static int v9fs_file_flock_dotl(struct file *filp, int cmd,
struct file_lock *fl)
{
struct inode *inode = file_inode(filp);
int ret = -ENOLCK;
p9_debug(P9_DEBUG_VFS, "filp: %p cmd:%d lock: %p name: %pD\n",
filp, cmd, fl, filp);
/* No mandatory locks */
if (__mandatory_lock(inode) && fl->fl_type != F_UNLCK)
goto out_err;
if (!(fl->fl_flags & FL_FLOCK))
goto out_err;
if ((IS_SETLK(cmd) || IS_SETLKW(cmd)) && fl->fl_type != F_UNLCK) {
filemap_write_and_wait(inode->i_mapping);
invalidate_mapping_pages(&inode->i_data, 0, -1);
}
/* Convert flock to posix lock */
fl->fl_flags |= FL_POSIX;
fl->fl_flags ^= FL_FLOCK;
if (IS_SETLK(cmd) | IS_SETLKW(cmd))
ret = v9fs_file_do_lock(filp, cmd, fl);
else
ret = -EINVAL;
out_err:
return ret;
}
static int v9fs_file_lock_dotl(struct file *filp, int cmd, struct file_lock *fl)
{
struct inode *inode = filp->f_path.dentry->d_inode;
int ret = -ENOLCK;
P9_DPRINTK(P9_DEBUG_VFS, "filp: %p cmd:%d lock: %p name: %s\n", filp,
cmd, fl, filp->f_path.dentry->d_name.name);
/* No mandatory locks */
if (__mandatory_lock(inode) && fl->fl_type != F_UNLCK)
goto out_err;
if ((IS_SETLK(cmd) || IS_SETLKW(cmd)) && fl->fl_type != F_UNLCK) {
filemap_write_and_wait(inode->i_mapping);
invalidate_mapping_pages(&inode->i_data, 0, -1);
}
if (IS_SETLK(cmd) || IS_SETLKW(cmd))
ret = v9fs_file_do_lock(filp, cmd, fl);
else if (IS_GETLK(cmd))
ret = v9fs_file_getlock(filp, fl);
else
ret = -EINVAL;
out_err:
return ret;
}
static int v9fs_file_flock_dotl(struct file *filp, int cmd,
struct file_lock *fl)
{
struct inode *inode = filp->f_path.dentry->d_inode;
int ret = -ENOLCK;
P9_DPRINTK(P9_DEBUG_VFS, "filp: %p cmd:%d lock: %p name: %s\n", filp,
cmd, fl, filp->f_path.dentry->d_name.name);
/* No mandatory locks */
if (__mandatory_lock(inode) && fl->fl_type != F_UNLCK)
goto out_err;
if (!(fl->fl_flags & FL_FLOCK))
goto out_err;
if ((IS_SETLK(cmd) || IS_SETLKW(cmd)) && fl->fl_type != F_UNLCK) {
filemap_write_and_wait(inode->i_mapping);
invalidate_mapping_pages(&inode->i_data, 0, -1);
}
/* Convert flock to posix lock */
fl->fl_owner = (fl_owner_t)filp;
fl->fl_start = 0;
fl->fl_end = OFFSET_MAX;
fl->fl_flags |= FL_POSIX;
fl->fl_flags ^= FL_FLOCK;
if (IS_SETLK(cmd) | IS_SETLKW(cmd))
ret = v9fs_file_do_lock(filp, cmd, fl);
else
ret = -EINVAL;
out_err:
return ret;
}
static int ocfs2_do_flock(struct file *file, struct inode *inode,
int cmd, struct file_lock *fl)
{
int ret = 0, level = 0, trylock = 0;
struct ocfs2_file_private *fp = file->private_data;
struct ocfs2_lock_res *lockres = &fp->fp_flock;
if (fl->fl_type == F_WRLCK)
level = 1;
if (!IS_SETLKW(cmd))
trylock = 1;
mutex_lock(&fp->fp_mutex);
if (lockres->l_flags & OCFS2_LOCK_ATTACHED &&
lockres->l_level > LKM_NLMODE) {
int old_level = 0;
if (lockres->l_level == LKM_EXMODE)
old_level = 1;
if (level == old_level)
goto out;
/*
* Converting an existing lock is not guaranteed to be
* atomic, so we can get away with simply unlocking
* here and allowing the lock code to try at the new
* level.
