static struct dentry *ocfs2_get_dentry(struct super_block *sb,
struct ocfs2_inode_handle *handle)
{
struct inode *inode;
struct dentry *result;
mlog_entry("(0x%p, 0x%p)\n", sb, handle);
if (handle->ih_blkno == 0) {
mlog_errno(-ESTALE);
return ERR_PTR(-ESTALE);
}
inode = ocfs2_iget(OCFS2_SB(sb), handle->ih_blkno, 0, 0);
if (IS_ERR(inode))
return (void *)inode;
if (handle->ih_generation != inode->i_generation) {
iput(inode);
return ERR_PTR(-ESTALE);
}
result = d_alloc_anon(inode);
if (!result) {
iput(inode);
mlog_errno(-ENOMEM);
return ERR_PTR(-ENOMEM);
}
result->d_op = &ocfs2_dentry_ops;
mlog_exit_ptr(result);
return result;
}
static int ocfs2_file_open(struct inode *inode, struct file *file)
{
int status;
int mode = file->f_flags;
struct ocfs2_inode_info *oi = OCFS2_I(inode);
mlog_entry("(0x%p, 0x%p, '%.*s')\n", inode, file,
file->f_path.dentry->d_name.len, file->f_path.dentry->d_name.name);
spin_lock(&oi->ip_lock);
/* Check that the inode hasn't been wiped from disk by another
* node. If it hasn't then we're safe as long as we hold the
* spin lock until our increment of open count. */
if (OCFS2_I(inode)->ip_flags & OCFS2_INODE_DELETED) {
spin_unlock(&oi->ip_lock);
status = -ENOENT;
goto leave;
}
if (mode & O_DIRECT)
oi->ip_flags |= OCFS2_INODE_OPEN_DIRECT;
oi->ip_open_count++;
spin_unlock(&oi->ip_lock);
status = 0;
leave:
mlog_exit(status);
return status;
}
/*
* NOTE: this should always be called with parent dir i_mutex taken.
*/
int ocfs2_find_files_on_disk(const char *name,
int namelen,
u64 *blkno,
struct inode *inode,
struct buffer_head **dirent_bh,
struct ocfs2_dir_entry **dirent)
{
int status = -ENOENT;
struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
mlog_entry("(osb=%p, parent=%"MLFu64", name='%.*s', blkno=%p, "
"inode=%p)\n",
osb, OCFS2_I(inode)->ip_blkno, namelen, name, blkno, inode);
*dirent_bh = ocfs2_find_entry(name, namelen, inode, dirent);
if (!*dirent_bh || !*dirent) {
status = -ENOENT;
goto leave;
}
*blkno = le64_to_cpu((*dirent)->inode);
status = 0;
leave:
if (status < 0) {
*dirent = NULL;
if (*dirent_bh) {
brelse(*dirent_bh);
*dirent_bh = NULL;
}
}
mlog_exit(status);
return status;
}
static struct page *ocfs2_nopage(struct vm_area_struct * area,
unsigned long address,
int *type)
{
struct page *page = NOPAGE_SIGBUS;
sigset_t blocked, oldset;
int ret;
mlog_entry("(area=%p, address=%lu, type=%p)\n", area, address,
type);
/* The best way to deal with signals in this path is
* to block them upfront, rather than allowing the
* locking paths to return -ERESTARTSYS. */
sigfillset(&blocked);
/* We should technically never get a bad ret return
* from sigprocmask */
ret = sigprocmask(SIG_BLOCK, &blocked, &oldset);
if (ret < 0) {
mlog_errno(ret);
goto out;
}
page = filemap_nopage(area, address, type);
ret = sigprocmask(SIG_SETMASK, &oldset, NULL);
if (ret < 0)
mlog_errno(ret);
out:
mlog_exit_ptr(page);
return page;
}
static ssize_t ocfs2_file_splice_read(struct file *in,
loff_t *ppos,
struct pipe_inode_info *pipe,
size_t len,
unsigned int flags)
{
int ret = 0;
struct inode *inode = in->f_path.dentry->d_inode;
mlog_entry("(0x%p, 0x%p, %u, '%.*s')\n", in, pipe,
(unsigned int)len,
in->f_path.dentry->d_name.len,
in->f_path.dentry->d_name.name);
/*
* See the comment in ocfs2_file_aio_read()
*/
ret = ocfs2_meta_lock(inode, NULL, 0);
if (ret < 0) {
mlog_errno(ret);
goto bail;
}
ocfs2_meta_unlock(inode, 0);
ret = generic_file_splice_read(in, ppos, pipe, len, flags);
bail:
mlog_exit(ret);
return ret;
}
void __dlm_lockres_calc_usage(struct dlm_ctxt *dlm,
struct dlm_lock_resource *res)
{
mlog_entry("%.*s\n", res->lockname.len, res->lockname.name);
