static struct file *__dentry_open(struct dentry *dentry, struct vfsmount *mnt,
struct file *f,
int (*open)(struct inode *, struct file *),
const struct cred *cred)
{
static const struct file_operations empty_fops = {};
struct inode *inode;
int error;
f->f_mode = OPEN_FMODE(f->f_flags) | FMODE_LSEEK |
FMODE_PREAD | FMODE_PWRITE;
if (unlikely(f->f_flags & O_PATH))
f->f_mode = FMODE_PATH;
inode = dentry->d_inode;
if (f->f_mode & FMODE_WRITE) {
error = __get_file_write_access(inode, mnt);
if (error)
goto cleanup_file;
if (!special_file(inode->i_mode))
file_take_write(f);
}
f->f_mapping = inode->i_mapping;
f->f_path.dentry = dentry;
f->f_path.mnt = mnt;
f->f_pos = 0;
file_sb_list_add(f, inode->i_sb);
if (unlikely(f->f_mode & FMODE_PATH)) {
f->f_op = &empty_fops;
return f;
}
f->f_op = fops_get(inode->i_fop);
error = security_dentry_open(f, cred);
if (error)
goto cleanup_all;
error = break_lease(inode, f->f_flags);
if (error)
goto cleanup_all;
if (!open && f->f_op)
open = f->f_op->open;
if (open) {
error = open(inode, f);
if (error)
goto cleanup_all;
}
if ((f->f_mode & (FMODE_READ | FMODE_WRITE)) == FMODE_READ)
i_readcount_inc(inode);
f->f_flags &= ~(O_CREAT | O_EXCL | O_NOCTTY | O_TRUNC);
file_ra_state_init(&f->f_ra, f->f_mapping->host->i_mapping);
/* NB: we're sure to have correct a_ops only after f_op->open */
if (f->f_flags & O_DIRECT) {
if (!f->f_mapping->a_ops ||
((!f->f_mapping->a_ops->direct_IO) &&
(!f->f_mapping->a_ops->get_xip_mem))) {
fput(f);
f = ERR_PTR(-EINVAL);
}
}
return f;
cleanup_all:
fops_put(f->f_op);
if (f->f_mode & FMODE_WRITE) {
put_write_access(inode);
if (!special_file(inode->i_mode)) {
/*
* We don't consider this a real
* mnt_want/drop_write() pair
* because it all happenend right
* here, so just reset the state.
*/
file_reset_write(f);
mnt_drop_write(mnt);
}
}
file_sb_list_del(f);
f->f_path.dentry = NULL;
f->f_path.mnt = NULL;
cleanup_file:
put_filp(f);
dput(dentry);
mntput(mnt);
return ERR_PTR(error);
}
/*
* Set various file attributes.
* N.B. After this call fhp needs an fh_put
*/
__be32
nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp, struct iattr *iap,
int check_guard, time_t guardtime)
{
struct dentry *dentry;
struct inode *inode;
int accmode = MAY_SATTR;
int ftype = 0;
int imode;
__be32 err;
int host_err;
int size_change = 0;
if (iap->ia_valid & (ATTR_ATIME | ATTR_MTIME | ATTR_SIZE))
accmode |= MAY_WRITE|MAY_OWNER_OVERRIDE;
if (iap->ia_valid & ATTR_SIZE)
ftype = S_IFREG;
/* Get inode */
err = fh_verify(rqstp, fhp, ftype, accmode);
if (err)
goto out;
dentry = fhp->fh_dentry;
inode = dentry->d_inode;
/* Ignore any mode updates on symlinks */
if (S_ISLNK(inode->i_mode))
iap->ia_valid &= ~ATTR_MODE;
if (!iap->ia_valid)
goto out;
/* NFSv2 does not differentiate between "set-[ac]time-to-now"
* which only requires access, and "set-[ac]time-to-X" which
* requires ownership.
* So if it looks like it might be "set both to the same time which
* is close to now", and if inode_change_ok fails, then we
* convert to "set to now" instead of "set to explicit time"
*
* We only call inode_change_ok as the last test as technically
* it is not an interface that we should be using. It is only
* valid if the filesystem does not define it's own i_op->setattr.
*/
#define BOTH_TIME_SET (ATTR_ATIME_SET | ATTR_MTIME_SET)
#define MAX_TOUCH_TIME_ERROR (30*60)
if ((iap->ia_valid & BOTH_TIME_SET) == BOTH_TIME_SET
&& iap->ia_mtime.tv_sec == iap->ia_atime.tv_sec
) {
/* Looks probable. Now just make sure time is in the right ballpark.
* Solaris, at least, doesn't seem to care what the time request is.
* We require it be within 30 minutes of now.
*/
time_t delta = iap->ia_atime.tv_sec - get_seconds();
if (delta<0) delta = -delta;
if (delta < MAX_TOUCH_TIME_ERROR &&
inode_change_ok(inode, iap) != 0) {
/* turn off ATTR_[AM]TIME_SET but leave ATTR_[AM]TIME
* this will cause notify_change to set these times to "now"
*/
iap->ia_valid &= ~BOTH_TIME_SET;
}
}
/* The size case is special. It changes the file as well as the attributes. */
if (iap->ia_valid & ATTR_SIZE) {
if (iap->ia_size < inode->i_size) {
err = nfsd_permission(fhp->fh_export, dentry, MAY_TRUNC|MAY_OWNER_OVERRIDE);
if (err)
goto out;
}
/*
* If we are changing the size of the file, then
* we need to break all leases.
