/*
* If we're not journaling and this is a just-created file, we have to
* sync our parent directory (if it was freshly created) since
* otherwise it will only be written by writeback, leaving a huge
* window during which a crash may lose the file. This may apply for
* the parent directory's parent as well, and so on recursively, if
* they are also freshly created.
*/
static int ext4_sync_parent(struct inode *inode)
{
struct dentry *dentry = NULL;
struct inode *next;
int ret = 0;
if (!ext4_test_inode_state(inode, EXT4_STATE_NEWENTRY))
return 0;
inode = igrab(inode);
while (ext4_test_inode_state(inode, EXT4_STATE_NEWENTRY)) {
ext4_clear_inode_state(inode, EXT4_STATE_NEWENTRY);
dentry = d_find_any_alias(inode);
if (!dentry)
break;
next = igrab(dentry->d_parent->d_inode);
dput(dentry);
if (!next)
break;
iput(inode);
inode = next;
ret = sync_mapping_buffers(inode->i_mapping);
if (ret)
break;
ret = sync_inode_metadata(inode, 1);
if (ret)
break;
}
iput(inode);
return ret;
}
static int gfs2_fsync(struct file *file, loff_t start, loff_t end,
int datasync)
{
struct address_space *mapping = file->f_mapping;
struct inode *inode = mapping->host;
int sync_state = inode->i_state & (I_DIRTY_SYNC|I_DIRTY_DATASYNC);
struct gfs2_inode *ip = GFS2_I(inode);
int ret = 0, ret1 = 0;
if (mapping->nrpages) {
ret1 = filemap_fdatawrite_range(mapping, start, end);
if (ret1 == -EIO)
return ret1;
}
if (datasync)
sync_state &= ~I_DIRTY_SYNC;
if (sync_state) {
ret = sync_inode_metadata(inode, 1);
if (ret)
return ret;
if (gfs2_is_jdata(ip))
filemap_write_and_wait(mapping);
gfs2_ail_flush(ip->i_gl, 1);
}
if (mapping->nrpages)
ret = filemap_fdatawait_range(mapping, start, end);
return ret ? ret : ret1;
}
/* exofs_file_fsync - flush the inode to disk
*
* Note, in exofs all metadata is written as part of inode, regardless.
* The writeout is synchronous
*/
static int exofs_file_fsync(struct file *filp, int datasync)
{
int ret;
ret = sync_inode_metadata(filp->f_mapping->host, 1);
return ret;
}
/* exofs_file_fsync - flush the inode to disk
*
* Note, in exofs all metadata is written as part of inode, regardless.
* The writeout is synchronous
*/
static int exofs_file_fsync(struct file *filp, loff_t start, loff_t end,
int datasync)
{
struct inode *inode = filp->f_mapping->host;
int ret;
ret = file_write_and_wait_range(filp, start, end);
if (ret)
return ret;
inode_lock(inode);
ret = sync_inode_metadata(filp->f_mapping->host, 1);
inode_unlock(inode);
return ret;
}
static int exofs_file_fsync(struct file *filp, int datasync)
{
int ret;
struct inode *inode = filp->f_mapping->host;
struct super_block *sb;
ret = sync_inode_metadata(inode, 1);
/* This is a good place to write the sb */
/* TODO: Sechedule an sb-sync on create */
sb = inode->i_sb;
if (sb->s_dirt)
exofs_sync_fs(sb, 1);
return ret;
}
/**
* __sync_file - generic_file_fsync without the locking and filemap_write
* @inode: inode to sync
* @datasync: only sync essential metadata if true
*
* This is just generic_file_fsync without the locking. This is needed for
* nojournal mode to make sure this inodes data/metadata makes it to disk
* properly. The i_mutex should be held already.
*/
static int __sync_inode(struct inode *inode, int datasync)
{
int err;
int ret;
ret = sync_mapping_buffers(inode->i_mapping);
if (!(inode->i_state & I_DIRTY))
return ret;
if (datasync && !(inode->i_state & I_DIRTY_DATASYNC))
return ret;
err = sync_inode_metadata(inode, 1);
if (ret == 0)
ret = err;
return ret;
}
/*
* If we're not journaling and this is a just-created file, we have to
* sync our parent directory (if it was freshly created) since
* otherwise it will only be written by writeback, leaving a huge
* window during which a crash may lose the file. This may apply for
* the parent directory's parent as well, and so on recursively, if
* they are also freshly created.