*/
flock_lock_file_wait(file,
&(struct file_lock){.fl_type = F_UNLCK});
static int ocfs2_do_flock(struct file *file, struct inode *inode,
int cmd, struct file_lock *fl)
{
int ret = 0, level = 0, trylock = 0;
struct ocfs2_file_private *fp = file->private_data;
struct ocfs2_lock_res *lockres = &fp->fp_flock;
if (fl->fl_type == F_WRLCK)
level = 1;
if (!IS_SETLKW(cmd))
trylock = 1;
mutex_lock(&fp->fp_mutex);
if (lockres->l_flags & OCFS2_LOCK_ATTACHED &&
lockres->l_level > LKM_NLMODE) {
int old_level = 0;
if (lockres->l_level == LKM_EXMODE)
old_level = 1;
if (level == old_level)
goto out;
flock_lock_file_wait(file,
&(struct file_lock){.fl_type = F_UNLCK});
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->f_path.dentry->d_inode);
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;
flock_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 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_lockname name =
{ .ln_number = ip->i_num.no_addr,
.ln_type = LM_TYPE_PLOCK };
if (!(fl->fl_flags & FL_POSIX))
return -ENOLCK;
if ((ip->i_inode.i_mode & (S_ISGID | S_IXGRP)) == S_ISGID)
return -ENOLCK;
if (sdp->sd_args.ar_localflocks) {
if (IS_GETLK(cmd)) {
struct file_lock tmp;
int ret;
ret = posix_test_lock(file, fl, &tmp);
fl->fl_type = F_UNLCK;
if (ret)
memcpy(fl, &tmp, sizeof(struct file_lock));
return 0;
} else {
return posix_lock_file_wait(file, fl);
}
}
if (IS_GETLK(cmd))
return gfs2_lm_plock_get(sdp, &name, file, fl);
else if (fl->fl_type == F_UNLCK)
return gfs2_lm_punlock(sdp, &name, file, fl);
else
return gfs2_lm_plock(sdp, &name, 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->f_path.dentry->d_inode);
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;
gfs2_glock_hold(gl);
flock_lock_file_wait(file,
&(struct file_lock){.fl_type = F_UNLCK});
gfs2_glock_dq_uninit(fl_gh);
} else {
/**
* Attempt to set an fcntl lock.
* For now, this just goes away to the server. Later it may be more awesome.
*/
int ceph_lock(struct file *file, int cmd, struct file_lock *fl)
{
u8 lock_cmd;
int err;
u8 wait = 0;
u16 op = CEPH_MDS_OP_SETFILELOCK;
if (!(fl->fl_flags & FL_POSIX))
return -ENOLCK;
/* No mandatory locks */
if (__mandatory_lock(file->f_mapping->host) && fl->fl_type != F_UNLCK)
return -ENOLCK;
dout("ceph_lock, fl_owner: %p", fl->fl_owner);
/* set wait bit as appropriate, then make command as Ceph expects it*/
if (IS_GETLK(cmd))
op = CEPH_MDS_OP_GETFILELOCK;
else if (IS_SETLKW(cmd))
wait = 1;
if (F_RDLCK == fl->fl_type)
lock_cmd = CEPH_LOCK_SHARED;
else if (F_WRLCK == fl->fl_type)
lock_cmd = CEPH_LOCK_EXCL;
else
lock_cmd = CEPH_LOCK_UNLOCK;
err = ceph_lock_message(CEPH_LOCK_FCNTL, op, file, lock_cmd, wait, fl);
if (!err) {
if (op != CEPH_MDS_OP_GETFILELOCK) {
dout("mds locked, locking locally");
err = posix_lock_file(file, fl, NULL);
if (err && (CEPH_MDS_OP_SETFILELOCK == op)) {
/* undo! This should only happen if
* the kernel detects local
* deadlock. */
ceph_lock_message(CEPH_LOCK_FCNTL, op, file,
CEPH_LOCK_UNLOCK, 0, fl);
dout("got %d on posix_lock_file, undid lock",
err);
}
}
} else if (err == -ERESTARTSYS) {
dout("undoing lock\n");
ceph_lock_message(CEPH_LOCK_FCNTL, op, file,
CEPH_LOCK_UNLOCK, 0, fl);
}
return err;
}
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_lockname name =
{ .ln_number = ip->i_no_addr,
.ln_type = LM_TYPE_PLOCK };
if (!(fl->fl_flags & FL_POSIX))
return -ENOLCK;
if (__mandatory_lock(&ip->i_inode))
return -ENOLCK;
