/*
* Truncate file. Must have write permission and not be a directory.
*/
int
xfs_setattr_size(
struct xfs_inode *ip,
struct iattr *iattr)
{
struct xfs_mount *mp = ip->i_mount;
struct inode *inode = VFS_I(ip);
xfs_off_t oldsize, newsize;
struct xfs_trans *tp;
int error;
uint lock_flags = 0;
uint commit_flags = 0;
trace_xfs_setattr(ip);
if (mp->m_flags & XFS_MOUNT_RDONLY)
return XFS_ERROR(EROFS);
if (XFS_FORCED_SHUTDOWN(mp))
return XFS_ERROR(EIO);
error = -inode_change_ok(inode, iattr);
if (error)
return XFS_ERROR(error);
ASSERT(xfs_isilocked(ip, XFS_IOLOCK_EXCL));
ASSERT(S_ISREG(ip->i_d.di_mode));
ASSERT((iattr->ia_valid & (ATTR_UID|ATTR_GID|ATTR_ATIME|ATTR_ATIME_SET|
ATTR_MTIME_SET|ATTR_KILL_PRIV|ATTR_TIMES_SET)) == 0);
oldsize = inode->i_size;
newsize = iattr->ia_size;
/*
* Short circuit the truncate case for zero length files.
*/
if (newsize == 0 && oldsize == 0 && ip->i_d.di_nextents == 0) {
if (!(iattr->ia_valid & (ATTR_CTIME|ATTR_MTIME)))
return 0;
/*
* Use the regular setattr path to update the timestamps.
*/
iattr->ia_valid &= ~ATTR_SIZE;
return xfs_setattr_nonsize(ip, iattr, 0);
}
/*
* Make sure that the dquots are attached to the inode.
*/
error = xfs_qm_dqattach(ip, 0);
if (error)
return error;
/*
* Now we can make the changes. Before we join the inode to the
* transaction, take care of the part of the truncation that must be
* done without the inode lock. This needs to be done before joining
* the inode to the transaction, because the inode cannot be unlocked
* once it is a part of the transaction.
*/
if (newsize > oldsize) {
/*
* Do the first part of growing a file: zero any data in the
* last block that is beyond the old EOF. We need to do this
* before the inode is joined to the transaction to modify
* i_size.
*/
error = xfs_zero_eof(ip, newsize, oldsize);
if (error)
return error;
}
/*
* We are going to log the inode size change in this transaction so
* any previous writes that are beyond the on disk EOF and the new
* EOF that have not been written out need to be written here. If we
* do not write the data out, we expose ourselves to the null files
* problem.
*
* Only flush from the on disk size to the smaller of the in memory
* file size or the new size as that's the range we really care about
* here and prevents waiting for other data not within the range we
* care about here.
*/
if (oldsize != ip->i_d.di_size && newsize > ip->i_d.di_size) {
error = -filemap_write_and_wait_range(VFS_I(ip)->i_mapping,
ip->i_d.di_size, newsize);
if (error)
return error;
}
/*
* Wait for all direct I/O to complete.
*/
inode_dio_wait(inode);
error = -block_truncate_page(inode->i_mapping, newsize, xfs_get_blocks);
if (error)
return error;
tp = xfs_trans_alloc(mp, XFS_TRANS_SETATTR_SIZE);
error = xfs_trans_reserve(tp, &M_RES(mp)->tr_itruncate, 0, 0);
if (error)
goto out_trans_cancel;
truncate_setsize(inode, newsize);
commit_flags = XFS_TRANS_RELEASE_LOG_RES;
lock_flags |= XFS_ILOCK_EXCL;
xfs_ilock(ip, XFS_ILOCK_EXCL);
xfs_trans_ijoin(tp, ip, 0);
/*
* Only change the c/mtime if we are changing the size or we are
* explicitly asked to change it. This handles the semantic difference
* between truncate() and ftruncate() as implemented in the VFS.
*
* The regular truncate() case without ATTR_CTIME and ATTR_MTIME is a
* special case where we need to update the times despite not having
* these flags set. For all other operations the VFS set these flags
* explicitly if it wants a timestamp update.
*/
if (newsize != oldsize &&
!(iattr->ia_valid & (ATTR_CTIME | ATTR_MTIME))) {
iattr->ia_ctime = iattr->ia_mtime =
current_fs_time(inode->i_sb);
iattr->ia_valid |= ATTR_CTIME | ATTR_MTIME;
}
/*
* The first thing we do is set the size to new_size permanently on
* disk. This way we don't have to worry about anyone ever being able
* to look at the data being freed even in the face of a crash.
* What we're getting around here is the case where we free a block, it
* is allocated to another file, it is written to, and then we crash.
* If the new data gets written to the file but the log buffers
* containing the free and reallocation don't, then we'd end up with
* garbage in the blocks being freed. As long as we make the new size
* permanent before actually freeing any blocks it doesn't matter if
* they get written to.
*/
ip->i_d.di_size = newsize;
xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
if (newsize <= oldsize) {
error = xfs_itruncate_extents(&tp, ip, XFS_DATA_FORK, newsize);
if (error)
goto out_trans_abort;
/*
* Truncated "down", so we're removing references to old data
* here - if we delay flushing for a long time, we expose
* ourselves unduly to the notorious NULL files problem. So,
* we mark this inode and flush it when the file is closed,
* and do not wait the usual (long) time for writeout.
*/
xfs_iflags_set(ip, XFS_ITRUNCATED);
/* A truncate down always removes post-EOF blocks. */
xfs_inode_clear_eofblocks_tag(ip);
}
if (iattr->ia_valid & ATTR_MODE)
xfs_setattr_mode(ip, iattr);
if (iattr->ia_valid & (ATTR_ATIME|ATTR_CTIME|ATTR_MTIME))
xfs_setattr_time(ip, iattr);
xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
XFS_STATS_INC(xs_ig_attrchg);
if (mp->m_flags & XFS_MOUNT_WSYNC)
xfs_trans_set_sync(tp);
error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
out_unlock:
if (lock_flags)
xfs_iunlock(ip, lock_flags);
return error;
out_trans_abort:
commit_flags |= XFS_TRANS_ABORT;
out_trans_cancel:
xfs_trans_cancel(tp, commit_flags);
goto out_unlock;
}
int f2fs_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
{
struct inode *inode = file->f_mapping->host;
struct f2fs_inode_info *fi = F2FS_I(inode);
struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
nid_t ino = inode->i_ino;
int ret = 0;
bool need_cp = false;
struct writeback_control wbc = {
.sync_mode = WB_SYNC_ALL,
.nr_to_write = LONG_MAX,
.for_reclaim = 0,
};
if (unlikely(f2fs_readonly(inode->i_sb)))
return 0;
trace_f2fs_sync_file_enter(inode);
/* if fdatasync is triggered, let's do in-place-update */
if (get_dirty_pages(inode) <= SM_I(sbi)->min_fsync_blocks)
set_inode_flag(fi, FI_NEED_IPU);
ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
clear_inode_flag(fi, FI_NEED_IPU);
if (ret) {
trace_f2fs_sync_file_exit(inode, need_cp, datasync, ret);
return ret;
}
/* if the inode is dirty, let's recover all the time */
if (!datasync) {
f2fs_write_inode(inode, NULL);
goto go_write;
}
/*
* if there is no written data, don't waste time to write recovery info.
*/
if (!is_inode_flag_set(fi, FI_APPEND_WRITE) &&
!exist_written_data(sbi, ino, APPEND_INO)) {
/* it may call write_inode just prior to fsync */
if (need_inode_page_update(sbi, ino))
goto go_write;
if (is_inode_flag_set(fi, FI_UPDATE_WRITE) ||
exist_written_data(sbi, ino, UPDATE_INO))
goto flush_out;
goto out;
}
go_write:
/* guarantee free sections for fsync */
f2fs_balance_fs(sbi);
/*
* Both of fdatasync() and fsync() are able to be recovered from
* sudden-power-off.
*/
down_read(&fi->i_sem);
need_cp = need_do_checkpoint(inode);
up_read(&fi->i_sem);
if (need_cp) {
/* all the dirty node pages should be flushed for POR */
ret = f2fs_sync_fs(inode->i_sb, 1);
/*
* We've secured consistency through sync_fs. Following pino
* will be used only for fsynced inodes after checkpoint.
*/
try_to_fix_pino(inode);
clear_inode_flag(fi, FI_APPEND_WRITE);
clear_inode_flag(fi, FI_UPDATE_WRITE);
goto out;
}
sync_nodes:
sync_node_pages(sbi, ino, &wbc);
/* if cp_error was enabled, we should avoid infinite loop */
if (unlikely(f2fs_cp_error(sbi)))
goto out;
if (need_inode_block_update(sbi, ino)) {
mark_inode_dirty_sync(inode);
f2fs_write_inode(inode, NULL);
goto sync_nodes;
}
ret = wait_on_node_pages_writeback(sbi, ino);
if (ret)
goto out;
/* once recovery info is written, don't need to tack this */
remove_dirty_inode(sbi, ino, APPEND_INO);
clear_inode_flag(fi, FI_APPEND_WRITE);
flush_out:
remove_dirty_inode(sbi, ino, UPDATE_INO);
clear_inode_flag(fi, FI_UPDATE_WRITE);
ret = f2fs_issue_flush(sbi);
out:
trace_f2fs_sync_file_exit(inode, need_cp, datasync, ret);
f2fs_trace_ios(NULL, 1);
return ret;
}
static pgoff_t __get_first_dirty_index(struct address_space *mapping,
pgoff_t pgofs, int whence)
{
struct pagevec pvec;
int nr_pages;
if (whence != SEEK_DATA)
return 0;
/* find first dirty page index */
pagevec_init(&pvec, 0);
nr_pages = pagevec_lookup_tag(&pvec, mapping, &pgofs,
PAGECACHE_TAG_DIRTY, 1);
pgofs = nr_pages ? pvec.pages[0]->index : LONG_MAX;
pagevec_release(&pvec);
return pgofs;
}
static bool __found_offset(block_t blkaddr, pgoff_t dirty, pgoff_t pgofs,
int whence)
{
switch (whence) {
case SEEK_DATA:
if ((blkaddr == NEW_ADDR && dirty == pgofs) ||
(blkaddr != NEW_ADDR && blkaddr != NULL_ADDR))
return true;
break;
case SEEK_HOLE:
if (blkaddr == NULL_ADDR)
return true;
break;
}
return false;
}
static inline int unsigned_offsets(struct file *file)
{
return file->f_mode & FMODE_UNSIGNED_OFFSET;
}
static loff_t vfs_setpos(struct file *file, loff_t offset, loff_t maxsize)
{
if (offset < 0 && !unsigned_offsets(file))
return -EINVAL;
if (offset > maxsize)
return -EINVAL;
if (offset != file->f_pos) {
file->f_pos = offset;
file->f_version = 0;
}
return offset;
}
static loff_t f2fs_seek_block(struct file *file, loff_t offset, int whence)
{
struct inode *inode = file->f_mapping->host;
loff_t maxbytes = inode->i_sb->s_maxbytes;
struct dnode_of_data dn;
pgoff_t pgofs, end_offset, dirty;
loff_t data_ofs = offset;
loff_t isize;
int err = 0;
mutex_lock(&inode->i_mutex);
isize = i_size_read(inode);
if (offset >= isize)
goto fail;
/* handle inline data case */
if (f2fs_has_inline_data(inode) || f2fs_has_inline_dentry(inode)) {
if (whence == SEEK_HOLE)
data_ofs = isize;
goto found;
}
pgofs = (pgoff_t)(offset >> PAGE_CACHE_SHIFT);
dirty = __get_first_dirty_index(inode->i_mapping, pgofs, whence);
for (; data_ofs < isize; data_ofs = (loff_t)pgofs << PAGE_CACHE_SHIFT) {
set_new_dnode(&dn, inode, NULL, NULL, 0);
err = get_dnode_of_data(&dn, pgofs, LOOKUP_NODE_RA);
if (err && err != -ENOENT) {
goto fail;
} else if (err == -ENOENT) {
/* direct node does not exists */
if (whence == SEEK_DATA) {
pgofs = PGOFS_OF_NEXT_DNODE(pgofs,
F2FS_I(inode));
continue;
} else {
goto found;
}
}
end_offset = ADDRS_PER_PAGE(dn.node_page, F2FS_I(inode));
/* find data/hole in dnode block */
for (; dn.ofs_in_node < end_offset;
dn.ofs_in_node++, pgofs++,
data_ofs = (loff_t)pgofs << PAGE_CACHE_SHIFT) {
block_t blkaddr;
blkaddr = datablock_addr(dn.node_page, dn.ofs_in_node);
if (__found_offset(blkaddr, dirty, pgofs, whence)) {
f2fs_put_dnode(&dn);
goto found;
}
}
f2fs_put_dnode(&dn);
}
if (whence == SEEK_DATA)
goto fail;
found:
if (whence == SEEK_HOLE && data_ofs > isize)
data_ofs = isize;
mutex_unlock(&inode->i_mutex);
return vfs_setpos(file, data_ofs, maxbytes);
fail:
mutex_unlock(&inode->i_mutex);
return -ENXIO;
}
static loff_t f2fs_llseek(struct file *file, loff_t offset, int whence)
{
struct inode *inode = file->f_mapping->host;
loff_t maxbytes = inode->i_sb->s_maxbytes;
switch (whence) {
case SEEK_SET:
case SEEK_CUR:
case SEEK_END:
return generic_file_llseek_size(file, offset, whence,
maxbytes, i_size_read(inode));
case SEEK_DATA:
case SEEK_HOLE:
if (offset < 0)
return -ENXIO;
return f2fs_seek_block(file, offset, whence);
}
return -EINVAL;
}
static int f2fs_file_mmap(struct file *file, struct vm_area_struct *vma)
{
struct inode *inode = file_inode(file);
if (f2fs_encrypted_inode(inode)) {
int err = f2fs_get_encryption_info(inode);
if (err)
return 0;
}
/* we don't need to use inline_data strictly */
if (f2fs_has_inline_data(inode)) {
int err = f2fs_convert_inline_inode(inode);
if (err)
return err;
}
file_accessed(file);
vma->vm_ops = &f2fs_file_vm_ops;
return 0;
}
static int f2fs_file_open(struct inode *inode, struct file *filp)
{
int ret = generic_file_open(inode, filp);
if (!ret && f2fs_encrypted_inode(inode)) {
ret = f2fs_get_encryption_info(inode);
if (ret)
ret = -EACCES;
}
return ret;
}
int ext3_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
{
struct timespec before,after;
printk(KERN_DEBUG"fsync start\n");
getnstimeofday(&before);
struct inode *inode = file->f_mapping->host;
struct ext3_inode_info *ei = EXT3_I(inode);
journal_t *journal = EXT3_SB(inode->i_sb)->s_journal;
int ret, needs_barrier = 0;
tid_t commit_tid;
trace_ext3_sync_file_enter(file, datasync);
if (inode->i_sb->s_flags & MS_RDONLY) {
/* Make sure that we read updated state */
smp_rmb();
if (EXT3_SB(inode->i_sb)->s_mount_state & EXT3_ERROR_FS)
{
getnstimeofday(&after);
printk(KERN_DEBUG"fsync time:%.01f us",after.tv_nsec-before.tv_nsec);
return -EROFS;
}
getnstimeofday(&after);
printk(KERN_DEBUG"fsync time:%.01f us",after.tv_nsec-before.tv_nsec);
return 0;
}
ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
if (ret)
goto out;
J_ASSERT(ext3_journal_current_handle() == NULL);
/*
* data=writeback,ordered:
* The caller's filemap_fdatawrite()/wait will sync the data.