*/
host_err = break_lease(inode, FMODE_WRITE | O_NONBLOCK);
if (host_err == -EWOULDBLOCK)
host_err = -ETIMEDOUT;
if (host_err) /* ENOMEM or EWOULDBLOCK */
goto out_nfserr;
host_err = get_write_access(inode);
if (host_err)
goto out_nfserr;
size_change = 1;
host_err = locks_verify_truncate(inode, NULL, iap->ia_size);
if (host_err) {
put_write_access(inode);
goto out_nfserr;
}
DQUOT_INIT(inode);
}
imode = inode->i_mode;
if (iap->ia_valid & ATTR_MODE) {
iap->ia_mode &= S_IALLUGO;
imode = iap->ia_mode |= (imode & ~S_IALLUGO);
}
/* Revoke setuid/setgid bit on chown/chgrp */
if ((iap->ia_valid & ATTR_UID) && iap->ia_uid != inode->i_uid)
iap->ia_valid |= ATTR_KILL_SUID;
if ((iap->ia_valid & ATTR_GID) && iap->ia_gid != inode->i_gid)
iap->ia_valid |= ATTR_KILL_SGID;
/* Change the attributes. */
iap->ia_valid |= ATTR_CTIME;
err = nfserr_notsync;
if (!check_guard || guardtime == inode->i_ctime.tv_sec) {
fh_lock(fhp);
host_err = notify_change(dentry, iap);
err = nfserrno(host_err);
fh_unlock(fhp);
}
if (size_change)
put_write_access(inode);
if (!err)
if (EX_ISSYNC(fhp->fh_export))
write_inode_now(inode, 1);
out:
return err;
out_nfserr:
err = nfserrno(host_err);
goto out;
}
static long do_sys_truncate(const char __user *pathname, loff_t length)
{
struct path path;
struct inode *inode;
int error;
error = -EINVAL;
if (length < 0) /* sorry, but loff_t says... */
goto out;
error = user_path(pathname, &path);
if (error)
goto out;
inode = path.dentry->d_inode;
/* For directories it's -EISDIR, for other non-regulars - -EINVAL */
error = -EISDIR;
if (S_ISDIR(inode->i_mode))
goto dput_and_out;
error = -EINVAL;
if (!S_ISREG(inode->i_mode))
goto dput_and_out;
error = mnt_want_write(path.mnt);
if (error)
goto dput_and_out;
error = inode_permission(inode, MAY_WRITE);
if (error)
goto mnt_drop_write_and_out;
error = -EPERM;
if (IS_APPEND(inode))
goto mnt_drop_write_and_out;
error = get_write_access(inode);
if (error)
goto mnt_drop_write_and_out;
/*
* Make sure that there are no leases. get_write_access() protects
* against the truncate racing with a lease-granting setlease().
*/
error = break_lease(inode, O_WRONLY);
if (error)
goto put_write_and_out;
error = locks_verify_truncate(inode, NULL, length);
if (!error)
error = security_path_truncate(&path);
if (!error)
error = do_truncate(path.dentry, length, 0, NULL);
put_write_and_out:
put_write_access(inode);
mnt_drop_write_and_out:
mnt_drop_write(path.mnt);
dput_and_out:
path_put(&path);
out:
return error;
}
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;
gfs2_write_calc_reserv(ip, PAGE_CACHE_SIZE, &data_blocks, &ind_blocks);
ap.target = data_blocks + ind_blocks;
ret = gfs2_quota_lock_check(ip, &ap);
if (ret)
goto out_unlock;
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,
};
/**
* 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_ALL;
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 | I_DIRTY_TIME);
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_write_iter - 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_write_iter(struct kiocb *iocb, struct iov_iter *from)
{
struct file *file = iocb->ki_filp;
struct gfs2_inode *ip = GFS2_I(file_inode(file));
int ret;
ret = gfs2_rs_alloc(ip);
if (ret)
return ret;
gfs2_size_hint(file, iocb->ki_pos, iov_iter_count(from));
if (iocb->ki_flags & IOCB_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_write_iter(iocb, from);
}
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;
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;
}
}
out:
brelse(dibh);
return error;
}
/**
* calc_max_reserv() - Reverse of write_calc_reserv. Given a number of
* blocks, determine how many bytes can be written.
* @ip: The inode in question.
* @len: Max cap of bytes. What we return in *len must be <= this.
* @data_blocks: Compute and return the number of data blocks needed
* @ind_blocks: Compute and return the number of indirect blocks needed
* @max_blocks: The total blocks available to work with.
*
* Returns: void, but @len, @data_blocks and @ind_blocks are filled in.