assert_spin_locked(&dlm->spinlock);
assert_spin_locked(&res->spinlock);
if (__dlm_lockres_unused(res)){
if (list_empty(&res->purge)) {
mlog(0, "putting lockres %.*s:%p onto purge list\n",
res->lockname.len, res->lockname.name, res);
res->last_used = jiffies;
dlm_lockres_get(res);
list_add_tail(&res->purge, &dlm->purge_list);
dlm->purge_count++;
}
} else if (!list_empty(&res->purge)) {
mlog(0, "removing lockres %.*s:%p from purge list, owner=%u\n",
res->lockname.len, res->lockname.name, res, res->owner);
list_del_init(&res->purge);
dlm_lockres_put(res);
dlm->purge_count--;
}
}
/*
* NOTE: this should always be called with parent dir i_mutex taken.
*/
int ocfs2_find_files_on_disk(const char *name,
int namelen,
u64 *blkno,
struct inode *inode,
struct buffer_head **dirent_bh,
struct ocfs2_dir_entry **dirent)
{
int status = -ENOENT;
mlog_entry("(name=%.*s, blkno=%p, inode=%p, dirent_bh=%p, dirent=%p)\n",
namelen, name, blkno, inode, dirent_bh, dirent);
*dirent_bh = ocfs2_find_entry(name, namelen, inode, dirent);
if (!*dirent_bh || !*dirent) {
status = -ENOENT;
goto leave;
}
*blkno = le64_to_cpu((*dirent)->inode);
status = 0;
leave:
if (status < 0) {
*dirent = NULL;
if (*dirent_bh) {
brelse(*dirent_bh);
*dirent_bh = NULL;
}
}
mlog_exit(status);
return status;
}
/*
* initialize the new inode, but don't do anything that would cause
* us to sleep.
* return 0 on success, 1 on failure
*/
static int ocfs2_init_locked_inode(struct inode *inode, void *opaque)
{
struct ocfs2_find_inode_args *args = opaque;
static struct lock_class_key ocfs2_quota_ip_alloc_sem_key,
ocfs2_file_ip_alloc_sem_key;
mlog_entry("inode = %p, opaque = %p\n", inode, opaque);
inode->i_ino = args->fi_ino;
OCFS2_I(inode)->ip_blkno = args->fi_blkno;
if (args->fi_sysfile_type != 0)
lockdep_set_class(&inode->i_mutex,
&ocfs2_sysfile_lock_key[args->fi_sysfile_type]);
if (args->fi_sysfile_type == USER_QUOTA_SYSTEM_INODE ||
args->fi_sysfile_type == GROUP_QUOTA_SYSTEM_INODE ||
args->fi_sysfile_type == LOCAL_USER_QUOTA_SYSTEM_INODE ||
args->fi_sysfile_type == LOCAL_GROUP_QUOTA_SYSTEM_INODE)
lockdep_set_class(&OCFS2_I(inode)->ip_alloc_sem,
&ocfs2_quota_ip_alloc_sem_key);
else
lockdep_set_class(&OCFS2_I(inode)->ip_alloc_sem,
&ocfs2_file_ip_alloc_sem_key);
mlog_exit(0);
return 0;
}
static int ocfs2_fault(struct vm_area_struct *area, struct vm_fault *vmf)
{
sigset_t blocked, oldset;
int error, ret;
mlog_entry("(area=%p, page offset=%lu)\n", area, vmf->pgoff);
/* The best way to deal with signals in this path is
* to block them upfront, rather than allowing the
* locking paths to return -ERESTARTSYS. */
sigfillset(&blocked);
/* We should technically never get a bad ret return
* from sigprocmask */
error = sigprocmask(SIG_BLOCK, &blocked, &oldset);
if (ret < 0) {
mlog_errno(error);
ret = VM_FAULT_SIGBUS;
goto out;
}
ret = filemap_fault(area, vmf);
error = sigprocmask(SIG_SETMASK, &oldset, NULL);
if (error < 0)
mlog_errno(error);
out:
mlog_exit_ptr(vmf->page);
return ret;
}
static int ocfs2_readpage(struct file *file, struct page *page)
{
struct inode *inode = page->mapping->host;
loff_t start = (loff_t)page->index << PAGE_CACHE_SHIFT;
int ret, unlock = 1;
mlog_entry("(0x%p, %lu)\n", file, (page ? page->index : 0));
ret = ocfs2_meta_lock_with_page(inode, NULL, 0, page);
if (ret != 0) {
if (ret == AOP_TRUNCATED_PAGE)
unlock = 0;
mlog_errno(ret);
goto out;
}
if (down_read_trylock(&OCFS2_I(inode)->ip_alloc_sem) == 0) {
ret = AOP_TRUNCATED_PAGE;
goto out_meta_unlock;
}
/*
* i_size might have just been updated as we grabed the meta lock. We
* might now be discovering a truncate that hit on another node.