if (cmd == F_CANCELLK) {
/* Hack: */
cmd = F_SETLK;
fl->fl_type = F_UNLCK;
}
if (IS_GETLK(cmd))
return gfs2_lm_plock_get(sdp, &name, file, fl);
else if (fl->fl_type == F_UNLCK)
return gfs2_lm_punlock(sdp, &name, file, fl);
else
return gfs2_lm_plock(sdp, &name, 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->f_path.dentry->d_inode);
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 {
static int v9fs_file_lock(struct file *filp, int cmd, struct file_lock *fl)
{
int res = 0;
struct inode *inode = filp->f_path.dentry->d_inode;
P9_DPRINTK(P9_DEBUG_VFS, "filp: %p lock: %p\n", filp, fl);
/* No mandatory locks */
if (__mandatory_lock(inode) && fl->fl_type != F_UNLCK)
return -ENOLCK;
if ((IS_SETLK(cmd) || IS_SETLKW(cmd)) && fl->fl_type != F_UNLCK) {
filemap_write_and_wait(inode->i_mapping);
invalidate_mapping_pages(&inode->i_data, 0, -1);
}
return res;
}
int gdlm_plock(void *lockspace, struct lm_lockname *name,
struct file *file, int cmd, struct file_lock *fl)
{
struct gdlm_ls *ls = lockspace;
struct plock_op *op;
int rv;
op = kzalloc(sizeof(*op), GFP_KERNEL);
if (!op)
return -ENOMEM;
op->info.optype = GDLM_PLOCK_OP_LOCK;
op->info.pid = fl->fl_pid;
op->info.ex = (fl->fl_type == F_WRLCK);
op->info.wait = IS_SETLKW(cmd);
op->info.fsid = ls->id;
op->info.number = name->ln_number;
op->info.start = fl->fl_start;
op->info.end = fl->fl_end;
op->info.owner = (__u64)(long) fl->fl_owner;
send_op(op);
wait_event(recv_wq, (op->done != 0));
spin_lock(&ops_lock);
if (!list_empty(&op->list)) {
printk(KERN_INFO "plock op on list\n");
list_del(&op->list);
}
spin_unlock(&ops_lock);
rv = op->info.rv;
if (!rv) {
if (posix_lock_file_wait(file, fl) < 0)
log_error("gdlm_plock: vfs lock error %x,%llx",
name->ln_type,
(unsigned long long)name->ln_number);
}
kfree(op);
return rv;
}
int ceph_flock(struct file *file, int cmd, struct file_lock *fl)
{
u8 lock_cmd;
int err;
u8 wait = 0;
if (!(fl->fl_flags & FL_FLOCK))
return -ENOLCK;
/* No mandatory locks */
if (__mandatory_lock(file->f_mapping->host) && fl->fl_type != F_UNLCK)
return -ENOLCK;
dout("ceph_flock, fl_file: %p", fl->fl_file);
if (IS_SETLKW(cmd))
wait = 1;
if (F_RDLCK == fl->fl_type)
lock_cmd = CEPH_LOCK_SHARED;
else if (F_WRLCK == fl->fl_type)
lock_cmd = CEPH_LOCK_EXCL;
else
lock_cmd = CEPH_LOCK_UNLOCK;
err = ceph_lock_message(CEPH_LOCK_FLOCK, CEPH_MDS_OP_SETFILELOCK,
file, lock_cmd, wait, fl);
if (!err) {
err = flock_lock_file_wait(file, fl);
if (err) {
ceph_lock_message(CEPH_LOCK_FLOCK,
CEPH_MDS_OP_SETFILELOCK,
file, CEPH_LOCK_UNLOCK, 0, fl);
dout("got %d on flock_lock_file_wait, undid lock", err);
}
} else if (err == -ERESTARTSYS) {
dout("undoing lock\n");
ceph_lock_message(CEPH_LOCK_FLOCK,
CEPH_MDS_OP_SETFILELOCK,
file, CEPH_LOCK_UNLOCK, 0, fl);
}
return err;
}
/*
* Lock a (portion of) a file
*/
int
nfs_lock(struct file *filp, int cmd, struct file_lock *fl)
{
struct inode * inode = filp->f_dentry->d_inode;
int status = 0;
int status2;
dprintk("NFS: nfs_lock(f=%4x/%ld, t=%x, fl=%x, r=%Ld:%Ld)\n",
inode->i_dev, inode->i_ino,
fl->fl_type, fl->fl_flags,
(long long)fl->fl_start, (long long)fl->fl_end);
if (!inode)
return -EINVAL;
/* No mandatory locks over NFS */
if ((inode->i_mode & (S_ISGID | S_IXGRP)) == S_ISGID)
return -ENOLCK;
/* Fake OK code if mounted without NLM support */
if (NFS_SERVER(inode)->flags & NFS_MOUNT_NONLM) {
if (IS_GETLK(cmd))
status = LOCK_USE_CLNT;
goto out_ok;
}
/*
* No BSD flocks over NFS allowed.