*/
static void calc_max_reserv(struct gfs2_inode *ip, loff_t *len,
unsigned int *data_blocks, unsigned int *ind_blocks,
unsigned int max_blocks)
{
loff_t max = *len;
const struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
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;
}
*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, max_blks = UINT_MAX;
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;
next = (next + 1) << sdp->sd_sb.sb_bsize_shift;
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;
gfs2_size_hint(file, offset, len);
gfs2_write_calc_reserv(ip, PAGE_SIZE, &data_blocks, &ind_blocks);
ap.min_target = data_blocks + ind_blocks;
while (len > 0) {
if (len < bytes)
bytes = len;
if (!gfs2_write_alloc_required(ip, offset, bytes)) {
len -= bytes;
offset += bytes;
continue;
}
/* We need to determine how many bytes we can actually
* fallocate without exceeding quota or going over the
* end of the fs. We start off optimistically by assuming
* we can write max_bytes */
max_bytes = (len > max_chunk_size) ? max_chunk_size : len;
/* Since max_bytes is most likely a theoretical max, we
* calculate a more realistic 'bytes' to serve as a good
* starting point for the number of bytes we may be able
* to write */
gfs2_write_calc_reserv(ip, bytes, &data_blocks, &ind_blocks);
ap.target = data_blocks + ind_blocks;
error = gfs2_quota_lock_check(ip, &ap);
if (error)
return error;
/* ap.allowed tells us how many blocks quota will allow
* us to write. Check if this reduces max_blks */
if (ap.allowed && ap.allowed < max_blks)
max_blks = ap.allowed;
error = gfs2_inplace_reserve(ip, &ap);
if (error)
goto out_qunlock;
/* check if the selected rgrp limits our max_blks further */
if (ap.allowed && ap.allowed < max_blks)
max_blks = ap.allowed;
/* Almost done. Calculate bytes that can be written using
* max_blks. We also recompute max_bytes, data_blocks and
* ind_blocks */
calc_max_reserv(ip, &max_bytes, &data_blocks,
&ind_blocks, max_blks);
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 (!(mode & FALLOC_FL_KEEP_SIZE) && (pos + count) > inode->i_size) {
i_size_write(inode, pos + count);
file_update_time(file);
mark_inode_dirty(inode);
}
return generic_write_sync(file, pos, count);
out_trans_fail:
gfs2_inplace_release(ip);
out_qunlock:
gfs2_quota_unlock(ip);
return error;
}
static long gfs2_fallocate(struct file *file, int mode, loff_t offset, loff_t len)
{
struct inode *inode = file_inode(file);
struct gfs2_inode *ip = GFS2_I(inode);
struct gfs2_holder gh;
int ret;
if ((mode & ~FALLOC_FL_KEEP_SIZE) || gfs2_is_jdata(ip))
return -EOPNOTSUPP;
mutex_lock(&inode->i_mutex);
gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh);
ret = gfs2_glock_nq(&gh);
if (ret)
goto out_uninit;
if (!(mode & FALLOC_FL_KEEP_SIZE) &&
(offset + len) > inode->i_size) {
ret = inode_newsize_ok(inode, offset + len);
if (ret)
goto out_unlock;
}
ret = get_write_access(inode);
if (ret)
goto out_unlock;
ret = gfs2_rs_alloc(ip);
if (ret)
goto out_putw;
ret = __gfs2_fallocate(file, mode, offset, len);
if (ret)
gfs2_rs_deltree(ip->i_res);
out_putw:
put_write_access(inode);
out_unlock:
gfs2_glock_dq(&gh);
out_uninit:
gfs2_holder_uninit(&gh);
mutex_unlock(&inode->i_mutex);
return ret;
}
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 gfs2_inode *ip = GFS2_I(out->f_mapping->host);
error = gfs2_rs_alloc(ip);
if (error)
return (ssize_t)error;
gfs2_size_hint(out, *ppos, len);
return iter_file_splice_write(pipe, out, ppos, len, flags);
}
#ifdef CONFIG_GFS2_FS_LOCKING_DLM
/**
* 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)
locks_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;
int sleeptime;
state = (fl->fl_type == F_WRLCK) ? LM_ST_EXCLUSIVE : LM_ST_SHARED;
flags = (IS_SETLKW(cmd) ? 0 : LM_FLAG_TRY_1CB) | GL_EXACT;
mutex_lock(&fp->f_fl_mutex);
gl = fl_gh->gh_gl;
if (gl) {
if (fl_gh->gh_state == state)
goto out;