*/
static int ext4_sync_parent(struct inode *inode)
{
struct dentry *dentry = NULL;
struct inode *next;
int ret = 0;
if (!ext4_test_inode_state(inode, EXT4_STATE_NEWENTRY))
return 0;
inode = igrab(inode);
while (ext4_test_inode_state(inode, EXT4_STATE_NEWENTRY)) {
ext4_clear_inode_state(inode, EXT4_STATE_NEWENTRY);
dentry = d_find_any_alias(inode);
if (!dentry)
break;
next = igrab(d_inode(dentry->d_parent));
dput(dentry);
if (!next)
break;
iput(inode);
inode = next;
/*
* The directory inode may have gone through rmdir by now. But
* the inode itself and its blocks are still allocated (we hold
* a reference to the inode so it didn't go through
* ext4_evict_inode()) and so we are safe to flush metadata
* blocks and the inode.
*/
ret = sync_mapping_buffers(inode->i_mapping);
if (ret)
break;
ret = sync_inode_metadata(inode, 1);
if (ret)
break;
}
iput(inode);
return ret;
}
static int gfs2_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
{
struct page *page = vmf->page;
struct inode *inode = vma->vm_file->f_path.dentry->d_inode;
struct gfs2_inode *ip = GFS2_I(inode);
struct gfs2_sbd *sdp = GFS2_SB(inode);
struct gfs2_alloc_parms ap = { .aflags = 0, };
unsigned long last_index;
u64 pos = page->index << PAGE_CACHE_SHIFT;
unsigned int data_blocks, ind_blocks, rblocks;
int alloc_required = 0;
struct gfs2_holder gh;
loff_t size;
int ret;
sb_start_pagefault(inode->i_sb);
/* Update file times before taking page lock */
file_update_time(vma->vm_file);
ret = get_write_access(inode);
if (ret)
goto out;
ret = gfs2_rs_alloc(ip);
if (ret)
goto out_write_access;
gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh);
ret = gfs2_glock_nq(&gh);
if (ret)
goto out_uninit;
set_bit(GLF_DIRTY, &ip->i_gl->gl_flags);
set_bit(GIF_SW_PAGED, &ip->i_flags);
gfs2_size_hint(inode, pos, PAGE_CACHE_SIZE);
ret = gfs2_write_alloc_required(ip, pos, PAGE_CACHE_SIZE, &alloc_required);
if (ret)
goto out_unlock;
if (!alloc_required) {
lock_page(page);
if (!PageUptodate(page) || page->mapping != inode->i_mapping) {
ret = -EAGAIN;
unlock_page(page);
}
goto out_unlock;
}
ret = gfs2_rindex_update(sdp);
if (ret)
goto out_unlock;
ret = gfs2_quota_lock_check(ip);
if (ret)
goto out_unlock;
gfs2_write_calc_reserv(ip, PAGE_CACHE_SIZE, &data_blocks, &ind_blocks);
ap.target = data_blocks + ind_blocks;
ret = gfs2_inplace_reserve(ip, &ap);
if (ret)
goto out_quota_unlock;
rblocks = RES_DINODE + ind_blocks;
if (gfs2_is_jdata(ip))
rblocks += data_blocks ? data_blocks : 1;
if (ind_blocks || data_blocks) {
rblocks += RES_STATFS + RES_QUOTA;
rblocks += gfs2_rg_blocks(ip, data_blocks + ind_blocks);
}
ret = gfs2_trans_begin(sdp, rblocks, 0);
if (ret)
goto out_trans_fail;
lock_page(page);
ret = -EINVAL;
size = i_size_read(inode);
last_index = (size - 1) >> PAGE_CACHE_SHIFT;
/* Check page index against inode size */
if (size == 0 || (page->index > last_index))
goto out_trans_end;
ret = -EAGAIN;
/* If truncated, we must retry the operation, we may have raced
* with the glock demotion code.