* Note: we could try to fake a POSIX lock request here by
* using ((u32) filp | 0x80000000) or some such as the pid.
* Not sure whether that would be unique, though, or whether
* that would break in other places.
*/
if (!fl->fl_owner || (fl->fl_flags & (FL_POSIX|FL_BROKEN)) != FL_POSIX)
return -ENOLCK;
/*
* Flush all pending writes before doing anything
* with locks..
*/
status = filemap_fdatasync(inode->i_mapping);
down(&inode->i_sem);
status2 = nfs_wb_all(inode);
if (status2 && !status)
status = status2;
up(&inode->i_sem);
status2 = filemap_fdatawait(inode->i_mapping);
if (status2 && !status)
status = status2;
if (status < 0)
return status;
lock_kernel();
status = nlmclnt_proc(inode, cmd, fl);
unlock_kernel();
if (status < 0)
return status;
status = 0;
/*
* Make sure we clear the cache whenever we try to get the lock.
* This makes locking act as a cache coherency point.
*/
out_ok:
if ((IS_SETLK(cmd) || IS_SETLKW(cmd)) && fl->fl_type != F_UNLCK) {
filemap_fdatasync(inode->i_mapping);
down(&inode->i_sem);
nfs_wb_all(inode); /* we may have slept */
up(&inode->i_sem);
filemap_fdatawait(inode->i_mapping);
nfs_zap_caches(inode);
}
return status;
}
static int v9fs_file_do_lock(struct file *filp, int cmd, struct file_lock *fl)
{
struct p9_flock flock;
struct p9_fid *fid;
uint8_t status;
int res = 0;
unsigned char fl_type;
fid = filp->private_data;
BUG_ON(fid == NULL);
if ((fl->fl_flags & FL_POSIX) != FL_POSIX)
BUG();
res = posix_lock_file_wait(filp, fl);
if (res < 0)
goto out;
/* convert posix lock to p9 tlock args */
memset(&flock, 0, sizeof(flock));
/* map the lock type */
switch (fl->fl_type) {
case F_RDLCK:
flock.type = P9_LOCK_TYPE_RDLCK;
break;
case F_WRLCK:
flock.type = P9_LOCK_TYPE_WRLCK;
break;
case F_UNLCK:
flock.type = P9_LOCK_TYPE_UNLCK;
break;
}
flock.start = fl->fl_start;
if (fl->fl_end == OFFSET_MAX)
flock.length = 0;
else
flock.length = fl->fl_end - fl->fl_start + 1;
flock.proc_id = fl->fl_pid;
flock.client_id = utsname()->nodename;
if (IS_SETLKW(cmd))
flock.flags = P9_LOCK_FLAGS_BLOCK;
/*
* if its a blocked request and we get P9_LOCK_BLOCKED as the status
* for lock request, keep on trying
*/
for (;;) {
res = p9_client_lock_dotl(fid, &flock, &status);
if (res < 0)
break;
if (status != P9_LOCK_BLOCKED)
break;
if (status == P9_LOCK_BLOCKED && !IS_SETLKW(cmd))
break;
schedule_timeout_interruptible(P9_LOCK_TIMEOUT);
}
/* map 9p status to VFS status */
switch (status) {
case P9_LOCK_SUCCESS:
res = 0;
break;
case P9_LOCK_BLOCKED:
res = -EAGAIN;
break;
case P9_LOCK_ERROR:
case P9_LOCK_GRACE:
res = -ENOLCK;
break;
default:
BUG();
}
/*
* incase server returned error for lock request, revert
* it locally
*/
if (res < 0 && fl->fl_type != F_UNLCK) {
fl_type = fl->fl_type;
fl->fl_type = F_UNLCK;
res = posix_lock_file_wait(filp, fl);
fl->fl_type = fl_type;
}
out:
return res;
}
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;
}
int dlm_posix_lock(dlm_lockspace_t *lockspace, u64 number, struct file *file,
int cmd, struct file_lock *fl)
{
struct dlm_ls *ls;