locks_lock_file_wait(file,
&(struct file_lock){.fl_type = F_UNLCK});
gfs2_glock_dq(fl_gh);
gfs2_holder_reinit(state, flags, fl_gh);
} else {
static int hpfs_unlink(struct inode *dir, struct dentry *dentry)
{
const unsigned char *name = dentry->d_name.name;
unsigned len = dentry->d_name.len;
struct quad_buffer_head qbh;
struct hpfs_dirent *de;
struct inode *inode = dentry->d_inode;
dnode_secno dno;
fnode_secno fno;
int r;
int rep = 0;
int err;
lock_kernel();
hpfs_adjust_length(name, &len);
again:
mutex_lock(&hpfs_i(inode)->i_parent_mutex);
mutex_lock(&hpfs_i(dir)->i_mutex);
err = -ENOENT;
de = map_dirent(dir, hpfs_i(dir)->i_dno, name, len, &dno, &qbh);
if (!de)
goto out;
err = -EPERM;
if (de->first)
goto out1;
err = -EISDIR;
if (de->directory)
goto out1;
fno = de->fnode;
r = hpfs_remove_dirent(dir, dno, de, &qbh, 1);
switch (r) {
case 1:
hpfs_error(dir->i_sb, "there was error when removing dirent");
err = -EFSERROR;
break;
case 2: /* no space for deleting, try to truncate file */
err = -ENOSPC;
if (rep++)
break;
mutex_unlock(&hpfs_i(dir)->i_mutex);
mutex_unlock(&hpfs_i(inode)->i_parent_mutex);
dentry_unhash(dentry);
if (!d_unhashed(dentry)) {
dput(dentry);
unlock_kernel();
return -ENOSPC;
}
if (generic_permission(inode, MAY_WRITE, 0, NULL) ||
!S_ISREG(inode->i_mode) ||
get_write_access(inode)) {
d_rehash(dentry);
dput(dentry);
} else {
struct iattr newattrs;
/*printk("HPFS: truncating file before delete.\n");*/
newattrs.ia_size = 0;
newattrs.ia_valid = ATTR_SIZE | ATTR_CTIME;
err = notify_change(dentry, &newattrs);
put_write_access(inode);
dput(dentry);
if (!err)
goto again;
}
unlock_kernel();
return -ENOSPC;
default:
drop_nlink(inode);
err = 0;
}
goto out;
out1:
hpfs_brelse4(&qbh);
out:
mutex_unlock(&hpfs_i(dir)->i_mutex);
mutex_unlock(&hpfs_i(inode)->i_parent_mutex);
unlock_kernel();
return err;
}
static int do_dentry_open(struct file *f,
struct inode *inode,
int (*open)(struct inode *, struct file *),
const struct cred *cred)
{
static const struct file_operations empty_fops = {};
int error;
f->f_mode = OPEN_FMODE(f->f_flags) | FMODE_LSEEK |
FMODE_PREAD | FMODE_PWRITE;
path_get(&f->f_path);
f->f_inode = inode;
f->f_mapping = inode->i_mapping;
if (unlikely(f->f_flags & O_PATH)) {
f->f_mode = FMODE_PATH;
f->f_op = &empty_fops;
return 0;
}
if (f->f_mode & FMODE_WRITE && !special_file(inode->i_mode)) {
error = get_write_access(inode);
if (unlikely(error))
goto cleanup_file;
error = __mnt_want_write(f->f_path.mnt);
if (unlikely(error)) {
put_write_access(inode);
goto cleanup_file;
}
f->f_mode |= FMODE_WRITER;
}
/* POSIX.1-2008/SUSv4 Section XSI 2.9.7 */
if (S_ISREG(inode->i_mode))
f->f_mode |= FMODE_ATOMIC_POS;
f->f_op = fops_get(inode->i_fop);
if (unlikely(WARN_ON(!f->f_op))) {
error = -ENODEV;
goto cleanup_all;
}
error = security_file_open(f, cred);
if (error)
goto cleanup_all;
error = break_lease(inode, f->f_flags);
if (error)
goto cleanup_all;
if (!open)
open = f->f_op->open;
if (open) {
error = open(inode, f);
if (error)
goto cleanup_all;
}
if ((f->f_mode & (FMODE_READ | FMODE_WRITE)) == FMODE_READ)
i_readcount_inc(inode);
if ((f->f_mode & FMODE_READ) &&
likely(f->f_op->read || f->f_op->read_iter))
f->f_mode |= FMODE_CAN_READ;
if ((f->f_mode & FMODE_WRITE) &&
likely(f->f_op->write || f->f_op->write_iter))
f->f_mode |= FMODE_CAN_WRITE;
f->f_flags &= ~(O_CREAT | O_EXCL | O_NOCTTY | O_TRUNC);
file_ra_state_init(&f->f_ra, f->f_mapping->host->i_mapping);
return 0;
cleanup_all:
fops_put(f->f_op);
if (f->f_mode & FMODE_WRITER) {
put_write_access(inode);
__mnt_drop_write(f->f_path.mnt);
}
cleanup_file:
path_put(&f->f_path);
f->f_path.mnt = NULL;
f->f_path.dentry = NULL;
f->f_inode = NULL;
return error;
}
static inline int
do_aout_core_dump(long signr, struct pt_regs * regs)
{
struct inode * inode = NULL;
struct file file;
unsigned short fs;
int has_dumped = 0;
char corefile[6+sizeof(current->comm)];