*/
if (!PageUptodate(page) || page->mapping != inode->i_mapping)
goto out_trans_end;
/* Unstuff, if required, and allocate backing blocks for page */
ret = 0;
if (gfs2_is_stuffed(ip))
ret = gfs2_unstuff_dinode(ip, page);
if (ret == 0)
ret = gfs2_allocate_page_backing(page);
out_trans_end:
if (ret)
unlock_page(page);
gfs2_trans_end(sdp);
out_trans_fail:
gfs2_inplace_release(ip);
out_quota_unlock:
gfs2_quota_unlock(ip);
out_unlock:
gfs2_glock_dq(&gh);
out_uninit:
gfs2_holder_uninit(&gh);
if (ret == 0) {
set_page_dirty(page);
wait_for_stable_page(page);
}
out_write_access:
put_write_access(inode);
out:
sb_end_pagefault(inode->i_sb);
return block_page_mkwrite_return(ret);
}
static const struct vm_operations_struct gfs2_vm_ops = {
.fault = filemap_fault,
.page_mkwrite = gfs2_page_mkwrite,
};
/**
* gfs2_mmap -
* @file: The file to map
* @vma: The VMA which described the mapping
*
* There is no need to get a lock here unless we should be updating
* atime. We ignore any locking errors since the only consequence is
* a missed atime update (which will just be deferred until later).
*
* Returns: 0
*/
static int gfs2_mmap(struct file *file, struct vm_area_struct *vma)
{
struct gfs2_inode *ip = GFS2_I(file->f_mapping->host);
if (!(file->f_flags & O_NOATIME) &&
!IS_NOATIME(&ip->i_inode)) {
struct gfs2_holder i_gh;
int error;
error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY,
&i_gh);
if (error)
return error;
/* grab lock to update inode */
gfs2_glock_dq_uninit(&i_gh);
file_accessed(file);
}
vma->vm_ops = &gfs2_vm_ops;
vma->vm_flags |= VM_CAN_NONLINEAR;
return 0;
}
/**
* gfs2_open - open a file
* @inode: the inode to open
* @file: the struct file for this opening
*
* Returns: errno
*/
static int gfs2_open(struct inode *inode, struct file *file)
{
struct gfs2_inode *ip = GFS2_I(inode);
struct gfs2_holder i_gh;
struct gfs2_file *fp;
int error;
fp = kzalloc(sizeof(struct gfs2_file), GFP_KERNEL);
if (!fp)
return -ENOMEM;
mutex_init(&fp->f_fl_mutex);
gfs2_assert_warn(GFS2_SB(inode), !file->private_data);
file->private_data = fp;
if (S_ISREG(ip->i_inode.i_mode)) {
error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY,
&i_gh);
if (error)
goto fail;
if (!(file->f_flags & O_LARGEFILE) &&
i_size_read(inode) > MAX_NON_LFS) {
error = -EOVERFLOW;
goto fail_gunlock;
}
gfs2_glock_dq_uninit(&i_gh);
}
return 0;
fail_gunlock:
gfs2_glock_dq_uninit(&i_gh);
fail:
file->private_data = NULL;
kfree(fp);
return error;
}
/**
* gfs2_release - called to close a struct file
* @inode: the inode the struct file belongs to
* @file: the struct file being closed
*
* Returns: errno
*/
static int gfs2_release(struct inode *inode, struct file *file)
{
struct gfs2_inode *ip = GFS2_I(inode);
kfree(file->private_data);
file->private_data = NULL;
if (!(file->f_mode & FMODE_WRITE))
return 0;
gfs2_rs_delete(ip);
return 0;
}
/**
* gfs2_fsync - sync the dirty data for a file (across the cluster)
* @file: the file that points to the dentry
* @start: the start position in the file to sync
* @end: the end position in the file to sync
* @datasync: set if we can ignore timestamp changes
*
* The VFS will flush data for us. We only need to worry
* about metadata here.
*
* Returns: errno
*/
static int gfs2_fsync(struct file *file, struct dentry *dentry, int datasync)
{
struct inode *inode = dentry->d_inode;
int sync_state = inode->i_state & I_DIRTY;
struct gfs2_inode *ip = GFS2_I(inode);
int ret;
if (!gfs2_is_jdata(ip))
sync_state &= ~I_DIRTY_PAGES;
if (datasync)
sync_state &= ~I_DIRTY_SYNC;
if (sync_state) {
ret = sync_inode_metadata(inode, 1);
if (ret)
return ret;
if (gfs2_is_jdata(ip))
filemap_write_and_wait(inode->i_mapping);
gfs2_ail_flush(ip->i_gl, 1);
}
return 0;
}
/**
* gfs2_file_aio_write - Perform a write to a file
* @iocb: The io context
* @iov: The data to write
* @nr_segs: Number of @iov segments
* @pos: The file position
*
* We have to do a lock/unlock here to refresh the inode size for
* O_APPEND writes, otherwise we can land up writing at the wrong
* offset. There is still a race, but provided the app is using its
* own file locking, this will make O_APPEND work as expected.
