int
xfs_qm_scall_quotaoff(
xfs_mount_t *mp,
uint flags)
{
struct xfs_quotainfo *q = mp->m_quotainfo;
uint dqtype;
int error;
uint inactivate_flags;
xfs_qoff_logitem_t *qoffstart;
if ((mp->m_qflags & flags) == 0)
return XFS_ERROR(EEXIST);
error = 0;
flags &= (XFS_ALL_QUOTA_ACCT | XFS_ALL_QUOTA_ENFD);
ASSERT(q);
mutex_lock(&q->qi_quotaofflock);
if ((flags & XFS_ALL_QUOTA_ACCT) == 0) {
mp->m_qflags &= ~(flags);
spin_lock(&mp->m_sb_lock);
mp->m_sb.sb_qflags = mp->m_qflags;
spin_unlock(&mp->m_sb_lock);
mutex_unlock(&q->qi_quotaofflock);
error = xfs_qm_write_sb_changes(mp, XFS_SB_QFLAGS);
return (error);
}
dqtype = 0;
inactivate_flags = 0;
if (flags & XFS_UQUOTA_ACCT) {
dqtype |= XFS_QMOPT_UQUOTA;
flags |= (XFS_UQUOTA_CHKD | XFS_UQUOTA_ENFD);
inactivate_flags |= XFS_UQUOTA_ACTIVE;
}
if (flags & XFS_GQUOTA_ACCT) {
dqtype |= XFS_QMOPT_GQUOTA;
flags |= (XFS_OQUOTA_CHKD | XFS_OQUOTA_ENFD);
inactivate_flags |= XFS_GQUOTA_ACTIVE;
} else if (flags & XFS_PQUOTA_ACCT) {
dqtype |= XFS_QMOPT_PQUOTA;
flags |= (XFS_OQUOTA_CHKD | XFS_OQUOTA_ENFD);
inactivate_flags |= XFS_PQUOTA_ACTIVE;
}
if ((mp->m_qflags & flags) == 0)
goto out_unlock;
error = xfs_qm_log_quotaoff(mp, &qoffstart, flags);
if (error)
goto out_unlock;
mp->m_qflags &= ~inactivate_flags;
xfs_qm_dqrele_all_inodes(mp, flags);
mp->m_qflags &= ~flags;
xfs_qm_dqpurge_all(mp, dqtype);
/*
* Transactions that had started before ACTIVE state bit was cleared
* could have logged many dquots, so they'd have higher LSNs than
* the first QUOTAOFF log record does. If we happen to crash when
* the tail of the log has gone past the QUOTAOFF record, but
* before the last dquot modification, those dquots __will__
* recover, and that's not good.
*
* So, we have QUOTAOFF start and end logitems; the start
* logitem won't get overwritten until the end logitem appears...
*/
error = xfs_qm_log_quotaoff_end(mp, qoffstart, flags);
if (error) {
xfs_force_shutdown(mp, SHUTDOWN_CORRUPT_INCORE);
goto out_unlock;
}
if (((flags & XFS_MOUNT_QUOTA_ALL) == XFS_MOUNT_QUOTA_SET1) ||
((flags & XFS_MOUNT_QUOTA_ALL) == XFS_MOUNT_QUOTA_SET2)) {
mutex_unlock(&q->qi_quotaofflock);
xfs_qm_destroy_quotainfo(mp);
return (0);
}
if ((dqtype & XFS_QMOPT_UQUOTA) && q->qi_uquotaip) {
IRELE(q->qi_uquotaip);
q->qi_uquotaip = NULL;
}
if ((dqtype & (XFS_QMOPT_GQUOTA|XFS_QMOPT_PQUOTA)) && q->qi_gquotaip) {
IRELE(q->qi_gquotaip);
q->qi_gquotaip = NULL;
}
out_unlock:
mutex_unlock(&q->qi_quotaofflock);
return 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;
}
ssize_t /* bytes written, or (-) error */
xfs_write(
struct xfs_inode *xip,
struct kiocb *iocb,
const struct iovec *iovp,
unsigned int nsegs,
loff_t *offset,
int ioflags)
{
struct file *file = iocb->ki_filp;
struct address_space *mapping = file->f_mapping;
struct inode *inode = mapping->host;
unsigned long segs = nsegs;
xfs_mount_t *mp;
ssize_t ret = 0, error = 0;
xfs_fsize_t isize, new_size;
int iolock;
int eventsent = 0;
size_t ocount = 0, count;
loff_t pos;
int need_i_mutex;
XFS_STATS_INC(xs_write_calls);
error = generic_segment_checks(iovp, &segs, &ocount, VERIFY_READ);
if (error)
return error;
count = ocount;
pos = *offset;
if (count == 0)
return 0;
mp = xip->i_mount;
xfs_wait_for_freeze(mp, SB_FREEZE_WRITE);
if (XFS_FORCED_SHUTDOWN(mp))
return -EIO;
relock:
if (ioflags & IO_ISDIRECT) {
iolock = XFS_IOLOCK_SHARED;
need_i_mutex = 0;
} else {
iolock = XFS_IOLOCK_EXCL;
need_i_mutex = 1;
mutex_lock(&inode->i_mutex);
}
xfs_ilock(xip, XFS_ILOCK_EXCL|iolock);
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_mutex;
}
if ((DM_EVENT_ENABLED(xip, DM_EVENT_WRITE) &&
!(ioflags & IO_INVIS) && !eventsent)) {
int dmflags = FILP_DELAY_FLAG(file);
if (need_i_mutex)
dmflags |= DM_FLAGS_IMUX;
xfs_iunlock(xip, XFS_ILOCK_EXCL);
error = XFS_SEND_DATA(xip->i_mount, DM_EVENT_WRITE, xip,
pos, count, dmflags, &iolock);
if (error) {
goto out_unlock_internal;
}
xfs_ilock(xip, XFS_ILOCK_EXCL);
eventsent = 1;
/*
* The iolock was dropped and reacquired 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) && pos != xip->i_size)
goto start;
}
if (ioflags & IO_ISDIRECT) {
xfs_buftarg_t *target =
XFS_IS_REALTIME_INODE(xip) ?
mp->m_rtdev_targp : mp->m_ddev_targp;
if ((pos & target->bt_smask) || (count & target->bt_smask)) {
xfs_iunlock(xip, XFS_ILOCK_EXCL|iolock);
return XFS_ERROR(-EINVAL);
}
if (!need_i_mutex && (mapping->nrpages || pos > xip->i_size)) {
xfs_iunlock(xip, XFS_ILOCK_EXCL|iolock);
iolock = XFS_IOLOCK_EXCL;
need_i_mutex = 1;
mutex_lock(&inode->i_mutex);
xfs_ilock(xip, XFS_ILOCK_EXCL|iolock);
goto start;
}
}
new_size = pos + count;
if (new_size > xip->i_size)
xip->i_new_size = new_size;
if (likely(!(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 (pos > xip->i_size) {
error = xfs_zero_eof(xip, pos, xip->i_size);
if (error) {
xfs_iunlock(xip, XFS_ILOCK_EXCL);
goto out_unlock_internal;
}
}
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 = -file_remove_suid(file);
if (unlikely(error)) {
goto out_unlock_internal;
}
}
/* We can write back this queue in page reclaim */
current->backing_dev_info = mapping->backing_dev_info;
if ((ioflags & IO_ISDIRECT)) {
if (mapping->nrpages) {
WARN_ON(need_i_mutex == 0);
xfs_inval_cached_trace(xip, pos, -1,
(pos & PAGE_CACHE_MASK), -1);
error = xfs_flushinval_pages(xip,
(pos & PAGE_CACHE_MASK),
-1, FI_REMAPF_LOCKED);
if (error)
goto out_unlock_internal;
}
if (need_i_mutex) {
/* demote the lock now the cached pages are gone */
xfs_ilock_demote(xip, XFS_IOLOCK_EXCL);
mutex_unlock(&inode->i_mutex);
iolock = XFS_IOLOCK_SHARED;
need_i_mutex = 0;
}
xfs_rw_enter_trace(XFS_DIOWR_ENTER, xip, (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;
ioflags &= ~IO_ISDIRECT;
xfs_iunlock(xip, iolock);
goto relock;
}
} else {
xfs_rw_enter_trace(XFS_WRITE_ENTER, xip, (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);
isize = i_size_read(inode);
if (unlikely(ret < 0 && ret != -EFAULT && *offset > isize))
*offset = isize;
if (*offset > xip->i_size) {
xfs_ilock(xip, XFS_ILOCK_EXCL);
if (*offset > xip->i_size)
xip->i_size = *offset;
xfs_iunlock(xip, XFS_ILOCK_EXCL);
}
if (ret == -ENOSPC &&
DM_EVENT_ENABLED(xip, DM_EVENT_NOSPACE) && !(ioflags & IO_INVIS)) {
xfs_iunlock(xip, iolock);
if (need_i_mutex)
mutex_unlock(&inode->i_mutex);
error = XFS_SEND_NAMESP(xip->i_mount, DM_EVENT_NOSPACE, xip,
DM_RIGHT_NULL, xip, DM_RIGHT_NULL, NULL, NULL,
0, 0, 0); /* Delay flag intentionally unused */
if (need_i_mutex)
mutex_lock(&inode->i_mutex);
xfs_ilock(xip, iolock);
if (error)
goto out_unlock_internal;
goto start;
}
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)) {
int error2;
xfs_iunlock(xip, iolock);
if (need_i_mutex)
mutex_unlock(&inode->i_mutex);
error2 = sync_page_range(inode, mapping, pos, ret);
if (!error)
error = error2;
if (need_i_mutex)
mutex_lock(&inode->i_mutex);
xfs_ilock(xip, iolock);
error2 = xfs_write_sync_logforce(mp, xip);
if (!error)
error = error2;
}
out_unlock_internal:
if (xip->i_new_size) {
xfs_ilock(xip, XFS_ILOCK_EXCL);
xip->i_new_size = 0;
/*
* If this was a direct or synchronous I/O that failed (such
* as ENOSPC) then part of the I/O may have been written to
* disk before the error occured. In this case the on-disk
* file size may have been adjusted beyond the in-memory file
* size and now needs to be truncated back.
*/
if (xip->i_d.di_size > xip->i_size)
xip->i_d.di_size = xip->i_size;
xfs_iunlock(xip, XFS_ILOCK_EXCL);
}
xfs_iunlock(xip, iolock);
out_unlock_mutex:
if (need_i_mutex)
mutex_unlock(&inode->i_mutex);
return -error;
}
/*
* This function fills in xfs_mount_t fields based on mount args.
* Note: the superblock has _not_ yet been read in.
*
* Note that this function leaks the various device name allocations on
* failure. The caller takes care of them.
*/
STATIC int
xfs_parseargs(
struct xfs_mount *mp,
char *options)
{
struct super_block *sb = mp->m_super;
char *this_char, *value, *eov;
int dsunit = 0;
int dswidth = 0;
int iosize = 0;
__uint8_t iosizelog = 0;
/*
* set up the mount name first so all the errors will refer to the
* correct device.
*/
mp->m_fsname = kstrndup(sb->s_id, MAXNAMELEN, GFP_KERNEL);
if (!mp->m_fsname)
return ENOMEM;
mp->m_fsname_len = strlen(mp->m_fsname) + 1;
/*
* Copy binary VFS mount flags we are interested in.
*/
if (sb->s_flags & MS_RDONLY)
mp->m_flags |= XFS_MOUNT_RDONLY;
if (sb->s_flags & MS_DIRSYNC)
mp->m_flags |= XFS_MOUNT_DIRSYNC;
if (sb->s_flags & MS_SYNCHRONOUS)
mp->m_flags |= XFS_MOUNT_WSYNC;
/*
* Set some default flags that could be cleared by the mount option
* parsing.
*/
mp->m_flags |= XFS_MOUNT_BARRIER;
mp->m_flags |= XFS_MOUNT_COMPAT_IOSIZE;
mp->m_flags |= XFS_MOUNT_SMALL_INUMS;
/*
* These can be overridden by the mount option parsing.