unsigned long dump_start, dump_size;
struct user dump;
#ifdef __alpha__
# define START_DATA(u) (u.start_data)
#elif defined(CONFIG_ARM)
# define START_DATA(u) ((u.u_tsize << PAGE_SHIFT) + u.start_code)
#else
# define START_DATA(u) (u.u_tsize << PAGE_SHIFT)
#endif
if (!current->dumpable || current->mm->count != 1)
return 0;
current->dumpable = 0;
/* See if we have enough room to write the upage. */
if (current->rlim[RLIMIT_CORE].rlim_cur < PAGE_SIZE)
return 0;
fs = get_fs();
set_fs(KERNEL_DS);
memcpy(corefile,"core.",5);
#if 0
memcpy(corefile+5,current->comm,sizeof(current->comm));
#else
corefile[4] = '\0';
#endif
if (open_namei(corefile,O_CREAT | 2 | O_TRUNC,0600,&inode,NULL)) {
inode = NULL;
goto end_coredump;
}
if (!S_ISREG(inode->i_mode))
goto end_coredump;
if (!inode->i_op || !inode->i_op->default_file_ops)
goto end_coredump;
if (get_write_access(inode))
goto end_coredump;
file.f_mode = 3;
file.f_flags = 0;
file.f_count = 1;
file.f_inode = inode;
file.f_pos = 0;
file.f_reada = 0;
file.f_op = inode->i_op->default_file_ops;
if (file.f_op->open)
if (file.f_op->open(inode,&file))
goto done_coredump;
if (!file.f_op->write)
goto close_coredump;
has_dumped = 1;
current->flags |= PF_DUMPCORE;
strncpy(dump.u_comm, current->comm, sizeof(current->comm));
dump.u_ar0 = (void *)(((unsigned long)(&dump.regs)) - ((unsigned long)(&dump)));
dump.signal = signr;
dump_thread(regs, &dump);
/* If the size of the dump file exceeds the rlimit, then see what would happen
if we wrote the stack, but not the data area. */
if ((dump.u_dsize+dump.u_ssize+1) * PAGE_SIZE >
current->rlim[RLIMIT_CORE].rlim_cur)
dump.u_dsize = 0;
/* Make sure we have enough room to write the stack and data areas. */
if ((dump.u_ssize+1) * PAGE_SIZE >
current->rlim[RLIMIT_CORE].rlim_cur)
dump.u_ssize = 0;
/* make sure we actually have a data and stack area to dump */
set_fs(USER_DS);
if (verify_area(VERIFY_READ, (void *) START_DATA(dump), dump.u_dsize << PAGE_SHIFT))
dump.u_dsize = 0;
if (verify_area(VERIFY_READ, (void *) dump.start_stack, dump.u_ssize << PAGE_SHIFT))
dump.u_ssize = 0;
set_fs(KERNEL_DS);
/* struct user */
DUMP_WRITE(&dump,sizeof(dump));
/* Now dump all of the user data. Include malloced stuff as well */
DUMP_SEEK(PAGE_SIZE);
/* now we start writing out the user space info */
set_fs(USER_DS);
/* Dump the data area */
if (dump.u_dsize != 0) {
dump_start = START_DATA(dump);
dump_size = dump.u_dsize << PAGE_SHIFT;
DUMP_WRITE(dump_start,dump_size);
}
/* Now prepare to dump the stack area */
if (dump.u_ssize != 0) {
dump_start = dump.start_stack;
dump_size = dump.u_ssize << PAGE_SHIFT;
DUMP_WRITE(dump_start,dump_size);
}
/* Finally dump the task struct. Not be used by gdb, but could be useful */
set_fs(KERNEL_DS);
DUMP_WRITE(current,sizeof(*current));
close_coredump:
if (file.f_op->release)
file.f_op->release(inode,&file);
done_coredump:
put_write_access(inode);
end_coredump:
set_fs(fs);
iput(inode);
return has_dumped;
}
static long do_sys_truncate(const char __user * path, loff_t length)
{
struct nameidata nd;
struct inode * inode;
int error;
error = -EINVAL;
if (length < 0) /* sorry, but loff_t says... */
goto out;
error = user_path_walk(path, &nd);
if (error)
goto out;
inode = nd.path.dentry->d_inode;
/* For directories it's -EISDIR, for other non-regulars - -EINVAL */
error = -EISDIR;
if (S_ISDIR(inode->i_mode))
goto dput_and_out;
error = -EINVAL;
if (!S_ISREG(inode->i_mode))
goto dput_and_out;
error = mnt_want_write(nd.path.mnt);
if (error)
goto dput_and_out;
error = vfs_permission(&nd, MAY_WRITE);
if (error)
goto mnt_drop_write_and_out;
error = -EPERM;
if (IS_IMMUTABLE(inode) || IS_APPEND(inode))
goto mnt_drop_write_and_out;
error = get_write_access(inode);
if (error)
goto mnt_drop_write_and_out;
/*
* Make sure that there are no leases. get_write_access() protects
* against the truncate racing with a lease-granting setlease().