*
*/
static ssize_t gfs2_file_aio_write(struct kiocb *iocb, const struct iovec *iov,
unsigned long nr_segs, loff_t pos)
{
struct file *file = iocb->ki_filp;
size_t writesize = iov_length(iov, nr_segs);
struct dentry *dentry = file->f_dentry;
struct gfs2_inode *ip = GFS2_I(dentry->d_inode);
struct gfs2_sbd *sdp;
int ret;
sdp = GFS2_SB(file->f_mapping->host);
ret = gfs2_rs_alloc(ip);
if (ret)
return ret;
gfs2_size_hint(file->f_dentry->d_inode, pos, writesize);
if (file->f_flags & O_APPEND) {
struct gfs2_holder gh;
ret = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, 0, &gh);
if (ret)
return ret;
gfs2_glock_dq_uninit(&gh);
}
return generic_file_aio_write(iocb, iov, nr_segs, pos);
}
static ssize_t gfs2_file_splice_write(struct pipe_inode_info *pipe,
struct file *out, loff_t *ppos,
size_t len, unsigned int flags)
{
int error;
struct inode *inode = out->f_mapping->host;
struct gfs2_inode *ip = GFS2_I(inode);
error = gfs2_rs_alloc(ip);
if (error)
return (ssize_t)error;
gfs2_size_hint(inode, *ppos, len);
return generic_file_splice_write(pipe, out, ppos, len, flags);
}
static int gfs2_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
{
struct page *page = vmf->page;
struct inode *inode = file_inode(vma->vm_file);
struct gfs2_inode *ip = GFS2_I(inode);
struct gfs2_sbd *sdp = GFS2_SB(inode);
struct gfs2_alloc_parms ap = { .aflags = 0, };
unsigned long last_index;
u64 pos = page->index << PAGE_CACHE_SHIFT;
unsigned int data_blocks, ind_blocks, rblocks;
struct gfs2_holder gh;
loff_t size;
int ret;
sb_start_pagefault(inode->i_sb);
/* Update file times before taking page lock */
file_update_time(vma->vm_file);
ret = get_write_access(inode);
if (ret)
goto out;
ret = gfs2_rs_alloc(ip);
if (ret)
goto out_write_access;
gfs2_size_hint(vma->vm_file, pos, PAGE_CACHE_SIZE);
gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh);
ret = gfs2_glock_nq(&gh);
if (ret)
goto out_uninit;
set_bit(GLF_DIRTY, &ip->i_gl->gl_flags);
set_bit(GIF_SW_PAGED, &ip->i_flags);
if (!gfs2_write_alloc_required(ip, pos, PAGE_CACHE_SIZE)) {
lock_page(page);
if (!PageUptodate(page) || page->mapping != inode->i_mapping) {
ret = -EAGAIN;
unlock_page(page);
}
goto out_unlock;
}
ret = gfs2_rindex_update(sdp);
if (ret)
goto out_unlock;
ret = gfs2_quota_lock_check(ip);
if (ret)
goto out_unlock;
gfs2_write_calc_reserv(ip, PAGE_CACHE_SIZE, &data_blocks, &ind_blocks);
ap.target = data_blocks + ind_blocks;
ret = gfs2_inplace_reserve(ip, &ap);
if (ret)
goto out_quota_unlock;
rblocks = RES_DINODE + ind_blocks;
if (gfs2_is_jdata(ip))
rblocks += data_blocks ? data_blocks : 1;
if (ind_blocks || data_blocks) {
rblocks += RES_STATFS + RES_QUOTA;
rblocks += gfs2_rg_blocks(ip, data_blocks + ind_blocks);
}
ret = gfs2_trans_begin(sdp, rblocks, 0);
if (ret)
goto out_trans_fail;
lock_page(page);
ret = -EINVAL;
size = i_size_read(inode);
last_index = (size - 1) >> PAGE_CACHE_SHIFT;
/* Check page index against inode size */
if (size == 0 || (page->index > last_index))
goto out_trans_end;
ret = -EAGAIN;
/* If truncated, we must retry the operation, we may have raced
* with the glock demotion code.
