTR_GID|ATTR_ATIME|ATTR_ATIME_SET|
ATTR_MTIME_SET|ATTR_KILL_SUID|ATTR_KILL_SGID|
ATTR_KILL_PRIV|ATTR_TIMES_SET)) == 0);
lock_flags = XFS_ILOCK_EXCL;
if (!(flags & XFS_ATTR_NOLOCK))
lock_flags |= XFS_IOLOCK_EXCL;
xfs_ilock(ip, lock_flags);
/*
* Short circuit the truncate case for zero length files.
*/
if (iattr->ia_size == 0 &&
ip->i_size == 0 && ip->i_d.di_nextents == 0) {
if (!(mask & (ATTR_CTIME|ATTR_MTIME)))
goto out_unlock;
/*
* Use the regular setattr path to update the timestamps.
*/
xfs_iunlock(ip, lock_flags);
iattr->ia_valid &= ~ATTR_SIZE;
return xfs_setattr_nonsize(ip, iattr, 0);
}
/*
* Make sure that the dquots are attached to the inode.
*/
error = xfs_qm_dqattach_locked(ip, 0);
if (error)
goto out_unlock;
/*
* Now we can make the changes. Before we join the inode to the
* transaction, take care of the part of the truncation that must be
* done without the inode lock. This needs to be done before joining
* the inode to the transaction, because the inode cannot be unlocked
* once it is a part of the transaction.
*/
if (iattr->ia_size > ip->i_size) {
/*
* Do the first part of growing a file: zero any data in the
* last block that is beyond the old EOF. We need to do this
* before the inode is joined to the transaction to modify
* i_size.
*/
error = xfs_zero_eof(ip, iattr->ia_size, ip->i_size);
if (error)
goto out_unlock;
}
xfs_iunlock(ip, XFS_ILOCK_EXCL);
lock_flags &= ~XFS_ILOCK_EXCL;
/*
* We are going to log the inode size change in this transaction so
* any previous writes that are beyond the on disk EOF and the new
* EOF that have not been written out need to be written here. If we
* do not write the data out, we expose ourselves to the null files
* problem.
*
* Only flush from the on disk size to the smaller of the in memory
* file size or the new size as that's the range we really care about
* here and prevents waiting for other data not within the range we
* care about here.
*/
if (ip->i_size != ip->i_d.di_size && iattr->ia_size > ip->i_d.di_size) {
error = xfs_flush_pages(ip, ip->i_d.di_size, iattr->ia_size,
XBF_ASYNC, FI_NONE);
if (error)
goto out_unlock;
}
/*
* Wait for all I/O to complete.
*/
xfs_ioend_wait(ip);
error = -block_truncate_page(inode->i_mapping, iattr->ia_size,
xfs_get_blocks);
if (error)
goto out_unlock;
tp = xfs_trans_alloc(mp, XFS_TRANS_SETATTR_SIZE);
error = xfs_trans_reserve(tp, 0, XFS_ITRUNCATE_LOG_RES(mp), 0,
XFS_TRANS_PERM_LOG_RES,
XFS_ITRUNCATE_LOG_COUNT);
if (error)
goto out_trans_cancel;
truncate_setsize(inode, iattr->ia_size);
commit_flags = XFS_TRANS_RELEASE_LOG_RES;
lock_flags |= XFS_ILOCK_EXCL;
xfs_ilock(ip, XFS_ILOCK_EXCL);
xfs_trans_ijoin(tp, ip);
/*
* Only change the c/mtime if we are changing the size or we are
* explicitly asked to change it. This handles the semantic difference
* between truncate() and ftruncate() as implemented in the VFS.
*
* The regular truncate() case without ATTR_CTIME and ATTR_MTIME is a
* special case where we need to update the times despite not having
* these flags set. For all other operations the VFS set these flags
* explicitly if it wants a timestamp update.
*/
if (iattr->ia_size != ip->i_size &&
(!(mask & (ATTR_CTIME | ATTR_MTIME)))) {
iattr->ia_ctime = iattr->ia_mtime =
current_fs_time(inode->i_sb);
mask |= ATTR_CTIME | ATTR_MTIME;
}
if (iattr->ia_size > ip->i_size) {
ip->i_d.di_size = iattr->ia_size;
ip->i_size = iattr->ia_size;
} else if (iattr->ia_size <= ip->i_size ||
(iattr->ia_size == 0 && ip->i_d.di_nextents)) {
error = xfs_itruncate_data(&tp, ip, iattr->ia_size);
if (error)
goto out_trans_abort;
/*
* Truncated "down", so we're removing references to old data
* here - if we delay flushing for a long time, we expose
* ourselves unduly to the notorious NULL files problem. So,
* we mark this inode and flush it when the file is closed,
* and do not wait the usual (long) time for writeout.
*/
xfs_iflags_set(ip, XFS_ITRUNCATED);
}
if (mask & ATTR_CTIME) {
inode->i_ctime = iattr->ia_ctime;
ip->i_d.di_ctime.t_sec = iattr->ia_ctime.tv_sec;
ip->i_d.di_ctime.t_nsec = iattr->ia_ctime.tv_nsec;
ip->i_update_core = 1;
}
if (mask & ATTR_MTIME) {
inode->i_mtime = iattr->ia_mtime;
ip->i_d.di_mtime.t_sec = iattr->ia_mtime.tv_sec;
ip->i_d.di_mtime.t_nsec = iattr->ia_mtime.tv_nsec;
ip->i_update_core = 1;
}
xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
XFS_STATS_INC(xs_ig_attrchg);
if (mp->m_flags & XFS_MOUNT_WSYNC)
xfs_trans_set_sync(tp);
error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
out_unlock:
if (lock_flags)
xfs_iunlock(ip, lock_flags);
return error;
out_trans_abort:
commit_flags |= XFS_TRANS_ABORT;
out_trans_cancel:
xfs_trans_cancel(tp, commit_flags);
goto out_unlock;
}
/*
* allocate an incore dquot from the kernel heap,
* and fill its core with quota information kept on disk.
* If XFS_QMOPT_DQALLOC is set, it'll allocate a dquot on disk
* if it wasn't already allocated.
*/
STATIC int
xfs_qm_idtodq(
xfs_mount_t *mp,
xfs_dqid_t id, /* gid or uid, depending on type */
uint type, /* UDQUOT or GDQUOT */
uint flags, /* DQALLOC, DQREPAIR */
xfs_dquot_t **O_dqpp)/* OUT : incore dquot, not locked */
{
xfs_dquot_t *dqp;
int error;
xfs_trans_t *tp;
int cancelflags=0;
dqp = xfs_qm_dqinit(mp, id, type);
tp = NULL;
if (flags & XFS_QMOPT_DQALLOC) {
tp = xfs_trans_alloc(mp, XFS_TRANS_QM_DQALLOC);
if ((error = xfs_trans_reserve(tp,
XFS_QM_DQALLOC_SPACE_RES(mp),
XFS_WRITE_LOG_RES(mp) +
BBTOB(XFS_QI_DQCHUNKLEN(mp)) - 1 +
128,
0,
XFS_TRANS_PERM_LOG_RES,
XFS_WRITE_LOG_COUNT))) {
cancelflags = 0;
goto error0;
}
cancelflags = XFS_TRANS_RELEASE_LOG_RES;
}
/*
* Read it from disk; xfs_dqread() takes care of
* all the necessary initialization of dquot's fields (locks, etc)
*/
if ((error = xfs_qm_dqread(&tp, id, dqp, flags))) {
/*
* This can happen if quotas got turned off (ESRCH),
* or if the dquot didn't exist on disk and we ask to
* allocate (ENOENT).
*/
xfs_dqtrace_entry(dqp, "DQREAD FAIL");
cancelflags |= XFS_TRANS_ABORT;
goto error0;
}
if (tp) {
if ((error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES)))
goto error1;
}
*O_dqpp = dqp;
return (0);
error0:
ASSERT(error);
if (tp)
xfs_trans_cancel(tp, cancelflags);
error1:
xfs_qm_dqdestroy(dqp);
*O_dqpp = NULL;
return (error);
}
/*
* Process a bmap update intent item that was recovered from the log.
* We need to update some inode's bmbt.
*/
int
xfs_bui_recover(
struct xfs_mount *mp,
struct xfs_bui_log_item *buip)
{
int error = 0;
unsigned int bui_type;
struct xfs_map_extent *bmap;
xfs_fsblock_t startblock_fsb;
xfs_fsblock_t inode_fsb;
bool op_ok;
struct xfs_bud_log_item *budp;
enum xfs_bmap_intent_type type;
int whichfork;
xfs_exntst_t state;
struct xfs_trans *tp;
struct xfs_inode *ip = NULL;
struct xfs_defer_ops dfops;
xfs_fsblock_t firstfsb;
ASSERT(!test_bit(XFS_BUI_RECOVERED, &buip->bui_flags));
/* Only one mapping operation per BUI... */
if (buip->bui_format.bui_nextents != XFS_BUI_MAX_FAST_EXTENTS) {
set_bit(XFS_BUI_RECOVERED, &buip->bui_flags);
xfs_bui_release(buip);
return -EIO;
}
/*
* First check the validity of the extent described by the
* BUI. If anything is bad, then toss the BUI.
*/
bmap = &buip->bui_format.bui_extents[0];
startblock_fsb = XFS_BB_TO_FSB(mp,
XFS_FSB_TO_DADDR(mp, bmap->me_startblock));
inode_fsb = XFS_BB_TO_FSB(mp, XFS_FSB_TO_DADDR(mp,
XFS_INO_TO_FSB(mp, bmap->me_owner)));
switch (bmap->me_flags & XFS_BMAP_EXTENT_TYPE_MASK) {
case XFS_BMAP_MAP:
case XFS_BMAP_UNMAP:
op_ok = true;
break;
default:
op_ok = false;
break;
}
if (!op_ok || startblock_fsb == 0 ||
bmap->me_len == 0 ||
inode_fsb == 0 ||
startblock_fsb >= mp->m_sb.sb_dblocks ||
bmap->me_len >= mp->m_sb.sb_agblocks ||
inode_fsb >= mp->m_sb.sb_dblocks ||
(bmap->me_flags & ~XFS_BMAP_EXTENT_FLAGS)) {
/*
* This will pull the BUI from the AIL and
* free the memory associated with it.
*/
set_bit(XFS_BUI_RECOVERED, &buip->bui_flags);
xfs_bui_release(buip);
return -EIO;
}
error = xfs_trans_alloc(mp, &M_RES(mp)->tr_itruncate, 0, 0, 0, &tp);
if (error)
return error;
budp = xfs_trans_get_bud(tp, buip);
/* Grab the inode. */
error = xfs_iget(mp, tp, bmap->me_owner, 0, XFS_ILOCK_EXCL, &ip);
if (error)
goto err_inode;
if (VFS_I(ip)->i_nlink == 0)
xfs_iflags_set(ip, XFS_IRECOVERY);
xfs_defer_init(&dfops, &firstfsb);
/* Process deferred bmap item. */
state = (bmap->me_flags & XFS_BMAP_EXTENT_UNWRITTEN) ?
XFS_EXT_UNWRITTEN : XFS_EXT_NORM;
whichfork = (bmap->me_flags & XFS_BMAP_EXTENT_ATTR_FORK) ?
XFS_ATTR_FORK : XFS_DATA_FORK;
bui_type = bmap->me_flags & XFS_BMAP_EXTENT_TYPE_MASK;
switch (bui_type) {
case XFS_BMAP_MAP:
case XFS_BMAP_UNMAP:
type = bui_type;
break;
default:
error = -EFSCORRUPTED;
goto err_dfops;
}
xfs_trans_ijoin(tp, ip, 0);
error = xfs_trans_log_finish_bmap_update(tp, budp, &dfops, type,
ip, whichfork, bmap->me_startoff,
bmap->me_startblock, bmap->me_len,
state);
if (error)
goto err_dfops;
/* Finish transaction, free inodes. */
error = xfs_defer_finish(&tp, &dfops, NULL);
if (error)
goto err_dfops;
set_bit(XFS_BUI_RECOVERED, &buip->bui_flags);
error = xfs_trans_commit(tp);
xfs_iunlock(ip, XFS_ILOCK_EXCL);
IRELE(ip);
return error;
err_dfops:
xfs_defer_cancel(&dfops);
err_inode:
xfs_trans_cancel(tp);
if (ip) {
xfs_iunlock(ip, XFS_ILOCK_EXCL);
IRELE(ip);
}
return error;
}
ssize_t /* bytes written, or (-) error */
xfs_write(
bhv_desc_t *bdp,
struct kiocb *iocb,
const struct iovec *iovp,
unsigned int nsegs,
loff_t *offset,
int ioflags,
cred_t *credp)
{
struct file *file = iocb->ki_filp;
struct address_space *mapping = file->f_mapping;
struct inode *inode = mapping->host;
unsigned long segs = nsegs;
xfs_inode_t *xip;
xfs_mount_t *mp;
ssize_t ret = 0, error = 0;
xfs_fsize_t isize, new_size;
xfs_iocore_t *io;
vnode_t *vp;
unsigned long seg;
int iolock;
int eventsent = 0;
vrwlock_t locktype;
size_t ocount = 0, count;
loff_t pos;
int need_isem = 1, need_flush = 0;
XFS_STATS_INC(xs_write_calls);
vp = BHV_TO_VNODE(bdp);
xip = XFS_BHVTOI(bdp);
for (seg = 0; seg < segs; seg++) {
const struct iovec *iv = &iovp[seg];
/*
* If any segment has a negative length, or the cumulative
* length ever wraps negative then return -EINVAL.
*/
ocount += iv->iov_len;
if (unlikely((ssize_t)(ocount|iv->iov_len) < 0))
return -EINVAL;
if (access_ok(VERIFY_READ, iv->iov_base, iv->iov_len))
continue;
if (seg == 0)
return -EFAULT;
segs = seg;
ocount -= iv->iov_len; /* This segment is no good */
break;
}
count = ocount;
pos = *offset;
if (count == 0)
return 0;
io = &xip->i_iocore;
mp = io->io_mount;
if (XFS_FORCED_SHUTDOWN(mp))
return -EIO;
fs_check_frozen(vp->v_vfsp, SB_FREEZE_WRITE);
if (ioflags & IO_ISDIRECT) {
xfs_buftarg_t *target =
(xip->i_d.di_flags & XFS_DIFLAG_REALTIME) ?
mp->m_rtdev_targp : mp->m_ddev_targp;
if (ioflags & IO_ISAIO)
return XFS_ERROR(-ENOSYS);
if ((pos & target->pbr_smask) || (count & target->pbr_smask))
return XFS_ERROR(-EINVAL);
if (!VN_CACHED(vp) && pos < i_size_read(inode))
need_isem = 0;
if (VN_CACHED(vp))
need_flush = 1;
}
relock:
if (need_isem) {
iolock = XFS_IOLOCK_EXCL;
locktype = VRWLOCK_WRITE;
down(&inode->i_sem);
} else {
iolock = XFS_IOLOCK_SHARED;
locktype = VRWLOCK_WRITE_DIRECT;
}
xfs_ilock(xip, XFS_ILOCK_EXCL|iolock);
isize = i_size_read(inode);
if (file->f_flags & O_APPEND)
*offset = isize;
start:
error = -generic_write_checks(file, &pos, &count,
S_ISBLK(inode->i_mode));
if (error) {
xfs_iunlock(xip, XFS_ILOCK_EXCL|iolock);
goto out_unlock_isem;
}
new_size = pos + count;
if (new_size > isize)
io->io_new_size = new_size;
if ((DM_EVENT_ENABLED(vp->v_vfsp, xip, DM_EVENT_WRITE) &&
!(ioflags & IO_INVIS) && !eventsent)) {
loff_t savedsize = pos;
int dmflags = FILP_DELAY_FLAG(file);
if (need_isem)
dmflags |= DM_FLAGS_ISEM;
xfs_iunlock(xip, XFS_ILOCK_EXCL);
error = XFS_SEND_DATA(xip->i_mount, DM_EVENT_WRITE, vp,
pos, count,
dmflags, &locktype);
if (error) {
xfs_iunlock(xip, iolock);
goto out_unlock_isem;
}
xfs_ilock(xip, XFS_ILOCK_EXCL);
eventsent = 1;
/*
* The iolock was dropped and reaquired in XFS_SEND_DATA
* so we have to recheck the size when appending.
* We will only "goto start;" once, since having sent the
* event prevents another call to XFS_SEND_DATA, which is
* what allows the size to change in the first place.
*/
if ((file->f_flags & O_APPEND) && savedsize != isize) {
pos = isize = xip->i_d.di_size;
goto start;
}
}
/*
* On Linux, generic_file_write updates the times even if
* no data is copied in so long as the write had a size.
*
* We must update xfs' times since revalidate will overcopy xfs.
*/
if (!(ioflags & IO_INVIS)) {
xfs_ichgtime(xip, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
inode_update_time(inode, 1);
}
/*
* If the offset is beyond the size of the file, we have a couple
* of things to do. First, if there is already space allocated
* we need to either create holes or zero the disk or ...
*
* If there is a page where the previous size lands, we need
* to zero it out up to the new size.
*/
if (pos > isize) {
error = xfs_zero_eof(BHV_TO_VNODE(bdp), io, pos,
isize, pos + count);
if (error) {
xfs_iunlock(xip, XFS_ILOCK_EXCL|iolock);
goto out_unlock_isem;
}
}
xfs_iunlock(xip, XFS_ILOCK_EXCL);
/*
* If we're writing the file then make sure to clear the
* setuid and setgid bits if the process is not being run
* by root. This keeps people from modifying setuid and
* setgid binaries.
*/
if (((xip->i_d.di_mode & S_ISUID) ||
((xip->i_d.di_mode & (S_ISGID | S_IXGRP)) ==
(S_ISGID | S_IXGRP))) &&
!capable(CAP_FSETID)) {
error = xfs_write_clear_setuid(xip);
if (likely(!error))
error = -remove_suid(file->f_dentry);
if (unlikely(error)) {
xfs_iunlock(xip, iolock);
goto out_unlock_isem;
}
}
retry:
/* We can write back this queue in page reclaim */
current->backing_dev_info = mapping->backing_dev_info;
if ((ioflags & IO_ISDIRECT)) {
if (need_flush) {
xfs_inval_cached_trace(io, pos, -1,
ctooff(offtoct(pos)), -1);
VOP_FLUSHINVAL_PAGES(vp, ctooff(offtoct(pos)),
-1, FI_REMAPF_LOCKED);
}
if (need_isem) {
/* demote the lock now the cached pages are gone */
XFS_ILOCK_DEMOTE(mp, io, XFS_IOLOCK_EXCL);
up(&inode->i_sem);
iolock = XFS_IOLOCK_SHARED;
locktype = VRWLOCK_WRITE_DIRECT;
need_isem = 0;
}
xfs_rw_enter_trace(XFS_DIOWR_ENTER, io, (void *)iovp, segs,
*offset, ioflags);
ret = generic_file_direct_write(iocb, iovp,
&segs, pos, offset, count, ocount);
/*
* direct-io write to a hole: fall through to buffered I/O
* for completing the rest of the request.