* block_read_full_page->get_block freaks out if it is asked to read
* beyond the end of a file, so we check here. Callers
* (generic_file_read, vm_ops->fault) are clever enough to check i_size
* and notice that the page they just read isn't needed.
*
* XXX sys_readahead() seems to get that wrong?
*/
if (start >= i_size_read(inode)) {
zero_user_page(page, 0, PAGE_SIZE, KM_USER0);
SetPageUptodate(page);
ret = 0;
goto out_alloc;
}
ret = ocfs2_data_lock_with_page(inode, 0, page);
if (ret != 0) {
if (ret == AOP_TRUNCATED_PAGE)
unlock = 0;
mlog_errno(ret);
goto out_alloc;
}
ret = block_read_full_page(page, ocfs2_get_block);
unlock = 0;
ocfs2_data_unlock(inode, 0);
out_alloc:
up_read(&OCFS2_I(inode)->ip_alloc_sem);
out_meta_unlock:
ocfs2_meta_unlock(inode, 0);
out:
if (unlock)
unlock_page(page);
mlog_exit(ret);
return ret;
}
static int ocfs2_encode_fh(struct dentry *dentry, u32 *fh_in, int *max_len,
int connectable)
{
struct inode *inode = dentry->d_inode;
int len = *max_len;
int type = 1;
u64 blkno;
u32 generation;
__le32 *fh = (__force __le32 *) fh_in;
mlog_entry("(0x%p, '%.*s', 0x%p, %d, %d)\n", dentry,
dentry->d_name.len, dentry->d_name.name,
fh, len, connectable);
if (len < 3 || (connectable && len < 6)) {
mlog(ML_ERROR, "fh buffer is too small for encoding\n");
type = 255;
goto bail;
}
blkno = OCFS2_I(inode)->ip_blkno;
generation = inode->i_generation;
mlog(0, "Encoding fh: blkno: %llu, generation: %u\n",
(unsigned long long)blkno, generation);
len = 3;
fh[0] = cpu_to_le32((u32)(blkno >> 32));
fh[1] = cpu_to_le32((u32)(blkno & 0xffffffff));
fh[2] = cpu_to_le32(generation);
if (connectable && !S_ISDIR(inode->i_mode)) {
struct inode *parent;
spin_lock(&dentry->d_lock);
parent = dentry->d_parent->d_inode;
blkno = OCFS2_I(parent)->ip_blkno;
generation = parent->i_generation;
fh[3] = cpu_to_le32((u32)(blkno >> 32));
fh[4] = cpu_to_le32((u32)(blkno & 0xffffffff));
fh[5] = cpu_to_le32(generation);
spin_unlock(&dentry->d_lock);
len = 6;
type = 2;
mlog(0, "Encoding parent: blkno: %llu, generation: %u\n",
(unsigned long long)blkno, generation);
}
*max_len = len;
bail:
mlog_exit(type);
return type;
}
int ocfs2_journal_access(handle_t *handle,
struct inode *inode,
struct buffer_head *bh,
int type)
{
int status;
BUG_ON(!inode);
BUG_ON(!handle);
BUG_ON(!bh);
mlog_entry("bh->b_blocknr=%llu, type=%d (\"%s\"), bh->b_size = %zu\n",
(unsigned long long)bh->b_blocknr, type,
(type == OCFS2_JOURNAL_ACCESS_CREATE) ?