*/
if (!PageUptodate(page) || page->mapping != inode->i_mapping)
goto out_trans_end;
/* Unstuff, if required, and allocate backing blocks for page */
ret = 0;
if (gfs2_is_stuffed(ip))
ret = gfs2_unstuff_dinode(ip, page);
if (ret == 0)
ret = gfs2_allocate_page_backing(page);
out_trans_end:
if (ret)
unlock_page(page);
gfs2_trans_end(sdp);
out_trans_fail:
gfs2_inplace_release(ip);
out_quota_unlock:
gfs2_quota_unlock(ip);
out_unlock:
gfs2_glock_dq(&gh);
out_uninit:
gfs2_holder_uninit(&gh);
if (ret == 0) {
set_page_dirty(page);
wait_for_stable_page(page);
}
out_write_access:
put_write_access(inode);
out:
sb_end_pagefault(inode->i_sb);
return block_page_mkwrite_return(ret);
}
static const struct vm_operations_struct gfs2_vm_ops = {
.fault = filemap_fault,
.map_pages = filemap_map_pages,
.page_mkwrite = gfs2_page_mkwrite,
.remap_pages = generic_file_remap_pages,
};
/**
* gfs2_mmap -
* @file: The file to map
* @vma: The VMA which described the mapping
*
* There is no need to get a lock here unless we should be updating
* atime. We ignore any locking errors since the only consequence is
* a missed atime update (which will just be deferred until later).
*
* Returns: 0
*/
static int gfs2_mmap(struct file *file, struct vm_area_struct *vma)
{
struct gfs2_inode *ip = GFS2_I(file->f_mapping->host);
if (!(file->f_flags & O_NOATIME) &&
!IS_NOATIME(&ip->i_inode)) {
struct gfs2_holder i_gh;
int error;
error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY,
&i_gh);
if (error)
return error;
/* grab lock to update inode */
gfs2_glock_dq_uninit(&i_gh);
file_accessed(file);
}
vma->vm_ops = &gfs2_vm_ops;
return 0;
}
/**
* gfs2_open_common - This is common to open and atomic_open
* @inode: The inode being opened
* @file: The file being opened
*
* This maybe called under a glock or not depending upon how it has
* been called. We must always be called under a glock for regular
* files, however. For other file types, it does not matter whether
* we hold the glock or not.
*
* Returns: Error code or 0 for success
*/
int gfs2_open_common(struct inode *inode, struct file *file)
{
struct gfs2_file *fp;
int ret;
if (S_ISREG(inode->i_mode)) {
ret = generic_file_open(inode, file);
if (ret)
return ret;
}
fp = kzalloc(sizeof(struct gfs2_file), GFP_NOFS);
if (!fp)
return -ENOMEM;
mutex_init(&fp->f_fl_mutex);
gfs2_assert_warn(GFS2_SB(inode), !file->private_data);
file->private_data = fp;
return 0;
}
/**
* gfs2_open - open a file
* @inode: the inode to open
* @file: the struct file for this opening
*
* After atomic_open, this function is only used for opening files
* which are already cached. We must still get the glock for regular
* files to ensure that we have the file size uptodate for the large
* file check which is in the common code. That is only an issue for
* regular files though.
*
* Returns: errno
*/
static int gfs2_open(struct inode *inode, struct file *file)
{
struct gfs2_inode *ip = GFS2_I(inode);
struct gfs2_holder i_gh;
int error;
bool need_unlock = false;
if (S_ISREG(ip->i_inode.i_mode)) {
error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY,
&i_gh);
if (error)
return error;
need_unlock = true;
}
error = gfs2_open_common(inode, file);
if (need_unlock)
gfs2_glock_dq_uninit(&i_gh);
return error;
}
/**
* gfs2_release - called to close a struct file
* @inode: the inode the struct file belongs to
* @file: the struct file being closed
*
* Returns: errno
*/
static int gfs2_release(struct inode *inode, struct file *file)
{
struct gfs2_inode *ip = GFS2_I(inode);
kfree(file->private_data);
file->private_data = NULL;
if (!(file->f_mode & FMODE_WRITE))
return 0;
gfs2_rs_delete(ip, &inode->i_writecount);
return 0;
}
/**
* gfs2_fsync - sync the dirty data for a file (across the cluster)
* @file: the file that points to the dentry
* @start: the start position in the file to sync
* @end: the end position in the file to sync
* @datasync: set if we can ignore timestamp changes
*
* We split the data flushing here so that we don't wait for the data
* until after we've also sent the metadata to disk. Note that for
* data=ordered, we will write & wait for the data at the log flush
* stage anyway, so this is unlikely to make much of a difference
* except in the data=writeback case.