* Metadata is in the journal, we wait for a proper transaction
* to commit here.
*
* data=journal:
* filemap_fdatawrite won't do anything (the buffers are clean).
* ext3_force_commit will write the file data into the journal and
* will wait on that.
* filemap_fdatawait() will encounter a ton of newly-dirtied pages
* (they were dirtied by commit). But that's OK - the blocks are
* safe in-journal, which is all fsync() needs to ensure.
*/
if (ext3_should_journal_data(inode)) {
ret = ext3_force_commit(inode->i_sb);
goto out;
}
if (datasync)
commit_tid = atomic_read(&ei->i_datasync_tid);
else
commit_tid = atomic_read(&ei->i_sync_tid);
if (test_opt(inode->i_sb, BARRIER) &&
!journal_trans_will_send_data_barrier(journal, commit_tid))
needs_barrier = 1;
log_start_commit(journal, commit_tid);
ret = log_wait_commit(journal, commit_tid);
/*
* In case we didn't commit a transaction, we have to flush
* disk caches manually so that data really is on persistent
* storage
*/
if (needs_barrier) {
int err;
err = blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL);
if (!ret)
ret = err;
}
out:
getnstimeofday(&after);
printk(KERN_DEBUG"fsync time:%.01f us",after.tv_nsec-before.tv_nsec);
trace_ext3_sync_file_exit(inode, ret);
return ret;
}
/*
* Check that the two inodes are eligible for cloning, the ranges make
* sense, and then flush all dirty data. Caller must ensure that the
* inodes have been locked against any other modifications.
*
* Returns: 0 for "nothing to clone", 1 for "something to clone", or
* the usual negative error code.
*/
int vfs_clone_file_prep_inodes(struct inode *inode_in, loff_t pos_in,
struct inode *inode_out, loff_t pos_out,
u64 *len, bool is_dedupe)
{
loff_t bs = inode_out->i_sb->s_blocksize;
loff_t blen;
loff_t isize;
bool same_inode = (inode_in == inode_out);
int ret;
/* Don't touch certain kinds of inodes */
if (IS_IMMUTABLE(inode_out))
return -EPERM;
if (IS_SWAPFILE(inode_in) || IS_SWAPFILE(inode_out))
return -ETXTBSY;
/* Don't reflink dirs, pipes, sockets... */
if (S_ISDIR(inode_in->i_mode) || S_ISDIR(inode_out->i_mode))
return -EISDIR;
if (!S_ISREG(inode_in->i_mode) || !S_ISREG(inode_out->i_mode))
return -EINVAL;
/* Are we going all the way to the end? */
isize = i_size_read(inode_in);
if (isize == 0)
return 0;
/* Zero length dedupe exits immediately; reflink goes to EOF. */
if (*len == 0) {
if (is_dedupe || pos_in == isize)
return 0;
if (pos_in > isize)
return -EINVAL;
*len = isize - pos_in;
}
/* Ensure offsets don't wrap and the input is inside i_size */
if (pos_in + *len < pos_in || pos_out + *len < pos_out ||
pos_in + *len > isize)
return -EINVAL;
/* Don't allow dedupe past EOF in the dest file */
if (is_dedupe) {
loff_t disize;
disize = i_size_read(inode_out);
if (pos_out >= disize || pos_out + *len > disize)
return -EINVAL;
}
/* If we're linking to EOF, continue to the block boundary. */
if (pos_in + *len == isize)
blen = ALIGN(isize, bs) - pos_in;
else
blen = *len;
/* Only reflink if we're aligned to block boundaries */
if (!IS_ALIGNED(pos_in, bs) || !IS_ALIGNED(pos_in + blen, bs) ||
!IS_ALIGNED(pos_out, bs) || !IS_ALIGNED(pos_out + blen, bs))
return -EINVAL;
/* Don't allow overlapped reflink within the same file */
if (same_inode) {
if (pos_out + blen > pos_in && pos_out < pos_in + blen)
return -EINVAL;
}
/* Wait for the completion of any pending IOs on both files */
inode_dio_wait(inode_in);
if (!same_inode)
inode_dio_wait(inode_out);
ret = filemap_write_and_wait_range(inode_in->i_mapping,
pos_in, pos_in + *len - 1);
if (ret)
return ret;
ret = filemap_write_and_wait_range(inode_out->i_mapping,
pos_out, pos_out + *len - 1);
if (ret)
return ret;
/*
* Check that the extents are the same.
*/
if (is_dedupe) {
bool is_same = false;
ret = vfs_dedupe_file_range_compare(inode_in, pos_in,
inode_out, pos_out, *len, &is_same);
if (ret)
return ret;
if (!is_same)
return -EBADE;
}
return 1;
}
static int aio_submit(struct aio_output *output, struct aio_mref_aspect *mref_a, bool use_fdsync)
{
struct mref_object *mref = mref_a->object;
mm_segment_t oldfs;
int res;
struct iocb iocb = {
.aio_data = (__u64)mref_a,
.aio_lio_opcode = use_fdsync ? IOCB_CMD_FDSYNC : (mref->ref_rw != 0 ? IOCB_CMD_PWRITE : IOCB_CMD_PREAD),
.aio_fildes = output->fd,
.aio_buf = (unsigned long)mref->ref_data,
.aio_nbytes = mref->ref_len,
.aio_offset = mref->ref_pos,
// .aio_reqprio = something(mref->ref_prio) field exists, but not yet implemented in kernelspace :(
};
struct iocb *iocbp = &iocb;
unsigned long long latency;
mars_trace(mref, "aio_submit");
if (unlikely(output->fd < 0)) {
MARS_ERR("bad fd = %d\n", output->fd);
res = -EBADF;
goto done;
}
oldfs = get_fs();
set_fs(get_ds());
latency = TIME_STATS(&timings[mref->ref_rw & 1], res = sys_io_submit(output->ctxp, 1, &iocbp));
set_fs(oldfs);
threshold_check(&aio_submit_threshold, latency);
atomic_inc(&output->total_submit_count);
if (likely(res >= 0)) {
atomic_inc(&output->submit_count);
} else if (likely(res == -EAGAIN)) {
atomic_inc(&output->total_again_count);
} else {
MARS_ERR("error = %d\n", res);
}
done:
return res;
}
static int aio_submit_dummy(struct aio_output *output)
{
mm_segment_t oldfs;
int res;
int dummy;
struct iocb iocb = {
.aio_buf = (__u64)&dummy,
};
struct iocb *iocbp = &iocb;
oldfs = get_fs();
set_fs(get_ds());
res = sys_io_submit(output->ctxp, 1, &iocbp);
set_fs(oldfs);
if (likely(res >= 0)) {
atomic_inc(&output->submit_count);
}
return res;
}
static
int aio_start_thread(
struct aio_output *output,
struct aio_threadinfo *tinfo,
int(*fn)(void*),
char class)
{
int j;
for (j = 0; j < MARS_PRIO_NR; j++) {
INIT_LIST_HEAD(&tinfo->mref_list[j]);
}
tinfo->output = output;
spin_lock_init(&tinfo->lock);
init_waitqueue_head(&tinfo->event);
init_waitqueue_head(&tinfo->terminate_event);
tinfo->should_terminate = false;
tinfo->terminated = false;
tinfo->thread = brick_thread_create(fn, tinfo, "mars_aio_%c%d", class, output->index);
if (unlikely(!tinfo->thread)) {
MARS_ERR("cannot create thread\n");
return -ENOENT;
}
return 0;
}
static
void aio_stop_thread(struct aio_output *output, int i, bool do_submit_dummy)
{
struct aio_threadinfo *tinfo = &output->tinfo[i];
struct task_struct *thread = tinfo->thread;
if (thread) {
MARS_DBG("stopping thread %d ...\n", i);
tinfo->should_terminate = true;
// workaround for waking up the receiver thread. TODO: check whether signal handlong could do better.
if (do_submit_dummy) {
MARS_DBG("submitting dummy for wakeup %d...\n", i);
use_fake_mm();
aio_submit_dummy(output);
if (likely(current->mm)) {
unuse_fake_mm();
}
}
// wait for termination
MARS_DBG("waiting for thread %d ...\n", i);
wait_event_interruptible_timeout(
tinfo->terminate_event,
tinfo->terminated,
(60 - i * 2) * HZ);
if (likely(tinfo->terminated)) {
brick_thread_stop(tinfo->thread);
} else {
MARS_ERR("thread %d did not terminate - leaving a zombie\n", i);
}
}
}
static
int aio_sync(struct file *file)
{
int err;
switch (aio_sync_mode) {
case 1:
#if defined(S_BIAS) || (defined(RHEL_MAJOR) && (RHEL_MAJOR < 7))
err = vfs_fsync_range(file, file->f_path.dentry, 0, LLONG_MAX, 1);
#else
err = vfs_fsync_range(file, 0, LLONG_MAX, 1);
#endif
break;
case 2:
#if defined(S_BIAS) || (defined(RHEL_MAJOR) && (RHEL_MAJOR < 7))
err = vfs_fsync_range(file, file->f_path.dentry, 0, LLONG_MAX, 0);
#else
err = vfs_fsync_range(file, 0, LLONG_MAX, 0);
#endif
break;
default:
err = filemap_write_and_wait_range(file->f_mapping, 0, LLONG_MAX);
}
return err;
}
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;
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;
ret = generic_segment_checks(iovp, &nr_segs, &size, VERIFY_WRITE);
if (ret < 0)
return ret;
if (unlikely(ioflags & IO_ISDIRECT)) {
xfs_buftarg_t *target =
XFS_IS_REALTIME_INODE(ip) ?
mp->m_rtdev_targp : mp->m_ddev_targp;
/* DIO must be aligned to device logical sector size */
if ((pos | size) & target->bt_logical_sectormask) {
if (pos == i_size_read(inode))
return 0;
return -XFS_ERROR(EINVAL);
}
}
n = mp->m_super->s_maxbytes - 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 = -filemap_write_and_wait_range(
VFS_I(ip)->i_mapping,
pos, -1);
if (ret) {
xfs_rw_iunlock(ip, XFS_IOLOCK_EXCL);
return ret;
}
truncate_pagecache_range(VFS_I(ip), pos, -1);
}
xfs_rw_ilock_demote(ip, XFS_IOLOCK_EXCL);
}
trace_xfs_file_read(ip, size, pos, ioflags);
ret = generic_file_aio_read(iocb, iovp, nr_segs, pos);
if (ret > 0)
XFS_STATS_ADD(xs_read_bytes, ret);
xfs_rw_iunlock(ip, XFS_IOLOCK_SHARED);
return ret;
}
/*
* Notes about direct IO locking for write:
*
* 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. xfs_ioend_wait()).