"OCFS2_JOURNAL_ACCESS_CREATE" :
"OCFS2_JOURNAL_ACCESS_WRITE",
bh->b_size);
/* we can safely remove this assertion after testing. */
if (!buffer_uptodate(bh)) {
mlog(ML_ERROR, "giving me a buffer that's not uptodate!\n");
mlog(ML_ERROR, "b_blocknr=%llu\n",
(unsigned long long)bh->b_blocknr);
BUG();
}
/* Set the current transaction information on the inode so
* that the locking code knows whether it can drop it's locks
* on this inode or not. We're protected from the commit
* thread updating the current transaction id until
* ocfs2_commit_trans() because ocfs2_start_trans() took
* j_trans_barrier for us. */
ocfs2_set_inode_lock_trans(OCFS2_SB(inode->i_sb)->journal, inode);
mutex_lock(&OCFS2_I(inode)->ip_io_mutex);
switch (type) {
case OCFS2_JOURNAL_ACCESS_CREATE:
case OCFS2_JOURNAL_ACCESS_WRITE:
status = journal_get_write_access(handle, bh);
break;
case OCFS2_JOURNAL_ACCESS_UNDO:
status = journal_get_undo_access(handle, bh);
break;
default:
status = -EINVAL;
mlog(ML_ERROR, "Uknown access type!\n");
}
mutex_unlock(&OCFS2_I(inode)->ip_io_mutex);
if (status < 0)
mlog(ML_ERROR, "Error %d getting %d access to buffer!\n",
status, type);
mlog_exit(status);
return status;
}
static int ocfs2_local_alloc_find_clear_bits(struct ocfs2_super *osb,
struct ocfs2_dinode *alloc,
u32 numbits)
{
int numfound, bitoff, left, startoff, lastzero;
void *bitmap = NULL;
mlog_entry("(numbits wanted = %u)\n", numbits);
if (!alloc->id1.bitmap1.i_total) {
mlog(0, "No bits in my window!\n");
bitoff = -1;
goto bail;
}
bitmap = OCFS2_LOCAL_ALLOC(alloc)->la_bitmap;
numfound = bitoff = startoff = 0;
lastzero = -1;
left = le32_to_cpu(alloc->id1.bitmap1.i_total);
while ((bitoff = ocfs2_find_next_zero_bit(bitmap, left, startoff)) != -1) {
if (bitoff == left) {
/* mlog(0, "bitoff (%d) == left", bitoff); */
break;
}
/* mlog(0, "Found a zero: bitoff = %d, startoff = %d, "
"numfound = %d\n", bitoff, startoff, numfound);*/
/* Ok, we found a zero bit... is it contig. or do we
* start over?*/
if (bitoff == startoff) {
/* we found a zero */
numfound++;
startoff++;
} else {
/* got a zero after some ones */
numfound = 1;
startoff = bitoff+1;
}
/* we got everything we needed */
if (numfound == numbits) {
/* mlog(0, "Found it all!\n"); */
break;
}
}
mlog(0, "Exiting loop, bitoff = %d, numfound = %d\n", bitoff,
numfound);
if (numfound == numbits)
bitoff = startoff - numfound;
else
bitoff = -1;
bail:
mlog_exit(bitoff);
return bitoff;
}
static struct dentry *ocfs2_get_parent(struct dentry *child)
{
int status;
u64 blkno;
struct dentry *parent;
struct inode *inode;
struct inode *dir = child->d_inode;
struct buffer_head *dirent_bh = NULL;
struct ocfs2_dir_entry *dirent;
mlog_entry("(0x%p, '%.*s')\n", child,
child->d_name.len, child->d_name.name);
mlog(0, "find parent of directory %llu\n",
(unsigned long long)OCFS2_I(dir)->ip_blkno);
status = ocfs2_meta_lock(dir, NULL, 0);
if (status < 0) {
if (status != -ENOENT)
mlog_errno(status);
parent = ERR_PTR(status);
goto bail;
}
status = ocfs2_find_files_on_disk("..", 2, &blkno, dir, &dirent_bh,