*
* If the fdatawrite fails due to any reason except -EIO, we will
* continue the remainder of the fsync, although we'll still report
* the error at the end. This is to match filemap_write_and_wait_range()
* behaviour.
*
* Returns: errno
*/
static int gfs2_fsync(struct file *file, loff_t start, loff_t end,
int datasync)
{
struct address_space *mapping = file->f_mapping;
struct inode *inode = mapping->host;
int sync_state = inode->i_state & I_DIRTY;
struct gfs2_inode *ip = GFS2_I(inode);
int ret = 0, ret1 = 0;
if (mapping->nrpages) {
ret1 = filemap_fdatawrite_range(mapping, start, end);
if (ret1 == -EIO)
return ret1;
}
if (!gfs2_is_jdata(ip))
sync_state &= ~I_DIRTY_PAGES;
if (datasync)
sync_state &= ~I_DIRTY_SYNC;
if (sync_state) {
ret = sync_inode_metadata(inode, 1);
if (ret)
return ret;
if (gfs2_is_jdata(ip))
filemap_write_and_wait(mapping);
gfs2_ail_flush(ip->i_gl, 1);
}
if (mapping->nrpages)
ret = filemap_fdatawait_range(mapping, start, end);
return ret ? ret : ret1;
}
/**
* gfs2_file_aio_write - Perform a write to a file
* @iocb: The io context
* @iov: The data to write
* @nr_segs: Number of @iov segments
* @pos: The file position
*
* We have to do a lock/unlock here to refresh the inode size for
* O_APPEND writes, otherwise we can land up writing at the wrong
* offset. There is still a race, but provided the app is using its
* own file locking, this will make O_APPEND work as expected.
*
*/
static ssize_t gfs2_file_aio_write(struct kiocb *iocb, const struct iovec *iov,
unsigned long nr_segs, loff_t pos)
{
struct file *file = iocb->ki_filp;
size_t writesize = iov_length(iov, nr_segs);
struct gfs2_inode *ip = GFS2_I(file_inode(file));
int ret;
ret = gfs2_rs_alloc(ip);
if (ret)
return ret;
gfs2_size_hint(file, pos, writesize);
if (file->f_flags & O_APPEND) {
struct gfs2_holder gh;
ret = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, 0, &gh);
if (ret)
return ret;
gfs2_glock_dq_uninit(&gh);
}
return generic_file_aio_write(iocb, iov, nr_segs, pos);
}
static int fallocate_chunk(struct inode *inode, loff_t offset, loff_t len,
int mode)
{
struct gfs2_inode *ip = GFS2_I(inode);
struct buffer_head *dibh;
int error;
loff_t size = len;
unsigned int nr_blks;
sector_t lblock = offset >> inode->i_blkbits;
error = gfs2_meta_inode_buffer(ip, &dibh);
if (unlikely(error))
return error;
gfs2_trans_add_meta(ip->i_gl, dibh);
if (gfs2_is_stuffed(ip)) {
error = gfs2_unstuff_dinode(ip, NULL);
if (unlikely(error))
goto out;
}
while (len) {
struct buffer_head bh_map = { .b_state = 0, .b_blocknr = 0 };
bh_map.b_size = len;
set_buffer_zeronew(&bh_map);
error = gfs2_block_map(inode, lblock, &bh_map, 1);
if (unlikely(error))
goto out;
len -= bh_map.b_size;
nr_blks = bh_map.b_size >> inode->i_blkbits;
lblock += nr_blks;
if (!buffer_new(&bh_map))
continue;
if (unlikely(!buffer_zeronew(&bh_map))) {
error = -EIO;
goto out;
}
}
if (offset + size > inode->i_size && !(mode & FALLOC_FL_KEEP_SIZE))
i_size_write(inode, offset + size);
mark_inode_dirty(inode);
out:
brelse(dibh);
return error;
}
static void calc_max_reserv(struct gfs2_inode *ip, loff_t max, loff_t *len,
unsigned int *data_blocks, unsigned int *ind_blocks)
{
const struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
unsigned int max_blocks = ip->i_rgd->rd_free_clone;
unsigned int tmp, max_data = max_blocks - 3 * (sdp->sd_max_height - 1);
for (tmp = max_data; tmp > sdp->sd_diptrs;) {