*/
mp->m_logbufs = -1;
mp->m_logbsize = -1;
if (!options)
goto done;
while ((this_char = strsep(&options, ",")) != NULL) {
if (!*this_char)
continue;
if ((value = strchr(this_char, '=')) != NULL)
*value++ = 0;
if (!strcmp(this_char, MNTOPT_LOGBUFS)) {
if (!value || !*value) {
xfs_warn(mp, "%s option requires an argument",
this_char);
return EINVAL;
}
mp->m_logbufs = simple_strtoul(value, &eov, 10);
} else if (!strcmp(this_char, MNTOPT_LOGBSIZE)) {
if (!value || !*value) {
xfs_warn(mp, "%s option requires an argument",
this_char);
return EINVAL;
}
mp->m_logbsize = suffix_strtoul(value, &eov, 10);
} else if (!strcmp(this_char, MNTOPT_LOGDEV)) {
if (!value || !*value) {
xfs_warn(mp, "%s option requires an argument",
this_char);
return EINVAL;
}
mp->m_logname = kstrndup(value, MAXNAMELEN, GFP_KERNEL);
if (!mp->m_logname)
return ENOMEM;
} else if (!strcmp(this_char, MNTOPT_MTPT)) {
xfs_warn(mp, "%s option not allowed on this system",
this_char);
return EINVAL;
} else if (!strcmp(this_char, MNTOPT_RTDEV)) {
if (!value || !*value) {
xfs_warn(mp, "%s option requires an argument",
this_char);
return EINVAL;
}
mp->m_rtname = kstrndup(value, MAXNAMELEN, GFP_KERNEL);
if (!mp->m_rtname)
return ENOMEM;
} else if (!strcmp(this_char, MNTOPT_BIOSIZE)) {
if (!value || !*value) {
xfs_warn(mp, "%s option requires an argument",
this_char);
return EINVAL;
}
iosize = simple_strtoul(value, &eov, 10);
iosizelog = ffs(iosize) - 1;
} else if (!strcmp(this_char, MNTOPT_ALLOCSIZE)) {
if (!value || !*value) {
xfs_warn(mp, "%s option requires an argument",
this_char);
return EINVAL;
}
iosize = suffix_strtoul(value, &eov, 10);
iosizelog = ffs(iosize) - 1;
} else if (!strcmp(this_char, MNTOPT_GRPID) ||
!strcmp(this_char, MNTOPT_BSDGROUPS)) {
mp->m_flags |= XFS_MOUNT_GRPID;
} else if (!strcmp(this_char, MNTOPT_NOGRPID) ||
!strcmp(this_char, MNTOPT_SYSVGROUPS)) {
mp->m_flags &= ~XFS_MOUNT_GRPID;
} else if (!strcmp(this_char, MNTOPT_WSYNC)) {
mp->m_flags |= XFS_MOUNT_WSYNC;
} else if (!strcmp(this_char, MNTOPT_NORECOVERY)) {
mp->m_flags |= XFS_MOUNT_NORECOVERY;
} else if (!strcmp(this_char, MNTOPT_NOALIGN)) {
mp->m_flags |= XFS_MOUNT_NOALIGN;
} else if (!strcmp(this_char, MNTOPT_SWALLOC)) {
mp->m_flags |= XFS_MOUNT_SWALLOC;
} else if (!strcmp(this_char, MNTOPT_SUNIT)) {
if (!value || !*value) {
xfs_warn(mp, "%s option requires an argument",
this_char);
return EINVAL;
}
dsunit = simple_strtoul(value, &eov, 10);
} else if (!strcmp(this_char, MNTOPT_SWIDTH)) {
if (!value || !*value) {
xfs_warn(mp, "%s option requires an argument",
this_char);
return EINVAL;
}
dswidth = simple_strtoul(value, &eov, 10);
} else if (!strcmp(this_char, MNTOPT_64BITINODE)) {
mp->m_flags &= ~XFS_MOUNT_SMALL_INUMS;
#if !XFS_BIG_INUMS
xfs_warn(mp, "%s option not allowed on this system",
this_char);
return EINVAL;
#endif
} else if (!strcmp(this_char, MNTOPT_NOUUID)) {
mp->m_flags |= XFS_MOUNT_NOUUID;
} else if (!strcmp(this_char, MNTOPT_BARRIER)) {
mp->m_flags |= XFS_MOUNT_BARRIER;
} else if (!strcmp(this_char, MNTOPT_NOBARRIER)) {
mp->m_flags &= ~XFS_MOUNT_BARRIER;
} else if (!strcmp(this_char, MNTOPT_IKEEP)) {
mp->m_flags |= XFS_MOUNT_IKEEP;
} else if (!strcmp(this_char, MNTOPT_NOIKEEP)) {
mp->m_flags &= ~XFS_MOUNT_IKEEP;
} else if (!strcmp(this_char, MNTOPT_LARGEIO)) {
mp->m_flags &= ~XFS_MOUNT_COMPAT_IOSIZE;
} else if (!strcmp(this_char, MNTOPT_NOLARGEIO)) {
mp->m_flags |= XFS_MOUNT_COMPAT_IOSIZE;
} else if (!strcmp(this_char, MNTOPT_ATTR2)) {
mp->m_flags |= XFS_MOUNT_ATTR2;
} else if (!strcmp(this_char, MNTOPT_NOATTR2)) {
mp->m_flags &= ~XFS_MOUNT_ATTR2;
mp->m_flags |= XFS_MOUNT_NOATTR2;
} else if (!strcmp(this_char, MNTOPT_FILESTREAM)) {
mp->m_flags |= XFS_MOUNT_FILESTREAMS;
} else if (!strcmp(this_char, MNTOPT_NOQUOTA)) {
mp->m_qflags &= ~XFS_ALL_QUOTA_ACCT;
mp->m_qflags &= ~XFS_ALL_QUOTA_ENFD;
mp->m_qflags &= ~XFS_ALL_QUOTA_ACTIVE;
} else if (!strcmp(this_char, MNTOPT_QUOTA) ||
!strcmp(this_char, MNTOPT_UQUOTA) ||
!strcmp(this_char, MNTOPT_USRQUOTA)) {
mp->m_qflags |= (XFS_UQUOTA_ACCT | XFS_UQUOTA_ACTIVE |
XFS_UQUOTA_ENFD);
} else if (!strcmp(this_char, MNTOPT_QUOTANOENF) ||
!strcmp(this_char, MNTOPT_UQUOTANOENF)) {
mp->m_qflags |= (XFS_UQUOTA_ACCT | XFS_UQUOTA_ACTIVE);
mp->m_qflags &= ~XFS_UQUOTA_ENFD;
} else if (!strcmp(this_char, MNTOPT_PQUOTA) ||
!strcmp(this_char, MNTOPT_PRJQUOTA)) {
mp->m_qflags |= (XFS_PQUOTA_ACCT | XFS_PQUOTA_ACTIVE |
XFS_OQUOTA_ENFD);
} else if (!strcmp(this_char, MNTOPT_PQUOTANOENF)) {
mp->m_qflags |= (XFS_PQUOTA_ACCT | XFS_PQUOTA_ACTIVE);
mp->m_qflags &= ~XFS_OQUOTA_ENFD;
} else if (!strcmp(this_char, MNTOPT_GQUOTA) ||
!strcmp(this_char, MNTOPT_GRPQUOTA)) {
mp->m_qflags |= (XFS_GQUOTA_ACCT | XFS_GQUOTA_ACTIVE |
XFS_OQUOTA_ENFD);
} else if (!strcmp(this_char, MNTOPT_GQUOTANOENF)) {
mp->m_qflags |= (XFS_GQUOTA_ACCT | XFS_GQUOTA_ACTIVE);
mp->m_qflags &= ~XFS_OQUOTA_ENFD;
} else if (!strcmp(this_char, MNTOPT_DELAYLOG)) {
xfs_warn(mp,
"delaylog is the default now, option is deprecated.");
} else if (!strcmp(this_char, MNTOPT_NODELAYLOG)) {
xfs_warn(mp,
"nodelaylog support has been removed, option is deprecated.");
} else if (!strcmp(this_char, MNTOPT_DISCARD)) {
mp->m_flags |= XFS_MOUNT_DISCARD;
} else if (!strcmp(this_char, MNTOPT_NODISCARD)) {
mp->m_flags &= ~XFS_MOUNT_DISCARD;
} else if (!strcmp(this_char, "ihashsize")) {
xfs_warn(mp,
"ihashsize no longer used, option is deprecated.");
} else if (!strcmp(this_char, "osyncisdsync")) {
xfs_warn(mp,
"osyncisdsync has no effect, option is deprecated.");
} else if (!strcmp(this_char, "osyncisosync")) {
xfs_warn(mp,
"osyncisosync has no effect, option is deprecated.");
} else if (!strcmp(this_char, "irixsgid")) {
xfs_warn(mp,
"irixsgid is now a sysctl(2) variable, option is deprecated.");
} else {
xfs_warn(mp, "unknown mount option [%s].", this_char);
return EINVAL;
}
}
/*
* no recovery flag requires a read-only mount
*/
if ((mp->m_flags & XFS_MOUNT_NORECOVERY) &&
!(mp->m_flags & XFS_MOUNT_RDONLY)) {
xfs_warn(mp, "no-recovery mounts must be read-only.");
return EINVAL;
}
if ((mp->m_flags & XFS_MOUNT_NOALIGN) && (dsunit || dswidth)) {
xfs_warn(mp,
"sunit and swidth options incompatible with the noalign option");
return EINVAL;
}
#ifndef CONFIG_XFS_QUOTA
if (XFS_IS_QUOTA_RUNNING(mp)) {
xfs_warn(mp, "quota support not available in this kernel.");
return EINVAL;
}
#endif
if ((mp->m_qflags & (XFS_GQUOTA_ACCT | XFS_GQUOTA_ACTIVE)) &&
(mp->m_qflags & (XFS_PQUOTA_ACCT | XFS_PQUOTA_ACTIVE))) {
xfs_warn(mp, "cannot mount with both project and group quota");
return EINVAL;
}
if ((dsunit && !dswidth) || (!dsunit && dswidth)) {
xfs_warn(mp, "sunit and swidth must be specified together");
return EINVAL;
}
if (dsunit && (dswidth % dsunit != 0)) {
xfs_warn(mp,
"stripe width (%d) must be a multiple of the stripe unit (%d)",
dswidth, dsunit);
return EINVAL;
}
done:
if (!(mp->m_flags & XFS_MOUNT_NOALIGN)) {
/*
* At this point the superblock has not been read
* in, therefore we do not know the block size.
* Before the mount call ends we will convert
* these to FSBs.
*/
if (dsunit) {
mp->m_dalign = dsunit;
mp->m_flags |= XFS_MOUNT_RETERR;
}
if (dswidth)
mp->m_swidth = dswidth;
}
if (mp->m_logbufs != -1 &&
mp->m_logbufs != 0 &&
(mp->m_logbufs < XLOG_MIN_ICLOGS ||
mp->m_logbufs > XLOG_MAX_ICLOGS)) {
xfs_warn(mp, "invalid logbufs value: %d [not %d-%d]",
mp->m_logbufs, XLOG_MIN_ICLOGS, XLOG_MAX_ICLOGS);
return XFS_ERROR(EINVAL);
}
if (mp->m_logbsize != -1 &&
mp->m_logbsize != 0 &&
(mp->m_logbsize < XLOG_MIN_RECORD_BSIZE ||
mp->m_logbsize > XLOG_MAX_RECORD_BSIZE ||
!is_power_of_2(mp->m_logbsize))) {
xfs_warn(mp,
"invalid logbufsize: %d [not 16k,32k,64k,128k or 256k]",
mp->m_logbsize);
return XFS_ERROR(EINVAL);
}
if (iosizelog) {
if (iosizelog > XFS_MAX_IO_LOG ||
iosizelog < XFS_MIN_IO_LOG) {
xfs_warn(mp, "invalid log iosize: %d [not %d-%d]",
iosizelog, XFS_MIN_IO_LOG,
XFS_MAX_IO_LOG);
return XFS_ERROR(EINVAL);
}
mp->m_flags |= XFS_MOUNT_DFLT_IOSIZE;
mp->m_readio_log = iosizelog;
mp->m_writeio_log = iosizelog;
}
return 0;
}
int
xfs_open_by_handle(
struct file *parfilp,
xfs_fsop_handlereq_t *hreq)
{
const struct cred *cred = current_cred();
int error;
int fd;
int permflag;
struct file *filp;
struct inode *inode;
struct dentry *dentry;
if (!capable(CAP_SYS_ADMIN))
return -XFS_ERROR(EPERM);
dentry = xfs_handlereq_to_dentry(parfilp, hreq);
if (IS_ERR(dentry))
return PTR_ERR(dentry);
inode = dentry->d_inode;
/* Restrict xfs_open_by_handle to directories & regular files. */
if (!(S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode))) {
error = -XFS_ERROR(EPERM);
goto out_dput;
}
#if BITS_PER_LONG != 32
hreq->oflags |= O_LARGEFILE;
#endif
/* Put open permission in namei format. */
permflag = hreq->oflags;
if ((permflag+1) & O_ACCMODE)
permflag++;
if (permflag & O_TRUNC)
permflag |= 2;
if ((!(permflag & O_APPEND) || (permflag & O_TRUNC)) &&
(permflag & FMODE_WRITE) && IS_APPEND(inode)) {
error = -XFS_ERROR(EPERM);
goto out_dput;
}
if ((permflag & FMODE_WRITE) && IS_IMMUTABLE(inode)) {
error = -XFS_ERROR(EACCES);
goto out_dput;
}
/* Can't write directories. */
if (S_ISDIR(inode->i_mode) && (permflag & FMODE_WRITE)) {
error = -XFS_ERROR(EISDIR);
goto out_dput;
}
fd = get_unused_fd();
if (fd < 0) {
error = fd;
goto out_dput;
}
filp = dentry_open(dentry, mntget(parfilp->f_path.mnt),
hreq->oflags, cred);
if (IS_ERR(filp)) {
put_unused_fd(fd);
return PTR_ERR(filp);
}
if (inode->i_mode & S_IFREG) {
filp->f_flags |= O_NOATIME;
filp->f_mode |= FMODE_NOCMTIME;
}
fd_install(fd, filp);
return fd;
out_dput:
dput(dentry);
return error;
}
/*ARGSUSED*/
int
_xfs_trans_commit(
xfs_trans_t *tp,
uint flags,
int *log_flushed)
{
xfs_log_iovec_t *log_vector;
int nvec;
xfs_mount_t *mp;
xfs_lsn_t commit_lsn;
/* REFERENCED */
int error;
int log_flags;
int sync;
#define XFS_TRANS_LOGVEC_COUNT 16
xfs_log_iovec_t log_vector_fast[XFS_TRANS_LOGVEC_COUNT];
void *commit_iclog;
int shutdown;
commit_lsn = -1;
/*
* Determine whether this commit is releasing a permanent
* log reservation or not.
*/
if (flags & XFS_TRANS_RELEASE_LOG_RES) {
ASSERT(tp->t_flags & XFS_TRANS_PERM_LOG_RES);
log_flags = XFS_LOG_REL_PERM_RESERV;
} else {
log_flags = 0;
}
mp = tp->t_mountp;
/*
* If there is nothing to be logged by the transaction,
* then unlock all of the items associated with the
* transaction and free the transaction structure.
* Also make sure to return any reserved blocks to
* the free pool.
*/
shut_us_down:
shutdown = XFS_FORCED_SHUTDOWN(mp) ? EIO : 0;
if (!(tp->t_flags & XFS_TRANS_DIRTY) || shutdown) {
xfs_trans_unreserve_and_mod_sb(tp);
/*
* It is indeed possible for the transaction to be
* not dirty but the dqinfo portion to be. All that
* means is that we have some (non-persistent) quota
* reservations that need to be unreserved.
*/
xfs_trans_unreserve_and_mod_dquots(tp);
if (tp->t_ticket) {
commit_lsn = xfs_log_done(mp, tp->t_ticket,
NULL, log_flags);
if (commit_lsn == -1 && !shutdown)
shutdown = XFS_ERROR(EIO);
}
current_restore_flags_nested(&tp->t_pflags, PF_FSTRANS);
xfs_trans_free_items(tp, shutdown? XFS_TRANS_ABORT : 0);
xfs_trans_free_busy(tp);
xfs_trans_free(tp);
XFS_STATS_INC(xs_trans_empty);
return (shutdown);
}
ASSERT(tp->t_ticket != NULL);
/*
* If we need to update the superblock, then do it now.
*/
if (tp->t_flags & XFS_TRANS_SB_DIRTY)
xfs_trans_apply_sb_deltas(tp);
xfs_trans_apply_dquot_deltas(tp);
/*
* Ask each log item how many log_vector entries it will
* need so we can figure out how many to allocate.
* Try to avoid the kmem_alloc() call in the common case
* by using a vector from the stack when it fits.
*/
nvec = xfs_trans_count_vecs(tp);
if (nvec == 0) {
xfs_force_shutdown(mp, SHUTDOWN_LOG_IO_ERROR);
goto shut_us_down;
} else if (nvec <= XFS_TRANS_LOGVEC_COUNT) {
log_vector = log_vector_fast;
} else {
log_vector = (xfs_log_iovec_t *)kmem_alloc(nvec *
sizeof(xfs_log_iovec_t),
KM_SLEEP);
}
/*
* Fill in the log_vector and pin the logged items, and
* then write the transaction to the log.
*/
xfs_trans_fill_vecs(tp, log_vector);
error = xfs_log_write(mp, log_vector, nvec, tp->t_ticket, &(tp->t_lsn));
/*
* The transaction is committed incore here, and can go out to disk
* at any time after this call. However, all the items associated
* with the transaction are still locked and pinned in memory.
*/
commit_lsn = xfs_log_done(mp, tp->t_ticket, &commit_iclog, log_flags);
tp->t_commit_lsn = commit_lsn;
if (nvec > XFS_TRANS_LOGVEC_COUNT) {
kmem_free(log_vector);
}
/*
* If we got a log write error. Unpin the logitems that we
* had pinned, clean up, free trans structure, and return error.