&dirent);
if (status < 0) {
parent = ERR_PTR(-ENOENT);
goto bail_unlock;
}
inode = ocfs2_iget(OCFS2_SB(dir->i_sb), blkno, 0);
if (IS_ERR(inode)) {
mlog(ML_ERROR, "Unable to create inode %llu\n",
(unsigned long long)blkno);
parent = ERR_PTR(-EACCES);
goto bail_unlock;
}
parent = d_alloc_anon(inode);
if (!parent) {
iput(inode);
parent = ERR_PTR(-ENOMEM);
}
parent->d_op = &ocfs2_dentry_ops;
bail_unlock:
ocfs2_meta_unlock(dir, 0);
if (dirent_bh)
brelse(dirent_bh);
bail:
mlog_exit_ptr(parent);
return parent;
}
void dlm_lockres_calc_usage(struct dlm_ctxt *dlm,
struct dlm_lock_resource *res)
{
mlog_entry("%.*s\n", res->lockname.len, res->lockname.name);
spin_lock(&dlm->spinlock);
spin_lock(&res->spinlock);
__dlm_lockres_calc_usage(dlm, res);
spin_unlock(&res->spinlock);
spin_unlock(&dlm->spinlock);
}
struct inode *ocfs2_iget(struct ocfs2_super *osb, u64 blkno, unsigned flags,
int sysfile_type)
{
struct inode *inode = NULL;
struct super_block *sb = osb->sb;
struct ocfs2_find_inode_args args;
mlog_entry("(blkno = %llu)\n", (unsigned long long)blkno);
/* Ok. By now we've either got the offsets passed to us by the
* caller, or we just pulled them off the bh. Lets do some
* sanity checks to make sure they're OK. */
if (blkno == 0) {
inode = ERR_PTR(-EINVAL);
mlog_errno(PTR_ERR(inode));
goto bail;
}
args.fi_blkno = blkno;
args.fi_flags = flags;
args.fi_ino = ino_from_blkno(sb, blkno);
args.fi_sysfile_type = sysfile_type;
inode = iget5_locked(sb, args.fi_ino, ocfs2_find_actor,
ocfs2_init_locked_inode, &args);
/* inode was *not* in the inode cache. 2.6.x requires
* us to do our own read_inode call and unlock it
* afterwards. */
if (inode && inode->i_state & I_NEW) {
mlog(0, "Inode was not in inode cache, reading it.\n");
ocfs2_read_locked_inode(inode, &args);
unlock_new_inode(inode);
}
if (inode == NULL) {
inode = ERR_PTR(-ENOMEM);
mlog_errno(PTR_ERR(inode));
goto bail;
}
if (is_bad_inode(inode)) {
iput(inode);
inode = ERR_PTR(-ESTALE);
goto bail;
}
bail:
if (!IS_ERR(inode)) {
mlog(0, "returning inode with number %llu\n",
(unsigned long long)OCFS2_I(inode)->ip_blkno);
mlog_exit_ptr(inode);
}
return inode;
}
/*
* initialize the new inode, but don't do anything that would cause
* us to sleep.
* return 0 on success, 1 on failure
*/
static int ocfs2_init_locked_inode(struct inode *inode, void *opaque)
{
struct ocfs2_find_inode_args *args = opaque;
mlog_entry("inode = %p, opaque = %p\n", inode, opaque);
inode->i_ino = args->fi_ino;
OCFS2_I(inode)->ip_blkno = args->fi_blkno;
mlog_exit(0);
return 0;
}
/* Note: Because we don't support holes, our allocation has
* already happened (allocation writes zeros to the file data)
* so we don't have to worry about ordered writes in
* ocfs2_writepage.
*
* ->writepage is called during the process of invalidating the page cache
* during blocked lock processing. It can't block on any cluster locks
* to during block mapping. It's relying on the fact that the block
* mapping can't have disappeared under the dirty pages that it is
* being asked to write back.