*/
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;
xfs_fsize_t isize, new_size;
int iolock;
int eventsent = 0;
size_t ocount = 0, count;
loff_t pos;
int need_i_mutex;
int unaligned_io = 0;
XFS_STATS_INC(xs_write_calls);
ret = generic_segment_checks(iovp, &segs, &ocount, VERIFY_READ);
if (ret)
return ret;
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;
if ((ioflags & IO_ISDIRECT) &&
((pos & mp->m_blockmask) || ((pos + count) & mp->m_blockmask)))
unaligned_io = 1;
relock:
if ((ioflags & IO_ISDIRECT) && !unaligned_io) {
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:
ret = generic_write_checks(file, &pos, &count,
S_ISBLK(inode->i_mode));
if (ret) {
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);
ret = -XFS_SEND_DATA(xip->i_mount, DM_EVENT_WRITE, xip,
pos, count, dmflags, &iolock);
if (ret)
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)))
file_update_time(file);
/*
* 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) {
ret = -xfs_zero_eof(xip, pos, xip->i_size);
if (ret) {
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)) {
ret = -xfs_write_clear_setuid(xip);
if (likely(!ret))
ret = file_remove_suid(file);
if (unlikely(ret)) {
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);
ret = -xfs_flushinval_pages(xip,
(pos & PAGE_CACHE_MASK),
-1, FI_REMAPF_LOCKED);
if (ret)
goto out_unlock_internal;
}
/*
* 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)
xfs_ioend_wait(xip);
else 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;
}
trace_xfs_file_direct_write(xip, count, *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);
if (need_i_mutex)
mutex_unlock(&inode->i_mutex);
goto relock;
}
} else {
int enospc = 0;
write_retry:
trace_xfs_file_buffered_write(xip, count, *offset, ioflags);
ret = generic_file_buffered_write(iocb, iovp, segs,
pos, offset, count, ret);
/*
* if we just got an ENOSPC, flush the inode now we
* aren't holding any page locks and retry *once*
*/
if (ret == -ENOSPC && !enospc) {
ret = -xfs_flush_pages(xip, 0, -1, 0, FI_NONE);
if (ret)
goto out_unlock_internal;
enospc = 1;
goto write_retry;
}
}
current->backing_dev_info = NULL;
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);
ret = -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 (ret)
goto out_unlock_internal;
goto start;
}
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)) {
loff_t end = pos + ret - 1;
int error, error2;
xfs_iunlock(xip, iolock);
if (need_i_mutex)
mutex_unlock(&inode->i_mutex);
error = filemap_write_and_wait_range(mapping, pos, end);
if (need_i_mutex)
mutex_lock(&inode->i_mutex);
xfs_ilock(xip, iolock);
error2 = -xfs_fsync(xip);
if (error)
ret = error;
else if (error2)
ret = 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 ret;
}
STATIC ssize_t
xfs_file_aio_write(
struct kiocb *iocb,
const struct iovec *iovp,
unsigned long nr_segs,
loff_t pos)
{
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);
ssize_t ret;
int iolock;
size_t ocount = 0;
XFS_STATS_INC(xs_write_calls);
BUG_ON(iocb->ki_pos != pos);
ret = generic_segment_checks(iovp, &nr_segs, &ocount, VERIFY_READ);
if (ret)
return ret;
if (ocount == 0)
return 0;
xfs_wait_for_freeze(ip->i_mount, SB_FREEZE_WRITE);
if (XFS_FORCED_SHUTDOWN(ip->i_mount))
return -EIO;
if (unlikely(file->f_flags & O_DIRECT))
ret = xfs_file_dio_aio_write(iocb, iovp, nr_segs, pos,
ocount, &iolock);
else
ret = xfs_file_buffered_aio_write(iocb, iovp, nr_segs, pos,
ocount, &iolock);
xfs_aio_write_isize_update(inode, &iocb->ki_pos, ret);
if (ret <= 0)
goto out_unlock;
/* Handle various SYNC-type writes */
if ((file->f_flags & O_DSYNC) || IS_SYNC(inode)) {
loff_t end = pos + ret - 1;
int error, error2;
xfs_rw_iunlock(ip, iolock);
error = filemap_write_and_wait_range(mapping, pos, end);
xfs_rw_ilock(ip, iolock);
error2 = -xfs_file_fsync(file,
(file->f_flags & __O_SYNC) ? 0 : 1);
if (error)
ret = error;
else if (error2)
ret = error2;
}
out_unlock:
xfs_aio_write_newsize_update(ip);
xfs_rw_iunlock(ip, iolock);
return ret;
}
struct dentry *f2fs_get_parent(struct dentry *child)
{
struct qstr dotdot = {.len = 2, .name = ".."};
unsigned long ino = f2fs_inode_by_name(child->d_inode, &dotdot);
if (!ino)
return ERR_PTR(-ENOENT);
return d_obtain_alias(f2fs_iget(child->d_inode->i_sb, ino));
}
static int __recover_dot_dentries(struct inode *dir, nid_t pino)
{
struct f2fs_sb_info *sbi = F2FS_I_SB(dir);
struct qstr dot = {.len = 1, .name = "."};
struct qstr dotdot = {.len = 2, .name = ".."};
struct f2fs_dir_entry *de;
struct page *page;
int err = 0;
f2fs_lock_op(sbi);
de = f2fs_find_entry(dir, &dot, &page, 0);
if (de) {
f2fs_dentry_kunmap(dir, page);
f2fs_put_page(page, 0);
} else {
err = __f2fs_add_link(dir, &dot, NULL, dir->i_ino, S_IFDIR);
if (err)
goto out;
}
de = f2fs_find_entry(dir, &dotdot, &page, 0);
if (de) {
f2fs_dentry_kunmap(dir, page);
f2fs_put_page(page, 0);
} else {
err = __f2fs_add_link(dir, &dotdot, NULL, pino, S_IFDIR);
}
out:
if (!err) {
clear_inode_flag(F2FS_I(dir), FI_INLINE_DOTS);
mark_inode_dirty(dir);
}
f2fs_unlock_op(sbi);
return err;
}
static struct dentry *f2fs_lookup(struct inode *dir, struct dentry *dentry,
struct nameidata *nd)
{
struct inode *inode = NULL;
struct f2fs_dir_entry *de;
struct page *page;
nid_t ino;
int err = 0;
if (dentry->d_name.len > F2FS_NAME_LEN)
return ERR_PTR(-ENAMETOOLONG);
de = f2fs_find_entry(dir, &dentry->d_name, &page, nd ? nd->flags : 0);
if (!de)
return d_splice_alias(inode, dentry);
ino = le32_to_cpu(de->ino);
f2fs_dentry_kunmap(dir, page);
f2fs_put_page(page, 0);
inode = f2fs_iget(dir->i_sb, ino);
if (IS_ERR(inode))
return ERR_CAST(inode);
if (f2fs_has_inline_dots(inode)) {
err = __recover_dot_dentries(inode, dir->i_ino);
if (err)
goto err_out;
}
return d_splice_alias(inode, dentry);
err_out:
iget_failed(inode);
return ERR_PTR(err);
}
static int f2fs_unlink(struct inode *dir, struct dentry *dentry)
{
struct f2fs_sb_info *sbi = F2FS_I_SB(dir);
struct inode *inode = dentry->d_inode;
struct f2fs_dir_entry *de;
struct page *page;
int err = -ENOENT;
trace_f2fs_unlink_enter(dir, dentry);
f2fs_balance_fs(sbi);
de = f2fs_find_entry(dir, &dentry->d_name, &page, 0);
if (!de)
goto fail;
f2fs_lock_op(sbi);
err = acquire_orphan_inode(sbi);
if (err) {
f2fs_unlock_op(sbi);
f2fs_dentry_kunmap(dir, page);
f2fs_put_page(page, 0);
goto fail;
}
f2fs_delete_entry(de, page, dir, inode);
f2fs_unlock_op(sbi);
/* In order to evict this inode, we set it dirty */
mark_inode_dirty(inode);
if (IS_DIRSYNC(dir))
f2fs_sync_fs(sbi->sb, 1);
fail:
trace_f2fs_unlink_exit(inode, err);
return err;
}
static void *f2fs_follow_link(struct dentry *dentry, struct nameidata *nd)
{
struct page *page;
page = page_follow_link_light(dentry, nd);
if (IS_ERR(page))
return page;
/* this is broken symlink case */
if (*nd_get_link(nd) == 0) {
kunmap(page);
page_cache_release(page);
return ERR_PTR(-ENOENT);
}
return page;
}
static int f2fs_symlink(struct inode *dir, struct dentry *dentry,
const char *symname)
{
struct f2fs_sb_info *sbi = F2FS_I_SB(dir);
struct inode *inode;
size_t len = strlen(symname);
size_t p_len;
char *p_str;
struct f2fs_str disk_link = FSTR_INIT(NULL, 0);
struct f2fs_encrypted_symlink_data *sd = NULL;
int err;
if (len > dir->i_sb->s_blocksize)
return -ENAMETOOLONG;
f2fs_balance_fs(sbi);
inode = f2fs_new_inode(dir, S_IFLNK | S_IRWXUGO);
if (IS_ERR(inode))
return PTR_ERR(inode);
if (f2fs_encrypted_inode(inode))
inode->i_op = &f2fs_encrypted_symlink_inode_operations;
else
inode->i_op = &f2fs_symlink_inode_operations;
inode->i_mapping->a_ops = &f2fs_dblock_aops;
f2fs_lock_op(sbi);
err = f2fs_add_link(dentry, inode);
if (err)
goto out;
f2fs_unlock_op(sbi);
alloc_nid_done(sbi, inode->i_ino);
if (f2fs_encrypted_inode(dir)) {
struct qstr istr = QSTR_INIT(symname, len);
err = f2fs_get_encryption_info(inode);
if (err)
goto err_out;
err = f2fs_fname_crypto_alloc_buffer(inode, len, &disk_link);
if (err)
goto err_out;
err = f2fs_fname_usr_to_disk(inode, &istr, &disk_link);
if (err < 0)
goto err_out;
p_len = encrypted_symlink_data_len(disk_link.len) + 1;
if (p_len > dir->i_sb->s_blocksize) {
err = -ENAMETOOLONG;
goto err_out;
}
sd = kzalloc(p_len, GFP_NOFS);
if (!sd) {
err = -ENOMEM;
goto err_out;
}
memcpy(sd->encrypted_path, disk_link.name, disk_link.len);
sd->len = cpu_to_le16(disk_link.len);
p_str = (char *)sd;
} else {
p_len = len + 1;
p_str = (char *)symname;
}
err = page_symlink(inode, p_str, p_len);
err_out:
d_instantiate(dentry, inode);
unlock_new_inode(inode);
/*
* Let's flush symlink data in order to avoid broken symlink as much as
* possible. Nevertheless, fsyncing is the best way, but there is no
* way to get a file descriptor in order to flush that.
*
* Note that, it needs to do dir->fsync to make this recoverable.
* If the symlink path is stored into inline_data, there is no
* performance regression.