*/
if (error || commit_lsn == -1) {
current_restore_flags_nested(&tp->t_pflags, PF_FSTRANS);
xfs_trans_uncommit(tp, flags|XFS_TRANS_ABORT);
return XFS_ERROR(EIO);
}
/*
* Once the transaction has committed, unused
* reservations need to be released and changes to
* the superblock need to be reflected in the in-core
* version. Do that now.
*/
xfs_trans_unreserve_and_mod_sb(tp);
sync = tp->t_flags & XFS_TRANS_SYNC;
/*
* Tell the LM to call the transaction completion routine
* when the log write with LSN commit_lsn completes (e.g.
* when the transaction commit really hits the on-disk log).
* After this call we cannot reference tp, because the call
* can happen at any time and the call will free the transaction
* structure pointed to by tp. The only case where we call
* the completion routine (xfs_trans_committed) directly is
* if the log is turned off on a debug kernel or we're
* running in simulation mode (the log is explicitly turned
* off).
*/
tp->t_logcb.cb_func = (void(*)(void*, int))xfs_trans_committed;
tp->t_logcb.cb_arg = tp;
/*
* We need to pass the iclog buffer which was used for the
* transaction commit record into this function, and attach
* the callback to it. The callback must be attached before
* the items are unlocked to avoid racing with other threads
* waiting for an item to unlock.
*/
shutdown = xfs_log_notify(mp, commit_iclog, &(tp->t_logcb));
/*
* Mark this thread as no longer being in a transaction
*/
current_restore_flags_nested(&tp->t_pflags, PF_FSTRANS);
/*
* Once all the items of the transaction have been copied
* to the in core log and the callback is attached, the
* items can be unlocked.
*
* This will free descriptors pointing to items which were
* not logged since there is nothing more to do with them.
* For items which were logged, we will keep pointers to them
* so they can be unpinned after the transaction commits to disk.
* This will also stamp each modified meta-data item with
* the commit lsn of this transaction for dependency tracking
* purposes.
*/
xfs_trans_unlock_items(tp, commit_lsn);
/*
* If we detected a log error earlier, finish committing
* the transaction now (unpin log items, etc).
*
* Order is critical here, to avoid using the transaction
* pointer after its been freed (by xfs_trans_committed
* either here now, or as a callback). We cannot do this
* step inside xfs_log_notify as was done earlier because
* of this issue.
*/
if (shutdown)
xfs_trans_committed(tp, XFS_LI_ABORTED);
/*
* Now that the xfs_trans_committed callback has been attached,
* and the items are released we can finally allow the iclog to
* go to disk.
*/
error = xfs_log_release_iclog(mp, commit_iclog);
/*
* If the transaction needs to be synchronous, then force the
* log out now and wait for it.
*/
if (sync) {
if (!error) {
error = _xfs_log_force(mp, commit_lsn,
XFS_LOG_FORCE | XFS_LOG_SYNC,
log_flushed);
}
XFS_STATS_INC(xs_trans_sync);
} else {
XFS_STATS_INC(xs_trans_async);
}
return (error);
}
/*
* Walk the AGs and reclaim the inodes in them. Even if the filesystem is
* corrupted, we still want to try to reclaim all the inodes. If we don't,
* then a shut down during filesystem unmount reclaim walk leak all the
* unreclaimed inodes.
*/
STATIC int
xfs_reclaim_inodes_ag(
struct xfs_mount *mp,
int flags,
int *nr_to_scan)
{
struct xfs_perag *pag;
int error = 0;
int last_error = 0;
xfs_agnumber_t ag;
int trylock = flags & SYNC_TRYLOCK;
int skipped;
restart:
ag = 0;
skipped = 0;
while ((pag = xfs_perag_get_tag(mp, ag, XFS_ICI_RECLAIM_TAG))) {
unsigned long first_index = 0;
int done = 0;
int nr_found = 0;
ag = pag->pag_agno + 1;
if (trylock) {
if (!mutex_trylock(&pag->pag_ici_reclaim_lock)) {
skipped++;
xfs_perag_put(pag);
continue;
}
first_index = pag->pag_ici_reclaim_cursor;
} else
mutex_lock(&pag->pag_ici_reclaim_lock);
do {
struct xfs_inode *batch[XFS_LOOKUP_BATCH];
int i;
rcu_read_lock();
nr_found = radix_tree_gang_lookup_tag(
&pag->pag_ici_root,
(void **)batch, first_index,
XFS_LOOKUP_BATCH,
XFS_ICI_RECLAIM_TAG);
if (!nr_found) {
done = 1;
rcu_read_unlock();
break;
}
/*
* Grab the inodes before we drop the lock. if we found
* nothing, nr == 0 and the loop will be skipped.
*/
for (i = 0; i < nr_found; i++) {
struct xfs_inode *ip = batch[i];
if (done || xfs_reclaim_inode_grab(ip, flags))
batch[i] = NULL;
/*
* Update the index for the next lookup. Catch
* overflows into the next AG range which can
* occur if we have inodes in the last block of
* the AG and we are currently pointing to the
* last inode.
*
* Because we may see inodes that are from the
* wrong AG due to RCU freeing and
* reallocation, only update the index if it
* lies in this AG. It was a race that lead us
* to see this inode, so another lookup from
* the same index will not find it again.
*/
if (XFS_INO_TO_AGNO(mp, ip->i_ino) !=
pag->pag_agno)
continue;
first_index = XFS_INO_TO_AGINO(mp, ip->i_ino + 1);
if (first_index < XFS_INO_TO_AGINO(mp, ip->i_ino))
done = 1;
}
/* unlock now we've grabbed the inodes. */
rcu_read_unlock();
for (i = 0; i < nr_found; i++) {
if (!batch[i])
continue;
error = xfs_reclaim_inode(batch[i], pag, flags);
if (error && last_error != EFSCORRUPTED)
last_error = error;
}
*nr_to_scan -= XFS_LOOKUP_BATCH;
cond_resched();
} while (nr_found && !done && *nr_to_scan > 0);
if (trylock && !done)
pag->pag_ici_reclaim_cursor = first_index;
else
pag->pag_ici_reclaim_cursor = 0;
mutex_unlock(&pag->pag_ici_reclaim_lock);
xfs_perag_put(pag);
}
/*
* if we skipped any AG, and we still have scan count remaining, do
* another pass this time using blocking reclaim semantics (i.e
* waiting on the reclaim locks and ignoring the reclaim cursors). This
* ensure that when we get more reclaimers than AGs we block rather
* than spin trying to execute reclaim.
*/
if (skipped && (flags & SYNC_WAIT) && *nr_to_scan > 0) {
trylock = 0;
goto restart;
}
return XFS_ERROR(last_error);
}
/*
* Add an entry to a block directory.
*/
int /* error */
xfs_dir2_block_addname(
xfs_da_args_t *args) /* directory op arguments */
{
xfs_dir2_data_free_t *bf; /* bestfree table in block */
xfs_dir2_block_t *block; /* directory block structure */
xfs_dir2_leaf_entry_t *blp; /* block leaf entries */
xfs_dabuf_t *bp; /* buffer for block */
xfs_dir2_block_tail_t *btp; /* block tail */
int compact; /* need to compact leaf ents */
xfs_dir2_data_entry_t *dep; /* block data entry */
xfs_inode_t *dp; /* directory inode */
xfs_dir2_data_unused_t *dup; /* block unused entry */
int error; /* error return value */
xfs_dir2_data_unused_t *enddup=NULL; /* unused at end of data */
xfs_dahash_t hash; /* hash value of found entry */
int high; /* high index for binary srch */
int highstale; /* high stale index */
int lfloghigh=0; /* last final leaf to log */
int lfloglow=0; /* first final leaf to log */
int len; /* length of the new entry */
int low; /* low index for binary srch */
int lowstale; /* low stale index */
int mid=0; /* midpoint for binary srch */
xfs_mount_t *mp; /* filesystem mount point */
int needlog; /* need to log header */
int needscan; /* need to rescan freespace */
__be16 *tagp; /* pointer to tag value */
xfs_trans_t *tp; /* transaction structure */
xfs_dir2_trace_args("block_addname", args);
dp = args->dp;
tp = args->trans;
mp = dp->i_mount;
/*
* Read the (one and only) directory block into dabuf bp.
*/
if ((error =
xfs_da_read_buf(tp, dp, mp->m_dirdatablk, -1, &bp, XFS_DATA_FORK))) {
return error;
}
ASSERT(bp != NULL);
block = bp->data;
/*
* Check the magic number, corrupted if wrong.
*/
if (unlikely(be32_to_cpu(block->hdr.magic) != XFS_DIR2_BLOCK_MAGIC)) {
XFS_CORRUPTION_ERROR("xfs_dir2_block_addname",
XFS_ERRLEVEL_LOW, mp, block);
xfs_da_brelse(tp, bp);
return XFS_ERROR(EFSCORRUPTED);
}
len = xfs_dir2_data_entsize(args->namelen);
/*
* Set up pointers to parts of the block.
*/
bf = block->hdr.bestfree;
btp = xfs_dir2_block_tail_p(mp, block);
blp = xfs_dir2_block_leaf_p(btp);
/*
* No stale entries? Need space for entry and new leaf.
*/
if (!btp->stale) {
/*
* Tag just before the first leaf entry.
*/
tagp = (__be16 *)blp - 1;
/*
* Data object just before the first leaf entry.
*/
enddup = (xfs_dir2_data_unused_t *)((char *)block + be16_to_cpu(*tagp));
/*
* If it's not free then can't do this add without cleaning up:
* the space before the first leaf entry needs to be free so it
* can be expanded to hold the pointer to the new entry.
*/
if (be16_to_cpu(enddup->freetag) != XFS_DIR2_DATA_FREE_TAG)
dup = enddup = NULL;
/*
* Check out the biggest freespace and see if it's the same one.
*/
else {
dup = (xfs_dir2_data_unused_t *)
((char *)block + be16_to_cpu(bf[0].offset));
if (dup == enddup) {
/*
* It is the biggest freespace, is it too small
* to hold the new leaf too?
*/
if (be16_to_cpu(dup->length) < len + (uint)sizeof(*blp)) {
/*
* Yes, we use the second-largest
* entry instead if it works.
*/
if (be16_to_cpu(bf[1].length) >= len)
dup = (xfs_dir2_data_unused_t *)
((char *)block +
be16_to_cpu(bf[1].offset));
else
dup = NULL;
}
} else {
/*
* Not the same free entry,
* just check its length.
*/
if (be16_to_cpu(dup->length) < len) {
dup = NULL;
}
}
}
compact = 0;
}
/*
* If there are stale entries we'll use one for the leaf.
* Is the biggest entry enough to avoid compaction?
*/
else if (be16_to_cpu(bf[0].length) >= len) {
dup = (xfs_dir2_data_unused_t *)
((char *)block + be16_to_cpu(bf[0].offset));
compact = 0;
}
/*
* Will need to compact to make this work.
*/
else {
/*
* Tag just before the first leaf entry.
*/
tagp = (__be16 *)blp - 1;
/*
* Data object just before the first leaf entry.
*/
dup = (xfs_dir2_data_unused_t *)((char *)block + be16_to_cpu(*tagp));
/*
* If it's not free then the data will go where the
* leaf data starts now, if it works at all.
*/
if (be16_to_cpu(dup->freetag) == XFS_DIR2_DATA_FREE_TAG) {
if (be16_to_cpu(dup->length) + (be32_to_cpu(btp->stale) - 1) *
(uint)sizeof(*blp) < len)
dup = NULL;
} else if ((be32_to_cpu(btp->stale) - 1) * (uint)sizeof(*blp) < len)
dup = NULL;
else
dup = (xfs_dir2_data_unused_t *)blp;
compact = 1;
}
/*
* If this isn't a real add, we're done with the buffer.
*/
if (args->op_flags & XFS_DA_OP_JUSTCHECK)
xfs_da_brelse(tp, bp);
/*
* If we don't have space for the new entry & leaf ...
*/
if (!dup) {
/*
* Not trying to actually do anything, or don't have
* a space reservation: return no-space.
*/
if ((args->op_flags & XFS_DA_OP_JUSTCHECK) || args->total == 0)
return XFS_ERROR(ENOSPC);
/*
* Convert to the next larger format.
* Then add the new entry in that format.
*/
error = xfs_dir2_block_to_leaf(args, bp);
xfs_da_buf_done(bp);
if (error)
return error;
return xfs_dir2_leaf_addname(args);
}
/*
* Just checking, and it would work, so say so.
*/
if (args->op_flags & XFS_DA_OP_JUSTCHECK)
return 0;
needlog = needscan = 0;
/*
* If need to compact the leaf entries, do it now.
* Leave the highest-numbered stale entry stale.
* XXX should be the one closest to mid but mid is not yet computed.
*/
if (compact) {
int fromidx; /* source leaf index */
int toidx; /* target leaf index */
for (fromidx = toidx = be32_to_cpu(btp->count) - 1,
highstale = lfloghigh = -1;
fromidx >= 0;
fromidx--) {
if (be32_to_cpu(blp[fromidx].address) == XFS_DIR2_NULL_DATAPTR) {
if (highstale == -1)
highstale = toidx;
else {
if (lfloghigh == -1)
lfloghigh = toidx;
continue;
}
}
if (fromidx < toidx)
blp[toidx] = blp[fromidx];
toidx--;
}
lfloglow = toidx + 1 - (be32_to_cpu(btp->stale) - 1);
lfloghigh -= be32_to_cpu(btp->stale) - 1;
be32_add_cpu(&btp->count, -(be32_to_cpu(btp->stale) - 1));
xfs_dir2_data_make_free(tp, bp,
(xfs_dir2_data_aoff_t)((char *)blp - (char *)block),
(xfs_dir2_data_aoff_t)((be32_to_cpu(btp->stale) - 1) * sizeof(*blp)),
&needlog, &needscan);
blp += be32_to_cpu(btp->stale) - 1;
btp->stale = cpu_to_be32(1);
/*
* If we now need to rebuild the bestfree map, do so.
* This needs to happen before the next call to use_free.
*/
if (needscan) {
xfs_dir2_data_freescan(mp, (xfs_dir2_data_t *)block, &needlog);
needscan = 0;
}
}
/*
* Set leaf logging boundaries to impossible state.
* For the no-stale case they're set explicitly.
*/
else if (btp->stale) {
lfloglow = be32_to_cpu(btp->count);
lfloghigh = -1;
}
/*
* Find the slot that's first lower than our hash value, -1 if none.
*/
for (low = 0, high = be32_to_cpu(btp->count) - 1; low <= high; ) {
mid = (low + high) >> 1;
if ((hash = be32_to_cpu(blp[mid].hashval)) == args->hashval)
break;
if (hash < args->hashval)
low = mid + 1;
else
high = mid - 1;
}
while (mid >= 0 && be32_to_cpu(blp[mid].hashval) >= args->hashval) {
mid--;
}
/*
* No stale entries, will use enddup space to hold new leaf.
*/
if (!btp->stale) {
/*
* Mark the space needed for the new leaf entry, now in use.
*/
xfs_dir2_data_use_free(tp, bp, enddup,
(xfs_dir2_data_aoff_t)
((char *)enddup - (char *)block + be16_to_cpu(enddup->length) -
sizeof(*blp)),
(xfs_dir2_data_aoff_t)sizeof(*blp),
&needlog, &needscan);
/*
* Update the tail (entry count).