*/
if (ret >= 0 && ret != count) {
XFS_STATS_ADD(xs_write_bytes, ret);
pos += ret;
count -= ret;
need_isem = 1;
ioflags &= ~IO_ISDIRECT;
xfs_iunlock(xip, iolock);
goto relock;
}
} else {
xfs_rw_enter_trace(XFS_WRITE_ENTER, io, (void *)iovp, segs,
*offset, ioflags);
ret = generic_file_buffered_write(iocb, iovp, segs,
pos, offset, count, ret);
}
current->backing_dev_info = NULL;
if (ret == -EIOCBQUEUED && !(ioflags & IO_ISAIO))
ret = wait_on_sync_kiocb(iocb);
if ((ret == -ENOSPC) &&
DM_EVENT_ENABLED(vp->v_vfsp, xip, DM_EVENT_NOSPACE) &&
!(ioflags & IO_INVIS)) {
xfs_rwunlock(bdp, locktype);
if (need_isem)
up(&inode->i_sem);
error = XFS_SEND_NAMESP(xip->i_mount, DM_EVENT_NOSPACE, vp,
DM_RIGHT_NULL, vp, DM_RIGHT_NULL, NULL, NULL,
0, 0, 0); /* Delay flag intentionally unused */
if (error)
goto out_nounlocks;
if (need_isem)
down(&inode->i_sem);
xfs_rwlock(bdp, locktype);
pos = xip->i_d.di_size;
ret = 0;
goto retry;
}
if (*offset > xip->i_d.di_size) {
xfs_ilock(xip, XFS_ILOCK_EXCL);
if (*offset > xip->i_d.di_size) {
xip->i_d.di_size = *offset;
i_size_write(inode, *offset);
xip->i_update_core = 1;
xip->i_update_size = 1;
}
xfs_iunlock(xip, XFS_ILOCK_EXCL);
}
error = -ret;
if (ret <= 0)
goto out_unlock_internal;
XFS_STATS_ADD(xs_write_bytes, ret);
/* Handle various SYNC-type writes */
if ((file->f_flags & O_SYNC) || IS_SYNC(inode)) {
/*
* If we're treating this as O_DSYNC and we have not updated the
* size, force the log.
*/
if (!(mp->m_flags & XFS_MOUNT_OSYNCISOSYNC) &&
!(xip->i_update_size)) {
xfs_inode_log_item_t *iip = xip->i_itemp;
/*
* If an allocation transaction occurred
* without extending the size, then we have to force
* the log up the proper point to ensure that the
* allocation is permanent. We can't count on
* the fact that buffered writes lock out direct I/O
* writes - the direct I/O write could have extended
* the size nontransactionally, then finished before
* we started. xfs_write_file will think that the file
* didn't grow but the update isn't safe unless the
* size change is logged.
*
* Force the log if we've committed a transaction
* against the inode or if someone else has and
* the commit record hasn't gone to disk (e.g.
* the inode is pinned). This guarantees that
* all changes affecting the inode are permanent
* when we return.
*/
if (iip && iip->ili_last_lsn) {
xfs_log_force(mp, iip->ili_last_lsn,
XFS_LOG_FORCE | XFS_LOG_SYNC);
} else if (xfs_ipincount(xip) > 0) {
xfs_log_force(mp, (xfs_lsn_t)0,
XFS_LOG_FORCE | XFS_LOG_SYNC);
}
} else {
xfs_trans_t *tp;
/*
* O_SYNC or O_DSYNC _with_ a size update are handled
* the same way.
*
* If the write was synchronous then we need to make
* sure that the inode modification time is permanent.
* We'll have updated the timestamp above, so here
* we use a synchronous transaction to log the inode.
* It's not fast, but it's necessary.
*
* If this a dsync write and the size got changed
* non-transactionally, then we need to ensure that
* the size change gets logged in a synchronous
* transaction.
*/
tp = xfs_trans_alloc(mp, XFS_TRANS_WRITE_SYNC);
if ((error = xfs_trans_reserve(tp, 0,
XFS_SWRITE_LOG_RES(mp),
0, 0, 0))) {
/* Transaction reserve failed */
xfs_trans_cancel(tp, 0);
} else {
/* Transaction reserve successful */
xfs_ilock(xip, XFS_ILOCK_EXCL);
xfs_trans_ijoin(tp, xip, XFS_ILOCK_EXCL);
xfs_trans_ihold(tp, xip);
xfs_trans_log_inode(tp, xip, XFS_ILOG_CORE);
xfs_trans_set_sync(tp);
error = xfs_trans_commit(tp, 0, NULL);
xfs_iunlock(xip, XFS_ILOCK_EXCL);
}
if (error)
goto out_unlock_internal;
}
xfs_rwunlock(bdp, locktype);
if (need_isem)
up(&inode->i_sem);
error = sync_page_range(inode, mapping, pos, ret);
if (!error)
error = ret;
return error;
}
out_unlock_internal:
xfs_rwunlock(bdp, locktype);
out_unlock_isem:
if (need_isem)
up(&inode->i_sem);
out_nounlocks:
return -error;
}
/*
* xfs_rename
*/
int
xfs_rename(
xfs_inode_t *src_dp,
bhv_vname_t *src_vname,
bhv_vnode_t *target_dir_vp,
bhv_vname_t *target_vname)
{
bhv_vnode_t *src_dir_vp = XFS_ITOV(src_dp);
xfs_trans_t *tp;
xfs_inode_t *target_dp, *src_ip, *target_ip;
xfs_mount_t *mp = src_dp->i_mount;
int new_parent; /* moving to a new dir */
int src_is_directory; /* src_name is a directory */
int error;
xfs_bmap_free_t free_list;
xfs_fsblock_t first_block;
int cancel_flags;
int committed;
xfs_inode_t *inodes[4];
int target_ip_dropped = 0; /* dropped target_ip link? */
int spaceres;
int target_link_zero = 0;
int num_inodes;
char *src_name = VNAME(src_vname);
char *target_name = VNAME(target_vname);
int src_namelen = VNAMELEN(src_vname);
int target_namelen = VNAMELEN(target_vname);
vn_trace_entry(src_dp, "xfs_rename", (inst_t *)__return_address);
vn_trace_entry(xfs_vtoi(target_dir_vp), "xfs_rename", (inst_t *)__return_address);
/*
* Find the XFS behavior descriptor for the target directory
* vnode since it was not handed to us.
*/
target_dp = xfs_vtoi(target_dir_vp);
if (target_dp == NULL) {
return XFS_ERROR(EXDEV);
}
if (DM_EVENT_ENABLED(src_dp, DM_EVENT_RENAME) ||
DM_EVENT_ENABLED(target_dp, DM_EVENT_RENAME)) {
error = XFS_SEND_NAMESP(mp, DM_EVENT_RENAME,
src_dir_vp, DM_RIGHT_NULL,
target_dir_vp, DM_RIGHT_NULL,
src_name, target_name,
0, 0, 0);
if (error) {
return error;
}
}
/* Return through std_return after this point. */
/*
* Lock all the participating inodes. Depending upon whether
* the target_name exists in the target directory, and
* whether the target directory is the same as the source
* directory, we can lock from 2 to 4 inodes.
* xfs_lock_for_rename() will return ENOENT if src_name
* does not exist in the source directory.
*/
tp = NULL;
error = xfs_lock_for_rename(src_dp, target_dp, src_vname,
target_vname, &src_ip, &target_ip, inodes,
&num_inodes);
if (error) {
/*
* We have nothing locked, no inode references, and
* no transaction, so just get out.
*/
goto std_return;
}
ASSERT(src_ip != NULL);
if ((src_ip->i_d.di_mode & S_IFMT) == S_IFDIR) {
/*
* Check for link count overflow on target_dp
*/
if (target_ip == NULL && (src_dp != target_dp) &&
target_dp->i_d.di_nlink >= XFS_MAXLINK) {
error = XFS_ERROR(EMLINK);
xfs_rename_unlock4(inodes, XFS_ILOCK_SHARED);
goto rele_return;
}
}
/*
* If we are using project inheritance, we only allow renames
* into our tree when the project IDs are the same; else the
* tree quota mechanism would be circumvented.
*/
if (unlikely((target_dp->i_d.di_flags & XFS_DIFLAG_PROJINHERIT) &&
(target_dp->i_d.di_projid != src_ip->i_d.di_projid))) {
error = XFS_ERROR(EXDEV);
xfs_rename_unlock4(inodes, XFS_ILOCK_SHARED);
goto rele_return;
}
new_parent = (src_dp != target_dp);
src_is_directory = ((src_ip->i_d.di_mode & S_IFMT) == S_IFDIR);
/*
* Drop the locks on our inodes so that we can start the transaction.
*/
xfs_rename_unlock4(inodes, XFS_ILOCK_SHARED);
XFS_BMAP_INIT(&free_list, &first_block);
tp = xfs_trans_alloc(mp, XFS_TRANS_RENAME);
cancel_flags = XFS_TRANS_RELEASE_LOG_RES;
spaceres = XFS_RENAME_SPACE_RES(mp, target_namelen);
error = xfs_trans_reserve(tp, spaceres, XFS_RENAME_LOG_RES(mp), 0,
XFS_TRANS_PERM_LOG_RES, XFS_RENAME_LOG_COUNT);
if (error == ENOSPC) {
spaceres = 0;
error = xfs_trans_reserve(tp, 0, XFS_RENAME_LOG_RES(mp), 0,
XFS_TRANS_PERM_LOG_RES, XFS_RENAME_LOG_COUNT);
}
if (error) {
xfs_trans_cancel(tp, 0);
goto rele_return;
}
/*
* Attach the dquots to the inodes
*/
if ((error = XFS_QM_DQVOPRENAME(mp, inodes))) {
xfs_trans_cancel(tp, cancel_flags);
goto rele_return;
}
/*
* Reacquire the inode locks we dropped above.
*/
xfs_lock_inodes(inodes, num_inodes, 0, XFS_ILOCK_EXCL);
/*
* Join all the inodes to the transaction. From this point on,
* we can rely on either trans_commit or trans_cancel to unlock
* them. Note that we need to add a vnode reference to the
* directories since trans_commit & trans_cancel will decrement
* them when they unlock the inodes. Also, we need to be careful
* not to add an inode to the transaction more than once.
*/
VN_HOLD(src_dir_vp);
xfs_trans_ijoin(tp, src_dp, XFS_ILOCK_EXCL);
if (new_parent) {
VN_HOLD(target_dir_vp);
xfs_trans_ijoin(tp, target_dp, XFS_ILOCK_EXCL);
}
if ((src_ip != src_dp) && (src_ip != target_dp)) {
xfs_trans_ijoin(tp, src_ip, XFS_ILOCK_EXCL);
}
if ((target_ip != NULL) &&
(target_ip != src_ip) &&
(target_ip != src_dp) &&
(target_ip != target_dp)) {
xfs_trans_ijoin(tp, target_ip, XFS_ILOCK_EXCL);
}
/*
* Set up the target.
*/
if (target_ip == NULL) {
/*
* If there's no space reservation, check the entry will
* fit before actually inserting it.
*/
if (spaceres == 0 &&
(error = xfs_dir_canenter(tp, target_dp, target_name,
target_namelen)))
goto error_return;
/*
* If target does not exist and the rename crosses
* directories, adjust the target directory link count
* to account for the ".." reference from the new entry.
*/
error = xfs_dir_createname(tp, target_dp, target_name,
target_namelen, src_ip->i_ino,
&first_block, &free_list, spaceres);
if (error == ENOSPC)
goto error_return;
if (error)
goto abort_return;
xfs_ichgtime(target_dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
if (new_parent && src_is_directory) {
error = xfs_bumplink(tp, target_dp);
if (error)
goto abort_return;
}
} else { /* target_ip != NULL */
/*
* If target exists and it's a directory, check that both
* target and source are directories and that target can be
* destroyed, or that neither is a directory.
*/
if ((target_ip->i_d.di_mode & S_IFMT) == S_IFDIR) {
/*
* Make sure target dir is empty.
*/
if (!(xfs_dir_isempty(target_ip)) ||
(target_ip->i_d.di_nlink > 2)) {
error = XFS_ERROR(EEXIST);
goto error_return;
}
}
/*
* Link the source inode under the target name.
* If the source inode is a directory and we are moving
* it across directories, its ".." entry will be
* inconsistent until we replace that down below.
*
* In case there is already an entry with the same
* name at the destination directory, remove it first.
*/
error = xfs_dir_replace(tp, target_dp, target_name,
target_namelen, src_ip->i_ino,
&first_block, &free_list, spaceres);
if (error)
goto abort_return;
xfs_ichgtime(target_dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
/*
* Decrement the link count on the target since the target
* dir no longer points to it.
*/
error = xfs_droplink(tp, target_ip);
if (error)
goto abort_return;
target_ip_dropped = 1;
if (src_is_directory) {
/*
* Drop the link from the old "." entry.
*/
error = xfs_droplink(tp, target_ip);
if (error)
goto abort_return;
}
/* Do this test while we still hold the locks */
target_link_zero = (target_ip)->i_d.di_nlink==0;
} /* target_ip != NULL */
/*
* Remove the source.
*/
if (new_parent && src_is_directory) {
/*
* Rewrite the ".." entry to point to the new
* directory.
*/
error = xfs_dir_replace(tp, src_ip, "..", 2, target_dp->i_ino,
&first_block, &free_list, spaceres);
ASSERT(error != EEXIST);
if (error)
goto abort_return;
xfs_ichgtime(src_ip, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
} else {
/*
* We always want to hit the ctime on the source inode.
* We do it in the if clause above for the 'new_parent &&
* src_is_directory' case, and here we get all the other
* cases. This isn't strictly required by the standards
* since the source inode isn't really being changed,
* but old unix file systems did it and some incremental
* backup programs won't work without it.
*/
xfs_ichgtime(src_ip, XFS_ICHGTIME_CHG);
}
/*
* Adjust the link count on src_dp. This is necessary when
* renaming a directory, either within one parent when
* the target existed, or across two parent directories.
*/
if (src_is_directory && (new_parent || target_ip != NULL)) {
/*
* Decrement link count on src_directory since the
* entry that's moved no longer points to it.
*/
error = xfs_droplink(tp, src_dp);
if (error)
goto abort_return;
}
error = xfs_dir_removename(tp, src_dp, src_name, src_namelen,
src_ip->i_ino, &first_block, &free_list, spaceres);
if (error)
goto abort_return;
xfs_ichgtime(src_dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
/*
* Update the generation counts on all the directory inodes
* that we're modifying.
*/
src_dp->i_gen++;
xfs_trans_log_inode(tp, src_dp, XFS_ILOG_CORE);
if (new_parent) {
target_dp->i_gen++;
xfs_trans_log_inode(tp, target_dp, XFS_ILOG_CORE);
}
/*
* If there was a target inode, take an extra reference on
* it here so that it doesn't go to xfs_inactive() from
* within the commit.
*/
if (target_ip != NULL) {
IHOLD(target_ip);
}
/*
* If this is a synchronous mount, make sure that the
* rename transaction goes to disk before returning to
* the user.
*/
if (mp->m_flags & (XFS_MOUNT_WSYNC|XFS_MOUNT_DIRSYNC)) {
xfs_trans_set_sync(tp);
}
/*
* Take refs. for vop_link_removed calls below. No need to worry
* about directory refs. because the caller holds them.
*
* Do holds before the xfs_bmap_finish since it might rele them down
* to zero.
*/
if (target_ip_dropped)
IHOLD(target_ip);
IHOLD(src_ip);
error = xfs_bmap_finish(&tp, &free_list, &committed);
if (error) {
xfs_bmap_cancel(&free_list);
xfs_trans_cancel(tp, (XFS_TRANS_RELEASE_LOG_RES |
XFS_TRANS_ABORT));
if (target_ip != NULL) {
IRELE(target_ip);
}
if (target_ip_dropped) {
IRELE(target_ip);
}
IRELE(src_ip);
goto std_return;
}
/*
* trans_commit will unlock src_ip, target_ip & decrement
* the vnode references.
*/
error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
if (target_ip != NULL) {
xfs_refcache_purge_ip(target_ip);
IRELE(target_ip);
}
/*
* Let interposed file systems know about removed links.
*/
if (target_ip_dropped)
IRELE(target_ip);
IRELE(src_ip);
/* Fall through to std_return with error = 0 or errno from
* xfs_trans_commit */
std_return:
if (DM_EVENT_ENABLED(src_dp, DM_EVENT_POSTRENAME) ||
DM_EVENT_ENABLED(target_dp, DM_EVENT_POSTRENAME)) {
(void) XFS_SEND_NAMESP (mp, DM_EVENT_POSTRENAME,
src_dir_vp, DM_RIGHT_NULL,
target_dir_vp, DM_RIGHT_NULL,
src_name, target_name,
0, error, 0);
}
return error;
abort_return:
cancel_flags |= XFS_TRANS_ABORT;
/* FALLTHROUGH */
error_return:
xfs_bmap_cancel(&free_list);
xfs_trans_cancel(tp, cancel_flags);
goto std_return;
rele_return:
IRELE(src_ip);
if (target_ip != NULL) {
IRELE(target_ip);
}
goto std_return;
}
STATIC int
xfs_file_fsync(
struct file *file,
loff_t start,
loff_t end,
int datasync)
{
struct inode *inode = file->f_mapping->host;
struct xfs_inode *ip = XFS_I(inode);
struct xfs_mount *mp = ip->i_mount;
struct xfs_trans *tp;
int error = 0;
int log_flushed = 0;
xfs_lsn_t lsn = 0;
trace_xfs_file_fsync(ip);
error = filemap_write_and_wait_range(inode->i_mapping, start, end);
if (error)
return error;
if (XFS_FORCED_SHUTDOWN(mp))
return -XFS_ERROR(EIO);
xfs_iflags_clear(ip, XFS_ITRUNCATED);
if (mp->m_flags & XFS_MOUNT_BARRIER) {
/*
* If we have an RT and/or log subvolume we need to make sure
* to flush the write cache the device used for file data
* first. This is to ensure newly written file data make
* it to disk before logging the new inode size in case of
* an extending write.
*/
if (XFS_IS_REALTIME_INODE(ip))
xfs_blkdev_issue_flush(mp->m_rtdev_targp);
else if (mp->m_logdev_targp != mp->m_ddev_targp)
xfs_blkdev_issue_flush(mp->m_ddev_targp);
}
/*
* We always need to make sure that the required inode state is safe on
* disk. The inode might be clean but we still might need to force the
* log because of committed transactions that haven't hit the disk yet.
* Likewise, there could be unflushed non-transactional changes to the
* inode core that have to go to disk and this requires us to issue
* a synchronous transaction to capture these changes correctly.
*
* This code relies on the assumption that if the i_update_core field
* of the inode is clear and the inode is unpinned then it is clean
* and no action is required.
*/
xfs_ilock(ip, XFS_ILOCK_SHARED);
/*
* First check if the VFS inode is marked dirty. All the dirtying
* of non-transactional updates no goes through mark_inode_dirty*,
* which allows us to distinguish beteeen pure timestamp updates
* and i_size updates which need to be caught for fdatasync.
* After that also theck for the dirty state in the XFS inode, which
* might gets cleared when the inode gets written out via the AIL
* or xfs_iflush_cluster.
*/
if (((inode->i_state & I_DIRTY_DATASYNC) ||
((inode->i_state & I_DIRTY_SYNC) && !datasync)) &&
ip->i_update_core) {
/*
* Kick off a transaction to log the inode core to get the
* updates. The sync transaction will also force the log.
*/
xfs_iunlock(ip, XFS_ILOCK_SHARED);
tp = xfs_trans_alloc(mp, XFS_TRANS_FSYNC_TS);
error = xfs_trans_reserve(tp, 0,
XFS_FSYNC_TS_LOG_RES(mp), 0, 0, 0);
if (error) {
xfs_trans_cancel(tp, 0);
return -error;
}
xfs_ilock(ip, XFS_ILOCK_EXCL);
/*
* Note - it's possible that we might have pushed ourselves out
* of the way during trans_reserve which would flush the inode.
* But there's no guarantee that the inode buffer has actually
* gone out yet (it's delwri). Plus the buffer could be pinned
* anyway if it's part of an inode in another recent
* transaction. So we play it safe and fire off the
* transaction anyway.
*/
xfs_trans_ijoin(tp, ip, 0);
xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
error = xfs_trans_commit(tp, 0);
lsn = ip->i_itemp->ili_last_lsn;
xfs_iunlock(ip, XFS_ILOCK_EXCL);
} else {
/*
* Timestamps/size haven't changed since last inode flush or
* inode transaction commit. That means either nothing got
* written or a transaction committed which caught the updates.