*/
error = break_lease(inode, FMODE_WRITE);
if (error)
goto put_write_and_out;
error = locks_verify_truncate(inode, NULL, length);
if (!error) {
DQUOT_INIT(inode);
error = do_truncate(nd.path.dentry, length, 0, NULL);
}
put_write_and_out:
put_write_access(inode);
mnt_drop_write_and_out:
mnt_drop_write(nd.path.mnt);
dput_and_out:
path_put(&nd.path);
out:
return error;
}
static struct file *__dentry_open(struct dentry *dentry, struct vfsmount *mnt,
int flags, struct file *f,
int (*open)(struct inode *, struct file *))
{
struct inode *inode;
int error;
f->f_flags = flags;
f->f_mode = ((flags+1) & O_ACCMODE) | FMODE_LSEEK |
FMODE_PREAD | FMODE_PWRITE;
inode = dentry->d_inode;
if (f->f_mode & FMODE_WRITE) {
error = __get_file_write_access(inode, mnt);
if (error)
goto cleanup_file;
if (!special_file(inode->i_mode))
file_take_write(f);
}
f->f_mapping = inode->i_mapping;
f->f_path.dentry = dentry;
f->f_path.mnt = mnt;
f->f_pos = 0;
f->f_op = fops_get(inode->i_fop);
file_move(f, &inode->i_sb->s_files);
error = security_dentry_open(f);
if (error)
goto cleanup_all;
if (!open && f->f_op)
open = f->f_op->open;
if (open) {
error = open(inode, f);
if (error)
goto cleanup_all;
}
f->f_flags &= ~(O_CREAT | O_EXCL | O_NOCTTY | O_TRUNC);
file_ra_state_init(&f->f_ra, f->f_mapping->host->i_mapping);
/* NB: we're sure to have correct a_ops only after f_op->open */
if (f->f_flags & O_DIRECT) {
if (!f->f_mapping->a_ops ||
((!f->f_mapping->a_ops->direct_IO) &&
(!f->f_mapping->a_ops->get_xip_mem))) {
fput(f);
f = ERR_PTR(-EINVAL);
}
}
return f;
cleanup_all:
fops_put(f->f_op);
if (f->f_mode & FMODE_WRITE) {
put_write_access(inode);
if (!special_file(inode->i_mode)) {
/*
* We don't consider this a real
* mnt_want/drop_write() pair
* because it all happenend right
* here, so just reset the state.
*/
file_reset_write(f);
mnt_drop_write(mnt);
}
}
file_kill(f);
f->f_path.dentry = NULL;
f->f_path.mnt = NULL;
cleanup_file:
put_filp(f);
dput(dentry);
mntput(mnt);
return ERR_PTR(error);
}
int sys_open(char *filename, int flags, int mode)
{
struct inode *inode;
register struct file *f;
int error, fd, flag;
f = get_empty_filp();
if (!f) {
printk("\nNo filps\n");
return -ENFILE;
}
f->f_flags = (unsigned short int) (flag = flags);
f->f_mode = (mode_t) ((flag + 1) & O_ACCMODE);
if (f->f_mode)
flag++;
if (flag & (O_TRUNC | O_CREAT))
flag |= 2;
error = open_namei(filename, flag, mode, &inode, NULL);
if (!error) {
#ifdef BLOAT_FS
if (f->f_mode & FMODE_WRITE) {
error = get_write_access(inode);
if (error)
goto cleanup_inode;
}
#endif
f->f_inode = inode;
f->f_pos = 0;
#ifdef BLOAT_FS
f->f_reada = 0;
#endif
f->f_op = NULL;
{
register struct inode_operations *iop = inode->i_op;
if (iop)
f->f_op = iop->default_file_ops;
}
{
register struct file_operations *fop = f->f_op;
if (fop && fop->open) {
error = fop->open(inode, f);
if (error) {
goto cleanup_all;
}
}
f->f_flags &= ~(O_CREAT | O_EXCL | O_NOCTTY | O_TRUNC);
/*
* We have to do this last, because we mustn't export
* an incomplete fd to other processes which may share
* the same file table with us.
*/
if ((fd = get_unused_fd()) > -1) {
current->files.fd[fd] = f;
return fd;
}
error = -EMFILE;
if (fop && fop->release)
fop->release(inode, f);
}
cleanup_all:
#ifdef BLOAT_FS
if (f->f_mode & FMODE_WRITE)
put_write_access(inode);
cleanup_inode:
#endif
iput(inode);
}
cleanup_file:
f->f_count--;
return error;
}
/*
* Don't grab the superblock read-lock in unionfs_permission, which prevents
* a deadlock with the branch-management "add branch" code (which grabbed
* the write lock). It is safe to not grab the read lock here, because even
* with branch management taking place, there is no chance that
* unionfs_permission, or anything it calls, will use stale branch
* information.
*/
static int unionfs_permission(struct inode *inode, int mask)
{
struct inode *lower_inode = NULL;
int err = 0;
int bindex, bstart, bend;
const int is_file = !S_ISDIR(inode->i_mode);
const int write_mask = (mask & MAY_WRITE) && !(mask & MAY_READ);
struct inode *inode_grabbed = igrab(inode);
struct dentry *dentry = d_find_alias(inode);
if (dentry)
unionfs_lock_dentry(dentry, UNIONFS_DMUTEX_CHILD);
if (!UNIONFS_I(inode)->lower_inodes) {
if (is_file) /* dirs can be unlinked but chdir'ed to */
err = -ESTALE; /* force revalidate */
goto out;
}
bstart = ibstart(inode);
bend = ibend(inode);
if (unlikely(bstart < 0 || bend < 0)) {
/*
* With branch-management, we can get a stale inode here.
* If so, we return ESTALE back to link_path_walk, which
* would discard the dcache entry and re-lookup the
* dentry+inode. This should be equivalent to issuing
* __unionfs_d_revalidate_chain on nd.dentry here.