*/
be32_add_cpu(&btp->count, 1);
/*
* If we now need to rebuild the bestfree map, do so.
* This needs to happen before the next call to use_free.
*/
if (needscan) {
xfs_dir2_data_freescan(mp, (xfs_dir2_data_t *)block,
&needlog);
needscan = 0;
}
/*
* Adjust pointer to the first leaf entry, we're about to move
* the table up one to open up space for the new leaf entry.
* Then adjust our index to match.
*/
blp--;
mid++;
if (mid)
memmove(blp, &blp[1], mid * sizeof(*blp));
lfloglow = 0;
lfloghigh = mid;
}
/*
* Use a stale leaf for our new entry.
*/
else {
for (lowstale = mid;
lowstale >= 0 &&
be32_to_cpu(blp[lowstale].address) != XFS_DIR2_NULL_DATAPTR;
lowstale--)
continue;
for (highstale = mid + 1;
highstale < be32_to_cpu(btp->count) &&
be32_to_cpu(blp[highstale].address) != XFS_DIR2_NULL_DATAPTR &&
(lowstale < 0 || mid - lowstale > highstale - mid);
highstale++)
continue;
/*
* Move entries toward the low-numbered stale entry.
*/
if (lowstale >= 0 &&
(highstale == be32_to_cpu(btp->count) ||
mid - lowstale <= highstale - mid)) {
if (mid - lowstale)
memmove(&blp[lowstale], &blp[lowstale + 1],
(mid - lowstale) * sizeof(*blp));
lfloglow = MIN(lowstale, lfloglow);
lfloghigh = MAX(mid, lfloghigh);
}
/*
* Move entries toward the high-numbered stale entry.
*/
else {
ASSERT(highstale < be32_to_cpu(btp->count));
mid++;
if (highstale - mid)
memmove(&blp[mid + 1], &blp[mid],
(highstale - mid) * sizeof(*blp));
lfloglow = MIN(mid, lfloglow);
lfloghigh = MAX(highstale, lfloghigh);
}
be32_add_cpu(&btp->stale, -1);
}
/*
* Point to the new data entry.
*/
dep = (xfs_dir2_data_entry_t *)dup;
/*
* Fill in the leaf entry.
*/
blp[mid].hashval = cpu_to_be32(args->hashval);
blp[mid].address = cpu_to_be32(xfs_dir2_byte_to_dataptr(mp,
(char *)dep - (char *)block));
xfs_dir2_block_log_leaf(tp, bp, lfloglow, lfloghigh);
/*
* Mark space for the data entry used.
*/
xfs_dir2_data_use_free(tp, bp, dup,
(xfs_dir2_data_aoff_t)((char *)dup - (char *)block),
(xfs_dir2_data_aoff_t)len, &needlog, &needscan);
/*
* Create the new data entry.
*/
dep->inumber = cpu_to_be64(args->inumber);
dep->namelen = args->namelen;
memcpy(dep->name, args->name, args->namelen);
tagp = xfs_dir2_data_entry_tag_p(dep);
*tagp = cpu_to_be16((char *)dep - (char *)block);
/*
* Clean up the bestfree array and log the header, tail, and entry.
*/
if (needscan)
xfs_dir2_data_freescan(mp, (xfs_dir2_data_t *)block, &needlog);
if (needlog)
xfs_dir2_data_log_header(tp, bp);
xfs_dir2_block_log_tail(tp, bp);
xfs_dir2_data_log_entry(tp, bp, dep);
xfs_dir2_data_check(dp, bp);
xfs_da_buf_done(bp);
return 0;
}
ssize_t /* bytes read, or (-) error */
xfs_read(
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;
ssize_t ret;
xfs_fsize_t n;
xfs_inode_t *ip;
xfs_mount_t *mp;
vnode_t *vp;
unsigned long seg;
ip = XFS_BHVTOI(bdp);
vp = BHV_TO_VNODE(bdp);
mp = ip->i_mount;
vn_trace_entry(vp, "xfs_read", (inst_t *)__return_address);
XFS_STATS_INC(xs_read_calls);
/* 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 (ioflags & IO_ISDIRECT) {
pb_target_t *target =
(ip->i_d.di_flags & XFS_DIFLAG_REALTIME) ?
mp->m_rtdev_targp : mp->m_ddev_targp;
if ((*offset & target->pbr_smask) ||
(size & target->pbr_smask)) {
if (*offset == ip->i_d.di_size) {
return (0);
}
return -XFS_ERROR(EINVAL);
}
}
n = XFS_MAXIOFFSET(mp) - *offset;
if ((n <= 0) || (size == 0))
return 0;
if (n < size)
size = n;
if (XFS_FORCED_SHUTDOWN(mp)) {
return -EIO;
}
/* OK so we are holding the I/O lock for the duration
* of the submission, then what happens if the I/O
* does not really happen here, but is scheduled
* later?
*/
xfs_ilock(ip, XFS_IOLOCK_SHARED);
if (DM_EVENT_ENABLED(vp->v_vfsp, ip, DM_EVENT_READ) &&
!(ioflags & IO_INVIS)) {
int error;
vrwlock_t locktype = VRWLOCK_READ;
error = XFS_SEND_DATA(mp, DM_EVENT_READ, BHV_TO_VNODE(bdp), *offset, size,
FILP_DELAY_FLAG(file), &locktype);
if (error) {
xfs_iunlock(ip, XFS_IOLOCK_SHARED);
return -error;
}
}
ret = __generic_file_aio_read(iocb, iovp, segs, offset);
xfs_iunlock(ip, XFS_IOLOCK_SHARED);
if (ret > 0)
XFS_STATS_ADD(xs_read_bytes, ret);
if (likely(!(ioflags & IO_INVIS)))
xfs_ichgtime(ip, XFS_ICHGTIME_ACC);
return ret;
}
/*
* Convert the shortform directory to block form.
*/
int /* error */
xfs_dir2_sf_to_block(
xfs_da_args_t *args) /* operation arguments */
{
xfs_dir2_db_t blkno; /* dir-relative block # (0) */
xfs_dir2_block_t *block; /* block structure */
xfs_dir2_leaf_entry_t *blp; /* block leaf entries */
xfs_dabuf_t *bp; /* block buffer */
xfs_dir2_block_tail_t *btp; /* block tail pointer */
char *buf; /* sf buffer */
int buf_len;
xfs_dir2_data_entry_t *dep; /* data entry pointer */
xfs_inode_t *dp; /* incore directory inode */
int dummy; /* trash */
xfs_dir2_data_unused_t *dup; /* unused entry pointer */
int endoffset; /* end of data objects */
int error; /* error return value */
int i; /* index */
xfs_mount_t *mp; /* filesystem mount point */
int needlog; /* need to log block header */
int needscan; /* need to scan block freespc */
int newoffset; /* offset from current entry */
int offset; /* target block offset */
xfs_dir2_sf_entry_t *sfep; /* sf entry pointer */
xfs_dir2_sf_t *sfp; /* shortform structure */
__be16 *tagp; /* end of data entry */
xfs_trans_t *tp; /* transaction pointer */
struct xfs_name name;
xfs_dir2_trace_args("sf_to_block", args);
dp = args->dp;
tp = args->trans;
mp = dp->i_mount;
ASSERT(dp->i_df.if_flags & XFS_IFINLINE);
/*
* Bomb out if the shortform directory is way too short.
*/
if (dp->i_d.di_size < offsetof(xfs_dir2_sf_hdr_t, parent)) {
ASSERT(XFS_FORCED_SHUTDOWN(mp));
return XFS_ERROR(EIO);
}
ASSERT(dp->i_df.if_bytes == dp->i_d.di_size);
ASSERT(dp->i_df.if_u1.if_data != NULL);
sfp = (xfs_dir2_sf_t *)dp->i_df.if_u1.if_data;
ASSERT(dp->i_d.di_size >= xfs_dir2_sf_hdr_size(sfp->hdr.i8count));
/*
* Copy the directory into the stack buffer.
* Then pitch the incore inode data so we can make extents.
*/
buf_len = dp->i_df.if_bytes;
buf = kmem_alloc(dp->i_df.if_bytes, KM_SLEEP);
memcpy(buf, sfp, dp->i_df.if_bytes);
xfs_idata_realloc(dp, -dp->i_df.if_bytes, XFS_DATA_FORK);
dp->i_d.di_size = 0;
xfs_trans_log_inode(tp, dp, XFS_ILOG_CORE);
/*
* Reset pointer - old sfp is gone.
*/
sfp = (xfs_dir2_sf_t *)buf;
/*
* Add block 0 to the inode.
*/
error = xfs_dir2_grow_inode(args, XFS_DIR2_DATA_SPACE, &blkno);
if (error) {
kmem_free(buf);
return error;
}
/*
* Initialize the data block.
*/
error = xfs_dir2_data_init(args, blkno, &bp);
if (error) {
kmem_free(buf);
return error;
}
block = bp->data;
block->hdr.magic = cpu_to_be32(XFS_DIR2_BLOCK_MAGIC);
/*
* Compute size of block "tail" area.
*/
i = (uint)sizeof(*btp) +
(sfp->hdr.count + 2) * (uint)sizeof(xfs_dir2_leaf_entry_t);
/*
* The whole thing is initialized to free by the init routine.
* Say we're using the leaf and tail area.
*/
dup = (xfs_dir2_data_unused_t *)block->u;
needlog = needscan = 0;
xfs_dir2_data_use_free(tp, bp, dup, mp->m_dirblksize - i, i, &needlog,
&needscan);
ASSERT(needscan == 0);
/*
* Fill in the tail.
*/
btp = xfs_dir2_block_tail_p(mp, block);
btp->count = cpu_to_be32(sfp->hdr.count + 2); /* ., .. */
btp->stale = 0;
blp = xfs_dir2_block_leaf_p(btp);
endoffset = (uint)((char *)blp - (char *)block);
/*
* Remove the freespace, we'll manage it.
*/
xfs_dir2_data_use_free(tp, bp, dup,
(xfs_dir2_data_aoff_t)((char *)dup - (char *)block),
be16_to_cpu(dup->length), &needlog, &needscan);
/*
* Create entry for .
*/
dep = (xfs_dir2_data_entry_t *)
((char *)block + XFS_DIR2_DATA_DOT_OFFSET);
dep->inumber = cpu_to_be64(dp->i_ino);
dep->namelen = 1;
dep->name[0] = '.';
tagp = xfs_dir2_data_entry_tag_p(dep);
*tagp = cpu_to_be16((char *)dep - (char *)block);
xfs_dir2_data_log_entry(tp, bp, dep);
blp[0].hashval = cpu_to_be32(xfs_dir_hash_dot);
blp[0].address = cpu_to_be32(xfs_dir2_byte_to_dataptr(mp,
(char *)dep - (char *)block));
/*
* Create entry for ..
*/
dep = (xfs_dir2_data_entry_t *)
((char *)block + XFS_DIR2_DATA_DOTDOT_OFFSET);
dep->inumber = cpu_to_be64(xfs_dir2_sf_get_inumber(sfp, &sfp->hdr.parent));
dep->namelen = 2;
dep->name[0] = dep->name[1] = '.';
tagp = xfs_dir2_data_entry_tag_p(dep);
*tagp = cpu_to_be16((char *)dep - (char *)block);
xfs_dir2_data_log_entry(tp, bp, dep);
blp[1].hashval = cpu_to_be32(xfs_dir_hash_dotdot);
blp[1].address = cpu_to_be32(xfs_dir2_byte_to_dataptr(mp,
(char *)dep - (char *)block));
offset = XFS_DIR2_DATA_FIRST_OFFSET;
/*
* Loop over existing entries, stuff them in.
*/
if ((i = 0) == sfp->hdr.count)
sfep = NULL;
else
sfep = xfs_dir2_sf_firstentry(sfp);
/*
* Need to preserve the existing offset values in the sf directory.
* Insert holes (unused entries) where necessary.
*/
while (offset < endoffset) {
/*
* sfep is null when we reach the end of the list.
*/
if (sfep == NULL)
newoffset = endoffset;
else
newoffset = xfs_dir2_sf_get_offset(sfep);
/*
* There should be a hole here, make one.
*/
if (offset < newoffset) {
dup = (xfs_dir2_data_unused_t *)
((char *)block + offset);
dup->freetag = cpu_to_be16(XFS_DIR2_DATA_FREE_TAG);
dup->length = cpu_to_be16(newoffset - offset);
*xfs_dir2_data_unused_tag_p(dup) = cpu_to_be16(
((char *)dup - (char *)block));
xfs_dir2_data_log_unused(tp, bp, dup);
(void)xfs_dir2_data_freeinsert((xfs_dir2_data_t *)block,
dup, &dummy);
offset += be16_to_cpu(dup->length);
continue;
}
/*
* Copy a real entry.
*/
dep = (xfs_dir2_data_entry_t *)((char *)block + newoffset);
dep->inumber = cpu_to_be64(xfs_dir2_sf_get_inumber(sfp,
xfs_dir2_sf_inumberp(sfep)));
dep->namelen = sfep->namelen;
memcpy(dep->name, sfep->name, dep->namelen);
tagp = xfs_dir2_data_entry_tag_p(dep);
*tagp = cpu_to_be16((char *)dep - (char *)block);
xfs_dir2_data_log_entry(tp, bp, dep);
name.name = sfep->name;
name.len = sfep->namelen;
blp[2 + i].hashval = cpu_to_be32(mp->m_dirnameops->
hashname(&name));
blp[2 + i].address = cpu_to_be32(xfs_dir2_byte_to_dataptr(mp,
(char *)dep - (char *)block));
offset = (int)((char *)(tagp + 1) - (char *)block);
if (++i == sfp->hdr.count)
sfep = NULL;
else
sfep = xfs_dir2_sf_nextentry(sfp, sfep);
}
/* Done with the temporary buffer */
kmem_free(buf);
/*
* Sort the leaf entries by hash value.
*/
xfs_sort(blp, be32_to_cpu(btp->count), sizeof(*blp), xfs_dir2_block_sort);
/*
* Log the leaf entry area and tail.
* Already logged the header in data_init, ignore needlog.
*/
ASSERT(needscan == 0);
xfs_dir2_block_log_leaf(tp, bp, 0, be32_to_cpu(btp->count) - 1);
xfs_dir2_block_log_tail(tp, bp);
xfs_dir2_data_check(dp, bp);
xfs_da_buf_done(bp);
return 0;
}
/*
* Internal block lookup routine.
*/
static int /* error */
xfs_dir2_block_lookup_int(
xfs_da_args_t *args, /* dir lookup arguments */
xfs_dabuf_t **bpp, /* returned block buffer */
int *entno) /* returned entry number */
{
xfs_dir2_dataptr_t addr; /* data entry address */
xfs_dir2_block_t *block; /* block structure */
xfs_dir2_leaf_entry_t *blp; /* block leaf entries */
xfs_dabuf_t *bp; /* block buffer */
xfs_dir2_block_tail_t *btp; /* block tail */
xfs_dir2_data_entry_t *dep; /* block data entry */
xfs_inode_t *dp; /* incore inode */
int error; /* error return value */
xfs_dahash_t hash; /* found hash value */
int high; /* binary search high index */
int low; /* binary search low index */
int mid; /* binary search current idx */
xfs_mount_t *mp; /* filesystem mount point */
xfs_trans_t *tp; /* transaction pointer */
enum xfs_dacmp cmp; /* comparison result */
dp = args->dp;
tp = args->trans;
mp = dp->i_mount;
/*
* Read the buffer, return error if we can't get it.