* If the latter happened and the transaction hasn't hit the
* disk yet, the inode will be still be pinned. If it is,
* force the log.
*/
if (xfs_ipincount(ip))
lsn = ip->i_itemp->ili_last_lsn;
xfs_iunlock(ip, XFS_ILOCK_SHARED);
}
if (!error && lsn)
error = _xfs_log_force_lsn(mp, lsn, XFS_LOG_SYNC, &log_flushed);
/*
* If we only have a single device, and the log force about was
* a no-op we might have to flush the data device cache here.
* This can only happen for fdatasync/O_DSYNC if we were overwriting
* an already allocated file and thus do not have any metadata to
* commit.
*/
if ((mp->m_flags & XFS_MOUNT_BARRIER) &&
mp->m_logdev_targp == mp->m_ddev_targp &&
!XFS_IS_REALTIME_INODE(ip) &&
!log_flushed)
xfs_blkdev_issue_flush(mp->m_ddev_targp);
return -error;
}
/*
* Make sure the blocks described by maps are stable on disk. This includes
* converting any unwritten extents, flushing the disk cache and updating the
* time stamps.
*
* Note that we rely on the caller to always send us a timestamp update so that
* we always commit a transaction here. If that stops being true we will have
* to manually flush the cache here similar to what the fsync code path does
* for datasyncs on files that have no dirty metadata.
*/
int
xfs_fs_commit_blocks(
struct inode *inode,
struct iomap *maps,
int nr_maps,
struct iattr *iattr)
{
struct xfs_inode *ip = XFS_I(inode);
struct xfs_mount *mp = ip->i_mount;
struct xfs_trans *tp;
bool update_isize = false;
int error, i;
loff_t size;
ASSERT(iattr->ia_valid & (ATTR_ATIME|ATTR_CTIME|ATTR_MTIME));
xfs_ilock(ip, XFS_IOLOCK_EXCL);
size = i_size_read(inode);
if ((iattr->ia_valid & ATTR_SIZE) && iattr->ia_size > size) {
update_isize = true;
size = iattr->ia_size;
}
for (i = 0; i < nr_maps; i++) {
u64 start, length, end;
start = maps[i].offset;
if (start > size)
continue;
end = start + maps[i].length;
if (end > size)
end = size;
length = end - start;
if (!length)
continue;
/*
* Make sure reads through the pagecache see the new data.
*/
error = invalidate_inode_pages2_range(inode->i_mapping,
start >> PAGE_SHIFT,
(end - 1) >> PAGE_SHIFT);
WARN_ON_ONCE(error);
error = xfs_iomap_write_unwritten(ip, start, length);
if (error)
goto out_drop_iolock;
}
if (update_isize) {
error = xfs_pnfs_validate_isize(ip, size);
if (error)
goto out_drop_iolock;
}
error = xfs_trans_alloc(mp, &M_RES(mp)->tr_ichange, 0, 0, 0, &tp);
if (error)
goto out_drop_iolock;
xfs_ilock(ip, XFS_ILOCK_EXCL);
xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
xfs_setattr_time(ip, iattr);
if (update_isize) {
i_size_write(inode, iattr->ia_size);
ip->i_d.di_size = iattr->ia_size;
}
xfs_trans_set_sync(tp);
error = xfs_trans_commit(tp);
out_drop_iolock:
xfs_iunlock(ip, XFS_IOLOCK_EXCL);
return error;
}
ssize_t /* bytes written, or (-) error */
xfs_write(
bhv_desc_t *bdp,
struct kiocb *iocb,
const struct iovec *iovp,
unsigned int segs,
loff_t *offset,
int ioflags,
cred_t *credp)
{
struct file *file = iocb->ki_filp;
size_t size = 0;
xfs_inode_t *xip;
xfs_mount_t *mp;
ssize_t ret;
int error = 0;
xfs_fsize_t isize, new_size;
xfs_fsize_t n, limit;
xfs_iocore_t *io;
vnode_t *vp;
unsigned long seg;
int iolock;
int eventsent = 0;
vrwlock_t locktype;
XFS_STATS_INC(xs_write_calls);
vp = BHV_TO_VNODE(bdp);
vn_trace_entry(vp, "xfs_write", (inst_t *)__return_address);
xip = XFS_BHVTOI(bdp);
/* START copy & waste from filemap.c */
for (seg = 0; seg < segs; seg++) {
const struct iovec *iv = &iovp[seg];
/*
* If any segment has a negative length, or the cumulative
* length ever wraps negative then return -EINVAL.
*/
size += iv->iov_len;
if (unlikely((ssize_t)(size|iv->iov_len) < 0))
return XFS_ERROR(-EINVAL);
}
/* END copy & waste from filemap.c */
if (size == 0)
return 0;
io = &(xip->i_iocore);
mp = io->io_mount;
xfs_check_frozen(mp, bdp, XFS_FREEZE_WRITE);
if (XFS_FORCED_SHUTDOWN(mp)) {
return -EIO;
}
if (ioflags & IO_ISDIRECT) {
pb_target_t *target =
(xip->i_d.di_flags & XFS_DIFLAG_REALTIME) ?
mp->m_rtdev_targp : mp->m_ddev_targp;
if ((*offset & target->pbr_smask) ||
(size & target->pbr_smask)) {
return XFS_ERROR(-EINVAL);
}
iolock = XFS_IOLOCK_SHARED;
locktype = VRWLOCK_WRITE_DIRECT;
} else {
iolock = XFS_IOLOCK_EXCL;
locktype = VRWLOCK_WRITE;
}
xfs_ilock(xip, XFS_ILOCK_EXCL|iolock);
isize = xip->i_d.di_size;
limit = XFS_MAXIOFFSET(mp);
if (file->f_flags & O_APPEND)
*offset = isize;
start:
n = limit - *offset;
if (n <= 0) {
xfs_iunlock(xip, XFS_ILOCK_EXCL|iolock);
return -EFBIG;
}
if (n < size)
size = n;
new_size = *offset + size;
if (new_size > isize) {
io->io_new_size = new_size;
}
if ((DM_EVENT_ENABLED(vp->v_vfsp, xip, DM_EVENT_WRITE) &&
!(ioflags & IO_INVIS) && !eventsent)) {
loff_t savedsize = *offset;
xfs_iunlock(xip, XFS_ILOCK_EXCL);
error = XFS_SEND_DATA(xip->i_mount, DM_EVENT_WRITE, vp,
*offset, size,
FILP_DELAY_FLAG(file), &locktype);
if (error) {
xfs_iunlock(xip, iolock);
return -error;
}
xfs_ilock(xip, XFS_ILOCK_EXCL);
eventsent = 1;
/*
* The iolock was dropped and reaquired in XFS_SEND_DATA
* so we have to recheck the size when appending.
* We will only "goto start;" once, since having sent the
* event prevents another call to XFS_SEND_DATA, which is
* what allows the size to change in the first place.
*/
if ((file->f_flags & O_APPEND) &&
savedsize != xip->i_d.di_size) {
*offset = isize = xip->i_d.di_size;
goto start;
}
}
/*
* On Linux, generic_file_write updates the times even if
* no data is copied in so long as the write had a size.
*
* We must update xfs' times since revalidate will overcopy xfs.
*/
if (size && !(ioflags & IO_INVIS))
xfs_ichgtime(xip, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
/*
* If the offset is beyond the size of the file, we have a couple
* of things to do. First, if there is already space allocated
* we need to either create holes or zero the disk or ...
*
* If there is a page where the previous size lands, we need
* to zero it out up to the new size.
*/
if (!(ioflags & IO_ISDIRECT) && (*offset > isize && isize)) {
error = xfs_zero_eof(BHV_TO_VNODE(bdp), io, *offset,
isize, *offset + size);
if (error) {
xfs_iunlock(xip, XFS_ILOCK_EXCL|iolock);
return(-error);
}
}
xfs_iunlock(xip, XFS_ILOCK_EXCL);
/*
* If we're writing the file then make sure to clear the
* setuid and setgid bits if the process is not being run
* by root. This keeps people from modifying setuid and
* setgid binaries.
*/
if (((xip->i_d.di_mode & S_ISUID) ||
((xip->i_d.di_mode & (S_ISGID | S_IXGRP)) ==
(S_ISGID | S_IXGRP))) &&
!capable(CAP_FSETID)) {
error = xfs_write_clear_setuid(xip);
if (error) {
xfs_iunlock(xip, iolock);
return -error;
}
}
retry:
if (ioflags & IO_ISDIRECT) {
xfs_inval_cached_pages(vp, &xip->i_iocore, *offset, 1, 1);
}
ret = generic_file_aio_write_nolock(iocb, iovp, segs, offset);
if ((ret == -ENOSPC) &&
DM_EVENT_ENABLED(vp->v_vfsp, xip, DM_EVENT_NOSPACE) &&
!(ioflags & IO_INVIS)) {
xfs_rwunlock(bdp, locktype);
error = XFS_SEND_NAMESP(xip->i_mount, DM_EVENT_NOSPACE, vp,
DM_RIGHT_NULL, vp, DM_RIGHT_NULL, NULL, NULL,
0, 0, 0); /* Delay flag intentionally unused */
if (error)
return -error;
xfs_rwlock(bdp, locktype);
*offset = xip->i_d.di_size;
goto retry;
}
if (*offset > xip->i_d.di_size) {
xfs_ilock(xip, XFS_ILOCK_EXCL);
if (*offset > xip->i_d.di_size) {
struct inode *inode = LINVFS_GET_IP(vp);
xip->i_d.di_size = *offset;
i_size_write(inode, *offset);
xip->i_update_core = 1;
xip->i_update_size = 1;
}
xfs_iunlock(xip, XFS_ILOCK_EXCL);
}
if (ret <= 0) {
xfs_rwunlock(bdp, locktype);
return ret;
}
XFS_STATS_ADD(xs_write_bytes, ret);
/* Handle various SYNC-type writes */
if ((file->f_flags & O_SYNC) || IS_SYNC(file->f_dentry->d_inode)) {
/*
* If we're treating this as O_DSYNC and we have not updated the
* size, force the log.
*/
if (!(mp->m_flags & XFS_MOUNT_OSYNCISOSYNC)
&& !(xip->i_update_size)) {
/*
* If an allocation transaction occurred
* without extending the size, then we have to force
* the log up the proper point to ensure that the
* allocation is permanent. We can't count on
* the fact that buffered writes lock out direct I/O
* writes - the direct I/O write could have extended
* the size nontransactionally, then finished before
* we started. xfs_write_file will think that the file
* didn't grow but the update isn't safe unless the
* size change is logged.
*
* Force the log if we've committed a transaction
* against the inode or if someone else has and
* the commit record hasn't gone to disk (e.g.
* the inode is pinned). This guarantees that
* all changes affecting the inode are permanent
* when we return.
*/
xfs_inode_log_item_t *iip;
xfs_lsn_t lsn;
iip = xip->i_itemp;
if (iip && iip->ili_last_lsn) {
lsn = iip->ili_last_lsn;
xfs_log_force(mp, lsn,
XFS_LOG_FORCE | XFS_LOG_SYNC);
} else if (xfs_ipincount(xip) > 0) {
xfs_log_force(mp, (xfs_lsn_t)0,
XFS_LOG_FORCE | XFS_LOG_SYNC);
}
} else {
xfs_trans_t *tp;
/*
* O_SYNC or O_DSYNC _with_ a size update are handled
* the same way.
*
* If the write was synchronous then we need to make
* sure that the inode modification time is permanent.
* We'll have updated the timestamp above, so here
* we use a synchronous transaction to log the inode.
* It's not fast, but it's necessary.
*
* If this a dsync write and the size got changed
* non-transactionally, then we need to ensure that
* the size change gets logged in a synchronous
* transaction.
*/
tp = xfs_trans_alloc(mp, XFS_TRANS_WRITE_SYNC);
if ((error = xfs_trans_reserve(tp, 0,
XFS_SWRITE_LOG_RES(mp),
0, 0, 0))) {
/* Transaction reserve failed */
xfs_trans_cancel(tp, 0);
} else {
/* Transaction reserve successful */
xfs_ilock(xip, XFS_ILOCK_EXCL);
xfs_trans_ijoin(tp, xip, XFS_ILOCK_EXCL);
xfs_trans_ihold(tp, xip);
xfs_trans_log_inode(tp, xip, XFS_ILOG_CORE);
xfs_trans_set_sync(tp);
error = xfs_trans_commit(tp, 0, (xfs_lsn_t)0);
xfs_iunlock(xip, XFS_ILOCK_EXCL);
}
}
} /* (ioflags & O_SYNC) */
xfs_rwunlock(bdp, locktype);
return(ret);
}
int
xfs_qm_scall_setqlim(
xfs_mount_t *mp,
xfs_dqid_t id,
uint type,
fs_disk_quota_t *newlim)
{
struct xfs_quotainfo *q = mp->m_quotainfo;
xfs_disk_dquot_t *ddq;
xfs_dquot_t *dqp;
xfs_trans_t *tp;
int error;
xfs_qcnt_t hard, soft;
if (newlim->d_fieldmask & ~XFS_DQ_MASK)
return EINVAL;
if ((newlim->d_fieldmask & XFS_DQ_MASK) == 0)
return 0;
tp = xfs_trans_alloc(mp, XFS_TRANS_QM_SETQLIM);
if ((error = xfs_trans_reserve(tp, 0, sizeof(xfs_disk_dquot_t) + 128,
0, 0, XFS_DEFAULT_LOG_COUNT))) {
xfs_trans_cancel(tp, 0);
return (error);
}
mutex_lock(&q->qi_quotaofflock);
if ((error = xfs_qm_dqget(mp, NULL, id, type, XFS_QMOPT_DQALLOC, &dqp))) {
xfs_trans_cancel(tp, XFS_TRANS_ABORT);
ASSERT(error != ENOENT);
goto out_unlock;
}
xfs_trans_dqjoin(tp, dqp);
ddq = &dqp->q_core;
hard = (newlim->d_fieldmask & FS_DQ_BHARD) ?
(xfs_qcnt_t) XFS_BB_TO_FSB(mp, newlim->d_blk_hardlimit) :
be64_to_cpu(ddq->d_blk_hardlimit);
soft = (newlim->d_fieldmask & FS_DQ_BSOFT) ?
(xfs_qcnt_t) XFS_BB_TO_FSB(mp, newlim->d_blk_softlimit) :
be64_to_cpu(ddq->d_blk_softlimit);
if (hard == 0 || hard >= soft) {
ddq->d_blk_hardlimit = cpu_to_be64(hard);
ddq->d_blk_softlimit = cpu_to_be64(soft);
if (id == 0) {
q->qi_bhardlimit = hard;
q->qi_bsoftlimit = soft;
}
} else {
xfs_debug(mp, "blkhard %Ld < blksoft %Ld\n", hard, soft);
}
hard = (newlim->d_fieldmask & FS_DQ_RTBHARD) ?
(xfs_qcnt_t) XFS_BB_TO_FSB(mp, newlim->d_rtb_hardlimit) :
be64_to_cpu(ddq->d_rtb_hardlimit);
soft = (newlim->d_fieldmask & FS_DQ_RTBSOFT) ?
(xfs_qcnt_t) XFS_BB_TO_FSB(mp, newlim->d_rtb_softlimit) :
be64_to_cpu(ddq->d_rtb_softlimit);
if (hard == 0 || hard >= soft) {
ddq->d_rtb_hardlimit = cpu_to_be64(hard);
ddq->d_rtb_softlimit = cpu_to_be64(soft);
if (id == 0) {
q->qi_rtbhardlimit = hard;
q->qi_rtbsoftlimit = soft;
}
} else {
xfs_debug(mp, "rtbhard %Ld < rtbsoft %Ld\n", hard, soft);
}
hard = (newlim->d_fieldmask & FS_DQ_IHARD) ?
(xfs_qcnt_t) newlim->d_ino_hardlimit :
be64_to_cpu(ddq->d_ino_hardlimit);
soft = (newlim->d_fieldmask & FS_DQ_ISOFT) ?
(xfs_qcnt_t) newlim->d_ino_softlimit :
be64_to_cpu(ddq->d_ino_softlimit);
if (hard == 0 || hard >= soft) {
ddq->d_ino_hardlimit = cpu_to_be64(hard);
ddq->d_ino_softlimit = cpu_to_be64(soft);
if (id == 0) {
q->qi_ihardlimit = hard;
q->qi_isoftlimit = soft;
}
} else {
xfs_debug(mp, "ihard %Ld < isoft %Ld\n", hard, soft);
}
if (newlim->d_fieldmask & FS_DQ_BWARNS)
ddq->d_bwarns = cpu_to_be16(newlim->d_bwarns);
if (newlim->d_fieldmask & FS_DQ_IWARNS)
ddq->d_iwarns = cpu_to_be16(newlim->d_iwarns);
if (newlim->d_fieldmask & FS_DQ_RTBWARNS)
ddq->d_rtbwarns = cpu_to_be16(newlim->d_rtbwarns);
if (id == 0) {
if (newlim->d_fieldmask & FS_DQ_BTIMER) {
q->qi_btimelimit = newlim->d_btimer;
ddq->d_btimer = cpu_to_be32(newlim->d_btimer);
}
if (newlim->d_fieldmask & FS_DQ_ITIMER) {
q->qi_itimelimit = newlim->d_itimer;
ddq->d_itimer = cpu_to_be32(newlim->d_itimer);
}
if (newlim->d_fieldmask & FS_DQ_RTBTIMER) {
q->qi_rtbtimelimit = newlim->d_rtbtimer;
ddq->d_rtbtimer = cpu_to_be32(newlim->d_rtbtimer);
}
if (newlim->d_fieldmask & FS_DQ_BWARNS)
q->qi_bwarnlimit = newlim->d_bwarns;
if (newlim->d_fieldmask & FS_DQ_IWARNS)
q->qi_iwarnlimit = newlim->d_iwarns;
if (newlim->d_fieldmask & FS_DQ_RTBWARNS)
q->qi_rtbwarnlimit = newlim->d_rtbwarns;
} else {
xfs_qm_adjust_dqtimers(mp, ddq);
}
dqp->dq_flags |= XFS_DQ_DIRTY;
xfs_trans_log_dquot(tp, dqp);
error = xfs_trans_commit(tp, 0);
xfs_qm_dqrele(dqp);
out_unlock:
mutex_unlock(&q->qi_quotaofflock);
return error;
}
/*
* Pass in a delayed allocate extent, convert it to real extents;
* return to the caller the extent we create which maps on top of
* the originating callers request.
*
* Called without a lock on the inode.
*
* We no longer bother to look at the incoming map - all we have to
* guarantee is that whatever we allocate fills the required range.
*/
int
xfs_iomap_write_allocate(
xfs_inode_t *ip,
xfs_off_t offset,
size_t count,
xfs_bmbt_irec_t *map,
int *retmap)
{
xfs_mount_t *mp = ip->i_mount;
xfs_fileoff_t offset_fsb, last_block;
xfs_fileoff_t end_fsb, map_start_fsb;
xfs_fsblock_t first_block;
xfs_bmap_free_t free_list;
xfs_filblks_t count_fsb;
xfs_bmbt_irec_t imap;
xfs_trans_t *tp;
int nimaps, committed;
int error = 0;
int nres;
*retmap = 0;
/*
* Make sure that the dquots are there.
*/
error = xfs_qm_dqattach(ip, 0);
if (error)
return XFS_ERROR(error);
offset_fsb = XFS_B_TO_FSBT(mp, offset);
count_fsb = map->br_blockcount;
map_start_fsb = map->br_startoff;
XFS_STATS_ADD(xs_xstrat_bytes, XFS_FSB_TO_B(mp, count_fsb));
while (count_fsb != 0) {
/*
* Set up a transaction with which to allocate the
* backing store for the file. Do allocations in a
* loop until we get some space in the range we are
* interested in. The other space that might be allocated
* is in the delayed allocation extent on which we sit
* but before our buffer starts.