*/
if (is_file) /* dirs can be unlinked but chdir'ed to */
err = -ESTALE; /* force revalidate */
goto out;
}
for (bindex = bstart; bindex <= bend; bindex++) {
lower_inode = unionfs_lower_inode_idx(inode, bindex);
if (!lower_inode)
continue;
/*
* check the condition for D-F-D underlying files/directories,
* we don't have to check for files, if we are checking for
* directories.
*/
if (!is_file && !S_ISDIR(lower_inode->i_mode))
continue;
/*
* We check basic permissions, but we ignore any conditions
* such as readonly file systems or branches marked as
* readonly, because those conditions should lead to a
* copyup taking place later on. However, if user never had
* access to the file, then no copyup could ever take place.
*/
err = __inode_permission(lower_inode, mask);
if (err && err != -EACCES && err != EPERM && bindex > 0) {
umode_t mode = lower_inode->i_mode;
if ((is_robranch_super(inode->i_sb, bindex) ||
__is_rdonly(lower_inode)) &&
(S_ISREG(mode) || S_ISDIR(mode) || S_ISLNK(mode)))
err = 0;
if (IS_COPYUP_ERR(err))
err = 0;
}
/*
* NFS HACK: NFSv2/3 return EACCES on readonly-exported,
* locally readonly-mounted file systems, instead of EROFS
* like other file systems do. So we have no choice here
* but to intercept this and ignore it for NFS branches
* marked readonly. Specifically, we avoid using NFS's own
* "broken" ->permission method, and rely on
* generic_permission() to do basic checking for us.
*/
if (err && err == -EACCES &&
is_robranch_super(inode->i_sb, bindex) &&
lower_inode->i_sb->s_magic == NFS_SUPER_MAGIC)
err = generic_permission(lower_inode, mask, NULL);
/*
* The permissions are an intersection of the overall directory
* permissions, so we fail if one fails.
*/
if (err)
goto out;
/* only the leftmost file matters. */
if (is_file || write_mask) {
if (is_file && write_mask) {
err = get_write_access(lower_inode);
if (!err)
put_write_access(lower_inode);
}
break;
}
}
/* sync times which may have changed (asynchronously) below */
unionfs_copy_attr_times(inode);
out:
unionfs_check_inode(inode);
if (dentry) {
unionfs_unlock_dentry(dentry);
dput(dentry);
}
iput(inode_grabbed);
return err;
}
static struct file *__dentry_open(struct dentry *dentry, struct vfsmount *mnt,
int flags, struct file *f,
int (*open)(struct inode *, struct file *))
{
struct inode *inode;
int error;
f->f_flags = flags;
f->f_mode = ((flags+1) & O_ACCMODE) | FMODE_LSEEK |
FMODE_PREAD | FMODE_PWRITE;
inode = dentry->d_inode;
if (f->f_mode & FMODE_WRITE) {
error = get_write_access(inode);
if (error)
goto cleanup_file;
}
f->f_mapping = inode->i_mapping;
f->f_dentry = dentry;
f->f_vfsmnt = mnt;
f->f_pos = 0;
f->f_op = fops_get(inode->i_fop);
file_move(f, &inode->i_sb->s_files);
if (!open && f->f_op)
open = f->f_op->open;
if (open) {
error = open(inode, f);
if (error)
goto cleanup_all;
}
f->f_flags &= ~(O_CREAT | O_EXCL | O_NOCTTY | O_TRUNC);
file_ra_state_init(&f->f_ra, f->f_mapping->host->i_mapping);
/* NB: we're sure to have correct a_ops only after f_op->open */
if (f->f_flags & O_DIRECT) {
if (!f->f_mapping->a_ops ||
((!f->f_mapping->a_ops->direct_IO) &&
(!f->f_mapping->a_ops->get_xip_page))) {
fput(f);
f = ERR_PTR(-EINVAL);
}
}
return f;
cleanup_all:
fops_put(f->f_op);
if (f->f_mode & FMODE_WRITE)
put_write_access(inode);
file_kill(f);
f->f_dentry = NULL;
f->f_vfsmnt = NULL;
cleanup_file:
put_filp(f);
dput(dentry);
mntput(mnt);
return ERR_PTR(error);
}
static int do_dentry_open(struct file *f,
int (*open)(struct inode *, struct file *),
const struct cred *cred)
{
static const struct file_operations empty_fops = {};
struct inode *inode;
int error;
f->f_mode = OPEN_FMODE(f->f_flags) | FMODE_LSEEK |
FMODE_PREAD | FMODE_PWRITE;
if (unlikely(f->f_flags & O_PATH))
f->f_mode = FMODE_PATH;
path_get(&f->f_path);
inode = f->f_inode = f->f_path.dentry->d_inode;
if (f->f_mode & FMODE_WRITE && !special_file(inode->i_mode)) {
error = __get_file_write_access(inode, f->f_path.mnt);
if (error)
goto cleanup_file;
file_take_write(f);
}
f->f_mapping = inode->i_mapping;
file_sb_list_add(f, inode->i_sb);
if (unlikely(f->f_mode & FMODE_PATH)) {
f->f_op = &empty_fops;
return 0;
}
f->f_op = fops_get(inode->i_fop);
error = security_file_open(f, cred);
if (error)
goto cleanup_all;
error = break_lease(inode, f->f_flags);
if (error)
goto cleanup_all;
if (!open && f->f_op)
open = f->f_op->open;
if (open) {
error = open(inode, f);
if (error)
goto cleanup_all;
}
if ((f->f_mode & (FMODE_READ | FMODE_WRITE)) == FMODE_READ)
i_readcount_inc(inode);
f->f_flags &= ~(O_CREAT | O_EXCL | O_NOCTTY | O_TRUNC);
file_ra_state_init(&f->f_ra, f->f_mapping->host->i_mapping);
return 0;
cleanup_all:
fops_put(f->f_op);
file_sb_list_del(f);
if (f->f_mode & FMODE_WRITE) {
if (!special_file(inode->i_mode)) {
/*
* We don't consider this a real
* mnt_want/drop_write() pair
* because it all happenend right
* here, so just reset the state.