*/
if ((error =
xfs_da_read_buf(tp, dp, mp->m_dirdatablk, -1, &bp, XFS_DATA_FORK))) {
return error;
}
ASSERT(bp != NULL);
block = bp->data;
xfs_dir2_data_check(dp, bp);
btp = xfs_dir2_block_tail_p(mp, block);
blp = xfs_dir2_block_leaf_p(btp);
/*
* Loop doing a binary search for our hash value.
* Find our entry, ENOENT if it's not there.
*/
for (low = 0, high = be32_to_cpu(btp->count) - 1; ; ) {
ASSERT(low <= high);
mid = (low + high) >> 1;
if ((hash = be32_to_cpu(blp[mid].hashval)) == args->hashval)
break;
if (hash < args->hashval)
low = mid + 1;
else
high = mid - 1;
if (low > high) {
ASSERT(args->op_flags & XFS_DA_OP_OKNOENT);
xfs_da_brelse(tp, bp);
return XFS_ERROR(ENOENT);
}
}
/*
* Back up to the first one with the right hash value.
*/
while (mid > 0 && be32_to_cpu(blp[mid - 1].hashval) == args->hashval) {
mid--;
}
/*
* Now loop forward through all the entries with the
* right hash value looking for our name.
*/
do {
if ((addr = be32_to_cpu(blp[mid].address)) == XFS_DIR2_NULL_DATAPTR)
continue;
/*
* Get pointer to the entry from the leaf.
*/
dep = (xfs_dir2_data_entry_t *)
((char *)block + xfs_dir2_dataptr_to_off(mp, addr));
/*
* Compare name and if it's an exact match, return the index
* and buffer. If it's the first case-insensitive match, store
* the index and buffer and continue looking for an exact match.
*/
cmp = mp->m_dirnameops->compname(args, dep->name, dep->namelen);
if (cmp != XFS_CMP_DIFFERENT && cmp != args->cmpresult) {
args->cmpresult = cmp;
*bpp = bp;
*entno = mid;
if (cmp == XFS_CMP_EXACT)
return 0;
}
} while (++mid < be32_to_cpu(btp->count) &&
be32_to_cpu(blp[mid].hashval) == hash);
ASSERT(args->op_flags & XFS_DA_OP_OKNOENT);
/*
* Here, we can only be doing a lookup (not a rename or replace).
* If a case-insensitive match was found earlier, return success.
*/
if (args->cmpresult == XFS_CMP_CASE)
return 0;
/*
* No match, release the buffer and return ENOENT.
*/
xfs_da_brelse(tp, bp);
return XFS_ERROR(ENOENT);
}
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;
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 ((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)
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);
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_unlock_isem;
xfs_rwlock(bdp, locktype);
pos = 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) {
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);
return -error;
}
ssize_t /* bytes read, or (-) error */
xfs_read(
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;
ssize_t ret;
xfs_fsize_t n;
xfs_inode_t *ip;
xfs_mount_t *mp;
vnode_t *vp;
unsigned long seg;
ip = XFS_BHVTOI(bdp);
vp = BHV_TO_VNODE(bdp);
mp = ip->i_mount;
XFS_STATS_INC(xs_read_calls);
/* 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 (ioflags & IO_ISDIRECT) {
xfs_buftarg_t *target =
(ip->i_d.di_flags & XFS_DIFLAG_REALTIME) ?
mp->m_rtdev_targp : mp->m_ddev_targp;
if ((*offset & target->pbr_smask) ||
(size & target->pbr_smask)) {
if (*offset == ip->i_d.di_size) {
return (0);
}
return -XFS_ERROR(EINVAL);
}
}
n = XFS_MAXIOFFSET(mp) - *offset;
if ((n <= 0) || (size == 0))
return 0;
if (n < size)
size = n;
if (XFS_FORCED_SHUTDOWN(mp)) {
return -EIO;
}
xfs_ilock(ip, XFS_IOLOCK_SHARED);
if (DM_EVENT_ENABLED(vp->v_vfsp, ip, DM_EVENT_READ) &&
!(ioflags & IO_INVIS)) {
vrwlock_t locktype = VRWLOCK_READ;
ret = XFS_SEND_DATA(mp, DM_EVENT_READ,
BHV_TO_VNODE(bdp), *offset, size,
FILP_DELAY_FLAG(file), &locktype);
if (ret) {
xfs_iunlock(ip, XFS_IOLOCK_SHARED);
return -ret;
}
}
xfs_rw_enter_trace(XFS_READ_ENTER, &ip->i_iocore,
(void *)iovp, segs, *offset, ioflags);
ret = __generic_file_aio_read(iocb, iovp, segs, offset);
if (ret == -EIOCBQUEUED)
ret = wait_on_sync_kiocb(iocb);
xfs_iunlock(ip, XFS_IOLOCK_SHARED);
if (ret > 0)
XFS_STATS_ADD(xs_read_bytes, ret);
if (likely(!(ioflags & IO_INVIS)))
xfs_ichgtime(ip, XFS_ICHGTIME_ACC);
return ret;
}
int
xfs_qm_scall_getquota(
struct xfs_mount *mp,
xfs_dqid_t id,
uint type,
struct fs_disk_quota *dst)
{
struct xfs_dquot *dqp;
int error;
error = xfs_qm_dqget(mp, NULL, id, type, 0, &dqp);
if (error)
return error;
if (XFS_IS_DQUOT_UNINITIALIZED(dqp)) {
error = XFS_ERROR(ENOENT);
goto out_put;
}
memset(dst, 0, sizeof(*dst));
dst->d_version = FS_DQUOT_VERSION;
dst->d_flags = xfs_qm_export_qtype_flags(dqp->q_core.d_flags);
dst->d_id = be32_to_cpu(dqp->q_core.d_id);
dst->d_blk_hardlimit =
XFS_FSB_TO_BB(mp, be64_to_cpu(dqp->q_core.d_blk_hardlimit));
dst->d_blk_softlimit =
XFS_FSB_TO_BB(mp, be64_to_cpu(dqp->q_core.d_blk_softlimit));
dst->d_ino_hardlimit = be64_to_cpu(dqp->q_core.d_ino_hardlimit);
dst->d_ino_softlimit = be64_to_cpu(dqp->q_core.d_ino_softlimit);
dst->d_bcount = XFS_FSB_TO_BB(mp, dqp->q_res_bcount);
dst->d_icount = dqp->q_res_icount;
dst->d_btimer = be32_to_cpu(dqp->q_core.d_btimer);
dst->d_itimer = be32_to_cpu(dqp->q_core.d_itimer);
dst->d_iwarns = be16_to_cpu(dqp->q_core.d_iwarns);
dst->d_bwarns = be16_to_cpu(dqp->q_core.d_bwarns);
dst->d_rtb_hardlimit =
XFS_FSB_TO_BB(mp, be64_to_cpu(dqp->q_core.d_rtb_hardlimit));
dst->d_rtb_softlimit =
XFS_FSB_TO_BB(mp, be64_to_cpu(dqp->q_core.d_rtb_softlimit));
dst->d_rtbcount = XFS_FSB_TO_BB(mp, dqp->q_res_rtbcount);
dst->d_rtbtimer = be32_to_cpu(dqp->q_core.d_rtbtimer);
dst->d_rtbwarns = be16_to_cpu(dqp->q_core.d_rtbwarns);
if ((!XFS_IS_UQUOTA_ENFORCED(mp) && dqp->q_core.d_flags == XFS_DQ_USER) ||
(!XFS_IS_OQUOTA_ENFORCED(mp) &&
(dqp->q_core.d_flags & (XFS_DQ_PROJ | XFS_DQ_GROUP)))) {
dst->d_btimer = 0;
dst->d_itimer = 0;
dst->d_rtbtimer = 0;
}
#ifdef DEBUG
if (((XFS_IS_UQUOTA_ENFORCED(mp) && dst->d_flags == FS_USER_QUOTA) ||
(XFS_IS_OQUOTA_ENFORCED(mp) &&
(dst->d_flags & (FS_PROJ_QUOTA | FS_GROUP_QUOTA)))) &&
dst->d_id != 0) {
if (((int) dst->d_bcount > (int) dst->d_blk_softlimit) &&
(dst->d_blk_softlimit > 0)) {
ASSERT(dst->d_btimer != 0);
}
if (((int) dst->d_icount > (int) dst->d_ino_softlimit) &&
(dst->d_ino_softlimit > 0)) {
ASSERT(dst->d_itimer != 0);
}
}
#endif
out_put:
xfs_qm_dqput(dqp);
return error;
}
/*
* xfs_file_dio_aio_write - handle direct IO writes
*
* Lock the inode appropriately to prepare for and issue a direct IO write.
* By separating it from the buffered write path we remove all the tricky to
* follow locking changes and looping.
*
* If there are cached pages or we're extending the file, we need IOLOCK_EXCL
* until we're sure the bytes at the new EOF have been zeroed and/or the cached
* pages are flushed out.
*
* In most cases the direct IO writes will be done holding IOLOCK_SHARED
* allowing them to be done in parallel with reads and other direct IO writes.
* However, if the IO is not aligned to filesystem blocks, the direct IO layer
* needs to do sub-block zeroing and that requires serialisation against other
* direct IOs to the same block. In this case we need to serialise the
* submission of the unaligned IOs so that we don't get racing block zeroing in
* the dio layer. To avoid the problem with aio, we also need to wait for
* outstanding IOs to complete so that unwritten extent conversion is completed
* before we try to map the overlapping block. This is currently implemented by
* hitting it with a big hammer (i.e. inode_dio_wait()).
*
* Returns with locks held indicated by @iolock and errors indicated by
* negative return values.
*/
STATIC ssize_t
xfs_file_dio_aio_write(
struct kiocb *iocb,
const struct iovec *iovp,
unsigned long nr_segs,
loff_t pos,
size_t ocount,
xfs_fsize_t *new_size,
int *iolock)
{
struct file *file = iocb->ki_filp;
struct address_space *mapping = file->f_mapping;
struct inode *inode = mapping->host;
struct xfs_inode *ip = XFS_I(inode);
struct xfs_mount *mp = ip->i_mount;
ssize_t ret = 0;
size_t count = ocount;
int unaligned_io = 0;
struct xfs_buftarg *target = XFS_IS_REALTIME_INODE(ip) ?
mp->m_rtdev_targp : mp->m_ddev_targp;
*iolock = 0;
if ((pos & target->bt_smask) || (count & target->bt_smask))
return -XFS_ERROR(EINVAL);
if ((pos & mp->m_blockmask) || ((pos + count) & mp->m_blockmask))
unaligned_io = 1;
/*
* We don't need to take an exclusive lock unless there page cache needs
* to be invalidated or unaligned IO is being executed. We don't need to
* consider the EOF extension case here because
* xfs_file_aio_write_checks() will relock the inode as necessary for
* EOF zeroing cases and fill out the new inode size as appropriate.
*/
if (unaligned_io || mapping->nrpages)
*iolock = XFS_IOLOCK_EXCL;
else
*iolock = XFS_IOLOCK_SHARED;
xfs_rw_ilock(ip, *iolock);
/*
* Recheck if there are cached pages that need invalidate after we got
* the iolock to protect against other threads adding new pages while
* we were waiting for the iolock.
*/
if (mapping->nrpages && *iolock == XFS_IOLOCK_SHARED) {
xfs_rw_iunlock(ip, *iolock);
*iolock = XFS_IOLOCK_EXCL;
xfs_rw_ilock(ip, *iolock);
}
ret = xfs_file_aio_write_checks(file, &pos, &count, new_size, iolock);
if (ret)
return ret;
if (mapping->nrpages) {
ret = -xfs_flushinval_pages(ip, (pos & PAGE_CACHE_MASK), -1,
FI_REMAPF_LOCKED);
if (ret)
return ret;
}
/*
* If we are doing unaligned IO, wait for all other IO to drain,
* otherwise demote the lock if we had to flush cached pages
*/
if (unaligned_io)
inode_dio_wait(inode);
else if (*iolock == XFS_IOLOCK_EXCL) {
xfs_rw_ilock_demote(ip, XFS_IOLOCK_EXCL);
*iolock = XFS_IOLOCK_SHARED;
}
trace_xfs_file_direct_write(ip, count, iocb->ki_pos, 0);
ret = generic_file_direct_write(iocb, iovp,
&nr_segs, pos, &iocb->ki_pos, count, ocount);
/* No fallback to buffered IO on errors for XFS. */
ASSERT(ret < 0 || ret == count);
return ret;
}
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);
}
/*
* This is called to reserve free disk blocks and log space for the
* given transaction. This must be done before allocating any resources
* within the transaction.
*
* This will return ENOSPC if there are not enough blocks available.
* It will sleep waiting for available log space.
* The only valid value for the flags parameter is XFS_RES_LOG_PERM, which
* is used by long running transactions. If any one of the reservations
* fails then they will all be backed out.
*
* This does not do quota reservations. That typically is done by the
* caller afterwards.
*/
int
xfs_trans_reserve(
xfs_trans_t *tp,
uint blocks,
uint logspace,
uint rtextents,
uint flags,
uint logcount)
{
int log_flags;
int error = 0;
int rsvd = (tp->t_flags & XFS_TRANS_RESERVE) != 0;
/* Mark this thread as being in a transaction */
current_set_flags_nested(&tp->t_pflags, PF_FSTRANS);
/*
* Attempt to reserve the needed disk blocks by decrementing
* the number needed from the number available. This will
* fail if the count would go below zero.
*/
if (blocks > 0) {
error = xfs_mod_incore_sb(tp->t_mountp, XFS_SBS_FDBLOCKS,
-((int64_t)blocks), rsvd);
if (error != 0) {
current_restore_flags_nested(&tp->t_pflags, PF_FSTRANS);
return (XFS_ERROR(ENOSPC));
}
tp->t_blk_res += blocks;
}
/*
* Reserve the log space needed for this transaction.
*/
if (logspace > 0) {
ASSERT((tp->t_log_res == 0) || (tp->t_log_res == logspace));
ASSERT((tp->t_log_count == 0) ||
(tp->t_log_count == logcount));
if (flags & XFS_TRANS_PERM_LOG_RES) {
log_flags = XFS_LOG_PERM_RESERV;
tp->t_flags |= XFS_TRANS_PERM_LOG_RES;
} else {
ASSERT(tp->t_ticket == NULL);
ASSERT(!(tp->t_flags & XFS_TRANS_PERM_LOG_RES));
log_flags = 0;
}
error = xfs_log_reserve(tp->t_mountp, logspace, logcount,
&tp->t_ticket,
XFS_TRANSACTION, log_flags, tp->t_type);
if (error) {
goto undo_blocks;
}
tp->t_log_res = logspace;
tp->t_log_count = logcount;
}
/*
* Attempt to reserve the needed realtime extents by decrementing
* the number needed from the number available. This will
* fail if the count would go below zero.