tmp = DIV_ROUND_UP(tmp, sdp->sd_inptrs);
max_data -= tmp;
}
/* This calculation isn't the exact reverse of gfs2_write_calc_reserve,
so it might end up with fewer data blocks */
if (max_data <= *data_blocks)
return;
*data_blocks = max_data;
*ind_blocks = max_blocks - max_data;
*len = ((loff_t)max_data - 3) << sdp->sd_sb.sb_bsize_shift;
if (*len > max) {
*len = max;
gfs2_write_calc_reserv(ip, max, data_blocks, ind_blocks);
}
}
static long gfs2_fallocate(struct file *file, int mode, loff_t offset,
loff_t len)
{
struct inode *inode = file_inode(file);
struct gfs2_sbd *sdp = GFS2_SB(inode);
struct gfs2_inode *ip = GFS2_I(inode);
struct gfs2_alloc_parms ap = { .aflags = 0, };
unsigned int data_blocks = 0, ind_blocks = 0, rblocks;
loff_t bytes, max_bytes;
int error;
const loff_t pos = offset;
const loff_t count = len;
loff_t bsize_mask = ~((loff_t)sdp->sd_sb.sb_bsize - 1);
loff_t next = (offset + len - 1) >> sdp->sd_sb.sb_bsize_shift;
loff_t max_chunk_size = UINT_MAX & bsize_mask;
struct gfs2_holder gh;
next = (next + 1) << sdp->sd_sb.sb_bsize_shift;
/* We only support the FALLOC_FL_KEEP_SIZE mode */
if (mode & ~FALLOC_FL_KEEP_SIZE)
return -EOPNOTSUPP;
offset &= bsize_mask;
len = next - offset;
bytes = sdp->sd_max_rg_data * sdp->sd_sb.sb_bsize / 2;
if (!bytes)
bytes = UINT_MAX;
bytes &= bsize_mask;
if (bytes == 0)
bytes = sdp->sd_sb.sb_bsize;
error = gfs2_rs_alloc(ip);
if (error)
return error;
mutex_lock(&inode->i_mutex);
gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh);
error = gfs2_glock_nq(&gh);
if (unlikely(error))
goto out_uninit;
gfs2_size_hint(file, offset, len);
while (len > 0) {
if (len < bytes)
bytes = len;
if (!gfs2_write_alloc_required(ip, offset, bytes)) {
len -= bytes;
offset += bytes;
continue;
}
error = gfs2_quota_lock_check(ip);
if (error)
goto out_unlock;
retry:
gfs2_write_calc_reserv(ip, bytes, &data_blocks, &ind_blocks);
ap.target = data_blocks + ind_blocks;
error = gfs2_inplace_reserve(ip, &ap);
if (error) {
if (error == -ENOSPC && bytes > sdp->sd_sb.sb_bsize) {
bytes >>= 1;
bytes &= bsize_mask;
if (bytes == 0)
bytes = sdp->sd_sb.sb_bsize;
goto retry;
}
goto out_qunlock;
}
max_bytes = bytes;
calc_max_reserv(ip, (len > max_chunk_size)? max_chunk_size: len,
&max_bytes, &data_blocks, &ind_blocks);
rblocks = RES_DINODE + ind_blocks + RES_STATFS + RES_QUOTA +
RES_RG_HDR + gfs2_rg_blocks(ip, data_blocks + ind_blocks);
if (gfs2_is_jdata(ip))
rblocks += data_blocks ? data_blocks : 1;
error = gfs2_trans_begin(sdp, rblocks,
PAGE_CACHE_SIZE/sdp->sd_sb.sb_bsize);
if (error)
goto out_trans_fail;
error = fallocate_chunk(inode, offset, max_bytes, mode);
gfs2_trans_end(sdp);
if (error)
goto out_trans_fail;
len -= max_bytes;
offset += max_bytes;
gfs2_inplace_release(ip);
gfs2_quota_unlock(ip);
}
if (error == 0)
error = generic_write_sync(file, pos, count);
goto out_unlock;
out_trans_fail:
gfs2_inplace_release(ip);
out_qunlock:
gfs2_quota_unlock(ip);
out_unlock:
gfs2_glock_dq(&gh);
out_uninit:
gfs2_holder_uninit(&gh);
mutex_unlock(&inode->i_mutex);
return error;
}
#ifdef CONFIG_GFS2_FS_LOCKING_DLM
/**
* gfs2_setlease - acquire/release a file lease
* @file: the file pointer
* @arg: lease type
* @fl: file lock
*
* We don't currently have a way to enforce a lease across the whole
* cluster; until we do, disable leases (by just returning -EINVAL),
* unless the administrator has requested purely local locking.