*/
if (!err)
filemap_write_and_wait_range(inode->i_mapping, 0, p_len - 1);
if (IS_DIRSYNC(dir))
f2fs_sync_fs(sbi->sb, 1);
kfree(sd);
f2fs_fname_crypto_free_buffer(&disk_link);
return err;
out:
handle_failed_inode(inode);
return err;
}
static int f2fs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode)
{
struct f2fs_sb_info *sbi = F2FS_I_SB(dir);
struct inode *inode;
int err;
f2fs_balance_fs(sbi);
inode = f2fs_new_inode(dir, S_IFDIR | mode);
if (IS_ERR(inode))
return PTR_ERR(inode);
inode->i_op = &f2fs_dir_inode_operations;
inode->i_fop = &f2fs_dir_operations;
inode->i_mapping->a_ops = &f2fs_dblock_aops;
mapping_set_gfp_mask(inode->i_mapping, GFP_F2FS_HIGH_ZERO);
set_inode_flag(F2FS_I(inode), FI_INC_LINK);
f2fs_lock_op(sbi);
err = f2fs_add_link(dentry, inode);
if (err)
goto out_fail;
f2fs_unlock_op(sbi);
alloc_nid_done(sbi, inode->i_ino);
d_instantiate(dentry, inode);
unlock_new_inode(inode);
if (IS_DIRSYNC(dir))
f2fs_sync_fs(sbi->sb, 1);
return 0;
out_fail:
clear_inode_flag(F2FS_I(inode), FI_INC_LINK);
handle_failed_inode(inode);
return err;
}
static int f2fs_rmdir(struct inode *dir, struct dentry *dentry)
{
struct inode *inode = dentry->d_inode;
if (f2fs_empty_dir(inode))
return f2fs_unlink(dir, dentry);
return -ENOTEMPTY;
}
static int f2fs_mknod(struct inode *dir, struct dentry *dentry,
umode_t mode, dev_t rdev)
{
struct f2fs_sb_info *sbi = F2FS_I_SB(dir);
struct inode *inode;
int err = 0;
if (!new_valid_dev(rdev))
return -EINVAL;
f2fs_balance_fs(sbi);
inode = f2fs_new_inode(dir, mode);
if (IS_ERR(inode))
return PTR_ERR(inode);
init_special_inode(inode, inode->i_mode, rdev);
inode->i_op = &f2fs_special_inode_operations;
f2fs_lock_op(sbi);
err = f2fs_add_link(dentry, inode);
if (err)
goto out;
f2fs_unlock_op(sbi);
alloc_nid_done(sbi, inode->i_ino);
d_instantiate(dentry, inode);
unlock_new_inode(inode);
if (IS_DIRSYNC(dir))
f2fs_sync_fs(sbi->sb, 1);
return 0;
out:
handle_failed_inode(inode);
return err;
}
static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry,
struct inode *new_dir, struct dentry *new_dentry)
{
struct f2fs_sb_info *sbi = F2FS_I_SB(old_dir);
struct inode *old_inode = old_dentry->d_inode;
struct inode *new_inode = new_dentry->d_inode;
struct page *old_dir_page;
struct page *old_page, *new_page;
struct f2fs_dir_entry *old_dir_entry = NULL;
struct f2fs_dir_entry *old_entry;
struct f2fs_dir_entry *new_entry;
int err = -ENOENT;
if ((old_dir != new_dir) && f2fs_encrypted_inode(new_dir) &&
!f2fs_is_child_context_consistent_with_parent(new_dir,
old_inode)) {
err = -EPERM;
goto out;
}
f2fs_balance_fs(sbi);
old_entry = f2fs_find_entry(old_dir, &old_dentry->d_name, &old_page, 0);
if (!old_entry)
goto out;
if (S_ISDIR(old_inode->i_mode)) {
err = -EIO;
old_dir_entry = f2fs_parent_dir(old_inode, &old_dir_page);
if (!old_dir_entry)
goto out_old;
}
if (new_inode) {
err = -ENOTEMPTY;
if (old_dir_entry && !f2fs_empty_dir(new_inode))
goto out_dir;
err = -ENOENT;
new_entry = f2fs_find_entry(new_dir, &new_dentry->d_name,
&new_page, 0);
if (!new_entry)
goto out_dir;
f2fs_lock_op(sbi);
err = acquire_orphan_inode(sbi);
if (err)
goto put_out_dir;
if (update_dent_inode(old_inode, new_inode,
&new_dentry->d_name)) {
release_orphan_inode(sbi);
goto put_out_dir;
}
f2fs_set_link(new_dir, new_entry, new_page, old_inode);
new_inode->i_ctime = CURRENT_TIME;
down_write(&F2FS_I(new_inode)->i_sem);
if (old_dir_entry)
drop_nlink(new_inode);
drop_nlink(new_inode);
up_write(&F2FS_I(new_inode)->i_sem);
mark_inode_dirty(new_inode);
if (!new_inode->i_nlink)
add_orphan_inode(sbi, new_inode->i_ino);
else
release_orphan_inode(sbi);
update_inode_page(old_inode);
update_inode_page(new_inode);
} else {
f2fs_lock_op(sbi);
err = f2fs_add_link(new_dentry, old_inode);
if (err) {
f2fs_unlock_op(sbi);
goto out_dir;
}
if (old_dir_entry) {
inc_nlink(new_dir);
update_inode_page(new_dir);
}
}
down_write(&F2FS_I(old_inode)->i_sem);
file_lost_pino(old_inode);
if (new_inode && file_enc_name(new_inode))
file_set_enc_name(old_inode);
up_write(&F2FS_I(old_inode)->i_sem);
old_inode->i_ctime = CURRENT_TIME;
mark_inode_dirty(old_inode);
f2fs_delete_entry(old_entry, old_page, old_dir, NULL);
if (old_dir_entry) {
if (old_dir != new_dir) {
f2fs_set_link(old_inode, old_dir_entry,
old_dir_page, new_dir);
update_inode_page(old_inode);
} else {
f2fs_dentry_kunmap(old_inode, old_dir_page);
f2fs_put_page(old_dir_page, 0);
}
drop_nlink(old_dir);
mark_inode_dirty(old_dir);
update_inode_page(old_dir);
}
f2fs_unlock_op(sbi);
if (IS_DIRSYNC(old_dir) || IS_DIRSYNC(new_dir))
f2fs_sync_fs(sbi->sb, 1);
return 0;
put_out_dir:
f2fs_unlock_op(sbi);
f2fs_dentry_kunmap(new_dir, new_page);
f2fs_put_page(new_page, 0);
out_dir:
if (old_dir_entry) {
f2fs_dentry_kunmap(old_inode, old_dir_page);
f2fs_put_page(old_dir_page, 0);
}
out_old:
f2fs_dentry_kunmap(old_dir, old_page);
f2fs_put_page(old_page, 0);
out:
return err;
}
#ifdef CONFIG_F2FS_FS_ENCRYPTION
static void *f2fs_encrypted_follow_link(struct dentry *dentry,
struct nameidata *nd)
{
struct page *cpage = NULL;
char *caddr, *paddr = NULL;
struct f2fs_str cstr;
struct f2fs_str pstr = FSTR_INIT(NULL, 0);
struct inode *inode = dentry->d_inode;
struct f2fs_encrypted_symlink_data *sd;
loff_t size = min_t(loff_t, i_size_read(inode), PAGE_SIZE - 1);
u32 max_size = inode->i_sb->s_blocksize;
int res;
res = f2fs_get_encryption_info(inode);
if (res)
return ERR_PTR(res);
cpage = read_mapping_page(inode->i_mapping, 0, NULL);
if (IS_ERR(cpage))
return cpage;
caddr = kmap(cpage);
caddr[size] = 0;
/* Symlink is encrypted */
sd = (struct f2fs_encrypted_symlink_data *)caddr;
cstr.name = sd->encrypted_path;
cstr.len = le16_to_cpu(sd->len);
/* this is broken symlink case */
if (cstr.name[0] == 0 && cstr.len == 0) {
res = -ENOENT;
goto errout;
}
if ((cstr.len + sizeof(struct f2fs_encrypted_symlink_data) - 1) >
max_size) {
/* Symlink data on the disk is corrupted */
res = -EIO;
goto errout;
}
res = f2fs_fname_crypto_alloc_buffer(inode, cstr.len, &pstr);
if (res)
goto errout;
res = f2fs_fname_disk_to_usr(inode, NULL, &cstr, &pstr);
if (res < 0)
goto errout;
paddr = pstr.name;
/* Null-terminate the name */
paddr[res] = '\0';
nd_set_link(nd, paddr);
kunmap(cpage);
page_cache_release(cpage);
return NULL;
errout:
f2fs_fname_crypto_free_buffer(&pstr);
kunmap(cpage);
page_cache_release(cpage);
return ERR_PTR(res);
}
void kfree_put_link(struct dentry *dentry, struct nameidata *nd,
void *cookie)
{
char *s = nd_get_link(nd);
if (!IS_ERR(s))
kfree(s);
}
const struct inode_operations f2fs_encrypted_symlink_inode_operations = {
.readlink = generic_readlink,
.follow_link = f2fs_encrypted_follow_link,
.put_link = kfree_put_link,
.getattr = f2fs_getattr,
.setattr = f2fs_setattr,
.setxattr = generic_setxattr,
.getxattr = generic_getxattr,
.listxattr = f2fs_listxattr,
.removexattr = generic_removexattr,
};
#endif
const struct inode_operations f2fs_dir_inode_operations = {
.create = f2fs_create,
.lookup = f2fs_lookup,
.link = f2fs_link,
.unlink = f2fs_unlink,
.symlink = f2fs_symlink,
.mkdir = f2fs_mkdir,
.rmdir = f2fs_rmdir,
.mknod = f2fs_mknod,
.rename = f2fs_rename,
.getattr = f2fs_getattr,
.setattr = f2fs_setattr,
.get_acl = f2fs_get_acl,
#ifdef CONFIG_F2FS_FS_XATTR
.setxattr = generic_setxattr,
.getxattr = generic_getxattr,
.listxattr = f2fs_listxattr,
.removexattr = generic_removexattr,
#endif
};
const struct inode_operations f2fs_symlink_inode_operations = {
.readlink = generic_readlink,
.follow_link = f2fs_follow_link,
.put_link = page_put_link,
.getattr = f2fs_getattr,
.setattr = f2fs_setattr,
#ifdef CONFIG_F2FS_FS_XATTR
.setxattr = generic_setxattr,
.getxattr = generic_getxattr,
.listxattr = f2fs_listxattr,
.removexattr = generic_removexattr,
#endif
};
const struct inode_operations f2fs_special_inode_operations = {
.getattr = f2fs_getattr,
.setattr = f2fs_setattr,
.get_acl = f2fs_get_acl,
#ifdef CONFIG_F2FS_FS_XATTR
.setxattr = generic_setxattr,
.getxattr = generic_getxattr,
.listxattr = f2fs_listxattr,
.removexattr = generic_removexattr,
#endif
};
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;
}
static ssize_t gfs2_direct_IO(int rw, struct kiocb *iocb,
struct iov_iter *iter, loff_t offset)
{
struct file *file = iocb->ki_filp;
struct inode *inode = file->f_mapping->host;
struct address_space *mapping = inode->i_mapping;
struct gfs2_inode *ip = GFS2_I(inode);
struct gfs2_holder gh;
int rv;
/*
* Deferred lock, even if its a write, since we do no allocation
* on this path. All we need change is atime, and this lock mode
* ensures that other nodes have flushed their buffered read caches
* (i.e. their page cache entries for this inode). We do not,
* unfortunately have the option of only flushing a range like
* the VFS does.
*/
gfs2_holder_init(ip->i_gl, LM_ST_DEFERRED, 0, &gh);
rv = gfs2_glock_nq(&gh);
if (rv)
return rv;
rv = gfs2_ok_for_dio(ip, rw, offset);
if (rv != 1)
goto out; /* dio not valid, fall back to buffered i/o */
/*
* Now since we are holding a deferred (CW) lock at this point, you
* might be wondering why this is ever needed. There is a case however
* where we've granted a deferred local lock against a cached exclusive
* glock. That is ok provided all granted local locks are deferred, but
* it also means that it is possible to encounter pages which are
* cached and possibly also mapped. So here we check for that and sort
* them out ahead of the dio. The glock state machine will take care of
* everything else.
*
* If in fact the cached glock state (gl->gl_state) is deferred (CW) in
* the first place, mapping->nr_pages will always be zero.