*/
if (rtextents > 0) {
error = xfs_mod_incore_sb(tp->t_mountp, XFS_SBS_FREXTENTS,
-((int64_t)rtextents), rsvd);
if (error) {
error = XFS_ERROR(ENOSPC);
goto undo_log;
}
tp->t_rtx_res += rtextents;
}
return 0;
/*
* Error cases jump to one of these labels to undo any
* reservations which have already been performed.
*/
undo_log:
if (logspace > 0) {
if (flags & XFS_TRANS_PERM_LOG_RES) {
log_flags = XFS_LOG_REL_PERM_RESERV;
} else {
log_flags = 0;
}
xfs_log_done(tp->t_mountp, tp->t_ticket, NULL, log_flags);
tp->t_ticket = NULL;
tp->t_log_res = 0;
tp->t_flags &= ~XFS_TRANS_PERM_LOG_RES;
}
undo_blocks:
if (blocks > 0) {
(void) xfs_mod_incore_sb(tp->t_mountp, XFS_SBS_FDBLOCKS,
(int64_t)blocks, rsvd);
tp->t_blk_res = 0;
}
current_restore_flags_nested(&tp->t_pflags, PF_FSTRANS);
return error;
}
/*
* Allocate new inodes in the allocation group specified by agbp.
* Return 0 for success, else error code.
*/
STATIC int /* error code or 0 */
xfs_ialloc_ag_alloc(
xfs_trans_t *tp, /* transaction pointer */
xfs_buf_t *agbp, /* alloc group buffer */
int *alloc)
{
xfs_agi_t *agi; /* allocation group header */
xfs_alloc_arg_t args; /* allocation argument structure */
int blks_per_cluster; /* fs blocks per inode cluster */
xfs_btree_cur_t *cur; /* inode btree cursor */
xfs_daddr_t d; /* disk addr of buffer */
int error;
xfs_buf_t *fbuf; /* new free inodes' buffer */
xfs_dinode_t *free; /* new free inode structure */
int i; /* inode counter */
int j; /* block counter */
int nbufs; /* num bufs of new inodes */
xfs_agino_t newino; /* new first inode's number */
xfs_agino_t newlen; /* new number of inodes */
int ninodes; /* num inodes per buf */
xfs_agino_t thisino; /* current inode number, for loop */
int version; /* inode version number to use */
int isaligned = 0; /* inode allocation at stripe unit */
/* boundary */
args.tp = tp;
args.mp = tp->t_mountp;
/*
* Locking will ensure that we don't have two callers in here
* at one time.
*/
newlen = XFS_IALLOC_INODES(args.mp);
if (args.mp->m_maxicount &&
args.mp->m_sb.sb_icount + newlen > args.mp->m_maxicount)
return XFS_ERROR(ENOSPC);
args.minlen = args.maxlen = XFS_IALLOC_BLOCKS(args.mp);
/*
* First try to allocate inodes contiguous with the last-allocated
* chunk of inodes. If the filesystem is striped, this will fill
* an entire stripe unit with inodes.
*/
agi = XFS_BUF_TO_AGI(agbp);
newino = be32_to_cpu(agi->agi_newino);
args.agbno = XFS_AGINO_TO_AGBNO(args.mp, newino) +
XFS_IALLOC_BLOCKS(args.mp);
if (likely(newino != NULLAGINO &&
(args.agbno < be32_to_cpu(agi->agi_length)))) {
args.fsbno = XFS_AGB_TO_FSB(args.mp,
be32_to_cpu(agi->agi_seqno), args.agbno);
args.type = XFS_ALLOCTYPE_THIS_BNO;
args.mod = args.total = args.wasdel = args.isfl =
args.userdata = args.minalignslop = 0;
args.prod = 1;
args.alignment = 1;
/*
* Allow space for the inode btree to split.
*/
args.minleft = XFS_IN_MAXLEVELS(args.mp) - 1;
if ((error = xfs_alloc_vextent(&args)))
return error;
} else
args.fsbno = NULLFSBLOCK;
if (unlikely(args.fsbno == NULLFSBLOCK)) {
/*
* Set the alignment for the allocation.
* If stripe alignment is turned on then align at stripe unit
* boundary.
* If the cluster size is smaller than a filesystem block
* then we're doing I/O for inodes in filesystem block size
* pieces, so don't need alignment anyway.
*/
isaligned = 0;
if (args.mp->m_sinoalign) {
ASSERT(!(args.mp->m_flags & XFS_MOUNT_NOALIGN));
args.alignment = args.mp->m_dalign;
isaligned = 1;
} else if (XFS_SB_VERSION_HASALIGN(&args.mp->m_sb) &&
args.mp->m_sb.sb_inoalignmt >=
XFS_B_TO_FSBT(args.mp,
XFS_INODE_CLUSTER_SIZE(args.mp)))
args.alignment = args.mp->m_sb.sb_inoalignmt;
else
args.alignment = 1;
/*
* Need to figure out where to allocate the inode blocks.
* Ideally they should be spaced out through the a.g.
* For now, just allocate blocks up front.
*/
args.agbno = be32_to_cpu(agi->agi_root);
args.fsbno = XFS_AGB_TO_FSB(args.mp,
be32_to_cpu(agi->agi_seqno), args.agbno);
/*
* Allocate a fixed-size extent of inodes.
*/
args.type = XFS_ALLOCTYPE_NEAR_BNO;
args.mod = args.total = args.wasdel = args.isfl =
args.userdata = args.minalignslop = 0;
args.prod = 1;
/*
* Allow space for the inode btree to split.
*/
args.minleft = XFS_IN_MAXLEVELS(args.mp) - 1;
if ((error = xfs_alloc_vextent(&args)))
return error;
}
/*
* If stripe alignment is turned on, then try again with cluster
* alignment.
*/
if (isaligned && args.fsbno == NULLFSBLOCK) {
args.type = XFS_ALLOCTYPE_NEAR_BNO;
args.agbno = be32_to_cpu(agi->agi_root);
args.fsbno = XFS_AGB_TO_FSB(args.mp,
be32_to_cpu(agi->agi_seqno), args.agbno);
if (XFS_SB_VERSION_HASALIGN(&args.mp->m_sb) &&
args.mp->m_sb.sb_inoalignmt >=
XFS_B_TO_FSBT(args.mp, XFS_INODE_CLUSTER_SIZE(args.mp)))
args.alignment = args.mp->m_sb.sb_inoalignmt;
else
args.alignment = 1;
if ((error = xfs_alloc_vextent(&args)))
return error;
}
if (args.fsbno == NULLFSBLOCK) {
*alloc = 0;
return 0;
}
ASSERT(args.len == args.minlen);
/*
* Convert the results.
*/
newino = XFS_OFFBNO_TO_AGINO(args.mp, args.agbno, 0);
/*
* Loop over the new block(s), filling in the inodes.
* For small block sizes, manipulate the inodes in buffers
* which are multiples of the blocks size.
*/
if (args.mp->m_sb.sb_blocksize >= XFS_INODE_CLUSTER_SIZE(args.mp)) {
blks_per_cluster = 1;
nbufs = (int)args.len;
ninodes = args.mp->m_sb.sb_inopblock;
} else {
blks_per_cluster = XFS_INODE_CLUSTER_SIZE(args.mp) /
args.mp->m_sb.sb_blocksize;
nbufs = (int)args.len / blks_per_cluster;
ninodes = blks_per_cluster * args.mp->m_sb.sb_inopblock;
}
/*
* Figure out what version number to use in the inodes we create.
* If the superblock version has caught up to the one that supports
* the new inode format, then use the new inode version. Otherwise
* use the old version so that old kernels will continue to be
* able to use the file system.
*/
if (XFS_SB_VERSION_HASNLINK(&args.mp->m_sb))
version = XFS_DINODE_VERSION_2;
else
version = XFS_DINODE_VERSION_1;
for (j = 0; j < nbufs; j++) {
/*
* Get the block.
*/
d = XFS_AGB_TO_DADDR(args.mp, be32_to_cpu(agi->agi_seqno),
args.agbno + (j * blks_per_cluster));
fbuf = xfs_trans_get_buf(tp, args.mp->m_ddev_targp, d,
args.mp->m_bsize * blks_per_cluster,
XFS_BUF_LOCK);
ASSERT(fbuf);
ASSERT(!XFS_BUF_GETERROR(fbuf));
/*
* Set initial values for the inodes in this buffer.
*/
xfs_biozero(fbuf, 0, ninodes << args.mp->m_sb.sb_inodelog);
for (i = 0; i < ninodes; i++) {
free = XFS_MAKE_IPTR(args.mp, fbuf, i);
INT_SET(free->di_core.di_magic, ARCH_CONVERT, XFS_DINODE_MAGIC);
INT_SET(free->di_core.di_version, ARCH_CONVERT, version);
INT_SET(free->di_next_unlinked, ARCH_CONVERT, NULLAGINO);
xfs_ialloc_log_di(tp, fbuf, i,
XFS_DI_CORE_BITS | XFS_DI_NEXT_UNLINKED);
}
xfs_trans_inode_alloc_buf(tp, fbuf);
}
be32_add(&agi->agi_count, newlen);
be32_add(&agi->agi_freecount, newlen);
down_read(&args.mp->m_peraglock);
args.mp->m_perag[be32_to_cpu(agi->agi_seqno)].pagi_freecount += newlen;
up_read(&args.mp->m_peraglock);
agi->agi_newino = cpu_to_be32(newino);
/*
* Insert records describing the new inode chunk into the btree.
*/
cur = xfs_btree_init_cursor(args.mp, tp, agbp,
be32_to_cpu(agi->agi_seqno),
XFS_BTNUM_INO, (xfs_inode_t *)0, 0);
for (thisino = newino;
thisino < newino + newlen;
thisino += XFS_INODES_PER_CHUNK) {
if ((error = xfs_inobt_lookup_eq(cur, thisino,
XFS_INODES_PER_CHUNK, XFS_INOBT_ALL_FREE, &i))) {
xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
return error;
}
ASSERT(i == 0);
if ((error = xfs_inobt_insert(cur, &i))) {
xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
return error;
}
ASSERT(i == 1);
}
xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR);
/*
* Log allocation group header fields
*/
xfs_ialloc_log_agi(tp, agbp,
XFS_AGI_COUNT | XFS_AGI_FREECOUNT | XFS_AGI_NEWINO);
/*
* Modify/log superblock values for inode count and inode free count.
*/
xfs_trans_mod_sb(tp, XFS_TRANS_SB_ICOUNT, (long)newlen);
xfs_trans_mod_sb(tp, XFS_TRANS_SB_IFREE, (long)newlen);
*alloc = 1;
return 0;
}
/* copied from xfs_ioctl.c */
STATIC int
xfs_ioc_bulkstat_compat(
xfs_mount_t *mp,
unsigned int cmd,
void __user *arg)
{
compat_xfs_fsop_bulkreq_t __user *p32 = (void __user *)arg;
u32 addr;
xfs_fsop_bulkreq_t bulkreq;
int count; /* # of records returned */
xfs_ino_t inlast; /* last inode number */
int done;
int error;
/* done = 1 if there are more stats to get and if bulkstat */
/* should be called again (unused here, but used in dmapi) */
if (!capable(CAP_SYS_ADMIN))
return -EPERM;
if (XFS_FORCED_SHUTDOWN(mp))
return -XFS_ERROR(EIO);
if (get_user(addr, &p32->lastip))
return -EFAULT;
bulkreq.lastip = compat_ptr(addr);
if (get_user(bulkreq.icount, &p32->icount) ||
get_user(addr, &p32->ubuffer))
return -EFAULT;
bulkreq.ubuffer = compat_ptr(addr);
if (get_user(addr, &p32->ocount))
return -EFAULT;
bulkreq.ocount = compat_ptr(addr);
if (copy_from_user(&inlast, bulkreq.lastip, sizeof(__s64)))
return -XFS_ERROR(EFAULT);
if ((count = bulkreq.icount) <= 0)
return -XFS_ERROR(EINVAL);
if (bulkreq.ubuffer == NULL)
return -XFS_ERROR(EINVAL);
if (cmd == XFS_IOC_FSINUMBERS)
error = xfs_inumbers(mp, &inlast, &count,
bulkreq.ubuffer, xfs_inumbers_fmt_compat);
else {
/* declare a var to get a warning in case the type changes */
bulkstat_one_fmt_pf formatter = xfs_bulkstat_one_fmt_compat;
error = xfs_bulkstat(mp, &inlast, &count,
xfs_bulkstat_one, formatter,
sizeof(compat_xfs_bstat_t), bulkreq.ubuffer,
BULKSTAT_FG_QUICK, &done);
}
if (error)
return -error;
if (bulkreq.ocount != NULL) {
if (copy_to_user(bulkreq.lastip, &inlast,
sizeof(xfs_ino_t)))
return -XFS_ERROR(EFAULT);
if (copy_to_user(bulkreq.ocount, &count, sizeof(count)))
return -XFS_ERROR(EFAULT);
}
return 0;
}
/*
* 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_map_blocks(
struct inode *inode,
loff_t offset,
struct xfs_bmbt_irec *imap,
int type,
int nonblocking)
{
struct xfs_inode *ip = XFS_I(inode);
struct xfs_mount *mp = ip->i_mount;
ssize_t count = 1 << inode->i_blkbits;
xfs_fileoff_t offset_fsb, end_fsb;
int error = 0;
int bmapi_flags = XFS_BMAPI_ENTIRE;
int nimaps = 1;
if (XFS_FORCED_SHUTDOWN(mp))
return -XFS_ERROR(EIO);
if (type == IO_UNWRITTEN)
bmapi_flags |= XFS_BMAPI_IGSTATE;
if (!xfs_ilock_nowait(ip, XFS_ILOCK_SHARED)) {
if (nonblocking)
return -XFS_ERROR(EAGAIN);
xfs_ilock(ip, XFS_ILOCK_SHARED);
}
ASSERT(ip->i_d.di_format != XFS_DINODE_FMT_BTREE ||
(ip->i_df.if_flags & XFS_IFEXTENTS));
ASSERT(offset <= mp->m_maxioffset);
if (offset + count > mp->m_maxioffset)
count = mp->m_maxioffset - offset;
end_fsb = XFS_B_TO_FSB(mp, (xfs_ufsize_t)offset + count);
offset_fsb = XFS_B_TO_FSBT(mp, offset);
error = xfs_bmapi_read(ip, offset_fsb, end_fsb - offset_fsb,
imap, &nimaps, bmapi_flags);
xfs_iunlock(ip, XFS_ILOCK_SHARED);
if (error)
return -XFS_ERROR(error);
if (type == IO_DELALLOC &&
(!nimaps || isnullstartblock(imap->br_startblock))) {
error = xfs_iomap_write_allocate(ip, offset, count, imap);
if (!error)
trace_xfs_map_blocks_alloc(ip, offset, count, type, imap);
return -XFS_ERROR(error);
}
#ifdef DEBUG
if (type == IO_UNWRITTEN) {
ASSERT(nimaps);
ASSERT(imap->br_startblock != HOLESTARTBLOCK);
ASSERT(imap->br_startblock != DELAYSTARTBLOCK);
}
#endif
if (nimaps)
trace_xfs_map_blocks_found(ip, offset, count, type, imap);
return 0;
}
/*
* Allocates a new inode from disk and return a pointer to the
* incore copy. This routine will internally commit the current
* transaction and allocate a new one if the Space Manager needed
* to do an allocation to replenish the inode free-list.