*/
put_write_access(inode);
file_reset_write(f);
__mnt_drop_write(f->f_path.mnt);
}
}
cleanup_file:
path_put(&f->f_path);
f->f_path.mnt = NULL;
f->f_path.dentry = NULL;
f->f_inode = NULL;
return error;
}
extern int
vnode_shadow_fop_open(
INODE_T *inode,
FILE_T *file
)
{
int err = 0;
INODE_T *real_inode;
DENT_T *rdentry = NULL;
DENT_T *oldent;
struct file_operations *oldfops;
struct vfsmount *oldmnt, *newmnt;
VNODE_T *cvp;
oldmnt = file->f_vfsmnt;
oldent = file->f_dentry;
ASSERT(D_COUNT(oldent));
/* The Linux kernel has stopped ignoring the O_DIRECT flag.
* The problem is that they wait until after they call the fop open
* function to check the inode to see if it will support direct I/O.
* But they get the inode pointer before they call us and check the
* inode after we return so they never check the actual inode we open
* but only the shadow one. Their error handling never comes back to
* us and they release their old pointers and not our new ones. The
* only choice we have is to not allow O_DIRECT on shadow files.
*/
if (file->f_flags & O_DIRECT) {
err = -EINVAL;
goto out_nolock;
}
/* Get the real dentry */
rdentry = REALDENTRY_LOCKED(oldent, &cvp);
if (rdentry == NULL) {
err = -ENOENT;
goto out_nolock;
}
VNODE_DGET(rdentry); /* protect rdentry->d_inode */
if (rdentry->d_inode == NULL) {
/* delete race */
err = -ENOENT;
goto out;
}
newmnt = MDKI_MNTGET(REALVFSMNT(oldent));
if (newmnt == NULL) {
err = -EOPNOTSUPP; /* XXX */
goto out;
}
/* Check that we can write to this file. Clean up the count on the
* shadow inode.
*/
if (file->f_mode & FMODE_WRITE) {
err = get_write_access(rdentry->d_inode);
if (err) {
MDKI_MNTPUT(newmnt);
goto out;
}
}
real_inode = rdentry->d_inode;
/*
* Swap the file structure contents to point at the underlying object.
*/
/* In Linux 2.6 they added the mapping stuff to the file so we have to set
** that up here, too.
*/
file->f_mapping = real_inode->i_mapping;
VNLAYER_RA_STATE_INIT(&(file->f_ra), file->f_mapping);
file->f_dentry = VNODE_DGET(rdentry);
oldfops = file->f_op;
file->f_vfsmnt = newmnt;
file->f_op = fops_get(real_inode->i_fop);
if (real_inode->i_fop && !file->f_op)
/* If we failed to get the reference to a non-NULL op, bail out */
err = -EIO; /* XXX? better error code */
if (!err) {
/* Move the file to the file list for the real superblock
* and remove it from the shadow list
*/
/* It would be better to use file_move() but it's not exported */
file_list_lock();
list_del(&file->f_list);
list_add(&file->f_list, &real_inode->i_sb->s_files);
file_list_unlock();
if (file->f_op && file->f_op->open) {
err = (*file->f_op->open)(real_inode, file);
if (err) {
/* restore our file to the list on our super block */
file_list_lock();
list_del(&file->f_list);
list_add(&file->f_list, &oldent->d_inode->i_sb->s_files);
file_list_unlock();
}
}
}
if (err) {
/* MUST put back old dentry/fops to get accounting right in upper
* layer. */
put_write_access(rdentry->d_inode);
if (file->f_dentry)
VNODE_DPUT(file->f_dentry);
if (file->f_op)
fops_put(file->f_op);
MDKI_MNTPUT(file->f_vfsmnt);
file->f_vfsmnt = oldmnt;
file->f_dentry = oldent;
file->f_op = oldfops;
} else {
put_write_access(oldent->d_inode);
VNODE_DPUT(oldent);
/* Drop reference now that we've dropped our use of the file ops */
fops_put(oldfops);
MDKI_MNTPUT(oldmnt);
}
out:
VNODE_DPUT(rdentry);
REALDENTRY_UNLOCK(oldent, cvp);
out_nolock:
MDKI_TRACE(TRACE_OPEN, "%s: opened vp=%p fp=%p rdent=%p rdcnt=%d fcnt=%d"
", err %d\n", __func__,
inode, file, rdentry, rdentry ? D_COUNT(rdentry) : 0, F_COUNT(file), -err);
return(err);
}