*/
if (mapping->nrpages) {
loff_t lstart = offset & (PAGE_CACHE_SIZE - 1);
loff_t len = iov_iter_count(iter);
loff_t end = PAGE_ALIGN(offset + len) - 1;
rv = 0;
if (len == 0)
goto out;
if (test_and_clear_bit(GIF_SW_PAGED, &ip->i_flags))
unmap_shared_mapping_range(ip->i_inode.i_mapping, offset, len);
rv = filemap_write_and_wait_range(mapping, lstart, end);
if (rv)
goto out;
if (rw == WRITE)
truncate_inode_pages_range(mapping, lstart, end);
}
rv = __blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev,
iter, offset,
gfs2_get_block_direct, NULL, NULL, 0);
out:
gfs2_glock_dq(&gh);
gfs2_holder_uninit(&gh);
return rv;
}
static int f2fs_symlink(struct inode *dir, struct dentry *dentry,
const char *symname)
{
struct f2fs_sb_info *sbi = F2FS_I_SB(dir);
struct inode *inode;
size_t len = strlen(symname);
struct fscrypt_str disk_link = FSTR_INIT((char *)symname, len + 1);
struct fscrypt_symlink_data *sd = NULL;
int err;
if (f2fs_encrypted_inode(dir)) {
err = fscrypt_get_encryption_info(dir);
if (err)
return err;
if (!fscrypt_has_encryption_key(dir))
return -ENOKEY;
disk_link.len = (fscrypt_fname_encrypted_size(dir, len) +
sizeof(struct fscrypt_symlink_data));
}
if (disk_link.len > dir->i_sb->s_blocksize)
return -ENAMETOOLONG;
err = dquot_initialize(dir);
if (err)
return err;
inode = f2fs_new_inode(dir, S_IFLNK | S_IRWXUGO);
if (IS_ERR(inode))
return PTR_ERR(inode);
if (f2fs_encrypted_inode(inode))
inode->i_op = &f2fs_encrypted_symlink_inode_operations;
else
inode->i_op = &f2fs_symlink_inode_operations;
inode_nohighmem(inode);
inode->i_mapping->a_ops = &f2fs_dblock_aops;
f2fs_lock_op(sbi);
err = f2fs_add_link(dentry, inode);
if (err)
goto out;
f2fs_unlock_op(sbi);
alloc_nid_done(sbi, inode->i_ino);
if (f2fs_encrypted_inode(inode)) {
struct qstr istr = QSTR_INIT(symname, len);
struct fscrypt_str ostr;
sd = kzalloc(disk_link.len, GFP_NOFS);
if (!sd) {
err = -ENOMEM;
goto err_out;
}
err = fscrypt_get_encryption_info(inode);
if (err)
goto err_out;
if (!fscrypt_has_encryption_key(inode)) {
err = -ENOKEY;
goto err_out;
}
ostr.name = sd->encrypted_path;
ostr.len = disk_link.len;
err = fscrypt_fname_usr_to_disk(inode, &istr, &ostr);
if (err)
goto err_out;
sd->len = cpu_to_le16(ostr.len);
disk_link.name = (char *)sd;
}
err = page_symlink(inode, disk_link.name, disk_link.len);
err_out:
d_instantiate(dentry, inode);
unlock_new_inode(inode);
/*
* Let's flush symlink data in order to avoid broken symlink as much as
* possible. Nevertheless, fsyncing is the best way, but there is no
* way to get a file descriptor in order to flush that.
*
* Note that, it needs to do dir->fsync to make this recoverable.
* If the symlink path is stored into inline_data, there is no
* performance regression.
*/
if (!err) {
filemap_write_and_wait_range(inode->i_mapping, 0,
disk_link.len - 1);
if (IS_DIRSYNC(dir))
f2fs_sync_fs(sbi->sb, 1);
} else {
f2fs_unlink(dir, dentry);
}
kfree(sd);
f2fs_balance_fs(sbi, true);
return err;
out:
handle_failed_inode(inode);
return err;
}
/*
* Completely synchronous read and write methods. Direct from __user
* buffer to osd, or directly to user pages (if O_DIRECT).
*
* If the read spans object boundary, just do multiple reads.
*/
static ssize_t ceph_sync_read(struct kiocb *iocb, struct iov_iter *to,
int *checkeof)
{
struct file *file = iocb->ki_filp;
struct inode *inode = file_inode(file);
struct page **pages;
u64 off = iocb->ki_pos;
int num_pages;
ssize_t ret;
size_t len = iov_iter_count(to);
dout("sync_read on file %p %llu~%u %s\n", file, off,
(unsigned)len,
(file->f_flags & O_DIRECT) ? "O_DIRECT" : "");
if (!len)
return 0;
/*
* flush any page cache pages in this range. this
* will make concurrent normal and sync io slow,
* but it will at least behave sensibly when they are
* in sequence.
*/
ret = filemap_write_and_wait_range(inode->i_mapping, off,
off + len);
if (ret < 0)
return ret;
if (unlikely(to->type & ITER_PIPE)) {
size_t page_off;
ret = iov_iter_get_pages_alloc(to, &pages, len,
&page_off);
if (ret <= 0)
return -ENOMEM;
num_pages = DIV_ROUND_UP(ret + page_off, PAGE_SIZE);
ret = striped_read(inode, off, ret, pages, num_pages,
page_off, checkeof);
if (ret > 0) {
iov_iter_advance(to, ret);
off += ret;
} else {
iov_iter_advance(to, 0);
}
ceph_put_page_vector(pages, num_pages, false);
} else {
num_pages = calc_pages_for(off, len);
pages = ceph_alloc_page_vector(num_pages, GFP_KERNEL);
if (IS_ERR(pages))
return PTR_ERR(pages);
ret = striped_read(inode, off, len, pages, num_pages,
(off & ~PAGE_MASK), checkeof);
if (ret > 0) {
int l, k = 0;
size_t left = ret;
while (left) {
size_t page_off = off & ~PAGE_MASK;
size_t copy = min_t(size_t, left,
PAGE_SIZE - page_off);
l = copy_page_to_iter(pages[k++], page_off,
copy, to);
off += l;
left -= l;
if (l < copy)
break;
}
}
ceph_release_page_vector(pages, num_pages);
}
if (off > iocb->ki_pos) {
ret = off - iocb->ki_pos;
iocb->ki_pos = off;
}
dout("sync_read result %zd\n", ret);
return ret;
}
static int f2fs_zero_range(struct inode *inode, loff_t offset, loff_t len,
int mode)
{
struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
struct address_space *mapping = inode->i_mapping;
pgoff_t index, pg_start, pg_end;
loff_t new_size = i_size_read(inode);
loff_t off_start, off_end;
int ret = 0;
ret = inode_newsize_ok(inode, (len + offset));
if (ret)
return ret;
f2fs_balance_fs(sbi);
if (f2fs_has_inline_data(inode)) {
ret = f2fs_convert_inline_inode(inode);
if (ret)
return ret;
}
ret = filemap_write_and_wait_range(mapping, offset, offset + len - 1);
if (ret)
return ret;
truncate_pagecache_range(inode, offset, offset + len - 1);
pg_start = ((unsigned long long) offset) >> PAGE_CACHE_SHIFT;
pg_end = ((unsigned long long) offset + len) >> PAGE_CACHE_SHIFT;
off_start = offset & (PAGE_CACHE_SIZE - 1);
off_end = (offset + len) & (PAGE_CACHE_SIZE - 1);
if (pg_start == pg_end) {
ret = fill_zero(inode, pg_start, off_start,
off_end - off_start);
if (ret)
return ret;
if (offset + len > new_size)
new_size = offset + len;
new_size = max_t(loff_t, new_size, offset + len);
} else {
if (off_start) {
ret = fill_zero(inode, pg_start++, off_start,
PAGE_CACHE_SIZE - off_start);
if (ret)
return ret;
new_size = max_t(loff_t, new_size,
(loff_t)pg_start << PAGE_CACHE_SHIFT);
}
for (index = pg_start; index < pg_end; index++) {
struct dnode_of_data dn;
struct page *ipage;
f2fs_lock_op(sbi);
ipage = get_node_page(sbi, inode->i_ino);
if (IS_ERR(ipage)) {
ret = PTR_ERR(ipage);
f2fs_unlock_op(sbi);
goto out;
}
set_new_dnode(&dn, inode, ipage, NULL, 0);
ret = f2fs_reserve_block(&dn, index);
if (ret) {
f2fs_unlock_op(sbi);
goto out;
}
if (dn.data_blkaddr != NEW_ADDR) {
invalidate_blocks(sbi, dn.data_blkaddr);
dn.data_blkaddr = NEW_ADDR;
set_data_blkaddr(&dn);
dn.data_blkaddr = NULL_ADDR;
f2fs_update_extent_cache(&dn);
}
f2fs_put_dnode(&dn);
f2fs_unlock_op(sbi);
new_size = max_t(loff_t, new_size,
(loff_t)(index + 1) << PAGE_CACHE_SHIFT);
}
if (off_end) {
ret = fill_zero(inode, pg_end, 0, off_end);
if (ret)
goto out;
new_size = max_t(loff_t, new_size, offset + len);
}
}
out:
if (!(mode & FALLOC_FL_KEEP_SIZE) && i_size_read(inode) < new_size) {
i_size_write(inode, new_size);
mark_inode_dirty(inode);
update_inode_page(inode);
}
return ret;
}
struct dentry *f2fs_get_parent(struct dentry *child)
{
struct qstr dotdot = {.len = 2, .name = ".."};
struct page *page;
unsigned long ino = f2fs_inode_by_name(d_inode(child), &dotdot, &page);
if (!ino) {
if (IS_ERR(page))
return ERR_CAST(page);
return ERR_PTR(-ENOENT);
}
return d_obtain_alias(f2fs_iget(child->d_sb, ino));
}
static int __recover_dot_dentries(struct inode *dir, nid_t pino)
{
struct f2fs_sb_info *sbi = F2FS_I_SB(dir);
struct qstr dot = QSTR_INIT(".", 1);
struct qstr dotdot = QSTR_INIT("..", 2);
struct f2fs_dir_entry *de;
struct page *page;
int err = 0;
if (f2fs_readonly(sbi->sb)) {
f2fs_msg(sbi->sb, KERN_INFO,
"skip recovering inline_dots inode (ino:%lu, pino:%u) "
"in readonly mountpoint", dir->i_ino, pino);
return 0;
}
f2fs_balance_fs(sbi, true);
f2fs_lock_op(sbi);
de = f2fs_find_entry(dir, &dot, &page);
if (de) {
f2fs_dentry_kunmap(dir, page);
f2fs_put_page(page, 0);
} else if (IS_ERR(page)) {
err = PTR_ERR(page);
goto out;
} else {
err = __f2fs_add_link(dir, &dot, NULL, dir->i_ino, S_IFDIR);
if (err)
goto out;
}
de = f2fs_find_entry(dir, &dotdot, &page);
if (de) {
f2fs_dentry_kunmap(dir, page);
f2fs_put_page(page, 0);
} else if (IS_ERR(page)) {
err = PTR_ERR(page);
} else {
err = __f2fs_add_link(dir, &dotdot, NULL, pino, S_IFDIR);
}
out:
if (!err)
clear_inode_flag(dir, FI_INLINE_DOTS);
f2fs_unlock_op(sbi);
return err;
}
static struct dentry *f2fs_lookup(struct inode *dir, struct dentry *dentry,
struct nameidata *nd)
{
struct inode *inode = NULL;
struct f2fs_dir_entry *de;
struct page *page;
nid_t ino;
int err = 0;
unsigned int root_ino = F2FS_ROOT_INO(F2FS_I_SB(dir));
if (f2fs_encrypted_inode(dir)) {
int res = fscrypt_get_encryption_info(dir);
/*
* DCACHE_ENCRYPTED_WITH_KEY is set if the dentry is
* created while the directory was encrypted and we
* don't have access to the key.
