static int ocfs2_readpages(struct file *filp, struct address_space *mapping,
struct list_head *pages, unsigned nr_pages)
{
int ret, err = -EIO;
struct inode *inode = mapping->host;
struct ocfs2_inode_info *oi = OCFS2_I(inode);
loff_t start;
struct page *last;
ret = ocfs2_inode_lock_full(inode, NULL, 0, OCFS2_LOCK_NONBLOCK);
if (ret)
return err;
if (down_read_trylock(&oi->ip_alloc_sem) == 0) {
ocfs2_inode_unlock(inode, 0);
return err;
}
if (oi->ip_dyn_features & OCFS2_INLINE_DATA_FL)
goto out_unlock;
last = list_entry(pages->prev, struct page, lru);
start = (loff_t)last->index << PAGE_CACHE_SHIFT;
if (start >= i_size_read(inode))
goto out_unlock;
err = mpage_readpages(mapping, pages, nr_pages, ocfs2_get_block);
out_unlock:
up_read(&oi->ip_alloc_sem);
ocfs2_inode_unlock(inode, 0);
return err;
}
static struct dentry *ocfs2_get_parent(struct dentry *child)
{
int status;
u64 blkno;
struct dentry *parent;
struct inode *dir = child->d_inode;
trace_ocfs2_get_parent(child, child->d_name.len, child->d_name.name,
(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));
bail_unlock:
ocfs2_inode_unlock(dir, 0);
bail:
trace_ocfs2_get_parent_end(parent);
return parent;
}
/*
* Get posix acl.
*/
static struct posix_acl *ocfs2_get_acl(struct inode *inode, int type)
{
struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
struct buffer_head *di_bh = NULL;
struct posix_acl *acl;
int ret;
if (!(osb->s_mount_opt & OCFS2_MOUNT_POSIX_ACL))
return NULL;
ret = ocfs2_inode_lock(inode, &di_bh, 0);
if (ret < 0) {
mlog_errno(ret);
acl = ERR_PTR(ret);
return acl;
}
acl = ocfs2_get_acl_nolock(inode, type, di_bh);
ocfs2_inode_unlock(inode, 0);
brelse(di_bh);
return acl;
}
/*
* Step 2: By now, we've completed the journal recovery, we've stamped
* a clean local alloc on disk and dropped the node out of the
* recovery map. Dlm locks will no longer stall, so lets clear out the
* main bitmap.
*/
int ocfs2_complete_local_alloc_recovery(struct ocfs2_super *osb,
struct ocfs2_dinode *alloc)
{
int status;
handle_t *handle;
struct buffer_head *main_bm_bh = NULL;
struct inode *main_bm_inode;
main_bm_inode = ocfs2_get_system_file_inode(osb,
GLOBAL_BITMAP_SYSTEM_INODE,
OCFS2_INVALID_SLOT);
if (!main_bm_inode) {
status = -EINVAL;
mlog_errno(status);
goto out;
}
inode_lock(main_bm_inode);
status = ocfs2_inode_lock(main_bm_inode, &main_bm_bh, 1);
if (status < 0) {
mlog_errno(status);
goto out_mutex;
}
handle = ocfs2_start_trans(osb, OCFS2_WINDOW_MOVE_CREDITS);
if (IS_ERR(handle)) {
status = PTR_ERR(handle);
handle = NULL;
mlog_errno(status);
goto out_unlock;
}
/* we want the bitmap change to be recorded on disk asap */
handle->h_sync = 1;
status = ocfs2_sync_local_to_main(osb, handle, alloc,
main_bm_inode, main_bm_bh);
if (status < 0)
mlog_errno(status);
ocfs2_commit_trans(osb, handle);
out_unlock:
ocfs2_inode_unlock(main_bm_inode, 1);
out_mutex:
inode_unlock(main_bm_inode);
brelse(main_bm_bh);
iput(main_bm_inode);
out:
if (!status)
ocfs2_init_steal_slots(osb);
if (status)
mlog_errno(status);
return status;
}
/*
* This is called from our getattr.