struct plock_op *op;
struct plock_xop *xop;
int rv;
ls = dlm_find_lockspace_local(lockspace);
if (!ls)
return -EINVAL;
xop = kzalloc(sizeof(*xop), GFP_KERNEL);
if (!xop) {
rv = -ENOMEM;
goto out;
}
op = &xop->xop;
op->info.optype = DLM_PLOCK_OP_LOCK;
op->info.pid = fl->fl_pid;
op->info.ex = (fl->fl_type == F_WRLCK);
op->info.wait = IS_SETLKW(cmd);
op->info.fsid = ls->ls_global_id;
op->info.number = number;
op->info.start = fl->fl_start;
op->info.end = fl->fl_end;
if (fl->fl_lmops && fl->fl_lmops->fl_grant) {
/* fl_owner is lockd which doesn't distinguish
processes on the nfs client */
op->info.owner = (__u64) fl->fl_pid;
xop->callback = fl->fl_lmops->fl_grant;
locks_init_lock(&xop->flc);
locks_copy_lock(&xop->flc, fl);
xop->fl = fl;
xop->file = file;
} else {
op->info.owner = (__u64)(long) fl->fl_owner;
xop->callback = NULL;
}
send_op(op);
if (xop->callback == NULL)
wait_event(recv_wq, (op->done != 0));
else {
rv = FILE_LOCK_DEFERRED;
goto out;
}
spin_lock(&ops_lock);
if (!list_empty(&op->list)) {
log_error(ls, "dlm_posix_lock: op on list %llx",
(unsigned long long)number);
list_del(&op->list);
}
spin_unlock(&ops_lock);
rv = op->info.rv;
if (!rv) {
if (posix_lock_file_wait(file, fl) < 0)
log_error(ls, "dlm_posix_lock: vfs lock error %llx",
(unsigned long long)number);
}
kfree(xop);
out:
dlm_put_lockspace(ls);
return rv;
}
/*
* This is the main entry point for the NLM client.
*/
int
nlmclnt_proc(struct inode *inode, int cmd, struct file_lock *fl)
{
struct nfs_server *nfssrv = NFS_SERVER(inode);
struct nlm_host *host;
struct nlm_rqst reqst, *call = &reqst;
sigset_t oldset;
unsigned long flags;
int status, proto, vers;
vers = (NFS_PROTO(inode)->version == 3) ? 4 : 1;
if (NFS_PROTO(inode)->version > 3) {
printk(KERN_NOTICE "NFSv4 file locking not implemented!\n");
return -ENOLCK;
}
/* Retrieve transport protocol from NFS client */
proto = NFS_CLIENT(inode)->cl_xprt->prot;
if (!(host = nlmclnt_lookup_host(NFS_ADDR(inode), proto, vers)))
return -ENOLCK;
/* Create RPC client handle if not there, and copy soft
* and intr flags from NFS client. */
if (host->h_rpcclnt == NULL) {
struct rpc_clnt *clnt;
/* Bind an rpc client to this host handle (does not
* perform a portmapper lookup) */
if (!(clnt = nlm_bind_host(host))) {
status = -ENOLCK;
goto done;
}
clnt->cl_softrtry = nfssrv->client->cl_softrtry;
clnt->cl_intr = nfssrv->client->cl_intr;
clnt->cl_chatty = nfssrv->client->cl_chatty;
}
/* Keep the old signal mask */
spin_lock_irqsave(¤t->sighand->siglock, flags);
oldset = current->blocked;
/* If we're cleaning up locks because the process is exiting,
* perform the RPC call asynchronously. */
if ((IS_SETLK(cmd) || IS_SETLKW(cmd))
&& fl->fl_type == F_UNLCK
&& (current->flags & PF_EXITING)) {
sigfillset(¤t->blocked); /* Mask all signals */
recalc_sigpending();
spin_unlock_irqrestore(¤t->sighand->siglock, flags);
call = nlmclnt_alloc_call();
if (!call) {
status = -ENOMEM;
goto out_restore;
}
call->a_flags = RPC_TASK_ASYNC;
} else {
spin_unlock_irqrestore(¤t->sighand->siglock, flags);