*
* This routine is designed to be called from xfs_create and
* xfs_create_dir.
*
*/
int
xfs_dir_ialloc(
xfs_trans_t **tpp, /* input: current transaction;
output: may be a new transaction. */
xfs_inode_t *dp, /* directory within whose allocate
the inode. */
umode_t mode,
xfs_nlink_t nlink,
xfs_dev_t rdev,
prid_t prid, /* project id */
int okalloc, /* ok to allocate new space */
xfs_inode_t **ipp, /* pointer to inode; it will be
locked. */
int *committed)
{
xfs_trans_t *tp;
xfs_trans_t *ntp;
xfs_inode_t *ip;
xfs_buf_t *ialloc_context = NULL;
int code;
uint log_res;
uint log_count;
void *dqinfo;
uint tflags;
tp = *tpp;
ASSERT(tp->t_flags & XFS_TRANS_PERM_LOG_RES);
/*
* xfs_ialloc will return a pointer to an incore inode if
* the Space Manager has an available inode on the free
* list. Otherwise, it will do an allocation and replenish
* the freelist. Since we can only do one allocation per
* transaction without deadlocks, we will need to commit the
* current transaction and start a new one. We will then
* need to call xfs_ialloc again to get the inode.
*
* If xfs_ialloc did an allocation to replenish the freelist,
* it returns the bp containing the head of the freelist as
* ialloc_context. We will hold a lock on it across the
* transaction commit so that no other process can steal
* the inode(s) that we've just allocated.
*/
code = xfs_ialloc(tp, dp, mode, nlink, rdev, prid, okalloc,
&ialloc_context, &ip);
/*
* Return an error if we were unable to allocate a new inode.
* This should only happen if we run out of space on disk or
* encounter a disk error.
*/
if (code) {
*ipp = NULL;
return code;
}
if (!ialloc_context && !ip) {
*ipp = NULL;
return XFS_ERROR(ENOSPC);
}
/*
* If the AGI buffer is non-NULL, then we were unable to get an
* inode in one operation. We need to commit the current
* transaction and call xfs_ialloc() again. It is guaranteed
* to succeed the second time.
*/
if (ialloc_context) {
/*
* Normally, xfs_trans_commit releases all the locks.
* We call bhold to hang on to the ialloc_context across
* the commit. Holding this buffer prevents any other
* processes from doing any allocations in this
* allocation group.
*/
xfs_trans_bhold(tp, ialloc_context);
/*
* Save the log reservation so we can use
* them in the next transaction.
*/
log_res = xfs_trans_get_log_res(tp);
log_count = xfs_trans_get_log_count(tp);
/*
* We want the quota changes to be associated with the next
* transaction, NOT this one. So, detach the dqinfo from this
* and attach it to the next transaction.
*/
dqinfo = NULL;
tflags = 0;
if (tp->t_dqinfo) {
dqinfo = (void *)tp->t_dqinfo;
tp->t_dqinfo = NULL;
tflags = tp->t_flags & XFS_TRANS_DQ_DIRTY;
tp->t_flags &= ~(XFS_TRANS_DQ_DIRTY);
}
ntp = xfs_trans_dup(tp);
code = xfs_trans_commit(tp, 0);
tp = ntp;
if (committed != NULL) {
*committed = 1;
}
/*
* If we get an error during the commit processing,
* release the buffer that is still held and return
* to the caller.
*/
if (code) {
xfs_buf_relse(ialloc_context);
if (dqinfo) {
tp->t_dqinfo = dqinfo;
xfs_trans_free_dqinfo(tp);
}
*tpp = ntp;
*ipp = NULL;
return code;
}
/*
* transaction commit worked ok so we can drop the extra ticket
* reference that we gained in xfs_trans_dup()
*/
xfs_log_ticket_put(tp->t_ticket);
code = xfs_trans_reserve(tp, 0, log_res, 0,
XFS_TRANS_PERM_LOG_RES, log_count);
/*
* Re-attach the quota info that we detached from prev trx.
*/
if (dqinfo) {
tp->t_dqinfo = dqinfo;
tp->t_flags |= tflags;
}
if (code) {
xfs_buf_relse(ialloc_context);
*tpp = ntp;
*ipp = NULL;
return code;
}
xfs_trans_bjoin(tp, ialloc_context);
/*
* Call ialloc again. Since we've locked out all
* other allocations in this allocation group,
* this call should always succeed.
*/
code = xfs_ialloc(tp, dp, mode, nlink, rdev, prid,
okalloc, &ialloc_context, &ip);
/*
* If we get an error at this point, return to the caller
* so that the current transaction can be aborted.
*/
if (code) {
*tpp = tp;
*ipp = NULL;
return code;
}
ASSERT(!ialloc_context && ip);
} else {
if (committed != NULL)
*committed = 0;
}
*ipp = ip;
*tpp = tp;
return 0;
}
/*
* Note: some of the ioctl's return positive numbers as a
* byte count indicating success, such as readlink_by_handle.
* So we don't "sign flip" like most other routines. This means
* true errors need to be returned as a negative value.
*/
long
xfs_file_ioctl(
struct file *filp,
unsigned int cmd,
unsigned long p)
{
struct inode *inode = filp->f_path.dentry->d_inode;
struct xfs_inode *ip = XFS_I(inode);
struct xfs_mount *mp = ip->i_mount;
void __user *arg = (void __user *)p;
int ioflags = 0;
int error;
if (filp->f_mode & FMODE_NOCMTIME)
ioflags |= IO_INVIS;
xfs_itrace_entry(ip);
switch (cmd) {
case XFS_IOC_ALLOCSP:
case XFS_IOC_FREESP:
case XFS_IOC_RESVSP:
case XFS_IOC_UNRESVSP:
case XFS_IOC_ALLOCSP64:
case XFS_IOC_FREESP64:
case XFS_IOC_RESVSP64:
case XFS_IOC_UNRESVSP64: {
xfs_flock64_t bf;
if (copy_from_user(&bf, arg, sizeof(bf)))
return -XFS_ERROR(EFAULT);
return xfs_ioc_space(ip, inode, filp, ioflags, cmd, &bf);
}
case XFS_IOC_DIOINFO: {
struct dioattr da;
xfs_buftarg_t *target =
XFS_IS_REALTIME_INODE(ip) ?
mp->m_rtdev_targp : mp->m_ddev_targp;
da.d_mem = da.d_miniosz = 1 << target->bt_sshift;
da.d_maxiosz = INT_MAX & ~(da.d_miniosz - 1);
if (copy_to_user(arg, &da, sizeof(da)))
return -XFS_ERROR(EFAULT);
return 0;
}
case XFS_IOC_FSBULKSTAT_SINGLE:
case XFS_IOC_FSBULKSTAT:
case XFS_IOC_FSINUMBERS:
return xfs_ioc_bulkstat(mp, cmd, arg);
case XFS_IOC_FSGEOMETRY_V1:
return xfs_ioc_fsgeometry_v1(mp, arg);
case XFS_IOC_FSGEOMETRY:
return xfs_ioc_fsgeometry(mp, arg);
case XFS_IOC_GETVERSION:
return put_user(inode->i_generation, (int __user *)arg);
case XFS_IOC_FSGETXATTR:
return xfs_ioc_fsgetxattr(ip, 0, arg);
case XFS_IOC_FSGETXATTRA:
return xfs_ioc_fsgetxattr(ip, 1, arg);
case XFS_IOC_FSSETXATTR:
return xfs_ioc_fssetxattr(ip, filp, arg);
case XFS_IOC_GETXFLAGS:
return xfs_ioc_getxflags(ip, arg);
case XFS_IOC_SETXFLAGS:
return xfs_ioc_setxflags(ip, filp, arg);
case XFS_IOC_FSSETDM: {
struct fsdmidata dmi;
if (copy_from_user(&dmi, arg, sizeof(dmi)))
return -XFS_ERROR(EFAULT);
error = xfs_set_dmattrs(ip, dmi.fsd_dmevmask,
dmi.fsd_dmstate);
return -error;
}
case XFS_IOC_GETBMAP:
case XFS_IOC_GETBMAPA:
return xfs_ioc_getbmap(ip, ioflags, cmd, arg);
case XFS_IOC_GETBMAPX:
return xfs_ioc_getbmapx(ip, arg);
case XFS_IOC_FD_TO_HANDLE:
case XFS_IOC_PATH_TO_HANDLE:
case XFS_IOC_PATH_TO_FSHANDLE: {
xfs_fsop_handlereq_t hreq;
if (copy_from_user(&hreq, arg, sizeof(hreq)))
return -XFS_ERROR(EFAULT);
return xfs_find_handle(cmd, &hreq);
}
case XFS_IOC_OPEN_BY_HANDLE: {
xfs_fsop_handlereq_t hreq;
if (copy_from_user(&hreq, arg, sizeof(xfs_fsop_handlereq_t)))
return -XFS_ERROR(EFAULT);
return xfs_open_by_handle(filp, &hreq);
}
case XFS_IOC_FSSETDM_BY_HANDLE:
return xfs_fssetdm_by_handle(filp, arg);
case XFS_IOC_READLINK_BY_HANDLE: {
xfs_fsop_handlereq_t hreq;
if (copy_from_user(&hreq, arg, sizeof(xfs_fsop_handlereq_t)))
return -XFS_ERROR(EFAULT);
return xfs_readlink_by_handle(filp, &hreq);
}
case XFS_IOC_ATTRLIST_BY_HANDLE:
return xfs_attrlist_by_handle(filp, arg);
case XFS_IOC_ATTRMULTI_BY_HANDLE:
return xfs_attrmulti_by_handle(filp, arg);
case XFS_IOC_SWAPEXT: {
struct xfs_swapext sxp;
if (copy_from_user(&sxp, arg, sizeof(xfs_swapext_t)))
return -XFS_ERROR(EFAULT);
error = xfs_swapext(&sxp);
return -error;
}
case XFS_IOC_FSCOUNTS: {
xfs_fsop_counts_t out;
error = xfs_fs_counts(mp, &out);
if (error)
return -error;
if (copy_to_user(arg, &out, sizeof(out)))
return -XFS_ERROR(EFAULT);
return 0;
}
case XFS_IOC_SET_RESBLKS: {
xfs_fsop_resblks_t inout;
__uint64_t in;
if (!capable(CAP_SYS_ADMIN))
return -EPERM;
if (copy_from_user(&inout, arg, sizeof(inout)))
return -XFS_ERROR(EFAULT);
/* input parameter is passed in resblks field of structure */
in = inout.resblks;
error = xfs_reserve_blocks(mp, &in, &inout);
if (error)
return -error;
if (copy_to_user(arg, &inout, sizeof(inout)))
return -XFS_ERROR(EFAULT);
return 0;
}
case XFS_IOC_GET_RESBLKS: {
xfs_fsop_resblks_t out;
if (!capable(CAP_SYS_ADMIN))
return -EPERM;
error = xfs_reserve_blocks(mp, NULL, &out);
if (error)
return -error;
if (copy_to_user(arg, &out, sizeof(out)))
return -XFS_ERROR(EFAULT);
return 0;
}
case XFS_IOC_FSGROWFSDATA: {
xfs_growfs_data_t in;
if (copy_from_user(&in, arg, sizeof(in)))
return -XFS_ERROR(EFAULT);
error = xfs_growfs_data(mp, &in);
return -error;
}
case XFS_IOC_FSGROWFSLOG: {
xfs_growfs_log_t in;
if (copy_from_user(&in, arg, sizeof(in)))
return -XFS_ERROR(EFAULT);
error = xfs_growfs_log(mp, &in);
return -error;
}
case XFS_IOC_FSGROWFSRT: {
xfs_growfs_rt_t in;
if (copy_from_user(&in, arg, sizeof(in)))
return -XFS_ERROR(EFAULT);
error = xfs_growfs_rt(mp, &in);
return -error;
}
case XFS_IOC_GOINGDOWN: {
__uint32_t in;
if (!capable(CAP_SYS_ADMIN))
return -EPERM;
if (get_user(in, (__uint32_t __user *)arg))
return -XFS_ERROR(EFAULT);
error = xfs_fs_goingdown(mp, in);
return -error;
}
case XFS_IOC_ERROR_INJECTION: {
xfs_error_injection_t in;
if (!capable(CAP_SYS_ADMIN))
return -EPERM;
if (copy_from_user(&in, arg, sizeof(in)))
return -XFS_ERROR(EFAULT);
error = xfs_errortag_add(in.errtag, mp);
return -error;
}
case XFS_IOC_ERROR_CLEARALL:
if (!capable(CAP_SYS_ADMIN))
return -EPERM;
error = xfs_errortag_clearall(mp, 1);
return -error;
default:
return -ENOTTY;
}
}
/*
* Add a block to the directory.
* This routine is for data and free blocks, not leaf/node blocks
* which are handled by xfs_da_grow_inode.
*/
int
xfs_dir2_grow_inode(
xfs_da_args_t *args,
int space, /* v2 dir's space XFS_DIR2_xxx_SPACE */
xfs_dir2_db_t *dbp) /* out: block number added */
{
xfs_fileoff_t bno; /* directory offset of new block */
int count; /* count of filesystem blocks */
xfs_inode_t *dp; /* incore directory inode */
int error;
int got; /* blocks actually mapped */
int i;
xfs_bmbt_irec_t map; /* single structure for bmap */
int mapi; /* mapping index */
xfs_bmbt_irec_t *mapp; /* bmap mapping structure(s) */
xfs_mount_t *mp;
int nmap; /* number of bmap entries */
xfs_trans_t *tp;
xfs_dir2_trace_args_s("grow_inode", args, space);
dp = args->dp;
tp = args->trans;
mp = dp->i_mount;
/*
* Set lowest possible block in the space requested.
*/
bno = XFS_B_TO_FSBT(mp, space * XFS_DIR2_SPACE_SIZE);
count = mp->m_dirblkfsbs;
/*
* Find the first hole for our block.
*/
if ((error = xfs_bmap_first_unused(tp, dp, count, &bno, XFS_DATA_FORK)))
return error;
nmap = 1;
ASSERT(args->firstblock != NULL);
/*
* Try mapping the new block contiguously (one extent).
*/
if ((error = xfs_bmapi(tp, dp, bno, count,
XFS_BMAPI_WRITE|XFS_BMAPI_METADATA|XFS_BMAPI_CONTIG,
args->firstblock, args->total, &map, &nmap,
args->flist, NULL)))
return error;
ASSERT(nmap <= 1);
if (nmap == 1) {
mapp = ↦
mapi = 1;
}
/*
* Didn't work and this is a multiple-fsb directory block.
* Try again with contiguous flag turned on.
*/
else if (nmap == 0 && count > 1) {
xfs_fileoff_t b; /* current file offset */
/*
* Space for maximum number of mappings.