*/
if (fscrypt_has_encryption_key(dir))
fscrypt_set_encrypted_dentry(dentry);
fscrypt_set_d_op(dentry);
if (res && res != -ENOKEY)
return ERR_PTR(res);
}
if (dentry->d_name.len > F2FS_NAME_LEN)
return ERR_PTR(-ENAMETOOLONG);
de = f2fs_find_entry(dir, &dentry->d_name, &page);
if (!de) {
if (IS_ERR(page))
return (struct dentry *)page;
return d_splice_alias(inode, dentry);
}
ino = le32_to_cpu(de->ino);
f2fs_dentry_kunmap(dir, page);
f2fs_put_page(page, 0);
inode = f2fs_iget(dir->i_sb, ino);
if (IS_ERR(inode))
return ERR_CAST(inode);
if ((dir->i_ino == root_ino) && f2fs_has_inline_dots(dir)) {
err = __recover_dot_dentries(dir, root_ino);
if (err)
goto err_out;
}
if (f2fs_has_inline_dots(inode)) {
err = __recover_dot_dentries(inode, dir->i_ino);
if (err)
goto err_out;
}
if (!IS_ERR(inode) && f2fs_encrypted_inode(dir) &&
(S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode)) &&
!fscrypt_has_permitted_context(dir, inode)) {
bool nokey = f2fs_encrypted_inode(inode) &&
!fscrypt_has_encryption_key(inode);
err = nokey ? -ENOKEY : -EPERM;
goto err_out;
}
return d_splice_alias(inode, dentry);
err_out:
iput(inode);
return ERR_PTR(err);
}
static int f2fs_unlink(struct inode *dir, struct dentry *dentry)
{
struct f2fs_sb_info *sbi = F2FS_I_SB(dir);
struct inode *inode = d_inode(dentry);
struct f2fs_dir_entry *de;
struct page *page;
int err = -ENOENT;
trace_f2fs_unlink_enter(dir, dentry);
de = f2fs_find_entry(dir, &dentry->d_name, &page);
if (!de) {
if (IS_ERR(page))
err = PTR_ERR(page);
goto fail;
}
f2fs_balance_fs(sbi, true);
f2fs_lock_op(sbi);
err = acquire_orphan_inode(sbi);
if (err) {
f2fs_unlock_op(sbi);
f2fs_dentry_kunmap(dir, page);
f2fs_put_page(page, 0);
goto fail;
}
f2fs_delete_entry(de, page, dir, inode);
f2fs_unlock_op(sbi);
if (IS_DIRSYNC(dir))
f2fs_sync_fs(sbi->sb, 1);
fail:
trace_f2fs_unlink_exit(inode, err);
return err;
}
static void *f2fs_follow_link(struct dentry *dentry, struct nameidata *nd)
{
struct page *page;
char *link;
page = page_follow_link_light(dentry, nd);
if (IS_ERR(page))
return page;
link = nd_get_link(nd);
if (IS_ERR(link))
return link;
/* this is broken symlink case */
if (*link == 0) {
kunmap(page);
page_cache_release(page);
return ERR_PTR(-ENOENT);
}
return page;
}
static int f2fs_symlink(struct inode *dir, struct dentry *dentry,
const char *symname)
{
struct f2fs_sb_info *sbi = F2FS_I_SB(dir);
struct inode *inode;
size_t len = strlen(symname);
struct fscrypt_str disk_link = FSTR_INIT((char *)symname, len + 1);
struct fscrypt_symlink_data *sd = NULL;
int err;
if (f2fs_encrypted_inode(dir)) {
err = fscrypt_get_encryption_info(dir);
if (err)
return err;
if (!fscrypt_has_encryption_key(dir))
return -EPERM;
disk_link.len = (fscrypt_fname_encrypted_size(dir, len) +
sizeof(struct fscrypt_symlink_data));
}
if (disk_link.len > dir->i_sb->s_blocksize)
return -ENAMETOOLONG;
inode = f2fs_new_inode(dir, S_IFLNK | S_IRWXUGO);
if (IS_ERR(inode))
return PTR_ERR(inode);
if (f2fs_encrypted_inode(inode))
inode->i_op = &f2fs_encrypted_symlink_inode_operations;
else
inode->i_op = &f2fs_symlink_inode_operations;
inode_nohighmem(inode);
inode->i_mapping->a_ops = &f2fs_dblock_aops;
f2fs_balance_fs(sbi, true);
f2fs_lock_op(sbi);
err = f2fs_add_link(dentry, inode);
if (err)
goto out;
f2fs_unlock_op(sbi);
alloc_nid_done(sbi, inode->i_ino);
if (f2fs_encrypted_inode(inode)) {
struct qstr istr = QSTR_INIT(symname, len);
struct fscrypt_str ostr;
sd = kzalloc(disk_link.len, GFP_NOFS);
if (!sd) {
err = -ENOMEM;
goto err_out;
}
err = fscrypt_get_encryption_info(inode);
if (err)
goto err_out;
if (!fscrypt_has_encryption_key(inode)) {
err = -EPERM;
goto err_out;
}
ostr.name = sd->encrypted_path;
ostr.len = disk_link.len;
err = fscrypt_fname_usr_to_disk(inode, &istr, &ostr);
if (err < 0)
goto err_out;
sd->len = cpu_to_le16(ostr.len);
disk_link.name = (char *)sd;
}
err = page_symlink(inode, disk_link.name, disk_link.len);
err_out:
d_instantiate(dentry, inode);
unlock_new_inode(inode);
/*
* Let's flush symlink data in order to avoid broken symlink as much as
* possible. Nevertheless, fsyncing is the best way, but there is no
* way to get a file descriptor in order to flush that.
*
* Note that, it needs to do dir->fsync to make this recoverable.
* If the symlink path is stored into inline_data, there is no
* performance regression.
*/
if (!err) {
filemap_write_and_wait_range(inode->i_mapping, 0,
disk_link.len - 1);
if (IS_DIRSYNC(dir))
f2fs_sync_fs(sbi->sb, 1);
} else {
f2fs_unlink(dir, dentry);
}
kfree(sd);
return err;
out:
handle_failed_inode(inode);
return err;
}
static int f2fs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode)
{
struct f2fs_sb_info *sbi = F2FS_I_SB(dir);
struct inode *inode;
int err;
inode = f2fs_new_inode(dir, S_IFDIR | mode);
if (IS_ERR(inode))
return PTR_ERR(inode);
inode->i_op = &f2fs_dir_inode_operations;
inode->i_fop = &f2fs_dir_operations;
inode->i_mapping->a_ops = &f2fs_dblock_aops;
mapping_set_gfp_mask(inode->i_mapping, GFP_F2FS_HIGH_ZERO);
f2fs_balance_fs(sbi, true);
set_inode_flag(inode, FI_INC_LINK);
f2fs_lock_op(sbi);
err = f2fs_add_link(dentry, inode);
if (err)
goto out_fail;
f2fs_unlock_op(sbi);
alloc_nid_done(sbi, inode->i_ino);
d_instantiate(dentry, inode);
unlock_new_inode(inode);
if (IS_DIRSYNC(dir))
f2fs_sync_fs(sbi->sb, 1);
return 0;
out_fail:
clear_inode_flag(inode, FI_INC_LINK);
handle_failed_inode(inode);
return err;
}
static int f2fs_rmdir(struct inode *dir, struct dentry *dentry)
{
struct inode *inode = d_inode(dentry);
if (f2fs_empty_dir(inode))
return f2fs_unlink(dir, dentry);
return -ENOTEMPTY;
}
static int f2fs_mknod(struct inode *dir, struct dentry *dentry,
umode_t mode, dev_t rdev)
{
struct f2fs_sb_info *sbi = F2FS_I_SB(dir);
struct inode *inode;
int err = 0;
if (!new_valid_dev(rdev))
return -EINVAL;
inode = f2fs_new_inode(dir, mode);
if (IS_ERR(inode))
return PTR_ERR(inode);
init_special_inode(inode, inode->i_mode, rdev);
inode->i_op = &f2fs_special_inode_operations;
f2fs_balance_fs(sbi, true);
f2fs_lock_op(sbi);
err = f2fs_add_link(dentry, inode);
if (err)
goto out;
f2fs_unlock_op(sbi);
alloc_nid_done(sbi, inode->i_ino);
d_instantiate(dentry, inode);
unlock_new_inode(inode);
if (IS_DIRSYNC(dir))
f2fs_sync_fs(sbi->sb, 1);
return 0;
out:
handle_failed_inode(inode);
return err;
}
static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry,
struct inode *new_dir, struct dentry *new_dentry)
{
struct f2fs_sb_info *sbi = F2FS_I_SB(old_dir);
struct inode *old_inode = d_inode(old_dentry);
struct inode *new_inode = d_inode(new_dentry);
struct page *old_dir_page;
struct page *old_page, *new_page;
struct f2fs_dir_entry *old_dir_entry = NULL;
struct f2fs_dir_entry *old_entry;
struct f2fs_dir_entry *new_entry;
bool is_old_inline = f2fs_has_inline_dentry(old_dir);
int err = -ENOENT;
if ((old_dir != new_dir) && f2fs_encrypted_inode(new_dir) &&
!fscrypt_has_permitted_context(new_dir, old_inode)) {
err = -EPERM;
goto out;
}
old_entry = f2fs_find_entry(old_dir, &old_dentry->d_name, &old_page);
if (!old_entry) {
if (IS_ERR(old_page))
err = PTR_ERR(old_page);
goto out;
}
if (S_ISDIR(old_inode->i_mode)) {
old_dir_entry = f2fs_parent_dir(old_inode, &old_dir_page);
if (!old_dir_entry) {
if (IS_ERR(old_dir_page))
err = PTR_ERR(old_dir_page);
goto out_old;
}
}
if (new_inode) {
err = -ENOTEMPTY;
if (old_dir_entry && !f2fs_empty_dir(new_inode))
goto out_dir;
err = -ENOENT;
new_entry = f2fs_find_entry(new_dir, &new_dentry->d_name,
&new_page);
if (!new_entry) {
if (IS_ERR(new_page))
err = PTR_ERR(new_page);
goto out_dir;
}
f2fs_balance_fs(sbi, true);
f2fs_lock_op(sbi);
err = acquire_orphan_inode(sbi);
if (err)
goto put_out_dir;
err = update_dent_inode(old_inode, new_inode,
&new_dentry->d_name);
if (err) {
release_orphan_inode(sbi);
goto put_out_dir;
}
f2fs_set_link(new_dir, new_entry, new_page, old_inode);
new_inode->i_ctime = CURRENT_TIME;
down_write(&F2FS_I(new_inode)->i_sem);
if (old_dir_entry)
f2fs_i_links_write(new_inode, false);
f2fs_i_links_write(new_inode, false);
up_write(&F2FS_I(new_inode)->i_sem);
if (!new_inode->i_nlink)
add_orphan_inode(new_inode);
else
release_orphan_inode(sbi);
} else {
f2fs_balance_fs(sbi, true);
f2fs_lock_op(sbi);
err = f2fs_add_link(new_dentry, old_inode);
if (err) {
f2fs_unlock_op(sbi);
goto out_dir;
}
if (old_dir_entry)
f2fs_i_links_write(new_dir, true);
/*
* old entry and new entry can locate in the same inline
* dentry in inode, when attaching new entry in inline dentry,
* it could force inline dentry conversion, after that,
* old_entry and old_page will point to wrong address, in
* order to avoid this, let's do the check and update here.