*/
nimaps = 0;
while (nimaps == 0) {
tp = xfs_trans_alloc(mp, XFS_TRANS_STRAT_WRITE);
tp->t_flags |= XFS_TRANS_RESERVE;
nres = XFS_EXTENTADD_SPACE_RES(mp, XFS_DATA_FORK);
error = xfs_trans_reserve(tp, nres,
XFS_WRITE_LOG_RES(mp),
0, XFS_TRANS_PERM_LOG_RES,
XFS_WRITE_LOG_COUNT);
if (error) {
xfs_trans_cancel(tp, 0);
return XFS_ERROR(error);
}
xfs_ilock(ip, XFS_ILOCK_EXCL);
xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
xfs_trans_ihold(tp, ip);
xfs_bmap_init(&free_list, &first_block);
/*
* it is possible that the extents have changed since
* we did the read call as we dropped the ilock for a
* while. We have to be careful about truncates or hole
* punchs here - we are not allowed to allocate
* non-delalloc blocks here.
*
* The only protection against truncation is the pages
* for the range we are being asked to convert are
* locked and hence a truncate will block on them
* first.
*
* As a result, if we go beyond the range we really
* need and hit an delalloc extent boundary followed by
* a hole while we have excess blocks in the map, we
* will fill the hole incorrectly and overrun the
* transaction reservation.
*
* Using a single map prevents this as we are forced to
* check each map we look for overlap with the desired
* range and abort as soon as we find it. Also, given
* that we only return a single map, having one beyond
* what we can return is probably a bit silly.
*
* We also need to check that we don't go beyond EOF;
* this is a truncate optimisation as a truncate sets
* the new file size before block on the pages we
* currently have locked under writeback. Because they
* are about to be tossed, we don't need to write them
* back....
*/
nimaps = 1;
end_fsb = XFS_B_TO_FSB(mp, ip->i_size);
error = xfs_bmap_last_offset(NULL, ip, &last_block,
XFS_DATA_FORK);
if (error)
goto trans_cancel;
last_block = XFS_FILEOFF_MAX(last_block, end_fsb);
if ((map_start_fsb + count_fsb) > last_block) {
count_fsb = last_block - map_start_fsb;
if (count_fsb == 0) {
error = EAGAIN;
goto trans_cancel;
}
}
/* Go get the actual blocks */
error = xfs_bmapi(tp, ip, map_start_fsb, count_fsb,
XFS_BMAPI_WRITE, &first_block, 1,
&imap, &nimaps, &free_list, NULL);
if (error)
goto trans_cancel;
error = xfs_bmap_finish(&tp, &free_list, &committed);
if (error)
goto trans_cancel;
error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
if (error)
goto error0;
xfs_iunlock(ip, XFS_ILOCK_EXCL);
}
/*
* See if we were able to allocate an extent that
* covers at least part of the callers request
*/
if (!(imap.br_startblock || XFS_IS_REALTIME_INODE(ip)))
return xfs_cmn_err_fsblock_zero(ip, &imap);
if ((offset_fsb >= imap.br_startoff) &&
(offset_fsb < (imap.br_startoff +
imap.br_blockcount))) {
*map = imap;
*retmap = 1;
XFS_STATS_INC(xs_xstrat_quick);
return 0;
}
/*
* So far we have not mapped the requested part of the
* file, just surrounding data, try again.
*/
count_fsb -= imap.br_blockcount;
map_start_fsb = imap.br_startoff + imap.br_blockcount;
}
trans_cancel:
xfs_bmap_cancel(&free_list);
xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES | XFS_TRANS_ABORT);
error0:
xfs_iunlock(ip, XFS_ILOCK_EXCL);
return XFS_ERROR(error);
}
int
xfs_iomap_write_direct(
xfs_inode_t *ip,
xfs_off_t offset,
size_t count,
int flags,
xfs_bmbt_irec_t *ret_imap,
int *nmaps,
int found)
{
xfs_mount_t *mp = ip->i_mount;
xfs_fileoff_t offset_fsb;
xfs_fileoff_t last_fsb;
xfs_filblks_t count_fsb, resaligned;
xfs_fsblock_t firstfsb;
xfs_extlen_t extsz, temp;
int nimaps;
int bmapi_flag;
int quota_flag;
int rt;
xfs_trans_t *tp;
xfs_bmbt_irec_t imap;
xfs_bmap_free_t free_list;
uint qblocks, resblks, resrtextents;
int committed;
int error;
/*
* Make sure that the dquots are there. This doesn't hold
* the ilock across a disk read.
*/
error = xfs_qm_dqattach_locked(ip, 0);
if (error)
return XFS_ERROR(error);
rt = XFS_IS_REALTIME_INODE(ip);
extsz = xfs_get_extsz_hint(ip);
offset_fsb = XFS_B_TO_FSBT(mp, offset);
last_fsb = XFS_B_TO_FSB(mp, ((xfs_ufsize_t)(offset + count)));
if ((offset + count) > ip->i_size) {
error = xfs_iomap_eof_align_last_fsb(mp, ip, extsz, &last_fsb);
if (error)
goto error_out;
} else {
if (found && (ret_imap->br_startblock == HOLESTARTBLOCK))
last_fsb = MIN(last_fsb, (xfs_fileoff_t)
ret_imap->br_blockcount +
ret_imap->br_startoff);
}
count_fsb = last_fsb - offset_fsb;
ASSERT(count_fsb > 0);
resaligned = count_fsb;
if (unlikely(extsz)) {
if ((temp = do_mod(offset_fsb, extsz)))
resaligned += temp;
if ((temp = do_mod(resaligned, extsz)))
resaligned += extsz - temp;
}
if (unlikely(rt)) {
resrtextents = qblocks = resaligned;
resrtextents /= mp->m_sb.sb_rextsize;
resblks = XFS_DIOSTRAT_SPACE_RES(mp, 0);
quota_flag = XFS_QMOPT_RES_RTBLKS;
} else {
resrtextents = 0;
resblks = qblocks = XFS_DIOSTRAT_SPACE_RES(mp, resaligned);
quota_flag = XFS_QMOPT_RES_REGBLKS;
}
/*
* Allocate and setup the transaction
*/
xfs_iunlock(ip, XFS_ILOCK_EXCL);
tp = xfs_trans_alloc(mp, XFS_TRANS_DIOSTRAT);
error = xfs_trans_reserve(tp, resblks,
XFS_WRITE_LOG_RES(mp), resrtextents,
XFS_TRANS_PERM_LOG_RES,
XFS_WRITE_LOG_COUNT);
/*
* Check for running out of space, note: need lock to return
*/
if (error)
xfs_trans_cancel(tp, 0);
xfs_ilock(ip, XFS_ILOCK_EXCL);
if (error)
goto error_out;
error = xfs_trans_reserve_quota_nblks(tp, ip, qblocks, 0, quota_flag);
if (error)
goto error1;
xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
xfs_trans_ihold(tp, ip);
bmapi_flag = XFS_BMAPI_WRITE;
if ((flags & BMAPI_DIRECT) && (offset < ip->i_size || extsz))
bmapi_flag |= XFS_BMAPI_PREALLOC;
/*
* Issue the xfs_bmapi() call to allocate the blocks
*/
xfs_bmap_init(&free_list, &firstfsb);
nimaps = 1;
error = xfs_bmapi(tp, ip, offset_fsb, count_fsb, bmapi_flag,
&firstfsb, 0, &imap, &nimaps, &free_list, NULL);
if (error)
goto error0;
/*
* Complete the transaction
*/
error = xfs_bmap_finish(&tp, &free_list, &committed);
if (error)
goto error0;
error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
if (error)
goto error_out;
/*
* Copy any maps to caller's array and return any error.
*/
if (nimaps == 0) {
error = ENOSPC;
goto error_out;
}
if (!(imap.br_startblock || XFS_IS_REALTIME_INODE(ip))) {
error = xfs_cmn_err_fsblock_zero(ip, &imap);
goto error_out;
}
*ret_imap = imap;
*nmaps = 1;
return 0;
error0: /* Cancel bmap, unlock inode, unreserve quota blocks, cancel trans */
xfs_bmap_cancel(&free_list);
xfs_trans_unreserve_quota_nblks(tp, ip, qblocks, 0, quota_flag);
error1: /* Just cancel transaction */
xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES | XFS_TRANS_ABORT);
*nmaps = 0; /* nothing set-up here */
error_out:
return XFS_ERROR(error);
}
STATIC int
xfs_ioctl_setattr(
xfs_inode_t *ip,
struct fsxattr *fa,
int mask)
{
struct xfs_mount *mp = ip->i_mount;
struct xfs_trans *tp;
unsigned int lock_flags = 0;
struct xfs_dquot *udqp = NULL;
struct xfs_dquot *gdqp = NULL;
struct xfs_dquot *olddquot = NULL;
int code;
xfs_itrace_entry(ip);
if (mp->m_flags & XFS_MOUNT_RDONLY)
return XFS_ERROR(EROFS);
if (XFS_FORCED_SHUTDOWN(mp))
return XFS_ERROR(EIO);
/*
* If disk quotas is on, we make sure that the dquots do exist on disk,
* before we start any other transactions. Trying to do this later
* is messy. We don't care to take a readlock to look at the ids
* in inode here, because we can't hold it across the trans_reserve.
* If the IDs do change before we take the ilock, we're covered
* because the i_*dquot fields will get updated anyway.
*/
if (XFS_IS_QUOTA_ON(mp) && (mask & FSX_PROJID)) {
code = xfs_qm_vop_dqalloc(ip, ip->i_d.di_uid,
ip->i_d.di_gid, fa->fsx_projid,
XFS_QMOPT_PQUOTA, &udqp, &gdqp);
if (code)
return code;
}
/*
* For the other attributes, we acquire the inode lock and
* first do an error checking pass.
*/
tp = xfs_trans_alloc(mp, XFS_TRANS_SETATTR_NOT_SIZE);
code = xfs_trans_reserve(tp, 0, XFS_ICHANGE_LOG_RES(mp), 0, 0, 0);
if (code)
goto error_return;
lock_flags = XFS_ILOCK_EXCL;
xfs_ilock(ip, lock_flags);
/*
* CAP_FOWNER overrides the following restrictions:
*
* The user ID of the calling process must be equal
* to the file owner ID, except in cases where the
* CAP_FSETID capability is applicable.
*/
if (current_fsuid() != ip->i_d.di_uid && !capable(CAP_FOWNER)) {
code = XFS_ERROR(EPERM);
goto error_return;
}
/*
* Do a quota reservation only if projid is actually going to change.
*/
if (mask & FSX_PROJID) {
if (XFS_IS_QUOTA_RUNNING(mp) &&
XFS_IS_PQUOTA_ON(mp) &&
ip->i_d.di_projid != fa->fsx_projid) {
ASSERT(tp);
code = xfs_qm_vop_chown_reserve(tp, ip, udqp, gdqp,
capable(CAP_FOWNER) ?
XFS_QMOPT_FORCE_RES : 0);
if (code) /* out of quota */
goto error_return;
}
}
if (mask & FSX_EXTSIZE) {
/*
* Can't change extent size if any extents are allocated.
*/
if (ip->i_d.di_nextents &&
((ip->i_d.di_extsize << mp->m_sb.sb_blocklog) !=
fa->fsx_extsize)) {
code = XFS_ERROR(EINVAL); /* EFBIG? */
goto error_return;
}
/*
* Extent size must be a multiple of the appropriate block
* size, if set at all.
*/
if (fa->fsx_extsize != 0) {
xfs_extlen_t size;
if (XFS_IS_REALTIME_INODE(ip) ||
((mask & FSX_XFLAGS) &&
(fa->fsx_xflags & XFS_XFLAG_REALTIME))) {
size = mp->m_sb.sb_rextsize <<
mp->m_sb.sb_blocklog;
} else {
size = mp->m_sb.sb_blocksize;
}
if (fa->fsx_extsize % size) {
code = XFS_ERROR(EINVAL);
goto error_return;
}
}
}
if (mask & FSX_XFLAGS) {
/*
* Can't change realtime flag if any extents are allocated.
*/
if ((ip->i_d.di_nextents || ip->i_delayed_blks) &&
(XFS_IS_REALTIME_INODE(ip)) !=
(fa->fsx_xflags & XFS_XFLAG_REALTIME)) {
code = XFS_ERROR(EINVAL); /* EFBIG? */
goto error_return;
}
/*
* If realtime flag is set then must have realtime data.
*/
if ((fa->fsx_xflags & XFS_XFLAG_REALTIME)) {
if ((mp->m_sb.sb_rblocks == 0) ||
(mp->m_sb.sb_rextsize == 0) ||
(ip->i_d.di_extsize % mp->m_sb.sb_rextsize)) {
code = XFS_ERROR(EINVAL);
goto error_return;
}
}
/*
* Can't modify an immutable/append-only file unless
* we have appropriate permission.
*/
if ((ip->i_d.di_flags &
(XFS_DIFLAG_IMMUTABLE|XFS_DIFLAG_APPEND) ||
(fa->fsx_xflags &
(XFS_XFLAG_IMMUTABLE | XFS_XFLAG_APPEND))) &&
!capable(CAP_LINUX_IMMUTABLE)) {
code = XFS_ERROR(EPERM);
goto error_return;
}
}
xfs_trans_ijoin(tp, ip, lock_flags);
xfs_trans_ihold(tp, ip);
/*
* Change file ownership. Must be the owner or privileged.
*/
if (mask & FSX_PROJID) {
/*
* CAP_FSETID overrides the following restrictions:
*
* The set-user-ID and set-group-ID bits of a file will be
* cleared upon successful return from chown()
*/
if ((ip->i_d.di_mode & (S_ISUID|S_ISGID)) &&
!capable(CAP_FSETID))
ip->i_d.di_mode &= ~(S_ISUID|S_ISGID);
/*
* Change the ownerships and register quota modifications
* in the transaction.
*/
if (ip->i_d.di_projid != fa->fsx_projid) {
if (XFS_IS_QUOTA_RUNNING(mp) && XFS_IS_PQUOTA_ON(mp)) {
olddquot = xfs_qm_vop_chown(tp, ip,
&ip->i_gdquot, gdqp);
}
ip->i_d.di_projid = fa->fsx_projid;
/*
* We may have to rev the inode as well as
* the superblock version number since projids didn't
* exist before DINODE_VERSION_2 and SB_VERSION_NLINK.
*/
if (ip->i_d.di_version == 1)
xfs_bump_ino_vers2(tp, ip);
}
}
if (mask & FSX_EXTSIZE)
ip->i_d.di_extsize = fa->fsx_extsize >> mp->m_sb.sb_blocklog;
if (mask & FSX_XFLAGS) {
xfs_set_diflags(ip, fa->fsx_xflags);
xfs_diflags_to_linux(ip);
}
xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
xfs_ichgtime(ip, XFS_ICHGTIME_CHG);
XFS_STATS_INC(xs_ig_attrchg);
/*
* If this is a synchronous mount, make sure that the
* transaction goes to disk before returning to the user.
* This is slightly sub-optimal in that truncates require
* two sync transactions instead of one for wsync filesystems.
* One for the truncate and one for the timestamps since we
* don't want to change the timestamps unless we're sure the
* truncate worked. Truncates are less than 1% of the laddis
* mix so this probably isn't worth the trouble to optimize.
*/
if (mp->m_flags & XFS_MOUNT_WSYNC)
xfs_trans_set_sync(tp);
code = xfs_trans_commit(tp, 0);
xfs_iunlock(ip, lock_flags);
/*
* Release any dquot(s) the inode had kept before chown.
*/
xfs_qm_dqrele(olddquot);
xfs_qm_dqrele(udqp);
xfs_qm_dqrele(gdqp);
if (code)
return code;
if (DM_EVENT_ENABLED(ip, DM_EVENT_ATTRIBUTE)) {
XFS_SEND_NAMESP(mp, DM_EVENT_ATTRIBUTE, ip, DM_RIGHT_NULL,
NULL, DM_RIGHT_NULL, NULL, NULL, 0, 0,
(mask & FSX_NONBLOCK) ? DM_FLAGS_NDELAY : 0);
}
return 0;
error_return:
xfs_qm_dqrele(udqp);
xfs_qm_dqrele(gdqp);
xfs_trans_cancel(tp, 0);
if (lock_flags)
xfs_iunlock(ip, lock_flags);
return code;
}
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;
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 -EIO;
/*
* Check component lengths of the target path name.
*/
pathlen = strlen(target_path);
if (pathlen >= MAXPATHLEN) /* total string too long */
return -ENAMETOOLONG;
udqp = gdqp = NULL;
prid = xfs_get_initial_prid(dp);
/*
* 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)
return error;
/*
* 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_alloc(mp, &M_RES(mp)->tr_symlink, resblks, 0, 0, &tp);
if (error == -ENOSPC && fs_blocks == 0) {
resblks = 0;
error = xfs_trans_alloc(mp, &M_RES(mp)->tr_symlink, 0, 0, 0,
&tp);
}
if (error)
goto out_release_inode;
xfs_ilock(dp, XFS_IOLOCK_EXCL | XFS_ILOCK_EXCL |
XFS_IOLOCK_PARENT | 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 = -EPERM;
goto out_trans_cancel;
}
/*
* Reserve disk quota : blocks and inode.
*/
error = xfs_trans_reserve_quota(tp, mp, udqp, gdqp,
pdqp, resblks, 1, 0);
if (error)
goto out_trans_cancel;
/*
* Check for ability to enter directory entry, if no space reserved.
*/
if (!resblks) {
error = xfs_dir_canenter(tp, dp, link_name);
if (error)
goto out_trans_cancel;
}
/*
* 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)
goto out_trans_cancel;
/*
* Now we join the directory inode to the transaction. We do not do it
* earlier because xfs_dir_ialloc might commit the previous transaction
* (and release all the locks). An error from here on will result in
* the transaction cancel unlocking dp so don't do it explicitly in the
* error path.
*/
xfs_trans_ijoin(tp, dp, XFS_IOLOCK_EXCL | 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_init_local_fork(ip, XFS_DATA_FORK, target_path, pathlen);
ip->i_d.di_size = pathlen;
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 out_bmap_cancel;
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 out_bmap_cancel;
}
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_buf_set_type(tp, bp, XFS_BLFT_SYMLINK_BUF);
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 out_bmap_cancel;
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, NULL);
if (error)
goto out_bmap_cancel;
error = xfs_trans_commit(tp);
if (error)
goto out_release_inode;
xfs_qm_dqrele(udqp);
xfs_qm_dqrele(gdqp);
xfs_qm_dqrele(pdqp);
*ipp = ip;
return 0;
out_bmap_cancel:
xfs_bmap_cancel(&free_list);
out_trans_cancel:
xfs_trans_cancel(tp);
out_release_inode:
/*
* Wait until after the current transaction is aborted to finish the
* setup of the inode and release the inode. This prevents recursive
* transactions and deadlocks from xfs_inactive.
*/
if (ip) {
xfs_finish_inode_setup(ip);
IRELE(ip);
}
xfs_qm_dqrele(udqp);
xfs_qm_dqrele(gdqp);
xfs_qm_dqrele(pdqp);
if (unlock_dp_on_error)
xfs_iunlock(dp, XFS_IOLOCK_EXCL | XFS_ILOCK_EXCL);
return error;
}
/*
* xfs_attr_inactive kills all traces of an attribute fork on an inode. It
* removes both the on-disk and in-memory inode fork. Note that this also has to
* handle the condition of inodes without attributes but with an attribute fork
* configured, so we can't use xfs_inode_hasattr() here.
*
* The in-memory attribute fork is removed even on error.
*/
int
xfs_attr_inactive(
struct xfs_inode *dp)
{
struct xfs_trans *trans;
struct xfs_mount *mp;
int cancel_flags = 0;
int lock_mode = XFS_ILOCK_SHARED;
int error = 0;
mp = dp->i_mount;
ASSERT(! XFS_NOT_DQATTACHED(mp, dp));
xfs_ilock(dp, lock_mode);
if (!XFS_IFORK_Q(dp))
goto out_destroy_fork;
xfs_iunlock(dp, lock_mode);
/*
* Start our first transaction of the day.