*/
static int ocfs2_writepage(struct page *page, struct writeback_control *wbc)
{
int ret;
mlog_entry("(0x%p)\n", page);
ret = block_write_full_page(page, ocfs2_get_block, wbc);
mlog_exit(ret);
return ret;
}
static int ocfs2_write_remove_suid(struct inode *inode)
{
int ret;
struct buffer_head *bh = NULL;
struct ocfs2_inode_info *oi = OCFS2_I(inode);
handle_t *handle;
struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
struct ocfs2_dinode *di;
mlog_entry("(Inode %llu, mode 0%o)\n",
(unsigned long long)oi->ip_blkno, inode->i_mode);
handle = ocfs2_start_trans(osb, OCFS2_INODE_UPDATE_CREDITS);
if (handle == NULL) {
ret = -ENOMEM;
mlog_errno(ret);
goto out;
}
ret = ocfs2_read_block(osb, oi->ip_blkno, &bh, OCFS2_BH_CACHED, inode);
if (ret < 0) {
mlog_errno(ret);
goto out_trans;
}
ret = ocfs2_journal_access(handle, inode, bh,
OCFS2_JOURNAL_ACCESS_WRITE);
if (ret < 0) {
mlog_errno(ret);
goto out_bh;
}
inode->i_mode &= ~S_ISUID;
if ((inode->i_mode & S_ISGID) && (inode->i_mode & S_IXGRP))
inode->i_mode &= ~S_ISGID;
di = (struct ocfs2_dinode *) bh->b_data;
di->i_mode = cpu_to_le16(inode->i_mode);
ret = ocfs2_journal_dirty(handle, bh);
if (ret < 0)
mlog_errno(ret);
out_bh:
brelse(bh);
out_trans:
ocfs2_commit_trans(osb, handle);
out:
mlog_exit(ret);
return ret;
}
static int ocfs2_fault(struct vm_area_struct *area, struct vm_fault *vmf)
{
sigset_t oldset;
int ret;
mlog_entry("(area=%p, page offset=%lu)\n", area, vmf->pgoff);
ocfs2_block_signals(&oldset);
ret = filemap_fault(area, vmf);
ocfs2_unblock_signals(&oldset);
mlog_exit_ptr(vmf->page);
return ret;
}
/*
* initialize the new inode, but don't do anything that would cause
* us to sleep.
* return 0 on success, 1 on failure
*/
static int ocfs2_init_locked_inode(struct inode *inode, void *opaque)
{
struct ocfs2_find_inode_args *args = opaque;
mlog_entry("inode = %p, opaque = %p\n", inode, opaque);
inode->i_ino = args->fi_ino;
OCFS2_I(inode)->ip_blkno = args->fi_blkno;
if (args->fi_sysfile_type != 0)
lockdep_set_class(&inode->i_mutex,
&ocfs2_sysfile_lock_key[args->fi_sysfile_type]);
mlog_exit(0);
return 0;
}
int ocfs2_write_block(struct ocfs2_super *osb, struct buffer_head *bh,
struct inode *inode)
{
int ret = 0;
mlog_entry("(bh->b_blocknr = %llu, inode=%p)\n",
(unsigned long long)bh->b_blocknr, inode);
BUG_ON(bh->b_blocknr < OCFS2_SUPER_BLOCK_BLKNO);
BUG_ON(buffer_jbd(bh));
/* No need to check for a soft readonly file system here. non
* journalled writes are only ever done on system files which
* can get modified during recovery even if read-only. */
if (ocfs2_is_hard_readonly(osb)) {
ret = -EROFS;
goto out;
}
mutex_lock(&OCFS2_I(inode)->ip_io_mutex);
lock_buffer(bh);
set_buffer_uptodate(bh);
/* remove from dirty list before I/O. */
clear_buffer_dirty(bh);
get_bh(bh); /* for end_buffer_write_sync() */
bh->b_end_io = end_buffer_write_sync;
submit_bh(WRITE, bh);
wait_on_buffer(bh);
if (buffer_uptodate(bh)) {
ocfs2_set_buffer_uptodate(inode, bh);
} else {
/* We don't need to remove the clustered uptodate
* information for this bh as it's not marked locally
* uptodate. */
ret = -EIO;
put_bh(bh);
}
mutex_unlock(&OCFS2_I(inode)->ip_io_mutex);
out:
mlog_exit(ret);
return ret;
}
/*
* Updates a disk inode from a
* struct inode.