memset(call, 0, sizeof(*call));
locks_init_lock(&call->a_args.lock.fl);
locks_init_lock(&call->a_res.lock.fl);
}
call->a_host = host;
/* Set up the argument struct */
nlmclnt_setlockargs(call, fl);
if (IS_SETLK(cmd) || IS_SETLKW(cmd)) {
if (fl->fl_type != F_UNLCK) {
call->a_args.block = IS_SETLKW(cmd) ? 1 : 0;
status = nlmclnt_lock(call, fl);
} else
status = nlmclnt_unlock(call, fl);
} else if (IS_GETLK(cmd))
status = nlmclnt_test(call, fl);
else
status = -EINVAL;
if (status < 0 && (call->a_flags & RPC_TASK_ASYNC))
kfree(call);
out_restore:
spin_lock_irqsave(¤t->sighand->siglock, flags);
current->blocked = oldset;
recalc_sigpending();
spin_unlock_irqrestore(¤t->sighand->siglock, flags);
done:
dprintk("lockd: clnt proc returns %d\n", status);
nlm_release_host(host);
return status;
}
static int v9fs_file_do_lock(struct file *filp, int cmd, struct file_lock *fl)
{
struct p9_flock flock;
struct p9_fid *fid;
uint8_t status = P9_LOCK_ERROR;
int res = 0;
unsigned char fl_type;
struct v9fs_session_info *v9ses;
fid = filp->private_data;
BUG_ON(fid == NULL);
if ((fl->fl_flags & FL_POSIX) != FL_POSIX)
BUG();
res = locks_lock_file_wait(filp, fl);
if (res < 0)
goto out;
/* convert posix lock to p9 tlock args */
memset(&flock, 0, sizeof(flock));
/* map the lock type */
switch (fl->fl_type) {
case F_RDLCK:
flock.type = P9_LOCK_TYPE_RDLCK;
break;
case F_WRLCK:
flock.type = P9_LOCK_TYPE_WRLCK;
break;
case F_UNLCK:
flock.type = P9_LOCK_TYPE_UNLCK;
break;
}
flock.start = fl->fl_start;
if (fl->fl_end == OFFSET_MAX)
flock.length = 0;
else
flock.length = fl->fl_end - fl->fl_start + 1;
flock.proc_id = fl->fl_pid;
flock.client_id = fid->clnt->name;
if (IS_SETLKW(cmd))
flock.flags = P9_LOCK_FLAGS_BLOCK;
v9ses = v9fs_inode2v9ses(file_inode(filp));
/*
* if its a blocked request and we get P9_LOCK_BLOCKED as the status
* for lock request, keep on trying
*/
for (;;) {
res = p9_client_lock_dotl(fid, &flock, &status);
if (res < 0)
goto out_unlock;
if (status != P9_LOCK_BLOCKED)
break;
if (status == P9_LOCK_BLOCKED && !IS_SETLKW(cmd))
break;
if (schedule_timeout_interruptible(v9ses->session_lock_timeout)
!= 0)
break;
/*
* p9_client_lock_dotl overwrites flock.client_id with the
* server message, free and reuse the client name
*/
if (flock.client_id != fid->clnt->name) {
kfree(flock.client_id);
flock.client_id = fid->clnt->name;
}
}
/* map 9p status to VFS status */
switch (status) {
case P9_LOCK_SUCCESS:
res = 0;
break;
case P9_LOCK_BLOCKED:
res = -EAGAIN;
break;
default:
WARN_ONCE(1, "unknown lock status code: %d\n", status);
/* fall through */
case P9_LOCK_ERROR:
case P9_LOCK_GRACE:
res = -ENOLCK;
break;
}
out_unlock:
/*
* incase server returned error for lock request, revert
* it locally
*/
if (res < 0 && fl->fl_type != F_UNLCK) {
fl_type = fl->fl_type;
fl->fl_type = F_UNLCK;
/* Even if this fails we want to return the remote error */
locks_lock_file_wait(filp, fl);
fl->fl_type = fl_type;
}
if (flock.client_id != fid->clnt->name)
kfree(flock.client_id);
out:
return res;
}