*
* All future transactions during this code must be "chained" off
* this one via the trans_dup() call. All transactions will contain
* the inode, and the inode will always be marked with trans_ihold().
* Since the inode will be locked in all transactions, we must log
* the inode in every transaction to let it float upward through
* the log.
*/
lock_mode = 0;
trans = xfs_trans_alloc(mp, XFS_TRANS_ATTRINVAL);
error = xfs_trans_reserve(trans, &M_RES(mp)->tr_attrinval, 0, 0);
if (error)
goto out_cancel;
lock_mode = XFS_ILOCK_EXCL;
cancel_flags = XFS_TRANS_RELEASE_LOG_RES | XFS_TRANS_ABORT;
xfs_ilock(dp, lock_mode);
if (!XFS_IFORK_Q(dp))
goto out_cancel;
/*
* No need to make quota reservations here. We expect to release some
* blocks, not allocate, in the common case.
*/
xfs_trans_ijoin(trans, dp, 0);
/*
* Invalidate and truncate the attribute fork extents. Make sure the
* fork actually has attributes as otherwise the invalidation has no
* blocks to read and returns an error. In this case, just do the fork
* removal below.
*/
if (xfs_inode_hasattr(dp) &&
dp->i_d.di_aformat != XFS_DINODE_FMT_LOCAL) {
error = xfs_attr3_root_inactive(&trans, dp);
if (error)
goto out_cancel;
error = xfs_itruncate_extents(&trans, dp, XFS_ATTR_FORK, 0);
if (error)
goto out_cancel;
}
/* Reset the attribute fork - this also destroys the in-core fork */
xfs_attr_fork_remove(dp, trans);
error = xfs_trans_commit(trans, XFS_TRANS_RELEASE_LOG_RES);
xfs_iunlock(dp, lock_mode);
return error;
out_cancel:
xfs_trans_cancel(trans, cancel_flags);
out_destroy_fork:
/* kill the in-core attr fork before we drop the inode lock */
if (dp->i_afp)
xfs_idestroy_fork(dp, XFS_ATTR_FORK);
if (lock_mode)
xfs_iunlock(dp, lock_mode);
return error;
}
/*
* Free a symlink that has blocks associated with it.
*/
STATIC int
xfs_inactive_symlink_rmt(
struct xfs_inode *ip)
{
xfs_buf_t *bp;
int done;
int error;
xfs_fsblock_t first_block;
xfs_bmap_free_t free_list;
int i;
xfs_mount_t *mp;
xfs_bmbt_irec_t mval[XFS_SYMLINK_MAPS];
int nmaps;
int size;
xfs_trans_t *tp;
mp = ip->i_mount;
ASSERT(ip->i_df.if_flags & XFS_IFEXTENTS);
/*
* We're freeing a symlink that has some
* blocks allocated to it. Free the
* blocks here. We know that we've got
* either 1 or 2 extents and that we can
* free them all in one bunmapi call.
*/
ASSERT(ip->i_d.di_nextents > 0 && ip->i_d.di_nextents <= 2);
error = xfs_trans_alloc(mp, &M_RES(mp)->tr_itruncate, 0, 0, 0, &tp);
if (error)
return error;
xfs_ilock(ip, XFS_ILOCK_EXCL);
xfs_trans_ijoin(tp, ip, 0);
/*
* Lock the inode, fix the size, and join it to the transaction.
* Hold it so in the normal path, we still have it locked for
* the second transaction. In the error paths we need it
* held so the cancel won't rele it, see below.
*/
size = (int)ip->i_d.di_size;
ip->i_d.di_size = 0;
xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
/*
* Find the block(s) so we can inval and unmap them.
*/
done = 0;
xfs_bmap_init(&free_list, &first_block);
nmaps = ARRAY_SIZE(mval);
error = xfs_bmapi_read(ip, 0, xfs_symlink_blocks(mp, size),
mval, &nmaps, 0);
if (error)
goto error_trans_cancel;
/*
* Invalidate the block(s). No validation is done.
*/
for (i = 0; i < nmaps; i++) {
bp = xfs_trans_get_buf(tp, mp->m_ddev_targp,
XFS_FSB_TO_DADDR(mp, mval[i].br_startblock),
XFS_FSB_TO_BB(mp, mval[i].br_blockcount), 0);
if (!bp) {
error = -ENOMEM;
goto error_bmap_cancel;
}
xfs_trans_binval(tp, bp);
}
/*
* Unmap the dead block(s) to the free_list.
*/
error = xfs_bunmapi(tp, ip, 0, size, 0, nmaps,
&first_block, &free_list, &done);
if (error)
goto error_bmap_cancel;
ASSERT(done);
/*
* Commit the first transaction. This logs the EFI and the inode.
*/
error = xfs_bmap_finish(&tp, &free_list, ip);
if (error)
goto error_bmap_cancel;
/*
* The first xact was committed, so add the inode to the new one.
* Mark it dirty so it will be logged and moved forward in the log as
* part of every commit.
*/
xfs_trans_ijoin(tp, ip, 0);
xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
/*
* Commit the transaction containing extent freeing and EFDs.
*/
error = xfs_trans_commit(tp);
if (error) {
ASSERT(XFS_FORCED_SHUTDOWN(mp));
goto error_unlock;
}
/*
* Remove the memory for extent descriptions (just bookkeeping).
*/
if (ip->i_df.if_bytes)
xfs_idata_realloc(ip, -ip->i_df.if_bytes, XFS_DATA_FORK);
ASSERT(ip->i_df.if_bytes == 0);
xfs_iunlock(ip, XFS_ILOCK_EXCL);
return 0;
error_bmap_cancel:
xfs_bmap_cancel(&free_list);
error_trans_cancel:
xfs_trans_cancel(tp);
error_unlock:
xfs_iunlock(ip, XFS_ILOCK_EXCL);
return error;
}
static int
xfs_swap_extents(
xfs_inode_t *ip, /* target inode */
xfs_inode_t *tip, /* tmp inode */
xfs_swapext_t *sxp)
{
xfs_mount_t *mp = ip->i_mount;
xfs_trans_t *tp;
xfs_bstat_t *sbp = &sxp->sx_stat;
xfs_ifork_t *tempifp, *ifp, *tifp;
int ilf_fields, tilf_fields;
int error = 0;
int aforkblks = 0;
int taforkblks = 0;
__uint64_t tmp;
tempifp = kmem_alloc(sizeof(xfs_ifork_t), KM_MAYFAIL);
if (!tempifp) {
error = XFS_ERROR(ENOMEM);
goto out;
}
/*
* we have to do two separate lock calls here to keep lockdep
* happy. If we try to get all the locks in one call, lock will
* report false positives when we drop the ILOCK and regain them
* below.
*/
xfs_lock_two_inodes(ip, tip, XFS_IOLOCK_EXCL);
xfs_lock_two_inodes(ip, tip, XFS_ILOCK_EXCL);
/* Verify that both files have the same format */
if ((ip->i_d.di_mode & S_IFMT) != (tip->i_d.di_mode & S_IFMT)) {
error = XFS_ERROR(EINVAL);
goto out_unlock;
}
/* Verify both files are either real-time or non-realtime */
if (XFS_IS_REALTIME_INODE(ip) != XFS_IS_REALTIME_INODE(tip)) {
error = XFS_ERROR(EINVAL);
goto out_unlock;
}
if (VN_CACHED(VFS_I(tip)) != 0) {
error = xfs_flushinval_pages(tip, 0, -1,
FI_REMAPF_LOCKED);
if (error)
goto out_unlock;
}
/* Verify O_DIRECT for ftmp */
if (VN_CACHED(VFS_I(tip)) != 0) {
error = XFS_ERROR(EINVAL);
goto out_unlock;
}
/* Verify all data are being swapped */
if (sxp->sx_offset != 0 ||
sxp->sx_length != ip->i_d.di_size ||
sxp->sx_length != tip->i_d.di_size) {
error = XFS_ERROR(EFAULT);
goto out_unlock;
}
trace_xfs_swap_extent_before(ip, 0);
trace_xfs_swap_extent_before(tip, 1);
/* check inode formats now that data is flushed */
error = xfs_swap_extents_check_format(ip, tip);
if (error) {
xfs_notice(mp,
"%s: inode 0x%llx format is incompatible for exchanging.",
__func__, ip->i_ino);
goto out_unlock;
}
/*
* Compare the current change & modify times with that
* passed in. If they differ, we abort this swap.
* This is the mechanism used to ensure the calling
* process that the file was not changed out from
* under it.
*/
if ((sbp->bs_ctime.tv_sec != VFS_I(ip)->i_ctime.tv_sec) ||
(sbp->bs_ctime.tv_nsec != VFS_I(ip)->i_ctime.tv_nsec) ||
(sbp->bs_mtime.tv_sec != VFS_I(ip)->i_mtime.tv_sec) ||
(sbp->bs_mtime.tv_nsec != VFS_I(ip)->i_mtime.tv_nsec)) {
error = XFS_ERROR(EBUSY);
goto out_unlock;
}
/* We need to fail if the file is memory mapped. Once we have tossed
* all existing pages, the page fault will have no option
* but to go to the filesystem for pages. By making the page fault call
* vop_read (or write in the case of autogrow) they block on the iolock
* until we have switched the extents.
*/
if (VN_MAPPED(VFS_I(ip))) {
error = XFS_ERROR(EBUSY);
goto out_unlock;
}
xfs_iunlock(ip, XFS_ILOCK_EXCL);
xfs_iunlock(tip, XFS_ILOCK_EXCL);
/*
* There is a race condition here since we gave up the
* ilock. However, the data fork will not change since
* we have the iolock (locked for truncation too) so we
* are safe. We don't really care if non-io related
* fields change.
*/
xfs_tosspages(ip, 0, -1, FI_REMAPF);
tp = xfs_trans_alloc(mp, XFS_TRANS_SWAPEXT);
if ((error = xfs_trans_reserve(tp, 0,
XFS_ICHANGE_LOG_RES(mp), 0,
0, 0))) {
xfs_iunlock(ip, XFS_IOLOCK_EXCL);
xfs_iunlock(tip, XFS_IOLOCK_EXCL);
xfs_trans_cancel(tp, 0);
goto out;
}
xfs_lock_two_inodes(ip, tip, XFS_ILOCK_EXCL);
/*
* Count the number of extended attribute blocks
*/
if ( ((XFS_IFORK_Q(ip) != 0) && (ip->i_d.di_anextents > 0)) &&
(ip->i_d.di_aformat != XFS_DINODE_FMT_LOCAL)) {
error = xfs_bmap_count_blocks(tp, ip, XFS_ATTR_FORK, &aforkblks);
if (error)
goto out_trans_cancel;
}
if ( ((XFS_IFORK_Q(tip) != 0) && (tip->i_d.di_anextents > 0)) &&
(tip->i_d.di_aformat != XFS_DINODE_FMT_LOCAL)) {
error = xfs_bmap_count_blocks(tp, tip, XFS_ATTR_FORK,
&taforkblks);
if (error)
goto out_trans_cancel;
}
/*
* Swap the data forks of the inodes
*/
ifp = &ip->i_df;
tifp = &tip->i_df;
*tempifp = *ifp; /* struct copy */
*ifp = *tifp; /* struct copy */
*tifp = *tempifp; /* struct copy */
/*
* Fix the on-disk inode values
*/
tmp = (__uint64_t)ip->i_d.di_nblocks;
ip->i_d.di_nblocks = tip->i_d.di_nblocks - taforkblks + aforkblks;
tip->i_d.di_nblocks = tmp + taforkblks - aforkblks;
tmp = (__uint64_t) ip->i_d.di_nextents;
ip->i_d.di_nextents = tip->i_d.di_nextents;
tip->i_d.di_nextents = tmp;
tmp = (__uint64_t) ip->i_d.di_format;
ip->i_d.di_format = tip->i_d.di_format;
tip->i_d.di_format = tmp;
/*
* The extents in the source inode could still contain speculative
* preallocation beyond EOF (e.g. the file is open but not modified
* while defrag is in progress). In that case, we need to copy over the
* number of delalloc blocks the data fork in the source inode is
* tracking beyond EOF so that when the fork is truncated away when the
* temporary inode is unlinked we don't underrun the i_delayed_blks
* counter on that inode.
*/
ASSERT(tip->i_delayed_blks == 0);
tip->i_delayed_blks = ip->i_delayed_blks;
ip->i_delayed_blks = 0;
ilf_fields = XFS_ILOG_CORE;
switch(ip->i_d.di_format) {
case XFS_DINODE_FMT_EXTENTS:
/* If the extents fit in the inode, fix the
* pointer. Otherwise it's already NULL or
* pointing to the extent.
*/
if (ip->i_d.di_nextents <= XFS_INLINE_EXTS) {
ifp->if_u1.if_extents =
ifp->if_u2.if_inline_ext;
}
ilf_fields |= XFS_ILOG_DEXT;
break;
case XFS_DINODE_FMT_BTREE:
ilf_fields |= XFS_ILOG_DBROOT;
break;
}
tilf_fields = XFS_ILOG_CORE;
switch(tip->i_d.di_format) {
case XFS_DINODE_FMT_EXTENTS:
/* If the extents fit in the inode, fix the
* pointer. Otherwise it's already NULL or
* pointing to the extent.
*/
if (tip->i_d.di_nextents <= XFS_INLINE_EXTS) {
tifp->if_u1.if_extents =
tifp->if_u2.if_inline_ext;
}
tilf_fields |= XFS_ILOG_DEXT;
break;
case XFS_DINODE_FMT_BTREE:
tilf_fields |= XFS_ILOG_DBROOT;
break;
}
xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL);
xfs_trans_ijoin(tp, tip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL);
xfs_trans_log_inode(tp, ip, ilf_fields);
xfs_trans_log_inode(tp, tip, tilf_fields);
/*
* If this is a synchronous mount, make sure that the
* transaction goes to disk before returning to the user.
*/
if (mp->m_flags & XFS_MOUNT_WSYNC)
xfs_trans_set_sync(tp);
error = xfs_trans_commit(tp, 0);
trace_xfs_swap_extent_after(ip, 0);
trace_xfs_swap_extent_after(tip, 1);
out:
kmem_free(tempifp);
return error;
out_unlock:
xfs_iunlock(ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL);
xfs_iunlock(tip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL);
goto out;
out_trans_cancel:
xfs_trans_cancel(tp, 0);
goto out_unlock;
}
/*
* xfs_rename
*/
int
xfs_rename(
xfs_inode_t *src_dp,
struct xfs_name *src_name,
xfs_inode_t *src_ip,
xfs_inode_t *target_dp,
struct xfs_name *target_name,
xfs_inode_t *target_ip)
{
xfs_trans_t *tp = NULL;
xfs_mount_t *mp = src_dp->i_mount;
int new_parent; /* moving to a new dir */
int src_is_directory; /* src_name is a directory */
int error;
xfs_bmap_free_t free_list;
xfs_fsblock_t first_block;
int cancel_flags;
int committed;
xfs_inode_t *inodes[4];
int spaceres;
int num_inodes;
trace_xfs_rename(src_dp, target_dp, src_name, target_name);
new_parent = (src_dp != target_dp);
src_is_directory = S_ISDIR(src_ip->i_d.di_mode);
xfs_sort_for_rename(src_dp, target_dp, src_ip, target_ip,
inodes, &num_inodes);
xfs_bmap_init(&free_list, &first_block);
tp = xfs_trans_alloc(mp, XFS_TRANS_RENAME);
cancel_flags = XFS_TRANS_RELEASE_LOG_RES;
spaceres = XFS_RENAME_SPACE_RES(mp, target_name->len);
error = xfs_trans_reserve(tp, spaceres, XFS_RENAME_LOG_RES(mp), 0,
XFS_TRANS_PERM_LOG_RES, XFS_RENAME_LOG_COUNT);
if (error == ENOSPC) {
spaceres = 0;
error = xfs_trans_reserve(tp, 0, XFS_RENAME_LOG_RES(mp), 0,
XFS_TRANS_PERM_LOG_RES, XFS_RENAME_LOG_COUNT);
}
if (error) {
xfs_trans_cancel(tp, 0);
goto std_return;
}
/*
* Attach the dquots to the inodes
*/
error = xfs_qm_vop_rename_dqattach(inodes);
if (error) {
xfs_trans_cancel(tp, cancel_flags);
goto std_return;
}
/*
* Lock all the participating inodes. Depending upon whether
* the target_name exists in the target directory, and
* whether the target directory is the same as the source
* directory, we can lock from 2 to 4 inodes.
*/
xfs_lock_inodes(inodes, num_inodes, XFS_ILOCK_EXCL);
/*
* Join all the inodes to the transaction. From this point on,
* we can rely on either trans_commit or trans_cancel to unlock
* them.
*/
xfs_trans_ijoin(tp, src_dp, XFS_ILOCK_EXCL);
if (new_parent)
xfs_trans_ijoin(tp, target_dp, XFS_ILOCK_EXCL);
xfs_trans_ijoin(tp, src_ip, XFS_ILOCK_EXCL);
if (target_ip)
xfs_trans_ijoin(tp, target_ip, XFS_ILOCK_EXCL);
/*
* If we are using project inheritance, we only allow renames
* into our tree when the project IDs are the same; else the
* tree quota mechanism would be circumvented.
*/
if (unlikely((target_dp->i_d.di_flags & XFS_DIFLAG_PROJINHERIT) &&
(xfs_get_projid(target_dp) != xfs_get_projid(src_ip)))) {
error = XFS_ERROR(EXDEV);
goto error_return;
}
/*
* Set up the target.
*/
if (target_ip == NULL) {
/*
* If there's no space reservation, check the entry will
* fit before actually inserting it.
*/
error = xfs_dir_canenter(tp, target_dp, target_name, spaceres);
if (error)
goto error_return;
/*
* If target does not exist and the rename crosses
* directories, adjust the target directory link count
* to account for the ".." reference from the new entry.
*/
error = xfs_dir_createname(tp, target_dp, target_name,
src_ip->i_ino, &first_block,
&free_list, spaceres);
if (error == ENOSPC)
goto error_return;
if (error)
goto abort_return;
xfs_trans_ichgtime(tp, target_dp,
XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
if (new_parent && src_is_directory) {
error = xfs_bumplink(tp, target_dp);
if (error)
goto abort_return;
}
} else { /* target_ip != NULL */
/*
* If target exists and it's a directory, check that both
* target and source are directories and that target can be
* destroyed, or that neither is a directory.
*/
if (S_ISDIR(target_ip->i_d.di_mode)) {
/*
* Make sure target dir is empty.
*/
if (!(xfs_dir_isempty(target_ip)) ||
(target_ip->i_d.di_nlink > 2)) {
error = XFS_ERROR(EEXIST);
goto error_return;
}
}
/*
* Link the source inode under the target name.
* If the source inode is a directory and we are moving
* it across directories, its ".." entry will be
* inconsistent until we replace that down below.
*
* In case there is already an entry with the same
* name at the destination directory, remove it first.
*/
error = xfs_dir_replace(tp, target_dp, target_name,
src_ip->i_ino,
&first_block, &free_list, spaceres);
if (error)
goto abort_return;
xfs_trans_ichgtime(tp, target_dp,
XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
/*
* Decrement the link count on the target since the target
* dir no longer points to it.
*/
error = xfs_droplink(tp, target_ip);
if (error)
goto abort_return;
if (src_is_directory) {
/*
* Drop the link from the old "." entry.
*/
error = xfs_droplink(tp, target_ip);
if (error)
goto abort_return;
}
} /* target_ip != NULL */
/*
* Remove the source.
*/
if (new_parent && src_is_directory) {
/*
* Rewrite the ".." entry to point to the new
* directory.
*/
error = xfs_dir_replace(tp, src_ip, &xfs_name_dotdot,
target_dp->i_ino,
&first_block, &free_list, spaceres);
ASSERT(error != EEXIST);
if (error)
goto abort_return;
}
/*
* We always want to hit the ctime on the source inode.