* Only takes ip_lock.
*/
int ocfs2_mark_inode_dirty(handle_t *handle,
struct inode *inode,
struct buffer_head *bh)
{
int status;
struct ocfs2_dinode *fe = (struct ocfs2_dinode *) bh->b_data;
mlog_entry("(inode %llu)\n",
(unsigned long long)OCFS2_I(inode)->ip_blkno);
status = ocfs2_journal_access(handle, inode, bh,
OCFS2_JOURNAL_ACCESS_WRITE);
if (status < 0) {
mlog_errno(status);
goto leave;
}
spin_lock(&OCFS2_I(inode)->ip_lock);
fe->i_clusters = cpu_to_le32(OCFS2_I(inode)->ip_clusters);
ocfs2_get_inode_flags(OCFS2_I(inode));
fe->i_attr = cpu_to_le32(OCFS2_I(inode)->ip_attr);
fe->i_dyn_features = cpu_to_le16(OCFS2_I(inode)->ip_dyn_features);
spin_unlock(&OCFS2_I(inode)->ip_lock);
fe->i_size = cpu_to_le64(i_size_read(inode));
fe->i_links_count = cpu_to_le16(inode->i_nlink);
fe->i_uid = cpu_to_le32(inode->i_uid);
fe->i_gid = cpu_to_le32(inode->i_gid);
fe->i_mode = cpu_to_le16(inode->i_mode);
fe->i_atime = cpu_to_le64(inode->i_atime.tv_sec);
fe->i_atime_nsec = cpu_to_le32(inode->i_atime.tv_nsec);
fe->i_ctime = cpu_to_le64(inode->i_ctime.tv_sec);
fe->i_ctime_nsec = cpu_to_le32(inode->i_ctime.tv_nsec);
fe->i_mtime = cpu_to_le64(inode->i_mtime.tv_sec);
fe->i_mtime_nsec = cpu_to_le32(inode->i_mtime.tv_nsec);
status = ocfs2_journal_dirty(handle, bh);
if (status < 0)
mlog_errno(status);
status = 0;
leave:
mlog_exit(status);
return status;
}
int ocfs2_journal_dirty(handle_t *handle,
struct buffer_head *bh)
{
int status;
mlog_entry("(bh->b_blocknr=%llu)\n",
(unsigned long long)bh->b_blocknr);
status = journal_dirty_metadata(handle, bh);
if (status < 0)
mlog(ML_ERROR, "Could not dirty metadata buffer. "
"(bh->b_blocknr=%llu)\n",
(unsigned long long)bh->b_blocknr);
mlog_exit(status);
return status;
}
static int ocfs2_file_release(struct inode *inode, struct file *file)
{
struct ocfs2_inode_info *oi = OCFS2_I(inode);
mlog_entry("(0x%p, 0x%p, '%.*s')\n", inode, file,
file->f_path.dentry->d_name.len,
file->f_path.dentry->d_name.name);
spin_lock(&oi->ip_lock);
if (!--oi->ip_open_count)
oi->ip_flags &= ~OCFS2_INODE_OPEN_DIRECT;
spin_unlock(&oi->ip_lock);
mlog_exit(0);
return 0;
}
static sector_t ocfs2_bmap(struct address_space *mapping, sector_t block)
{
sector_t status;
u64 p_blkno = 0;
int err = 0;
struct inode *inode = mapping->host;
mlog_entry("(block = %llu)\n", (unsigned long long)block);
/* We don't need to lock journal system files, since they aren't
* accessed concurrently from multiple nodes.