*/
if (is_old_inline && !f2fs_has_inline_dentry(old_dir)) {
f2fs_put_page(old_page, 0);
old_page = NULL;
old_entry = f2fs_find_entry(old_dir,
&old_dentry->d_name, &old_page);
if (!old_entry) {
err = -ENOENT;
if (IS_ERR(old_page))
err = PTR_ERR(old_page);
f2fs_unlock_op(sbi);
goto out_dir;
}
}
}
down_write(&F2FS_I(old_inode)->i_sem);
file_lost_pino(old_inode);
if (new_inode && file_enc_name(new_inode))
file_set_enc_name(old_inode);
up_write(&F2FS_I(old_inode)->i_sem);
old_inode->i_ctime = CURRENT_TIME;
f2fs_mark_inode_dirty_sync(old_inode);
f2fs_delete_entry(old_entry, old_page, old_dir, NULL);
if (old_dir_entry) {
if (old_dir != new_dir) {
f2fs_set_link(old_inode, old_dir_entry,
old_dir_page, new_dir);
} else {
f2fs_dentry_kunmap(old_inode, old_dir_page);
f2fs_put_page(old_dir_page, 0);
}
f2fs_i_links_write(old_dir, false);
}
f2fs_unlock_op(sbi);
if (IS_DIRSYNC(old_dir) || IS_DIRSYNC(new_dir))
f2fs_sync_fs(sbi->sb, 1);
return 0;
put_out_dir:
f2fs_unlock_op(sbi);
f2fs_dentry_kunmap(new_dir, new_page);
f2fs_put_page(new_page, 0);
out_dir:
if (old_dir_entry) {
f2fs_dentry_kunmap(old_inode, old_dir_page);
f2fs_put_page(old_dir_page, 0);
}
out_old:
f2fs_dentry_kunmap(old_dir, old_page);
f2fs_put_page(old_page, 0);
out:
return err;
}
static void *f2fs_encrypted_follow_link(struct dentry *dentry,
struct nameidata *nd)
{
struct page *cpage = NULL;
char *caddr, *paddr = NULL;
struct fscrypt_str cstr = FSTR_INIT(NULL, 0);
struct fscrypt_str pstr = FSTR_INIT(NULL, 0);
struct fscrypt_symlink_data *sd;
struct inode *inode = d_inode(dentry);
loff_t size = min_t(loff_t, i_size_read(inode), PAGE_SIZE - 1);
u32 max_size = inode->i_sb->s_blocksize;
int res;
res = fscrypt_get_encryption_info(inode);
if (res)
return ERR_PTR(res);
cpage = read_mapping_page(inode->i_mapping, 0, NULL);
if (IS_ERR(cpage))
return cpage;
caddr = kmap(cpage);
caddr[size] = 0;
/* Symlink is encrypted */
sd = (struct fscrypt_symlink_data *)caddr;
cstr.name = sd->encrypted_path;
cstr.len = le16_to_cpu(sd->len);
/* this is broken symlink case */
if (unlikely(cstr.len == 0)) {
res = -ENOENT;
goto errout;
}
if ((cstr.len + sizeof(struct fscrypt_symlink_data) - 1) > max_size) {
/* Symlink data on the disk is corrupted */
res = -EIO;
goto errout;
}
res = fscrypt_fname_alloc_buffer(inode, cstr.len, &pstr);
if (res)
goto errout;
res = fscrypt_fname_disk_to_usr(inode, 0, 0, &cstr, &pstr);
if (res < 0)
goto errout;
/* this is broken symlink case */
if (unlikely(pstr.name[0] == 0)) {
res = -ENOENT;
goto errout;
}
paddr = pstr.name;
/* Null-terminate the name */
paddr[res] = '\0';
nd_set_link(nd, paddr);
kunmap(cpage);
page_cache_release(cpage);
return NULL;
errout:
fscrypt_fname_free_buffer(&pstr);
kunmap(cpage);
page_cache_release(cpage);
return ERR_PTR(res);
}
void kfree_put_link(struct dentry *dentry, struct nameidata *nd,
void *cookie)
{
char *s = nd_get_link(nd);
if (!IS_ERR(s))
kfree(s);
}
static int ncp_fsync(struct file *file, loff_t start, loff_t end, int datasync)
{
return filemap_write_and_wait_range(file->f_mapping, start, end);
}
STATIC ssize_t
xfs_file_read_iter(
struct kiocb *iocb,
struct iov_iter *to)
{
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 = iov_iter_count(to);
ssize_t ret = 0;
int ioflags = 0;
xfs_fsize_t n;
loff_t pos = iocb->ki_pos;
XFS_STATS_INC(xs_read_calls);
if (unlikely(file->f_flags & O_DIRECT))
ioflags |= XFS_IO_ISDIRECT;
if (file->f_mode & FMODE_NOCMTIME)
ioflags |= XFS_IO_INVIS;
if (unlikely(ioflags & XFS_IO_ISDIRECT)) {
xfs_buftarg_t *target =
XFS_IS_REALTIME_INODE(ip) ?
mp->m_rtdev_targp : mp->m_ddev_targp;
/* DIO must be aligned to device logical sector size */
if ((pos | size) & target->bt_logical_sectormask) {
if (pos == i_size_read(inode))
return 0;
return -EINVAL;
}
}
n = mp->m_super->s_maxbytes - 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 & XFS_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 = filemap_write_and_wait_range(
VFS_I(ip)->i_mapping,
pos, pos + size - 1);
if (ret) {
xfs_rw_iunlock(ip, XFS_IOLOCK_EXCL);
return ret;
}
/*
* Invalidate whole pages. This can return an error if
* we fail to invalidate a page, but this should never
* happen on XFS. Warn if it does fail.
*/
ret = invalidate_inode_pages2_range(VFS_I(ip)->i_mapping,
pos >> PAGE_CACHE_SHIFT,
(pos + size - 1) >> PAGE_CACHE_SHIFT);
WARN_ON_ONCE(ret);
ret = 0;
}
int ext4_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
{
struct inode *inode = file->f_mapping->host;
struct ext4_inode_info *ei = EXT4_I(inode);
journal_t *journal = EXT4_SB(inode->i_sb)->s_journal;
int ret, err;
tid_t commit_tid;
bool needs_barrier = false;
J_ASSERT(ext4_journal_current_handle() == NULL);
trace_ext4_sync_file_enter(file, datasync);
ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
if (ret)
return ret;
if (!mutex_trylock(&inode->i_mutex)) {
ret = blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL);
trace_ext4_sync_file_exit(inode, ret);
return ret;
}
if (inode->i_sb->s_flags & MS_RDONLY)
goto out;
ret = ext4_flush_unwritten_io(inode);
if (ret < 0)
goto out;
if (!journal) {
ret = __sync_inode(inode, datasync);
if (!ret && !hlist_empty(&inode->i_dentry))
ret = ext4_sync_parent(inode);
goto out;
}
/*
* data=writeback,ordered:
* The caller's filemap_fdatawrite()/wait will sync the data.
* Metadata is in the journal, we wait for proper transaction to
* commit here.
*
* data=journal:
* filemap_fdatawrite won't do anything (the buffers are clean).
* ext4_force_commit will write the file data into the journal and
* will wait on that.
* filemap_fdatawait() will encounter a ton of newly-dirtied pages
* (they were dirtied by commit). But that's OK - the blocks are
* safe in-journal, which is all fsync() needs to ensure.
*/
if (ext4_should_journal_data(inode)) {
ret = ext4_force_commit(inode->i_sb);
goto out;
}
commit_tid = datasync ? ei->i_datasync_tid : ei->i_sync_tid;
if (journal->j_flags & JBD2_BARRIER &&
!jbd2_trans_will_send_data_barrier(journal, commit_tid))
needs_barrier = true;
ret = jbd2_complete_transaction(journal, commit_tid);
if (needs_barrier) {
err = blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL);
if (!ret)
ret = err;
}
out:
mutex_unlock(&inode->i_mutex);
trace_ext4_sync_file_exit(inode, ret);
return ret;
}
int vfs_fsync_range(struct file *file, loff_t start, loff_t end, int datasync)
{
struct address_space *mapping = file->f_mapping;
int err, ret;
#ifdef FEATURE_PRINT_FSYNC_PID
pid_t curr_pid;
unsigned int i;
unsigned long long time1=0;
bool ptr_flag=false;
#endif
if (!file->f_op || !file->f_op->fsync) {
ret = -EINVAL;
goto out;
}
ret = filemap_write_and_wait_range(mapping, start, end);
/*
* We need to protect against concurrent writers, which could cause
* livelocks in fsync_buffers_list().
*/
mutex_lock(&mapping->host->i_mutex);
#ifdef FEATURE_PRINT_FSYNC_PID
time1 = sched_clock();
mutex_lock(&fsync_mutex);
if(fsync_last_t == 0)
{
fsync_last_t = time1;
}
if (time1 - fsync_last_t >= (unsigned long long)PRT_TIME_PERIOD)
{
sprintf(xlog_buf, "Fsync [(PID):cnt] -- ");
for(i=0;i<ID_CNT;i++)
{
if(fsync[i].pid==0)
break;
else
{
sprintf(xlog_buf+21+i*9, "(%4d):%d ", fsync[i].pid, fsync[i].cnt); //21=strlen("Fsync [(PID):cnt] -- "), 9=strlen("(%4d):%d ")
ptr_flag = true;
}
}
if(ptr_flag)
{
xlog_printk(ANDROID_LOG_INFO, "BLOCK_TAG", "Fsync statistic in timeline %lld\n", fsync_last_t);
xlog_printk(ANDROID_LOG_INFO, "BLOCK_TAG", "%s", xlog_buf);
}
for (i=0;i<ID_CNT;i++) //clear
{
fsync[i].pid=0;
fsync[i].cnt=0;
}
fsync_last_t = time1;
}
curr_pid = task_pid_nr(current);
do{
if(fsync[0].pid ==0)
{
fsync[0].pid= curr_pid;
fsync[0].cnt ++;
break;
}
if(curr_pid == fsync[idx].pid)
{
fsync[idx].cnt++;
break;
}
for(i=0;i<ID_CNT;i++)
{
if (curr_pid == fsync[i].pid) //found in the array
{
fsync[i].cnt++;
idx =i;
break;
}
else if (fsync[i+1].pid== 0) //find the empty space
{
if(i+1==ID_CNT) //check buf full, record in the last element
{
i -= 1;
fsync[i+1].cnt =0;
}
fsync[i+1].pid = curr_pid;
fsync[i+1].cnt++;
idx=i+1;
break;
}
}
}while(0);
mutex_unlock(&fsync_mutex);
#endif
err = file->f_op->fsync(file, datasync);
if (!ret)
ret = err;
mutex_unlock(&mapping->host->i_mutex);
out:
return ret;
}
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;
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);
}
/*
* All metadata updates are logged, which means that we just have
* to flush the log up to the latest LSN that touched the inode.
*/
xfs_ilock(ip, XFS_ILOCK_SHARED);
if (xfs_ipincount(ip)) {
if (!datasync ||
(ip->i_itemp->ili_fields & ~XFS_ILOG_TIMESTAMP))
lsn = ip->i_itemp->ili_last_lsn;
}
xfs_iunlock(ip, XFS_ILOCK_SHARED);
if (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;
}
static int f2fs_insert_range(struct inode *inode, loff_t offset, loff_t len)
{
struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
pgoff_t pg_start, pg_end, delta, nrpages, idx;
loff_t new_size;
int ret;
if (!S_ISREG(inode->i_mode))
return -EINVAL;
new_size = i_size_read(inode) + len;
if (new_size > inode->i_sb->s_maxbytes)
return -EFBIG;
if (offset >= i_size_read(inode))
return -EINVAL;
/* insert range should be aligned to block size of f2fs. */
if (offset & (F2FS_BLKSIZE - 1) || len & (F2FS_BLKSIZE - 1))
return -EINVAL;
f2fs_balance_fs(sbi);
ret = truncate_blocks(inode, i_size_read(inode), true);
if (ret)
return ret;
/* write out all dirty pages from offset */
ret = filemap_write_and_wait_range(inode->i_mapping, offset, LLONG_MAX);
if (ret)
return ret;
truncate_pagecache(inode, offset);
pg_start = offset >> PAGE_CACHE_SHIFT;
pg_end = (offset + len) >> PAGE_CACHE_SHIFT;
delta = pg_end - pg_start;
nrpages = (i_size_read(inode) + PAGE_SIZE - 1) / PAGE_SIZE;
for (idx = nrpages - 1; idx >= pg_start && idx != -1; idx--) {
struct dnode_of_data dn;
struct page *ipage;
block_t new_addr, old_addr;
f2fs_lock_op(sbi);
set_new_dnode(&dn, inode, NULL, NULL, 0);
ret = get_dnode_of_data(&dn, idx, LOOKUP_NODE_RA);
if (ret && ret != -ENOENT) {
goto out;
} else if (ret == -ENOENT) {
goto next;
} else if (dn.data_blkaddr == NULL_ADDR) {
f2fs_put_dnode(&dn);
goto next;
} else {
new_addr = dn.data_blkaddr;
truncate_data_blocks_range(&dn, 1);
f2fs_put_dnode(&dn);
}
ipage = get_node_page(sbi, inode->i_ino);
if (IS_ERR(ipage)) {
ret = PTR_ERR(ipage);
goto out;
}
set_new_dnode(&dn, inode, ipage, NULL, 0);
ret = f2fs_reserve_block(&dn, idx + delta);
if (ret)
goto out;
old_addr = dn.data_blkaddr;
f2fs_bug_on(sbi, old_addr != NEW_ADDR);
if (new_addr != NEW_ADDR) {
struct node_info ni;
get_node_info(sbi, dn.nid, &ni);
f2fs_replace_block(sbi, &dn, old_addr, new_addr,
ni.version, true);
}
f2fs_put_dnode(&dn);
next:
f2fs_unlock_op(sbi);
}
i_size_write(inode, new_size);
return 0;
out:
f2fs_unlock_op(sbi);
return ret;
}
/*
* 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)
{
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;
int iolock;
struct xfs_buftarg *target = XFS_IS_REALTIME_INODE(ip) ?
mp->m_rtdev_targp : mp->m_ddev_targp;
/* DIO must be aligned to device logical sector size */
if ((pos | count) & target->bt_logical_sectormask)
return -XFS_ERROR(EINVAL);
/* "unaligned" here means not aligned to a filesystem block */
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, &iolock);
if (ret)
goto out;
if (mapping->nrpages) {
ret = -filemap_write_and_wait_range(VFS_I(ip)->i_mapping,
pos, -1);
if (ret)
goto out;
truncate_pagecache_range(VFS_I(ip), pos, -1);
}
/*
* 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);
out:
xfs_rw_iunlock(ip, iolock);
/* No fallback to buffered IO on errors for XFS. */
ASSERT(ret < 0 || ret == count);
return ret;
}
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;
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 -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);
}
/*
* All metadata updates are logged, which means that we just have to
* flush the log up to the latest LSN that touched the inode. If we have
* concurrent fsync/fdatasync() calls, we need them to all block on the
* log force before we clear the ili_fsync_fields field. This ensures
* that we don't get a racing sync operation that does not wait for the
* metadata to hit the journal before returning. If we race with
* clearing the ili_fsync_fields, then all that will happen is the log
* force will do nothing as the lsn will already be on disk. We can't
* race with setting ili_fsync_fields because that is done under
* XFS_ILOCK_EXCL, and that can't happen because we hold the lock shared
* until after the ili_fsync_fields is cleared.