*
* This isn't strictly required by the standards since the source
* inode isn't really being changed, but old unix file systems did
* it and some incremental backup programs won't work without it.
*/
xfs_trans_ichgtime(tp, src_ip, XFS_ICHGTIME_CHG);
xfs_trans_log_inode(tp, src_ip, XFS_ILOG_CORE);
/*
* Adjust the link count on src_dp. This is necessary when
* renaming a directory, either within one parent when
* the target existed, or across two parent directories.
*/
if (src_is_directory && (new_parent || target_ip != NULL)) {
/*
* Decrement link count on src_directory since the
* entry that's moved no longer points to it.
*/
error = xfs_droplink(tp, src_dp);
if (error)
goto abort_return;
}
error = xfs_dir_removename(tp, src_dp, src_name, src_ip->i_ino,
&first_block, &free_list, spaceres);
if (error)
goto abort_return;
xfs_trans_ichgtime(tp, src_dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
xfs_trans_log_inode(tp, src_dp, XFS_ILOG_CORE);
if (new_parent)
xfs_trans_log_inode(tp, target_dp, XFS_ILOG_CORE);
/*
* If this is a synchronous mount, make sure that the
* rename 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) {
xfs_bmap_cancel(&free_list);
xfs_trans_cancel(tp, (XFS_TRANS_RELEASE_LOG_RES |
XFS_TRANS_ABORT));
goto std_return;
}
/*
* trans_commit will unlock src_ip, target_ip & decrement
* the vnode references.
*/
return xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
abort_return:
cancel_flags |= XFS_TRANS_ABORT;
error_return:
xfs_bmap_cancel(&free_list);
xfs_trans_cancel(tp, cancel_flags);
std_return:
return error;
}
int
xfs_attr_inactive(xfs_inode_t *dp)
{
xfs_trans_t *trans;
xfs_mount_t *mp;
int error;
mp = dp->i_mount;
ASSERT(! XFS_NOT_DQATTACHED(mp, dp));
xfs_ilock(dp, XFS_ILOCK_SHARED);
if (!xfs_inode_hasattr(dp) ||
dp->i_d.di_aformat == XFS_DINODE_FMT_LOCAL) {
xfs_iunlock(dp, XFS_ILOCK_SHARED);
return 0;
}
xfs_iunlock(dp, XFS_ILOCK_SHARED);
/*
* Start our first transaction of the day.
*
* All future transactions during this code must be "chained" off
* this one via the trans_dup() call. All transactions will contain
* the inode, and the inode will always be marked with trans_ihold().
* Since the inode will be locked in all transactions, we must log
* the inode in every transaction to let it float upward through
* the log.
*/
trans = xfs_trans_alloc(mp, XFS_TRANS_ATTRINVAL);
error = xfs_trans_reserve(trans, &M_RES(mp)->tr_attrinval, 0, 0);
if (error) {
xfs_trans_cancel(trans, 0);
return(error);
}
xfs_ilock(dp, XFS_ILOCK_EXCL);
/*
* No need to make quota reservations here. We expect to release some
* blocks, not allocate, in the common case.
*/
xfs_trans_ijoin(trans, dp, 0);
/*
* Decide on what work routines to call based on the inode size.
*/
if (!xfs_inode_hasattr(dp) ||
dp->i_d.di_aformat == XFS_DINODE_FMT_LOCAL) {
error = 0;
goto out;
}
error = xfs_attr3_root_inactive(&trans, dp);
if (error)
goto out;
error = xfs_itruncate_extents(&trans, dp, XFS_ATTR_FORK, 0);
if (error)
goto out;
error = xfs_trans_commit(trans, XFS_TRANS_RELEASE_LOG_RES);
xfs_iunlock(dp, XFS_ILOCK_EXCL);
return(error);
out:
xfs_trans_cancel(trans, XFS_TRANS_RELEASE_LOG_RES|XFS_TRANS_ABORT);
xfs_iunlock(dp, XFS_ILOCK_EXCL);
return(error);
}
/*
* Syssgi interface for swapext
*/
int
xfs_swapext(
xfs_swapext_t __user *sxu)
{
xfs_swapext_t *sxp;
xfs_inode_t *ip=NULL, *tip=NULL, *ips[2];
xfs_trans_t *tp;
xfs_mount_t *mp;
xfs_bstat_t *sbp;
struct file *fp = NULL, *tfp = NULL;
vnode_t *vp, *tvp;
static uint lock_flags = XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL;
int ilf_fields, tilf_fields;
int error = 0;
xfs_ifork_t *tempifp, *ifp, *tifp;
__uint64_t tmp;
int aforkblks = 0;
int taforkblks = 0;
char locked = 0;
sxp = kmem_alloc(sizeof(xfs_swapext_t), KM_MAYFAIL);
tempifp = kmem_alloc(sizeof(xfs_ifork_t), KM_MAYFAIL);
if (!sxp || !tempifp) {
error = XFS_ERROR(ENOMEM);
goto error0;
}
if (copy_from_user(sxp, sxu, sizeof(xfs_swapext_t))) {
error = XFS_ERROR(EFAULT);
goto error0;
}
/* Pull information for the target fd */
if (((fp = fget((int)sxp->sx_fdtarget)) == NULL) ||
((vp = vn_from_inode(fp->f_dentry->d_inode)) == NULL)) {
error = XFS_ERROR(EINVAL);
goto error0;
}
ip = xfs_vtoi(vp);
if (ip == NULL) {
error = XFS_ERROR(EBADF);
goto error0;
}
if (((tfp = fget((int)sxp->sx_fdtmp)) == NULL) ||
((tvp = vn_from_inode(tfp->f_dentry->d_inode)) == NULL)) {
error = XFS_ERROR(EINVAL);
goto error0;
}
tip = xfs_vtoi(tvp);
if (tip == NULL) {
error = XFS_ERROR(EBADF);
goto error0;
}
if (ip->i_mount != tip->i_mount) {
error = XFS_ERROR(EINVAL);
goto error0;
}
if (ip->i_ino == tip->i_ino) {
error = XFS_ERROR(EINVAL);
goto error0;
}
mp = ip->i_mount;
sbp = &sxp->sx_stat;
if (XFS_FORCED_SHUTDOWN(mp)) {
error = XFS_ERROR(EIO);
goto error0;
}
locked = 1;
/* Lock in i_ino order */
if (ip->i_ino < tip->i_ino) {
ips[0] = ip;
ips[1] = tip;
} else {
ips[0] = tip;
ips[1] = ip;
}
xfs_lock_inodes(ips, 2, 0, lock_flags);
/* Check permissions */
error = xfs_iaccess(ip, S_IWUSR, NULL);
if (error)
goto error0;
error = xfs_iaccess(tip, S_IWUSR, NULL);
if (error)
goto error0;
/* Verify that both files have the same format */
if ((ip->i_d.di_mode & S_IFMT) != (tip->i_d.di_mode & S_IFMT)) {
error = XFS_ERROR(EINVAL);
goto error0;
}
/* Verify both files are either real-time or non-realtime */
if ((ip->i_d.di_flags & XFS_DIFLAG_REALTIME) !=
(tip->i_d.di_flags & XFS_DIFLAG_REALTIME)) {
error = XFS_ERROR(EINVAL);
goto error0;
}
/* Should never get a local format */
if (ip->i_d.di_format == XFS_DINODE_FMT_LOCAL ||
tip->i_d.di_format == XFS_DINODE_FMT_LOCAL) {
error = XFS_ERROR(EINVAL);
goto error0;
}
if (VN_CACHED(tvp) != 0) {
xfs_inval_cached_trace(&tip->i_iocore, 0, -1, 0, -1);
VOP_FLUSHINVAL_PAGES(tvp, 0, -1, FI_REMAPF_LOCKED);
}
/* Verify O_DIRECT for ftmp */
if (VN_CACHED(tvp) != 0) {
error = XFS_ERROR(EINVAL);
goto error0;
}
/* Verify all data are being swapped */
if (sxp->sx_offset != 0 ||
sxp->sx_length != ip->i_d.di_size ||
sxp->sx_length != tip->i_d.di_size) {
error = XFS_ERROR(EFAULT);
goto error0;
}
/*
* If the target has extended attributes, the tmp file
* must also in order to ensure the correct data fork
* format.
*/
if ( XFS_IFORK_Q(ip) != XFS_IFORK_Q(tip) ) {
error = XFS_ERROR(EINVAL);
goto error0;
}
/*
* Compare the current change & modify times with that
* passed in. If they differ, we abort this swap.
* This is the mechanism used to ensure the calling
* process that the file was not changed out from
* under it.
*/
if ((sbp->bs_ctime.tv_sec != ip->i_d.di_ctime.t_sec) ||
(sbp->bs_ctime.tv_nsec != ip->i_d.di_ctime.t_nsec) ||
(sbp->bs_mtime.tv_sec != ip->i_d.di_mtime.t_sec) ||
(sbp->bs_mtime.tv_nsec != ip->i_d.di_mtime.t_nsec)) {
error = XFS_ERROR(EBUSY);
goto error0;
}
/* We need to fail if the file is memory mapped. Once we have tossed
* all existing pages, the page fault will have no option
* but to go to the filesystem for pages. By making the page fault call
* VOP_READ (or write in the case of autogrow) they block on the iolock
* until we have switched the extents.
*/
if (VN_MAPPED(vp)) {
error = XFS_ERROR(EBUSY);
goto error0;
}
xfs_iunlock(ip, XFS_ILOCK_EXCL);
xfs_iunlock(tip, XFS_ILOCK_EXCL);
/*
* There is a race condition here since we gave up the
* ilock. However, the data fork will not change since
* we have the iolock (locked for truncation too) so we
* are safe. We don't really care if non-io related
* fields change.
*/
VOP_TOSS_PAGES(vp, 0, -1, FI_REMAPF);
tp = xfs_trans_alloc(mp, XFS_TRANS_SWAPEXT);
if ((error = xfs_trans_reserve(tp, 0,
XFS_ICHANGE_LOG_RES(mp), 0,
0, 0))) {
xfs_iunlock(ip, XFS_IOLOCK_EXCL);
xfs_iunlock(tip, XFS_IOLOCK_EXCL);
xfs_trans_cancel(tp, 0);
locked = 0;
goto error0;
}
xfs_lock_inodes(ips, 2, 0, XFS_ILOCK_EXCL);
/*
* Count the number of extended attribute blocks
*/
if ( ((XFS_IFORK_Q(ip) != 0) && (ip->i_d.di_anextents > 0)) &&
(ip->i_d.di_aformat != XFS_DINODE_FMT_LOCAL)) {
error = xfs_bmap_count_blocks(tp, ip, XFS_ATTR_FORK, &aforkblks);
if (error) {
xfs_trans_cancel(tp, 0);
goto error0;
}
}
if ( ((XFS_IFORK_Q(tip) != 0) && (tip->i_d.di_anextents > 0)) &&
(tip->i_d.di_aformat != XFS_DINODE_FMT_LOCAL)) {
error = xfs_bmap_count_blocks(tp, tip, XFS_ATTR_FORK,
&taforkblks);
if (error) {
xfs_trans_cancel(tp, 0);
goto error0;
}
}
/*
* Swap the data forks of the inodes
*/
ifp = &ip->i_df;
tifp = &tip->i_df;
*tempifp = *ifp; /* struct copy */
*ifp = *tifp; /* struct copy */
*tifp = *tempifp; /* struct copy */
/*
* Fix the on-disk inode values
*/
tmp = (__uint64_t)ip->i_d.di_nblocks;
ip->i_d.di_nblocks = tip->i_d.di_nblocks - taforkblks + aforkblks;
tip->i_d.di_nblocks = tmp + taforkblks - aforkblks;
tmp = (__uint64_t) ip->i_d.di_nextents;
ip->i_d.di_nextents = tip->i_d.di_nextents;
tip->i_d.di_nextents = tmp;
tmp = (__uint64_t) ip->i_d.di_format;
ip->i_d.di_format = tip->i_d.di_format;
tip->i_d.di_format = tmp;
ilf_fields = XFS_ILOG_CORE;
switch(ip->i_d.di_format) {
case XFS_DINODE_FMT_EXTENTS:
/* If the extents fit in the inode, fix the
* pointer. Otherwise it's already NULL or
* pointing to the extent.
*/
if (ip->i_d.di_nextents <= XFS_INLINE_EXTS) {
ifp->if_u1.if_extents =
ifp->if_u2.if_inline_ext;
}
ilf_fields |= XFS_ILOG_DEXT;
break;
case XFS_DINODE_FMT_BTREE:
ilf_fields |= XFS_ILOG_DBROOT;
break;
}
tilf_fields = XFS_ILOG_CORE;
switch(tip->i_d.di_format) {
case XFS_DINODE_FMT_EXTENTS:
/* If the extents fit in the inode, fix the
* pointer. Otherwise it's already NULL or
* pointing to the extent.
*/
if (tip->i_d.di_nextents <= XFS_INLINE_EXTS) {
tifp->if_u1.if_extents =
tifp->if_u2.if_inline_ext;
}
tilf_fields |= XFS_ILOG_DEXT;
break;
case XFS_DINODE_FMT_BTREE:
tilf_fields |= XFS_ILOG_DBROOT;
break;
}
/*
* Increment vnode ref counts since xfs_trans_commit &
* xfs_trans_cancel will both unlock the inodes and
* decrement the associated ref counts.
*/
VN_HOLD(vp);
VN_HOLD(tvp);
xfs_trans_ijoin(tp, ip, lock_flags);
xfs_trans_ijoin(tp, tip, lock_flags);
xfs_trans_log_inode(tp, ip, ilf_fields);
xfs_trans_log_inode(tp, tip, tilf_fields);
/*
* If this is a synchronous mount, make sure that the
* transaction goes to disk before returning to the user.
*/
if (mp->m_flags & XFS_MOUNT_WSYNC) {
xfs_trans_set_sync(tp);
}
error = xfs_trans_commit(tp, XFS_TRANS_SWAPEXT, NULL);
locked = 0;
error0:
if (locked) {
xfs_iunlock(ip, lock_flags);
xfs_iunlock(tip, lock_flags);
}
if (fp != NULL)
fput(fp);
if (tfp != NULL)
fput(tfp);
if (sxp != NULL)
kmem_free(sxp, sizeof(xfs_swapext_t));
if (tempifp != NULL)
kmem_free(tempifp, sizeof(xfs_ifork_t));
return error;
}
ssize_t /* bytes written, or (-) error */
xfs_write(
bhv_desc_t *bdp,
uio_t *uio,
int ioflag,
cred_t *credp)
{
xfs_inode_t *xip;
xfs_mount_t *mp;
ssize_t ret = 0;
int error = 0;
xfs_fsize_t isize, new_size;
xfs_fsize_t n, limit;
xfs_fsize_t size;
xfs_iocore_t *io;
xfs_vnode_t *vp;
int iolock;
//int eventsent = 0;
vrwlock_t locktype;
xfs_off_t offset_c;
xfs_off_t *offset;
xfs_off_t pos;
XFS_STATS_INC(xs_write_calls);
vp = BHV_TO_VNODE(bdp);
xip = XFS_BHVTOI(bdp);
io = &xip->i_iocore;
mp = io->io_mount;
if (XFS_FORCED_SHUTDOWN(xip->i_mount)) {
return EIO;
}
size = uio->uio_resid;
pos = offset_c = uio->uio_offset;
offset = &offset_c;
if (unlikely(ioflag & IO_ISDIRECT)) {
if (((__psint_t)buf & BBMASK) ||
(*offset & mp->m_blockmask) ||
(size & mp->m_blockmask)) {
return EINVAL;
}
iolock = XFS_IOLOCK_SHARED;
locktype = VRWLOCK_WRITE_DIRECT;
} else {
if (io->io_flags & XFS_IOCORE_RT)
return EINVAL;
iolock = XFS_IOLOCK_EXCL;
locktype = VRWLOCK_WRITE;
}
iolock = XFS_IOLOCK_EXCL;
locktype = VRWLOCK_WRITE;
xfs_ilock(xip, XFS_ILOCK_EXCL|iolock);
isize = xip->i_d.di_size;
limit = XFS_MAXIOFFSET(mp);
if (ioflag & O_APPEND)
*offset = isize;
//start:
n = limit - *offset;
if (n <= 0) {
xfs_iunlock(xip, XFS_ILOCK_EXCL|iolock);
return EFBIG;
}
if (n < size)
size = n;
new_size = *offset + size;
if (new_size > isize) {
io->io_new_size = new_size;
}
#ifdef RMC
/* probably be a long time before if ever that we do dmapi */
if ((DM_EVENT_ENABLED(vp->v_vfsp, xip, DM_EVENT_WRITE) &&
!(ioflags & IO_INVIS) && !eventsent)) {
loff_t savedsize = *offset;
int dmflags = FILP_DELAY_FLAG(file) | DM_SEM_FLAG_RD(ioflags);
xfs_iunlock(xip, XFS_ILOCK_EXCL);
error = XFS_SEND_DATA(xip->i_mount, DM_EVENT_WRITE, vp,
*offset, size,
dmflags, &locktype);
if (error) {
if (iolock) xfs_iunlock(xip, iolock);
return -error;
}
xfs_ilock(xip, XFS_ILOCK_EXCL);
eventsent = 1;
/*
* The iolock was dropped and reaquired in XFS_SEND_DATA
* so we have to recheck the size when appending.
* We will only "goto start;" once, since having sent the
* event prevents another call to XFS_SEND_DATA, which is
* what allows the size to change in the first place.
*/
if ((file->f_flags & O_APPEND) &&
savedsize != xip->i_d.di_size) {
*offset = isize = xip->i_d.di_size;
goto start;
}
}
#endif
/*
* If the offset is beyond the size of the file, we have a couple
* of things to do. First, if there is already space allocated
* we need to either create holes or zero the disk or ...
*
* If there is a page where the previous size lands, we need
* to zero it out up to the new size.
*/
if (!(ioflag & IO_ISDIRECT) && (*offset > isize && isize)) {
error = xfs_zero_eof(BHV_TO_VNODE(bdp), io, *offset,
isize, *offset + size);
if (error) {
xfs_iunlock(xip, XFS_ILOCK_EXCL|iolock);
return(-error);
}
}
xfs_iunlock(xip, XFS_ILOCK_EXCL);
#if 0
/*
* If we're writing the file then make sure to clear the
* setuid and setgid bits if the process is not being run
* by root. This keeps people from modifying setuid and
* setgid binaries.
*/
if (((xip->i_d.di_mode & S_ISUID) ||
((xip->i_d.di_mode & (S_ISGID | S_IXGRP)) ==
(S_ISGID | S_IXGRP))) &&
!capable(CAP_FSETID)) {
error = xfs_write_clear_setuid(xip);
if (likely(!error))
error = -remove_suid(file->f_dentry);
if (unlikely(error)) {
xfs_iunlock(xip, iolock);
goto out_unlock_mutex;
}
}
#endif
//retry:
if (unlikely(ioflag & IO_ISDIRECT)) {
#ifdef RMC
xfs_off_t pos = *offset;
struct address_space *mapping = file->f_dentry->d_inode->i_mapping;
struct inode *inode = mapping->host;
ret = precheck_file_write(file, inode, &size, &pos);
if (ret || size == 0)
goto error;
xfs_inval_cached_pages(vp, io, pos, 1, 1);
inode->i_ctime = inode->i_mtime = CURRENT_TIME;
/* mark_inode_dirty_sync(inode); - we do this later */
xfs_rw_enter_trace(XFS_DIOWR_ENTER, io, buf, size, pos, ioflags);
ret = generic_file_direct_IO(WRITE, file, (char *)buf, size, pos);
xfs_inval_cached_pages(vp, io, pos, 1, 1);
if (ret > 0)
*offset += ret;
#endif
} else {
xfs_rw_enter_trace(XFS_WRITE_ENTER, io, buf, size, *offset, ioflags);
ret = xfs_write_file(xip,uio,ioflag);
}
xfs_ichgtime(xip, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
//error:
if (ret <= 0) {
if (iolock)
xfs_rwunlock(bdp, locktype);
return ret;
}
XFS_STATS_ADD(xs_write_bytes, ret);
if (*offset > xip->i_d.di_size) {
xfs_ilock(xip, XFS_ILOCK_EXCL);
if (*offset > xip->i_d.di_size) {
printf("xfs_write look at doing more here %s:%d\n",__FILE__,__LINE__);
#ifdef RMC
struct inode *inode = LINVFS_GET_IP(vp);
i_size_write(inode, *offset);
mark_inode_dirty_sync(inode);
#endif
xip->i_d.di_size = *offset;
xip->i_update_core = 1;
xip->i_update_size = 1;
}
xfs_iunlock(xip, XFS_ILOCK_EXCL);
}
/* Handle various SYNC-type writes */
#if 0
// if ((file->f_flags & O_SYNC) || IS_SYNC(inode)) {
#endif
if (ioflag & IO_SYNC) {
/*
* If we're treating this as O_DSYNC and we have not updated the
* size, force the log.