*/
if (!INODE_JOURNAL(inode)) {
err = ocfs2_inode_lock(inode, NULL, 0);
if (err) {
if (err != -ENOENT)
mlog_errno(err);
goto bail;
}
down_read(&OCFS2_I(inode)->ip_alloc_sem);
}
if (!(OCFS2_I(inode)->ip_dyn_features & OCFS2_INLINE_DATA_FL))
err = ocfs2_extent_map_get_blocks(inode, block, &p_blkno, NULL,
NULL);
if (!INODE_JOURNAL(inode)) {
up_read(&OCFS2_I(inode)->ip_alloc_sem);
ocfs2_inode_unlock(inode, 0);
}
if (err) {
mlog(ML_ERROR, "get_blocks() failed, block = %llu\n",
(unsigned long long)block);
mlog_errno(err);
goto bail;
}
bail:
status = err ? 0 : p_blkno;
mlog_exit((int)status);
return status;
}
static struct dentry *ocfs2_decode_fh(struct super_block *sb, u32 *fh_in,
int fh_len, int fileid_type,
int (*acceptable)(void *context,
struct dentry *de),
void *context)
{
struct ocfs2_inode_handle handle, parent;
struct dentry *ret = NULL;
__le32 *fh = (__force __le32 *) fh_in;
mlog_entry("(0x%p, 0x%p, %d, %d, 0x%p, 0x%p)\n",
sb, fh, fh_len, fileid_type, acceptable, context);
if (fh_len < 3 || fileid_type > 2)
goto bail;
if (fileid_type == 2) {
if (fh_len < 6)
goto bail;
parent.ih_blkno = (u64)le32_to_cpu(fh[3]) << 32;
parent.ih_blkno |= (u64)le32_to_cpu(fh[4]);
parent.ih_generation = le32_to_cpu(fh[5]);
mlog(0, "Decoding parent: blkno: %llu, generation: %u\n",
(unsigned long long)parent.ih_blkno,
parent.ih_generation);
}
handle.ih_blkno = (u64)le32_to_cpu(fh[0]) << 32;
handle.ih_blkno |= (u64)le32_to_cpu(fh[1]);
handle.ih_generation = le32_to_cpu(fh[2]);
mlog(0, "Encoding fh: blkno: %llu, generation: %u\n",
(unsigned long long)handle.ih_blkno, handle.ih_generation);
ret = ocfs2_export_ops.find_exported_dentry(sb, &handle, &parent,
acceptable, context);
bail:
mlog_exit_ptr(ret);
return ret;
}
static int ocfs2_find_actor(struct inode *inode, void *opaque)
{
struct ocfs2_find_inode_args *args = NULL;
struct ocfs2_inode_info *oi = OCFS2_I(inode);
int ret = 0;
mlog_entry("(0x%p, %lu, 0x%p)\n", inode, inode->i_ino, opaque);
args = opaque;
mlog_bug_on_msg(!inode, "No inode in find actor!\n");
if (oi->ip_blkno != args->fi_blkno)
goto bail;
ret = 1;
bail:
mlog_exit(ret);
return ret;
}
static struct dentry *ocfs2_get_parent(struct dentry *child)
{
int status;
u64 blkno;
struct dentry *parent;
struct inode *dir = child->d_inode;
mlog_entry("(0x%p, '%.*s')\n", child,
child->d_name.len, child->d_name.name);
mlog(0, "find parent of directory %llu\n",
(unsigned long long)OCFS2_I(dir)->ip_blkno);
status = ocfs2_inode_lock(dir, NULL, 0);
if (status < 0) {
if (status != -ENOENT)
mlog_errno(status);
parent = ERR_PTR(status);
goto bail;
}
status = ocfs2_lookup_ino_from_name(dir, "..", 2, &blkno);
if (status < 0) {
parent = ERR_PTR(-ENOENT);
goto bail_unlock;
}
parent = d_obtain_alias(ocfs2_iget(OCFS2_SB(dir->i_sb), blkno, 0, 0));
if (!IS_ERR(parent))
parent->d_op = &ocfs2_dentry_ops;
bail_unlock:
ocfs2_inode_unlock(dir, 0);
bail:
mlog_exit_ptr(parent);
return parent;
}
static int ocfs2_fault(struct vm_area_struct *area, struct vm_fault *vmf)
{
sigset_t blocked, oldset;
int error, ret;
mlog_entry("(area=%p, page offset=%lu)\n", area, vmf->pgoff);
error = ocfs2_vm_op_block_sigs(&blocked, &oldset);
if (error < 0) {
mlog_errno(error);
ret = VM_FAULT_SIGBUS;
goto out;
}
ret = filemap_fault(area, vmf);
error = ocfs2_vm_op_unblock_sigs(&oldset);
if (error < 0)
mlog_errno(error);
out:
mlog_exit_ptr(vmf->page);
return ret;
}