*/
mapp = kmem_alloc(sizeof(*mapp) * count, KM_SLEEP);
/*
* Iterate until we get to the end of our block.
*/
for (b = bno, mapi = 0; b < bno + count; ) {
int c; /* current fsb count */
/*
* Can't map more than MAX_NMAP at once.
*/
nmap = MIN(XFS_BMAP_MAX_NMAP, count);
c = (int)(bno + count - b);
if ((error = xfs_bmapi(tp, dp, b, c,
XFS_BMAPI_WRITE|XFS_BMAPI_METADATA,
args->firstblock, args->total,
&mapp[mapi], &nmap, args->flist,
NULL))) {
kmem_free(mapp, sizeof(*mapp) * count);
return error;
}
if (nmap < 1)
break;
/*
* Add this bunch into our table, go to the next offset.
*/
mapi += nmap;
b = mapp[mapi - 1].br_startoff +
mapp[mapi - 1].br_blockcount;
}
}
/*
* Didn't work.
*/
else {
mapi = 0;
mapp = NULL;
}
/*
* See how many fsb's we got.
*/
for (i = 0, got = 0; i < mapi; i++)
got += mapp[i].br_blockcount;
/*
* Didn't get enough fsb's, or the first/last block's are wrong.
*/
if (got != count || mapp[0].br_startoff != bno ||
mapp[mapi - 1].br_startoff + mapp[mapi - 1].br_blockcount !=
bno + count) {
if (mapp != &map)
kmem_free(mapp, sizeof(*mapp) * count);
return XFS_ERROR(ENOSPC);
}
/*
* Done with the temporary mapping table.
*/
if (mapp != &map)
kmem_free(mapp, sizeof(*mapp) * count);
*dbp = xfs_dir2_da_to_db(mp, (xfs_dablk_t)bno);
/*
* Update file's size if this is the data space and it grew.
*/
if (space == XFS_DIR2_DATA_SPACE) {
xfs_fsize_t size; /* directory file (data) size */
size = XFS_FSB_TO_B(mp, bno + count);
if (size > dp->i_d.di_size) {
dp->i_d.di_size = size;
xfs_trans_log_inode(tp, dp, XFS_ILOG_CORE);
}
}
return 0;
}
STATIC int
xfs_attrmulti_by_handle(
struct file *parfilp,
void __user *arg)
{
int error;
xfs_attr_multiop_t *ops;
xfs_fsop_attrmulti_handlereq_t am_hreq;
struct dentry *dentry;
unsigned int i, size;
char *attr_name;
if (!capable(CAP_SYS_ADMIN))
return -XFS_ERROR(EPERM);
if (copy_from_user(&am_hreq, arg, sizeof(xfs_fsop_attrmulti_handlereq_t)))
return -XFS_ERROR(EFAULT);
dentry = xfs_handlereq_to_dentry(parfilp, &am_hreq.hreq);
if (IS_ERR(dentry))
return PTR_ERR(dentry);
error = E2BIG;
size = am_hreq.opcount * sizeof(xfs_attr_multiop_t);
if (!size || size > 16 * PAGE_SIZE)
goto out_dput;
ops = memdup_user(am_hreq.ops, size);
if (IS_ERR(ops)) {
error = PTR_ERR(ops);
goto out_dput;
}
attr_name = kmalloc(MAXNAMELEN, GFP_KERNEL);
if (!attr_name)
goto out_kfree_ops;
error = 0;
for (i = 0; i < am_hreq.opcount; i++) {
ops[i].am_error = strncpy_from_user(attr_name,
ops[i].am_attrname, MAXNAMELEN);
if (ops[i].am_error == 0 || ops[i].am_error == MAXNAMELEN)
error = -ERANGE;
if (ops[i].am_error < 0)
break;
switch (ops[i].am_opcode) {
case ATTR_OP_GET:
ops[i].am_error = xfs_attrmulti_attr_get(
dentry->d_inode, attr_name,
ops[i].am_attrvalue, &ops[i].am_length,
ops[i].am_flags);
break;
case ATTR_OP_SET:
ops[i].am_error = mnt_want_write(parfilp->f_path.mnt);
if (ops[i].am_error)
break;
ops[i].am_error = xfs_attrmulti_attr_set(
dentry->d_inode, attr_name,
ops[i].am_attrvalue, ops[i].am_length,
ops[i].am_flags);
mnt_drop_write(parfilp->f_path.mnt);
break;
case ATTR_OP_REMOVE:
ops[i].am_error = mnt_want_write(parfilp->f_path.mnt);
if (ops[i].am_error)
break;
ops[i].am_error = xfs_attrmulti_attr_remove(
dentry->d_inode, attr_name,
ops[i].am_flags);
mnt_drop_write(parfilp->f_path.mnt);
break;
default:
ops[i].am_error = EINVAL;
}
}
if (copy_to_user(am_hreq.ops, ops, size))
error = XFS_ERROR(EFAULT);
kfree(attr_name);
out_kfree_ops:
kfree(ops);
out_dput:
dput(dentry);
return -error;
}
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;
}
ssize_t /* bytes read, or (-) error */
xfs_read(
xfs_inode_t *ip,
struct kiocb *iocb,
const struct iovec *iovp,
unsigned int segs,
loff_t *offset,
int ioflags)
{
struct file *file = iocb->ki_filp;
struct inode *inode = file->f_mapping->host;
xfs_mount_t *mp = ip->i_mount;
size_t size = 0;
ssize_t ret = 0;
xfs_fsize_t n;
unsigned long seg;
XFS_STATS_INC(xs_read_calls);
/* 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 (unlikely(ioflags & IO_ISDIRECT)) {
xfs_buftarg_t *target =
XFS_IS_REALTIME_INODE(ip) ?
mp->m_rtdev_targp : mp->m_ddev_targp;
if ((*offset & target->bt_smask) ||
(size & target->bt_smask)) {
if (*offset == ip->i_size) {
return (0);
}
return -XFS_ERROR(EINVAL);
}
}
n = XFS_MAXIOFFSET(mp) - *offset;
if ((n <= 0) || (size == 0))
return 0;
if (n < size)
size = n;
if (XFS_FORCED_SHUTDOWN(mp))
return -EIO;
if (unlikely(ioflags & IO_ISDIRECT))
mutex_lock(&inode->i_mutex);
xfs_ilock(ip, XFS_IOLOCK_SHARED);
if (DM_EVENT_ENABLED(ip, DM_EVENT_READ) && !(ioflags & IO_INVIS)) {
int dmflags = FILP_DELAY_FLAG(file) | DM_SEM_FLAG_RD(ioflags);
int iolock = XFS_IOLOCK_SHARED;
ret = -XFS_SEND_DATA(mp, DM_EVENT_READ, ip, *offset, size,
dmflags, &iolock);
if (ret) {
xfs_iunlock(ip, XFS_IOLOCK_SHARED);
if (unlikely(ioflags & IO_ISDIRECT))
mutex_unlock(&inode->i_mutex);
return ret;
}
}
if (unlikely(ioflags & IO_ISDIRECT)) {
if (inode->i_mapping->nrpages)
ret = -xfs_flushinval_pages(ip, (*offset & PAGE_CACHE_MASK),
-1, FI_REMAPF_LOCKED);
mutex_unlock(&inode->i_mutex);
if (ret) {
xfs_iunlock(ip, XFS_IOLOCK_SHARED);
return ret;
}
}
xfs_rw_enter_trace(XFS_READ_ENTER, ip,
(void *)iovp, segs, *offset, ioflags);
iocb->ki_pos = *offset;
ret = generic_file_aio_read(iocb, iovp, segs, *offset);
if (ret == -EIOCBQUEUED && !(ioflags & IO_ISAIO))
ret = wait_on_sync_kiocb(iocb);
if (ret > 0)
XFS_STATS_ADD(xs_read_bytes, ret);
xfs_iunlock(ip, XFS_IOLOCK_SHARED);
return ret;
}
STATIC ssize_t
xfs_file_aio_read(
struct kiocb *iocb,
const struct iovec *iovp,
unsigned long nr_segs,
loff_t pos)
{
struct file *file = iocb->ki_filp;
struct inode *inode = file->f_mapping->host;
struct xfs_inode *ip = XFS_I(inode);
struct xfs_mount *mp = ip->i_mount;
size_t size = 0;
ssize_t ret = 0;
int ioflags = 0;
xfs_fsize_t n;
unsigned long seg;
XFS_STATS_INC(xs_read_calls);
BUG_ON(iocb->ki_pos != pos);
if (unlikely(file->f_flags & O_DIRECT))
ioflags |= IO_ISDIRECT;
if (file->f_mode & FMODE_NOCMTIME)
ioflags |= IO_INVIS;
/* START copy & waste from filemap.c */
for (seg = 0; seg < nr_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 (unlikely(ioflags & IO_ISDIRECT)) {
xfs_buftarg_t *target =
XFS_IS_REALTIME_INODE(ip) ?
mp->m_rtdev_targp : mp->m_ddev_targp;
if ((iocb->ki_pos & target->bt_smask) ||
(size & target->bt_smask)) {
if (iocb->ki_pos == ip->i_size)
return 0;
return -XFS_ERROR(EINVAL);
}
}
n = XFS_MAXIOFFSET(mp) - iocb->ki_pos;
if (n <= 0 || size == 0)
return 0;
if (n < size)
size = n;
if (XFS_FORCED_SHUTDOWN(mp))
return -EIO;
/*
* Locking is a bit tricky here. If we take an exclusive lock
* for direct IO, we effectively serialise all new concurrent
* read IO to this file and block it behind IO that is currently in
* progress because IO in progress holds the IO lock shared. We only
* need to hold the lock exclusive to blow away the page cache, so
* only take lock exclusively if the page cache needs invalidation.
* This allows the normal direct IO case of no page cache pages to
* proceeed concurrently without serialisation.
*/
xfs_rw_ilock(ip, XFS_IOLOCK_SHARED);
if ((ioflags & IO_ISDIRECT) && inode->i_mapping->nrpages) {
xfs_rw_iunlock(ip, XFS_IOLOCK_SHARED);
xfs_rw_ilock(ip, XFS_IOLOCK_EXCL);
if (inode->i_mapping->nrpages) {
ret = -xfs_flushinval_pages(ip,
(iocb->ki_pos & PAGE_CACHE_MASK),
-1, FI_REMAPF_LOCKED);
if (ret) {
xfs_rw_iunlock(ip, XFS_IOLOCK_EXCL);
return ret;
}
}
xfs_rw_ilock_demote(ip, XFS_IOLOCK_EXCL);
}
trace_xfs_file_read(ip, size, iocb->ki_pos, ioflags);
ret = generic_file_aio_read(iocb, iovp, nr_segs, iocb->ki_pos);
if (ret > 0)
XFS_STATS_ADD(xs_read_bytes, ret);
xfs_rw_iunlock(ip, XFS_IOLOCK_SHARED);
return ret;
}
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;
xfs_itrace_entry(src_dp);
xfs_itrace_entry(target_dp);
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_dp, DM_RIGHT_NULL,
target_dp, DM_RIGHT_NULL,
src_name->name, target_name->name,
0, 0, 0);
if (error)
return error;
}
/* Return through std_return after this point. */
new_parent = (src_dp != target_dp);
src_is_directory = ((src_ip->i_d.di_mode & S_IFMT) == S_IFDIR);
if (src_is_directory) {
/*
* Check for link count overflow on target_dp
*/
if (target_ip == NULL && new_parent &&
target_dp->i_d.di_nlink >= XFS_MAXLINK) {
error = XFS_ERROR(EMLINK);
goto std_return;
}
}
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. 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.
*/
IHOLD(src_dp);
xfs_trans_ijoin(tp, src_dp, XFS_ILOCK_EXCL);
if (new_parent) {
IHOLD(target_dp);
xfs_trans_ijoin(tp, target_dp, XFS_ILOCK_EXCL);
}
IHOLD(src_ip);
xfs_trans_ijoin(tp, src_ip, XFS_ILOCK_EXCL);
if (target_ip) {
IHOLD(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) &&
(target_dp->i_d.di_projid != src_ip->i_d.di_projid))) {
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_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,
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;
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_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_ip->i_ino,
&first_block, &free_list, spaceres);
if (error)
goto abort_return;
xfs_ichgtime(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.
*/
error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
/* 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_dp, DM_RIGHT_NULL,
target_dp, DM_RIGHT_NULL,
src_name->name, target_name->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;
}
int
xfs_qm_scall_quotaon(
xfs_mount_t *mp,
uint flags)
{
int error;
uint qf;
__int64_t sbflags;
flags &= (XFS_ALL_QUOTA_ACCT | XFS_ALL_QUOTA_ENFD);
flags &= ~(XFS_ALL_QUOTA_ACCT);
sbflags = 0;
if (flags == 0) {
xfs_debug(mp, "%s: zero flags, m_qflags=%x\n",
__func__, mp->m_qflags);
return XFS_ERROR(EINVAL);
}
ASSERT((flags & XFS_ALL_QUOTA_ACCT) == 0);
if (((flags & XFS_UQUOTA_ACCT) == 0 &&
(mp->m_sb.sb_qflags & XFS_UQUOTA_ACCT) == 0 &&
(flags & XFS_UQUOTA_ENFD))
||
((flags & XFS_PQUOTA_ACCT) == 0 &&
(mp->m_sb.sb_qflags & XFS_PQUOTA_ACCT) == 0 &&
(flags & XFS_GQUOTA_ACCT) == 0 &&
(mp->m_sb.sb_qflags & XFS_GQUOTA_ACCT) == 0 &&
(flags & XFS_OQUOTA_ENFD))) {
xfs_debug(mp,
"%s: Can't enforce without acct, flags=%x sbflags=%x\n",
__func__, flags, mp->m_sb.sb_qflags);
return XFS_ERROR(EINVAL);
}
if ((mp->m_qflags & flags) == flags)
return XFS_ERROR(EEXIST);
spin_lock(&mp->m_sb_lock);
qf = mp->m_sb.sb_qflags;
mp->m_sb.sb_qflags = qf | flags;
spin_unlock(&mp->m_sb_lock);
if ((qf & flags) == flags && sbflags == 0)
return XFS_ERROR(EEXIST);
sbflags |= XFS_SB_QFLAGS;
if ((error = xfs_qm_write_sb_changes(mp, sbflags)))
return (error);
if (((mp->m_sb.sb_qflags & XFS_UQUOTA_ACCT) !=
(mp->m_qflags & XFS_UQUOTA_ACCT)) ||
((mp->m_sb.sb_qflags & XFS_PQUOTA_ACCT) !=
(mp->m_qflags & XFS_PQUOTA_ACCT)) ||
((mp->m_sb.sb_qflags & XFS_GQUOTA_ACCT) !=
(mp->m_qflags & XFS_GQUOTA_ACCT)) ||
(flags & XFS_ALL_QUOTA_ENFD) == 0)
return (0);
if (! XFS_IS_QUOTA_RUNNING(mp))
return XFS_ERROR(ESRCH);
mutex_lock(&mp->m_quotainfo->qi_quotaofflock);
mp->m_qflags |= (flags & XFS_ALL_QUOTA_ENFD);
mutex_unlock(&mp->m_quotainfo->qi_quotaofflock);
return (0);
}