*
* Locking: called under i_lock
*
* Returns: errno
*/
static int gfs2_setlease(struct file *file, long arg, struct file_lock **fl)
{
return -EINVAL;
}
/**
* gfs2_lock - acquire/release a posix lock on a file
* @file: the file pointer
* @cmd: either modify or retrieve lock state, possibly wait
* @fl: type and range of lock
*
* Returns: errno
*/
static int gfs2_lock(struct file *file, int cmd, struct file_lock *fl)
{
struct gfs2_inode *ip = GFS2_I(file->f_mapping->host);
struct gfs2_sbd *sdp = GFS2_SB(file->f_mapping->host);
struct lm_lockstruct *ls = &sdp->sd_lockstruct;
if (!(fl->fl_flags & FL_POSIX))
return -ENOLCK;
if (__mandatory_lock(&ip->i_inode) && fl->fl_type != F_UNLCK)
return -ENOLCK;
if (cmd == F_CANCELLK) {
/* Hack: */
cmd = F_SETLK;
fl->fl_type = F_UNLCK;
}
if (unlikely(test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) {
if (fl->fl_type == F_UNLCK)
posix_lock_file_wait(file, fl);
return -EIO;
}
if (IS_GETLK(cmd))
return dlm_posix_get(ls->ls_dlm, ip->i_no_addr, file, fl);
else if (fl->fl_type == F_UNLCK)
return dlm_posix_unlock(ls->ls_dlm, ip->i_no_addr, file, fl);
else
return dlm_posix_lock(ls->ls_dlm, ip->i_no_addr, file, cmd, fl);
}
static int do_flock(struct file *file, int cmd, struct file_lock *fl)
{
struct gfs2_file *fp = file->private_data;
struct gfs2_holder *fl_gh = &fp->f_fl_gh;
struct gfs2_inode *ip = GFS2_I(file_inode(file));
struct gfs2_glock *gl;
unsigned int state;
int flags;
int error = 0;
state = (fl->fl_type == F_WRLCK) ? LM_ST_EXCLUSIVE : LM_ST_SHARED;
flags = (IS_SETLKW(cmd) ? 0 : LM_FLAG_TRY) | GL_EXACT | GL_NOCACHE;
mutex_lock(&fp->f_fl_mutex);
gl = fl_gh->gh_gl;
if (gl) {
if (fl_gh->gh_state == state)
goto out;
flock_lock_file_wait(file,
&(struct file_lock){.fl_type = F_UNLCK});
gfs2_glock_dq_wait(fl_gh);
gfs2_holder_reinit(state, flags, fl_gh);
} else {
error = gfs2_glock_get(GFS2_SB(&ip->i_inode), ip->i_no_addr,
&gfs2_flock_glops, CREATE, &gl);
if (error)
goto out;
gfs2_holder_init(gl, state, flags, fl_gh);
gfs2_glock_put(gl);
}
error = gfs2_glock_nq(fl_gh);
if (error) {
gfs2_holder_uninit(fl_gh);
if (error == GLR_TRYFAILED)
error = -EAGAIN;
} else {
error = flock_lock_file_wait(file, fl);
gfs2_assert_warn(GFS2_SB(&ip->i_inode), !error);
}
out:
mutex_unlock(&fp->f_fl_mutex);
return error;
}