*/
if (!(mp->m_flags & XFS_MOUNT_OSYNCISOSYNC) &&
!(xip->i_update_size)) {
xfs_inode_log_item_t *iip = xip->i_itemp;
/*
* If an allocation transaction occurred
* without extending the size, then we have to force
* the log up the proper point to ensure that the
* allocation is permanent. We can't count on
* the fact that buffered writes lock out direct I/O
* writes - the direct I/O write could have extended
* the size nontransactionally, then finished before
* we started. xfs_write_file will think that the file
* didn't grow but the update isn't safe unless the
* size change is logged.
*
* Force the log if we've committed a transaction
* against the inode or if someone else has and
* the commit record hasn't gone to disk (e.g.
* the inode is pinned). This guarantees that
* all changes affecting the inode are permanent
* when we return.
*/
if (iip && iip->ili_last_lsn) {
xfs_log_force(mp, iip->ili_last_lsn,
XFS_LOG_FORCE | XFS_LOG_SYNC);
} else if (xfs_ipincount(xip) > 0) {
xfs_log_force(mp, (xfs_lsn_t)0,
XFS_LOG_FORCE | XFS_LOG_SYNC);
}
} else {
xfs_trans_t *tp;
/*
* O_SYNC or O_DSYNC _with_ a size update are handled
* the same way.
*
* If the write was synchronous then we need to make
* sure that the inode modification time is permanent.
* We'll have updated the timestamp above, so here
* we use a synchronous transaction to log the inode.
* It's not fast, but it's necessary.
*
* If this a dsync write and the size got changed
* non-transactionally, then we need to ensure that
* the size change gets logged in a synchronous
* transaction.
*/
tp = xfs_trans_alloc(mp, XFS_TRANS_WRITE_SYNC);
if ((error = xfs_trans_reserve(tp, 0,
XFS_SWRITE_LOG_RES(mp),
0, 0, 0))) {
/* Transaction reserve failed */
xfs_trans_cancel(tp, 0);
} else {
/* Transaction reserve successful */
xfs_ilock(xip, XFS_ILOCK_EXCL);
xfs_trans_ijoin(tp, xip, XFS_ILOCK_EXCL);
xfs_trans_ihold(tp, xip);
xfs_trans_log_inode(tp, xip, XFS_ILOG_CORE);
xfs_trans_set_sync(tp);
error = xfs_trans_commit(tp, 0, NULL);
xfs_iunlock(xip, XFS_ILOCK_EXCL);
}
if (error)
goto out_unlock_internal;
}
xfs_rwunlock(bdp, locktype);
return ret;
} /* (ioflags & O_SYNC) */
out_unlock_internal:
xfs_rwunlock(bdp, locktype);
#if 0
out_unlock_mutex:
if (need_i_mutex)
mutex_unlock(&inode->i_mutex);
#endif
//out_nounlocks:
return -error;
}
/*
* This routine allocates disk space for the given file.
* Originally derived from xfs_alloc_file_space().
*/
int
libxfs_alloc_file_space(
xfs_inode_t *ip,
xfs_off_t offset,
xfs_off_t len,
int alloc_type,
int attr_flags)
{
xfs_mount_t *mp;
xfs_off_t count;
xfs_filblks_t datablocks;
xfs_filblks_t allocated_fsb;
xfs_filblks_t allocatesize_fsb;
xfs_fsblock_t firstfsb;
xfs_bmap_free_t free_list;
xfs_bmbt_irec_t *imapp;
xfs_bmbt_irec_t imaps[1];
int reccount;
uint resblks;
xfs_fileoff_t startoffset_fsb;
xfs_trans_t *tp;
int xfs_bmapi_flags;
int committed;
int error;
if (len <= 0)
return EINVAL;
count = len;
error = 0;
imapp = &imaps[0];
reccount = 1;
xfs_bmapi_flags = XFS_BMAPI_WRITE | (alloc_type ? XFS_BMAPI_PREALLOC : 0);
mp = ip->i_mount;
startoffset_fsb = XFS_B_TO_FSBT(mp, offset);
allocatesize_fsb = XFS_B_TO_FSB(mp, count);
/* allocate file space until done or until there is an error */
while (allocatesize_fsb && !error) {
datablocks = allocatesize_fsb;
tp = xfs_trans_alloc(mp, XFS_TRANS_DIOSTRAT);
resblks = (uint)XFS_DIOSTRAT_SPACE_RES(mp, datablocks);
error = xfs_trans_reserve(tp, resblks, 0, 0, 0, 0);
if (error)
break;
xfs_trans_ijoin(tp, ip, 0);
xfs_trans_ihold(tp, ip);
XFS_BMAP_INIT(&free_list, &firstfsb);
error = xfs_bmapi(tp, ip, startoffset_fsb, allocatesize_fsb,
xfs_bmapi_flags, &firstfsb, 0, imapp,
&reccount, &free_list);
if (error)
break;
/* complete the transaction */
error = xfs_bmap_finish(&tp, &free_list, firstfsb, &committed);
if (error)
break;
error = xfs_trans_commit(tp, 0, NULL);
if (error)
break;
allocated_fsb = imapp->br_blockcount;
if (reccount == 0)
return ENOSPC;
startoffset_fsb += allocated_fsb;
allocatesize_fsb -= allocated_fsb;
}
return error;
}
/*
* Adjust quota limits, and start/stop timers accordingly.
*/
STATIC int
xfs_qm_scall_setqlim(
xfs_mount_t *mp,
xfs_dqid_t id,
uint type,
fs_disk_quota_t *newlim)
{
xfs_disk_dquot_t *ddq;
xfs_dquot_t *dqp;
xfs_trans_t *tp;
int error;
xfs_qcnt_t hard, soft;
if (!capable(CAP_SYS_ADMIN))
return XFS_ERROR(EPERM);
if ((newlim->d_fieldmask &
(FS_DQ_LIMIT_MASK|FS_DQ_TIMER_MASK|FS_DQ_WARNS_MASK)) == 0)
return (0);
tp = xfs_trans_alloc(mp, XFS_TRANS_QM_SETQLIM);
if ((error = xfs_trans_reserve(tp, 0, sizeof(xfs_disk_dquot_t) + 128,
0, 0, XFS_DEFAULT_LOG_COUNT))) {
xfs_trans_cancel(tp, 0);
return (error);
}
/*
* We don't want to race with a quotaoff so take the quotaoff lock.
* (We don't hold an inode lock, so there's nothing else to stop
* a quotaoff from happening). (XXXThis doesn't currently happen
* because we take the vfslock before calling xfs_qm_sysent).
*/
mutex_lock(&(XFS_QI_QOFFLOCK(mp)), PINOD);
/*
* Get the dquot (locked), and join it to the transaction.
* Allocate the dquot if this doesn't exist.
*/
if ((error = xfs_qm_dqget(mp, NULL, id, type, XFS_QMOPT_DQALLOC, &dqp))) {
xfs_trans_cancel(tp, XFS_TRANS_ABORT);
mutex_unlock(&(XFS_QI_QOFFLOCK(mp)));
ASSERT(error != ENOENT);
return (error);
}
xfs_dqtrace_entry(dqp, "Q_SETQLIM: AFT DQGET");
xfs_trans_dqjoin(tp, dqp);
ddq = &dqp->q_core;
/*
* Make sure that hardlimits are >= soft limits before changing.
*/
hard = (newlim->d_fieldmask & FS_DQ_BHARD) ?
(xfs_qcnt_t) XFS_BB_TO_FSB(mp, newlim->d_blk_hardlimit) :
be64_to_cpu(ddq->d_blk_hardlimit);
soft = (newlim->d_fieldmask & FS_DQ_BSOFT) ?
(xfs_qcnt_t) XFS_BB_TO_FSB(mp, newlim->d_blk_softlimit) :
be64_to_cpu(ddq->d_blk_softlimit);
if (hard == 0 || hard >= soft) {
ddq->d_blk_hardlimit = cpu_to_be64(hard);
ddq->d_blk_softlimit = cpu_to_be64(soft);
if (id == 0) {
mp->m_quotainfo->qi_bhardlimit = hard;
mp->m_quotainfo->qi_bsoftlimit = soft;
}
} else {
qdprintk("blkhard %Ld < blksoft %Ld\n", hard, soft);
}
hard = (newlim->d_fieldmask & FS_DQ_RTBHARD) ?
(xfs_qcnt_t) XFS_BB_TO_FSB(mp, newlim->d_rtb_hardlimit) :
be64_to_cpu(ddq->d_rtb_hardlimit);
soft = (newlim->d_fieldmask & FS_DQ_RTBSOFT) ?
(xfs_qcnt_t) XFS_BB_TO_FSB(mp, newlim->d_rtb_softlimit) :
be64_to_cpu(ddq->d_rtb_softlimit);
if (hard == 0 || hard >= soft) {
ddq->d_rtb_hardlimit = cpu_to_be64(hard);
ddq->d_rtb_softlimit = cpu_to_be64(soft);
if (id == 0) {
mp->m_quotainfo->qi_rtbhardlimit = hard;
mp->m_quotainfo->qi_rtbsoftlimit = soft;
}
} else {
qdprintk("rtbhard %Ld < rtbsoft %Ld\n", hard, soft);
}
hard = (newlim->d_fieldmask & FS_DQ_IHARD) ?
(xfs_qcnt_t) newlim->d_ino_hardlimit :
be64_to_cpu(ddq->d_ino_hardlimit);
soft = (newlim->d_fieldmask & FS_DQ_ISOFT) ?
(xfs_qcnt_t) newlim->d_ino_softlimit :
be64_to_cpu(ddq->d_ino_softlimit);
if (hard == 0 || hard >= soft) {
ddq->d_ino_hardlimit = cpu_to_be64(hard);
ddq->d_ino_softlimit = cpu_to_be64(soft);
if (id == 0) {
mp->m_quotainfo->qi_ihardlimit = hard;
mp->m_quotainfo->qi_isoftlimit = soft;
}
} else {
qdprintk("ihard %Ld < isoft %Ld\n", hard, soft);
}
/*
* Update warnings counter(s) if requested
*/
if (newlim->d_fieldmask & FS_DQ_BWARNS)
ddq->d_bwarns = cpu_to_be16(newlim->d_bwarns);
if (newlim->d_fieldmask & FS_DQ_IWARNS)
ddq->d_iwarns = cpu_to_be16(newlim->d_iwarns);
if (newlim->d_fieldmask & FS_DQ_RTBWARNS)
ddq->d_rtbwarns = cpu_to_be16(newlim->d_rtbwarns);
if (id == 0) {
/*
* Timelimits for the super user set the relative time
* the other users can be over quota for this file system.
* If it is zero a default is used. Ditto for the default
* soft and hard limit values (already done, above), and
* for warnings.
*/
if (newlim->d_fieldmask & FS_DQ_BTIMER) {
mp->m_quotainfo->qi_btimelimit = newlim->d_btimer;
ddq->d_btimer = cpu_to_be32(newlim->d_btimer);
}
if (newlim->d_fieldmask & FS_DQ_ITIMER) {
mp->m_quotainfo->qi_itimelimit = newlim->d_itimer;
ddq->d_itimer = cpu_to_be32(newlim->d_itimer);
}
if (newlim->d_fieldmask & FS_DQ_RTBTIMER) {
mp->m_quotainfo->qi_rtbtimelimit = newlim->d_rtbtimer;
ddq->d_rtbtimer = cpu_to_be32(newlim->d_rtbtimer);
}
if (newlim->d_fieldmask & FS_DQ_BWARNS)
mp->m_quotainfo->qi_bwarnlimit = newlim->d_bwarns;
if (newlim->d_fieldmask & FS_DQ_IWARNS)
mp->m_quotainfo->qi_iwarnlimit = newlim->d_iwarns;
if (newlim->d_fieldmask & FS_DQ_RTBWARNS)
mp->m_quotainfo->qi_rtbwarnlimit = newlim->d_rtbwarns;
} else {
/*
* If the user is now over quota, start the timelimit.
* The user will not be 'warned'.
* Note that we keep the timers ticking, whether enforcement
* is on or off. We don't really want to bother with iterating
* over all ondisk dquots and turning the timers on/off.
*/
xfs_qm_adjust_dqtimers(mp, ddq);
}
dqp->dq_flags |= XFS_DQ_DIRTY;
xfs_trans_log_dquot(tp, dqp);
xfs_dqtrace_entry(dqp, "Q_SETQLIM: COMMIT");
xfs_trans_commit(tp, 0, NULL);
xfs_qm_dqprint(dqp);
xfs_qm_dqrele(dqp);
mutex_unlock(&(XFS_QI_QOFFLOCK(mp)));
return (0);
}
STATIC long
xfs_file_fallocate(
struct file *file,
int mode,
loff_t offset,
loff_t len)
{
struct inode *inode = file_inode(file);
struct xfs_inode *ip = XFS_I(inode);
struct xfs_trans *tp;
long error;
loff_t new_size = 0;
if (!S_ISREG(inode->i_mode))
return -EINVAL;
if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE |
FALLOC_FL_COLLAPSE_RANGE | FALLOC_FL_ZERO_RANGE))
return -EOPNOTSUPP;
xfs_ilock(ip, XFS_IOLOCK_EXCL);
if (mode & FALLOC_FL_PUNCH_HOLE) {
error = xfs_free_file_space(ip, offset, len);
if (error)
goto out_unlock;
} else if (mode & FALLOC_FL_COLLAPSE_RANGE) {
unsigned blksize_mask = (1 << inode->i_blkbits) - 1;
if (offset & blksize_mask || len & blksize_mask) {
error = -EINVAL;
goto out_unlock;
}
/*
* There is no need to overlap collapse range with EOF,
* in which case it is effectively a truncate operation
*/
if (offset + len >= i_size_read(inode)) {
error = -EINVAL;
goto out_unlock;
}
new_size = i_size_read(inode) - len;
error = xfs_collapse_file_space(ip, offset, len);
if (error)
goto out_unlock;
} else {
if (!(mode & FALLOC_FL_KEEP_SIZE) &&
offset + len > i_size_read(inode)) {
new_size = offset + len;
error = inode_newsize_ok(inode, new_size);
if (error)
goto out_unlock;
}
if (mode & FALLOC_FL_ZERO_RANGE)
error = xfs_zero_file_space(ip, offset, len);
else
error = xfs_alloc_file_space(ip, offset, len,
XFS_BMAPI_PREALLOC);
if (error)
goto out_unlock;
}
tp = xfs_trans_alloc(ip->i_mount, XFS_TRANS_WRITEID);
error = xfs_trans_reserve(tp, &M_RES(ip->i_mount)->tr_writeid, 0, 0);
if (error) {
xfs_trans_cancel(tp, 0);
goto out_unlock;
}
xfs_ilock(ip, XFS_ILOCK_EXCL);
xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
ip->i_d.di_mode &= ~S_ISUID;
if (ip->i_d.di_mode & S_IXGRP)
ip->i_d.di_mode &= ~S_ISGID;
if (!(mode & (FALLOC_FL_PUNCH_HOLE | FALLOC_FL_COLLAPSE_RANGE)))
ip->i_d.di_flags |= XFS_DIFLAG_PREALLOC;
xfs_trans_ichgtime(tp, ip, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
if (file->f_flags & O_DSYNC)
xfs_trans_set_sync(tp);
error = xfs_trans_commit(tp, 0);
if (error)
goto out_unlock;
/* Change file size if needed */
if (new_size) {
struct iattr iattr;
iattr.ia_valid = ATTR_SIZE;
iattr.ia_size = new_size;
error = xfs_setattr_size(ip, &iattr);
}
out_unlock:
xfs_iunlock(ip, XFS_IOLOCK_EXCL);
return error;
}
int
xfs_setattr_nonsize(
struct xfs_inode *ip,
struct iattr *iattr,
int flags)
{
xfs_mount_t *mp = ip->i_mount;
struct inode *inode = VFS_I(ip);
int mask = iattr->ia_valid;
xfs_trans_t *tp;
int error;
kuid_t uid = GLOBAL_ROOT_UID, iuid = GLOBAL_ROOT_UID;
kgid_t gid = GLOBAL_ROOT_GID, igid = GLOBAL_ROOT_GID;
struct xfs_dquot *udqp = NULL, *gdqp = NULL;
struct xfs_dquot *olddquot1 = NULL, *olddquot2 = NULL;
trace_xfs_setattr(ip);
/* If acls are being inherited, we already have this checked */
if (!(flags & XFS_ATTR_NOACL)) {
if (mp->m_flags & XFS_MOUNT_RDONLY)
return XFS_ERROR(EROFS);
if (XFS_FORCED_SHUTDOWN(mp))
return XFS_ERROR(EIO);
error = -inode_change_ok(inode, iattr);
if (error)
return XFS_ERROR(error);
}
ASSERT((mask & ATTR_SIZE) == 0);
/*
* If disk quotas is on, we make sure that the dquots do exist on disk,
* before we start any other transactions. Trying to do this later
* is messy. We don't care to take a readlock to look at the ids
* in inode here, because we can't hold it across the trans_reserve.
* If the IDs do change before we take the ilock, we're covered
* because the i_*dquot fields will get updated anyway.
*/
if (XFS_IS_QUOTA_ON(mp) && (mask & (ATTR_UID|ATTR_GID))) {
uint qflags = 0;
if ((mask & ATTR_UID) && XFS_IS_UQUOTA_ON(mp)) {
uid = iattr->ia_uid;
qflags |= XFS_QMOPT_UQUOTA;
} else {
uid = inode->i_uid;
}
if ((mask & ATTR_GID) && XFS_IS_GQUOTA_ON(mp)) {
gid = iattr->ia_gid;
qflags |= XFS_QMOPT_GQUOTA;
} else {
gid = inode->i_gid;
}
/*
* We take a reference when we initialize udqp and gdqp,
* so it is important that we never blindly double trip on
* the same variable. See xfs_create() for an example.
*/
ASSERT(udqp == NULL);
ASSERT(gdqp == NULL);
error = xfs_qm_vop_dqalloc(ip, xfs_kuid_to_uid(uid),
xfs_kgid_to_gid(gid),
xfs_get_projid(ip),
qflags, &udqp, &gdqp, NULL);
if (error)
return error;
}
tp = xfs_trans_alloc(mp, XFS_TRANS_SETATTR_NOT_SIZE);
error = xfs_trans_reserve(tp, &M_RES(mp)->tr_ichange, 0, 0);
if (error)
goto out_dqrele;
xfs_ilock(ip, XFS_ILOCK_EXCL);
/*
* Change file ownership. Must be the owner or privileged.
*/
if (mask & (ATTR_UID|ATTR_GID)) {
/*
* These IDs could have changed since we last looked at them.
* But, we're assured that if the ownership did change
* while we didn't have the inode locked, inode's dquot(s)
* would have changed also.