*/
xfs_ilock(ip, XFS_ILOCK_SHARED);
if (xfs_ipincount(ip)) {
if (!datasync ||
(ip->i_itemp->ili_fsync_fields & ~XFS_ILOG_TIMESTAMP))
lsn = ip->i_itemp->ili_last_lsn;
}
if (lsn) {
error = _xfs_log_force_lsn(mp, lsn, XFS_LOG_SYNC, &log_flushed);
ip->i_itemp->ili_fsync_fields = 0;
}
xfs_iunlock(ip, XFS_ILOCK_SHARED);
/*
* 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;
}
/*
* Truncate file. Must have write permission and not be a directory.
*/
int
xfs_setattr_size(
struct xfs_inode *ip,
struct iattr *iattr)
{
struct xfs_mount *mp = ip->i_mount;
struct inode *inode = VFS_I(ip);
xfs_off_t oldsize, newsize;
struct xfs_trans *tp;
int error;
uint lock_flags = 0;
bool did_zeroing = false;
trace_xfs_setattr(ip);
if (mp->m_flags & XFS_MOUNT_RDONLY)
return -EROFS;
if (XFS_FORCED_SHUTDOWN(mp))
return -EIO;
error = inode_change_ok(inode, iattr);
if (error)
return error;
ASSERT(xfs_isilocked(ip, XFS_IOLOCK_EXCL));
ASSERT(xfs_isilocked(ip, XFS_MMAPLOCK_EXCL));
ASSERT(S_ISREG(ip->i_d.di_mode));
ASSERT((iattr->ia_valid & (ATTR_UID|ATTR_GID|ATTR_ATIME|ATTR_ATIME_SET|
ATTR_MTIME_SET|ATTR_KILL_PRIV|ATTR_TIMES_SET)) == 0);
oldsize = inode->i_size;
newsize = iattr->ia_size;
/*
* Short circuit the truncate case for zero length files.
*/
if (newsize == 0 && oldsize == 0 && ip->i_d.di_nextents == 0) {
if (!(iattr->ia_valid & (ATTR_CTIME|ATTR_MTIME)))
return 0;
/*
* Use the regular setattr path to update the timestamps.
*/
iattr->ia_valid &= ~ATTR_SIZE;
return xfs_setattr_nonsize(ip, iattr, 0);
}
/*
* Make sure that the dquots are attached to the inode.
*/
error = xfs_qm_dqattach(ip, 0);
if (error)
return error;
/*
* File data changes must be complete before we start the transaction to
* modify the inode. This needs to be done before joining the inode to
* the transaction because the inode cannot be unlocked once it is a
* part of the transaction.
*
* Start with zeroing any data block beyond EOF that we may expose on
* file extension.
*/
if (newsize > oldsize) {
error = xfs_zero_eof(ip, newsize, oldsize, &did_zeroing);
if (error)
return error;
}
/*
* We are going to log the inode size change in this transaction so
* any previous writes that are beyond the on disk EOF and the new
* EOF that have not been written out need to be written here. If we
* do not write the data out, we expose ourselves to the null files
* problem. Note that this includes any block zeroing we did above;
* otherwise those blocks may not be zeroed after a crash.
*/
if (newsize > ip->i_d.di_size &&
(oldsize != ip->i_d.di_size || did_zeroing)) {
error = filemap_write_and_wait_range(VFS_I(ip)->i_mapping,
ip->i_d.di_size, newsize);
if (error)
return error;
}
/* Now wait for all direct I/O to complete. */
inode_dio_wait(inode);
/*
* We've already locked out new page faults, so now we can safely remove
* pages from the page cache knowing they won't get refaulted until we
* drop the XFS_MMAP_EXCL lock after the extent manipulations are
* complete. The truncate_setsize() call also cleans partial EOF page
* PTEs on extending truncates and hence ensures sub-page block size
* filesystems are correctly handled, too.
*
* We have to do all the page cache truncate work outside the
* transaction context as the "lock" order is page lock->log space
* reservation as defined by extent allocation in the writeback path.
* Hence a truncate can fail with ENOMEM from xfs_trans_reserve(), but
* having already truncated the in-memory version of the file (i.e. made
* user visible changes). There's not much we can do about this, except
* to hope that the caller sees ENOMEM and retries the truncate
* operation.
*/
if (IS_DAX(inode))
error = dax_truncate_page(inode, newsize, xfs_get_blocks_direct);
else
error = block_truncate_page(inode->i_mapping, newsize,
xfs_get_blocks);
if (error)
return error;
truncate_setsize(inode, newsize);
tp = xfs_trans_alloc(mp, XFS_TRANS_SETATTR_SIZE);
error = xfs_trans_reserve(tp, &M_RES(mp)->tr_itruncate, 0, 0);
if (error)
goto out_trans_cancel;
lock_flags |= XFS_ILOCK_EXCL;
xfs_ilock(ip, XFS_ILOCK_EXCL);
xfs_trans_ijoin(tp, ip, 0);
/*
* Only change the c/mtime if we are changing the size or we are
* explicitly asked to change it. This handles the semantic difference
* between truncate() and ftruncate() as implemented in the VFS.
*
* The regular truncate() case without ATTR_CTIME and ATTR_MTIME is a
* special case where we need to update the times despite not having
* these flags set. For all other operations the VFS set these flags
* explicitly if it wants a timestamp update.
*/
if (newsize != oldsize &&
!(iattr->ia_valid & (ATTR_CTIME | ATTR_MTIME))) {
iattr->ia_ctime = iattr->ia_mtime =
current_fs_time(inode->i_sb);
iattr->ia_valid |= ATTR_CTIME | ATTR_MTIME;
}
/*
* The first thing we do is set the size to new_size permanently on
* disk. This way we don't have to worry about anyone ever being able
* to look at the data being freed even in the face of a crash.
* What we're getting around here is the case where we free a block, it
* is allocated to another file, it is written to, and then we crash.
* If the new data gets written to the file but the log buffers
* containing the free and reallocation don't, then we'd end up with
* garbage in the blocks being freed. As long as we make the new size
* permanent before actually freeing any blocks it doesn't matter if
* they get written to.
*/
ip->i_d.di_size = newsize;
xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
if (newsize <= oldsize) {
error = xfs_itruncate_extents(&tp, ip, XFS_DATA_FORK, newsize);
if (error)
goto out_trans_cancel;
/*
* Truncated "down", so we're removing references to old data
* here - if we delay flushing for a long time, we expose
* ourselves unduly to the notorious NULL files problem. So,
* we mark this inode and flush it when the file is closed,
* and do not wait the usual (long) time for writeout.
*/
xfs_iflags_set(ip, XFS_ITRUNCATED);
/* A truncate down always removes post-EOF blocks. */
xfs_inode_clear_eofblocks_tag(ip);
}
if (iattr->ia_valid & ATTR_MODE)
xfs_setattr_mode(ip, iattr);
if (iattr->ia_valid & (ATTR_ATIME|ATTR_CTIME|ATTR_MTIME))
xfs_setattr_time(ip, iattr);
xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
XFS_STATS_INC(mp, xs_ig_attrchg);
if (mp->m_flags & XFS_MOUNT_WSYNC)
xfs_trans_set_sync(tp);
error = xfs_trans_commit(tp);
out_unlock:
if (lock_flags)
xfs_iunlock(ip, lock_flags);
return error;
out_trans_cancel:
xfs_trans_cancel(tp);
goto out_unlock;
}
static int f2fs_defragment_range(struct f2fs_sb_info *sbi,
struct file *filp,
struct f2fs_defragment *range)
{
struct inode *inode = file_inode(filp);
struct f2fs_map_blocks map = { .m_next_pgofs = NULL };
struct extent_info ei;
pgoff_t pg_start, pg_end;
unsigned int blk_per_seg = sbi->blocks_per_seg;
unsigned int total = 0, sec_num;
unsigned int pages_per_sec = sbi->segs_per_sec * blk_per_seg;
block_t blk_end = 0;
bool fragmented = false;
int err;
/* if in-place-update policy is enabled, don't waste time here */
if (need_inplace_update(inode))
return -EINVAL;
pg_start = range->start >> PAGE_CACHE_SHIFT;
pg_end = (range->start + range->len) >> PAGE_CACHE_SHIFT;
f2fs_balance_fs(sbi, true);
inode_lock(inode);
/* writeback all dirty pages in the range */
err = filemap_write_and_wait_range(inode->i_mapping, range->start,
range->start + range->len - 1);
if (err)
goto out;
/*
* lookup mapping info in extent cache, skip defragmenting if physical
* block addresses are continuous.
*/
if (f2fs_lookup_extent_cache(inode, pg_start, &ei)) {
if (ei.fofs + ei.len >= pg_end)
goto out;
}
map.m_lblk = pg_start;
/*
* lookup mapping info in dnode page cache, skip defragmenting if all
* physical block addresses are continuous even if there are hole(s)
* in logical blocks.
*/
while (map.m_lblk < pg_end) {
map.m_len = pg_end - map.m_lblk;
err = f2fs_map_blocks(inode, &map, 0, F2FS_GET_BLOCK_READ);
if (err)
goto out;
if (!(map.m_flags & F2FS_MAP_FLAGS)) {
map.m_lblk++;
continue;
}
if (blk_end && blk_end != map.m_pblk) {
fragmented = true;
break;
}
blk_end = map.m_pblk + map.m_len;
map.m_lblk += map.m_len;
}
if (!fragmented)
goto out;
map.m_lblk = pg_start;
map.m_len = pg_end - pg_start;
sec_num = (map.m_len + pages_per_sec - 1) / pages_per_sec;
/*
* make sure there are enough free section for LFS allocation, this can
* avoid defragment running in SSR mode when free section are allocated
* intensively
*/
if (has_not_enough_free_secs(sbi, sec_num)) {
err = -EAGAIN;
goto out;
}
while (map.m_lblk < pg_end) {
pgoff_t idx;
int cnt = 0;
do_map:
map.m_len = pg_end - map.m_lblk;
err = f2fs_map_blocks(inode, &map, 0, F2FS_GET_BLOCK_READ);
if (err)
goto clear_out;
if (!(map.m_flags & F2FS_MAP_FLAGS)) {
map.m_lblk++;
continue;
}
set_inode_flag(F2FS_I(inode), FI_DO_DEFRAG);
idx = map.m_lblk;
while (idx < map.m_lblk + map.m_len && cnt < blk_per_seg) {
struct page *page;
page = get_lock_data_page(inode, idx, true);
if (IS_ERR(page)) {
err = PTR_ERR(page);
goto clear_out;
}
set_page_dirty(page);
f2fs_put_page(page, 1);
idx++;
cnt++;
total++;
}
map.m_lblk = idx;
if (idx < pg_end && cnt < blk_per_seg)
goto do_map;
clear_inode_flag(F2FS_I(inode), FI_DO_DEFRAG);
err = filemap_fdatawrite(inode->i_mapping);
if (err)
goto out;
}
clear_out:
clear_inode_flag(F2FS_I(inode), FI_DO_DEFRAG);
out:
inode_unlock(inode);
if (!err)
range->len = (u64)total << PAGE_CACHE_SHIFT;
return err;
}