*/
iuid = inode->i_uid;
igid = inode->i_gid;
gid = (mask & ATTR_GID) ? iattr->ia_gid : igid;
uid = (mask & ATTR_UID) ? iattr->ia_uid : iuid;
/*
* Do a quota reservation only if uid/gid is actually
* going to change.
*/
if (XFS_IS_QUOTA_RUNNING(mp) &&
((XFS_IS_UQUOTA_ON(mp) && !uid_eq(iuid, uid)) ||
(XFS_IS_GQUOTA_ON(mp) && !gid_eq(igid, gid)))) {
ASSERT(tp);
error = xfs_qm_vop_chown_reserve(tp, ip, udqp, gdqp,
NULL, capable(CAP_FOWNER) ?
XFS_QMOPT_FORCE_RES : 0);
if (error) /* out of quota */
goto out_trans_cancel;
}
}
xfs_trans_ijoin(tp, ip, 0);
/*
* Change file ownership. Must be the owner or privileged.
*/
if (mask & (ATTR_UID|ATTR_GID)) {
/*
* CAP_FSETID overrides the following restrictions:
*
* The set-user-ID and set-group-ID bits of a file will be
* cleared upon successful return from chown()
*/
if ((ip->i_d.di_mode & (S_ISUID|S_ISGID)) &&
!capable(CAP_FSETID))
ip->i_d.di_mode &= ~(S_ISUID|S_ISGID);
/*
* Change the ownerships and register quota modifications
* in the transaction.
*/
if (!uid_eq(iuid, uid)) {
if (XFS_IS_QUOTA_RUNNING(mp) && XFS_IS_UQUOTA_ON(mp)) {
ASSERT(mask & ATTR_UID);
ASSERT(udqp);
olddquot1 = xfs_qm_vop_chown(tp, ip,
&ip->i_udquot, udqp);
}
ip->i_d.di_uid = xfs_kuid_to_uid(uid);
inode->i_uid = uid;
}
if (!gid_eq(igid, gid)) {
if (XFS_IS_QUOTA_RUNNING(mp) && XFS_IS_GQUOTA_ON(mp)) {
ASSERT(xfs_sb_version_has_pquotino(&mp->m_sb) ||
!XFS_IS_PQUOTA_ON(mp));
ASSERT(mask & ATTR_GID);
ASSERT(gdqp);
olddquot2 = xfs_qm_vop_chown(tp, ip,
&ip->i_gdquot, gdqp);
}
ip->i_d.di_gid = xfs_kgid_to_gid(gid);
inode->i_gid = gid;
}
}
if (mask & ATTR_MODE)
xfs_setattr_mode(ip, iattr);
if (mask & (ATTR_ATIME|ATTR_CTIME|ATTR_MTIME))
xfs_setattr_time(ip, iattr);
xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
XFS_STATS_INC(xs_ig_attrchg);
if (mp->m_flags & XFS_MOUNT_WSYNC)
xfs_trans_set_sync(tp);
error = xfs_trans_commit(tp, 0);
xfs_iunlock(ip, XFS_ILOCK_EXCL);
/*
* Release any dquot(s) the inode had kept before chown.
*/
xfs_qm_dqrele(olddquot1);
xfs_qm_dqrele(olddquot2);
xfs_qm_dqrele(udqp);
xfs_qm_dqrele(gdqp);
if (error)
return XFS_ERROR(error);
/*
* XXX(hch): Updating the ACL entries is not atomic vs the i_mode
* update. We could avoid this with linked transactions
* and passing down the transaction pointer all the way
* to attr_set. No previous user of the generic
* Posix ACL code seems to care about this issue either.
*/
if ((mask & ATTR_MODE) && !(flags & XFS_ATTR_NOACL)) {
error = -posix_acl_chmod(inode, inode->i_mode);
if (error)
return XFS_ERROR(error);
}
return 0;
out_trans_cancel:
xfs_trans_cancel(tp, 0);
xfs_iunlock(ip, XFS_ILOCK_EXCL);
out_dqrele:
xfs_qm_dqrele(udqp);
xfs_qm_dqrele(gdqp);
return error;
}
/*
* Truncate file. Must have write permission and not be a directory.
*/
int
xfs_setattr_size(
struct xfs_inode *ip,
struct iattr *iattr)
{
struct xfs_mount *mp = ip->i_mount;
struct inode *inode = VFS_I(ip);
xfs_off_t oldsize, newsize;
struct xfs_trans *tp;
int error;
uint lock_flags = 0;
uint commit_flags = 0;
bool did_zeroing = false;
trace_xfs_setattr(ip);
if (mp->m_flags & XFS_MOUNT_RDONLY)
return -EROFS;
if (XFS_FORCED_SHUTDOWN(mp))
return -EIO;
error = inode_change_ok(inode, iattr);
if (error)
return error;
ASSERT(xfs_isilocked(ip, XFS_IOLOCK_EXCL));
ASSERT(xfs_isilocked(ip, XFS_MMAPLOCK_EXCL));
ASSERT(S_ISREG(ip->i_d.di_mode));
ASSERT((iattr->ia_valid & (ATTR_UID|ATTR_GID|ATTR_ATIME|ATTR_ATIME_SET|
ATTR_MTIME_SET|ATTR_KILL_PRIV|ATTR_TIMES_SET)) == 0);
oldsize = inode->i_size;
newsize = iattr->ia_size;
/*
* Short circuit the truncate case for zero length files.
*/
if (newsize == 0 && oldsize == 0 && ip->i_d.di_nextents == 0) {
if (!(iattr->ia_valid & (ATTR_CTIME|ATTR_MTIME)))
return 0;
/*
* Use the regular setattr path to update the timestamps.
*/
iattr->ia_valid &= ~ATTR_SIZE;
return xfs_setattr_nonsize(ip, iattr, 0);
}
/*
* Make sure that the dquots are attached to the inode.
*/
error = xfs_qm_dqattach(ip, 0);
if (error)
return error;
/*
* File data changes must be complete before we start the transaction to
* modify the inode. This needs to be done before joining the inode to
* the transaction because the inode cannot be unlocked once it is a
* part of the transaction.
*
* Start with zeroing any data block beyond EOF that we may expose on
* file extension.
*/
if (newsize > oldsize) {
error = xfs_zero_eof(ip, newsize, oldsize, &did_zeroing);
if (error)
return error;
}
/*
* We are going to log the inode size change in this transaction so
* any previous writes that are beyond the on disk EOF and the new
* EOF that have not been written out need to be written here. If we
* do not write the data out, we expose ourselves to the null files
* problem. Note that this includes any block zeroing we did above;
* otherwise those blocks may not be zeroed after a crash.
*/
if (newsize > ip->i_d.di_size &&
(oldsize != ip->i_d.di_size || did_zeroing)) {
error = filemap_write_and_wait_range(VFS_I(ip)->i_mapping,
ip->i_d.di_size, newsize);
if (error)
return error;
}
/* Now wait for all direct I/O to complete. */
inode_dio_wait(inode);
/*
* We've already locked out new page faults, so now we can safely remove
* pages from the page cache knowing they won't get refaulted until we
* drop the XFS_MMAP_EXCL lock after the extent manipulations are
* complete. The truncate_setsize() call also cleans partial EOF page
* PTEs on extending truncates and hence ensures sub-page block size
* filesystems are correctly handled, too.
*
* We have to do all the page cache truncate work outside the
* transaction context as the "lock" order is page lock->log space
* reservation as defined by extent allocation in the writeback path.
* Hence a truncate can fail with ENOMEM from xfs_trans_reserve(), but
* having already truncated the in-memory version of the file (i.e. made
* user visible changes). There's not much we can do about this, except
* to hope that the caller sees ENOMEM and retries the truncate
* operation.
*/
error = block_truncate_page(inode->i_mapping, newsize, xfs_get_blocks);
if (error)
return error;
truncate_setsize(inode, newsize);
tp = xfs_trans_alloc(mp, XFS_TRANS_SETATTR_SIZE);
error = xfs_trans_reserve(tp, &M_RES(mp)->tr_itruncate, 0, 0);
if (error)
goto out_trans_cancel;
commit_flags = XFS_TRANS_RELEASE_LOG_RES;
lock_flags |= XFS_ILOCK_EXCL;
xfs_ilock(ip, XFS_ILOCK_EXCL);
xfs_trans_ijoin(tp, ip, 0);
/*
* Only change the c/mtime if we are changing the size or we are
* explicitly asked to change it. This handles the semantic difference
* between truncate() and ftruncate() as implemented in the VFS.
*
* The regular truncate() case without ATTR_CTIME and ATTR_MTIME is a
* special case where we need to update the times despite not having
* these flags set. For all other operations the VFS set these flags
* explicitly if it wants a timestamp update.
*/
if (newsize != oldsize &&
!(iattr->ia_valid & (ATTR_CTIME | ATTR_MTIME))) {
iattr->ia_ctime = iattr->ia_mtime =
current_fs_time(inode->i_sb);
iattr->ia_valid |= ATTR_CTIME | ATTR_MTIME;
}
/*
* The first thing we do is set the size to new_size permanently on
* disk. This way we don't have to worry about anyone ever being able
* to look at the data being freed even in the face of a crash.
* What we're getting around here is the case where we free a block, it
* is allocated to another file, it is written to, and then we crash.
* If the new data gets written to the file but the log buffers
* containing the free and reallocation don't, then we'd end up with
* garbage in the blocks being freed. As long as we make the new size
* permanent before actually freeing any blocks it doesn't matter if
* they get written to.
*/
ip->i_d.di_size = newsize;
xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
if (newsize <= oldsize) {
error = xfs_itruncate_extents(&tp, ip, XFS_DATA_FORK, newsize);
if (error)
goto out_trans_abort;
/*
* Truncated "down", so we're removing references to old data
* here - if we delay flushing for a long time, we expose
* ourselves unduly to the notorious NULL files problem. So,
* we mark this inode and flush it when the file is closed,
* and do not wait the usual (long) time for writeout.
*/
xfs_iflags_set(ip, XFS_ITRUNCATED);
/* A truncate down always removes post-EOF blocks. */
xfs_inode_clear_eofblocks_tag(ip);
}
if (iattr->ia_valid & ATTR_MODE)
xfs_setattr_mode(ip, iattr);
if (iattr->ia_valid & (ATTR_ATIME|ATTR_CTIME|ATTR_MTIME))
xfs_setattr_time(ip, iattr);
xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
XFS_STATS_INC(xs_ig_attrchg);
if (mp->m_flags & XFS_MOUNT_WSYNC)
xfs_trans_set_sync(tp);
error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
out_unlock:
if (lock_flags)
xfs_iunlock(ip, lock_flags);
return error;
out_trans_abort:
commit_flags |= XFS_TRANS_ABORT;
out_trans_cancel:
xfs_trans_cancel(tp, commit_flags);
goto out_unlock;
}
/*
* Read in the ondisk dquot using dqtobp() then copy it to an incore version,
* and release the buffer immediately.
*
* If XFS_QMOPT_DQALLOC is set, allocate a dquot on disk if it needed.
*/
int
xfs_qm_dqread(
struct xfs_mount *mp,
xfs_dqid_t id,
uint type,
uint flags,
struct xfs_dquot **O_dqpp)
{
struct xfs_dquot *dqp;
struct xfs_disk_dquot *ddqp;
struct xfs_buf *bp;
struct xfs_trans *tp = NULL;
int error;
dqp = kmem_zone_zalloc(xfs_qm_dqzone, KM_SLEEP);
dqp->dq_flags = type;
dqp->q_core.d_id = cpu_to_be32(id);
dqp->q_mount = mp;
INIT_LIST_HEAD(&dqp->q_lru);
mutex_init(&dqp->q_qlock);
init_waitqueue_head(&dqp->q_pinwait);
/*
* Because we want to use a counting completion, complete
* the flush completion once to allow a single access to
* the flush completion without blocking.
*/
init_completion(&dqp->q_flush);
complete(&dqp->q_flush);
/*
* Make sure group quotas have a different lock class than user
* quotas.
*/
switch (type) {
case XFS_DQ_USER:
/* uses the default lock class */
break;
case XFS_DQ_GROUP:
lockdep_set_class(&dqp->q_qlock, &xfs_dquot_group_class);
break;
case XFS_DQ_PROJ:
lockdep_set_class(&dqp->q_qlock, &xfs_dquot_project_class);
break;
default:
ASSERT(0);
break;
}
XFS_STATS_INC(mp, xs_qm_dquot);
trace_xfs_dqread(dqp);
if (flags & XFS_QMOPT_DQALLOC) {
error = xfs_trans_alloc(mp, &M_RES(mp)->tr_qm_dqalloc,
XFS_QM_DQALLOC_SPACE_RES(mp), 0, 0, &tp);
if (error)
goto error0;
}
/*
* get a pointer to the on-disk dquot and the buffer containing it
* dqp already knows its own type (GROUP/USER).
*/
error = xfs_qm_dqtobp(&tp, dqp, &ddqp, &bp, flags);
if (error) {
/*
* This can happen if quotas got turned off (ESRCH),
* or if the dquot didn't exist on disk and we ask to
* allocate (ENOENT).
*/
trace_xfs_dqread_fail(dqp);
goto error1;
}
/* copy everything from disk dquot to the incore dquot */
memcpy(&dqp->q_core, ddqp, sizeof(xfs_disk_dquot_t));
xfs_qm_dquot_logitem_init(dqp);
/*
* Reservation counters are defined as reservation plus current usage
* to avoid having to add every time.
*/
dqp->q_res_bcount = be64_to_cpu(ddqp->d_bcount);
dqp->q_res_icount = be64_to_cpu(ddqp->d_icount);
dqp->q_res_rtbcount = be64_to_cpu(ddqp->d_rtbcount);
/* initialize the dquot speculative prealloc thresholds */
xfs_dquot_set_prealloc_limits(dqp);
/* Mark the buf so that this will stay incore a little longer */
xfs_buf_set_ref(bp, XFS_DQUOT_REF);
/*
* We got the buffer with a xfs_trans_read_buf() (in dqtobp())
* So we need to release with xfs_trans_brelse().
* The strategy here is identical to that of inodes; we lock
* the dquot in xfs_qm_dqget() before making it accessible to
* others. This is because dquots, like inodes, need a good level of
* concurrency, and we don't want to take locks on the entire buffers
* for dquot accesses.
* Note also that the dquot buffer may even be dirty at this point, if
* this particular dquot was repaired. We still aren't afraid to
* brelse it because we have the changes incore.
*/
ASSERT(xfs_buf_islocked(bp));
xfs_trans_brelse(tp, bp);
if (tp) {
error = xfs_trans_commit(tp);
if (error)
goto error0;
}
*O_dqpp = dqp;
return error;
error1:
if (tp)
xfs_trans_cancel(tp);
error0:
xfs_qm_dqdestroy(dqp);
*O_dqpp = NULL;
return error;
}
/*
* Pass in a delayed allocate extent, convert it to real extents;
* return to the caller the extent we create which maps on top of
* the originating callers request.
*
* Called without a lock on the inode.
*/
int
xfs_iomap_write_allocate(
xfs_inode_t *ip,
xfs_off_t offset,
size_t count,
xfs_bmbt_irec_t *map,
int *retmap)
{
xfs_mount_t *mp = ip->i_mount;
xfs_iocore_t *io = &ip->i_iocore;
xfs_fileoff_t offset_fsb, last_block;
xfs_fileoff_t end_fsb, map_start_fsb;
xfs_fsblock_t first_block;
xfs_bmap_free_t free_list;
xfs_filblks_t count_fsb;
xfs_bmbt_irec_t imap[XFS_STRAT_WRITE_IMAPS];
xfs_trans_t *tp;
int i, nimaps, committed;
int error = 0;
int nres;
*retmap = 0;
/*
* Make sure that the dquots are there.
*/
if ((error = XFS_QM_DQATTACH(mp, ip, 0)))
return XFS_ERROR(error);
offset_fsb = XFS_B_TO_FSBT(mp, offset);
count_fsb = map->br_blockcount;
map_start_fsb = map->br_startoff;
XFS_STATS_ADD(xs_xstrat_bytes, XFS_FSB_TO_B(mp, count_fsb));
while (count_fsb != 0) {
/*
* Set up a transaction with which to allocate the
* backing store for the file. Do allocations in a
* loop until we get some space in the range we are
* interested in. The other space that might be allocated
* is in the delayed allocation extent on which we sit
* but before our buffer starts.
*/
nimaps = 0;
while (nimaps == 0) {
tp = xfs_trans_alloc(mp, XFS_TRANS_STRAT_WRITE);
nres = XFS_EXTENTADD_SPACE_RES(mp, XFS_DATA_FORK);
error = xfs_trans_reserve(tp, nres,
XFS_WRITE_LOG_RES(mp),
0, XFS_TRANS_PERM_LOG_RES,
XFS_WRITE_LOG_COUNT);
if (error == ENOSPC) {
error = xfs_trans_reserve(tp, 0,
XFS_WRITE_LOG_RES(mp),
0,
XFS_TRANS_PERM_LOG_RES,
XFS_WRITE_LOG_COUNT);
}
if (error) {
xfs_trans_cancel(tp, 0);
return XFS_ERROR(error);
}
xfs_ilock(ip, XFS_ILOCK_EXCL);
xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
xfs_trans_ihold(tp, ip);
XFS_BMAP_INIT(&free_list, &first_block);
nimaps = XFS_STRAT_WRITE_IMAPS;
/*
* Ensure we don't go beyond eof - it is possible
* the extents changed since we did the read call,
* we dropped the ilock in the interim.
*/
end_fsb = XFS_B_TO_FSB(mp, ip->i_d.di_size);
xfs_bmap_last_offset(NULL, ip, &last_block,
XFS_DATA_FORK);
last_block = XFS_FILEOFF_MAX(last_block, end_fsb);
if ((map_start_fsb + count_fsb) > last_block) {
count_fsb = last_block - map_start_fsb;
if (count_fsb == 0) {
error = EAGAIN;
goto trans_cancel;
}
}
/* Go get the actual blocks */
error = xfs_bmapi(tp, ip, map_start_fsb, count_fsb,
XFS_BMAPI_WRITE, &first_block, 1,
imap, &nimaps, &free_list);
if (error)
goto trans_cancel;
error = xfs_bmap_finish(&tp, &free_list,
first_block, &committed);
if (error)
goto trans_cancel;
error = xfs_trans_commit(tp,
XFS_TRANS_RELEASE_LOG_RES, NULL);
if (error)
goto error0;
xfs_iunlock(ip, XFS_ILOCK_EXCL);
}
/*
* See if we were able to allocate an extent that
* covers at least part of the callers request
*/
for (i = 0; i < nimaps; i++) {
if ( !(io->io_flags & XFS_IOCORE_RT) &&
!imap[i].br_startblock) {
cmn_err(CE_PANIC,"Access to block zero: "
"fs <%s> inode: %lld "
"start_block : %llx start_off : %llx "
"blkcnt : %llx extent-state : %x \n",
(ip->i_mount)->m_fsname,
(long long)ip->i_ino,
imap[i].br_startblock,
imap[i].br_startoff,
imap[i].br_blockcount,imap[i].br_state);
}
if ((offset_fsb >= imap[i].br_startoff) &&
(offset_fsb < (imap[i].br_startoff +
imap[i].br_blockcount))) {
*map = imap[i];
*retmap = 1;
XFS_STATS_INC(xs_xstrat_quick);
return 0;
}
count_fsb -= imap[i].br_blockcount;
}
/* So far we have not mapped the requested part of the
* file, just surrounding data, try again.
*/
nimaps--;
map_start_fsb = imap[nimaps].br_startoff +
imap[nimaps].br_blockcount;
}
trans_cancel:
xfs_bmap_cancel(&free_list);
xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES | XFS_TRANS_ABORT);
error0:
xfs_iunlock(ip, XFS_ILOCK_EXCL);
return XFS_ERROR(error);
}