*/
int ocfs2_inode_revalidate(struct dentry *dentry)
{
struct inode *inode = dentry->d_inode;
int status = 0;
trace_ocfs2_inode_revalidate(inode,
inode ? (unsigned long long)OCFS2_I(inode)->ip_blkno : 0ULL,
inode ? (unsigned long long)OCFS2_I(inode)->ip_flags : 0);
if (!inode) {
status = -ENOENT;
goto bail;
}
spin_lock(&OCFS2_I(inode)->ip_lock);
if (OCFS2_I(inode)->ip_flags & OCFS2_INODE_DELETED) {
spin_unlock(&OCFS2_I(inode)->ip_lock);
status = -ENOENT;
goto bail;
}
spin_unlock(&OCFS2_I(inode)->ip_lock);
/* Let ocfs2_inode_lock do the work of updating our struct
* inode for us. */
status = ocfs2_inode_lock(inode, NULL, 0);
if (status < 0) {
if (status != -ENOENT)
mlog_errno(status);
goto bail;
}
ocfs2_inode_unlock(inode, 0);
bail:
return status;
}
/*
* This is used only for read-ahead. Failures or difficult to handle
* situations are safe to ignore.
*
* Right now, we don't bother with BH_Boundary - in-inode extent lists
* are quite large (243 extents on 4k blocks), so most inodes don't
* grow out to a tree. If need be, detecting boundary extents could
* trivially be added in a future version of ocfs2_get_block().
*/
static int ocfs2_readpages(struct file *filp, struct address_space *mapping,
struct list_head *pages, unsigned nr_pages)
{
int ret, err = -EIO;
struct inode *inode = mapping->host;
struct ocfs2_inode_info *oi = OCFS2_I(inode);
loff_t start;
struct page *last;
/*
* Use the nonblocking flag for the dlm code to avoid page
* lock inversion, but don't bother with retrying.
*/
ret = ocfs2_inode_lock_full(inode, NULL, 0, OCFS2_LOCK_NONBLOCK);
if (ret)
return err;
if (down_read_trylock(&oi->ip_alloc_sem) == 0) {
ocfs2_inode_unlock(inode, 0);
return err;
}
/*
* Don't bother with inline-data. There isn't anything
* to read-ahead in that case anyway...
*/
if (oi->ip_dyn_features & OCFS2_INLINE_DATA_FL)
goto out_unlock;
/*
* Check whether a remote node truncated this file - we just
* drop out in that case as it's not worth handling here.
*/
last = list_entry(pages->prev, struct page, lru);
start = (loff_t)last->index << PAGE_CACHE_SHIFT;
if (start >= i_size_read(inode))
goto out_unlock;
err = mpage_readpages(mapping, pages, nr_pages, ocfs2_get_block);
out_unlock:
up_read(&oi->ip_alloc_sem);
ocfs2_inode_unlock(inode, 0);
return err;
}
static int ocfs2_readpage(struct file *file, struct page *page)
{
struct inode *inode = page->mapping->host;
struct ocfs2_inode_info *oi = OCFS2_I(inode);
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_inode_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(&oi->ip_alloc_sem) == 0) {
ret = AOP_TRUNCATED_PAGE;
goto out_inode_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, 0, PAGE_SIZE);
SetPageUptodate(page);
ret = 0;
goto out_alloc;
}
if (oi->ip_dyn_features & OCFS2_INLINE_DATA_FL)
ret = ocfs2_readpage_inline(inode, page);
else
ret = block_read_full_page(page, ocfs2_get_block);
unlock = 0;
out_alloc:
up_read(&OCFS2_I(inode)->ip_alloc_sem);
out_inode_unlock:
ocfs2_inode_unlock(inode, 0);
out:
if (unlock)
unlock_page(page);
mlog_exit(ret);
return ret;
}
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;
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;
}
inode = ocfs2_iget(OCFS2_SB(dir->i_sb), blkno, 0, 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_inode_unlock(dir, 0);
bail:
mlog_exit_ptr(parent);
return parent;
}
static int ocfs2_readpage(struct file *file, struct page *page)
{
struct inode *inode = page->mapping->host;
struct ocfs2_inode_info *oi = OCFS2_I(inode);
loff_t start = (loff_t)page->index << PAGE_CACHE_SHIFT;
int ret, unlock = 1;
trace_ocfs2_readpage((unsigned long long)oi->ip_blkno,
(page ? page->index : 0));
ret = ocfs2_inode_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(&oi->ip_alloc_sem) == 0) {
ret = AOP_TRUNCATED_PAGE;
unlock_page(page);
unlock = 0;
down_read(&oi->ip_alloc_sem);
up_read(&oi->ip_alloc_sem);
goto out_inode_unlock;
}
if (start >= i_size_read(inode)) {
zero_user(page, 0, PAGE_SIZE);
SetPageUptodate(page);
ret = 0;
goto out_alloc;
}
if (oi->ip_dyn_features & OCFS2_INLINE_DATA_FL)
ret = ocfs2_readpage_inline(inode, page);
else
ret = block_read_full_page(page, ocfs2_get_block);
unlock = 0;
out_alloc:
up_read(&OCFS2_I(inode)->ip_alloc_sem);
out_inode_unlock:
ocfs2_inode_unlock(inode, 0);
out:
if (unlock)
unlock_page(page);
return ret;
}
int ocfs2_mmap(struct file *file, struct vm_area_struct *vma)
{
int ret = 0, lock_level = 0;
ret = ocfs2_inode_lock_atime(file_inode(file),
file->f_path.mnt, &lock_level);
if (ret < 0) {
mlog_errno(ret);
goto out;
}
ocfs2_inode_unlock(file_inode(file), lock_level);
out:
vma->vm_ops = &ocfs2_file_vm_ops;
return 0;
}
static int ocfs2_get_inode_attr(struct inode *inode, unsigned *flags)
{
int status;
status = ocfs2_inode_lock(inode, NULL, 0);
if (status < 0) {
mlog_errno(status);
return status;
}
ocfs2_get_inode_flags(OCFS2_I(inode));
*flags = OCFS2_I(inode)->ip_attr;
ocfs2_inode_unlock(inode, 0);
return status;
}
int ocfs2_iop_set_acl(struct inode *inode, struct posix_acl *acl, int type)
{
struct buffer_head *bh = NULL;
int status = 0;
status = ocfs2_inode_lock(inode, &bh, 1);
if (status < 0) {
if (status != -ENOENT)
mlog_errno(status);
return status;
}
status = ocfs2_set_acl(NULL, inode, bh, type, acl, NULL, NULL);
ocfs2_inode_unlock(inode, 1);
brelse(bh);
return status;
}
int ocfs2_mmap(struct file *file, struct vm_area_struct *vma)
{
int ret = 0, lock_level = 0;
ret = ocfs2_inode_lock_atime(file->f_dentry->d_inode,
file->f_vfsmnt, &lock_level);
if (ret < 0) {
mlog_errno(ret);
goto out;
}
ocfs2_inode_unlock(file->f_dentry->d_inode, lock_level);
out:
vma->vm_ops = &ocfs2_file_vm_ops;
vma->vm_flags |= VM_CAN_NONLINEAR;
return 0;
}
void ocfs2_free_ac_resource(struct ocfs2_alloc_context *ac)
{
struct inode *inode = ac->ac_inode;
if (inode) {
if (ac->ac_which != OCFS2_AC_USE_LOCAL)
ocfs2_inode_unlock(inode, 1);
mutex_unlock(&inode->i_mutex);
iput(inode);
ac->ac_inode = NULL;
}
brelse(ac->ac_bh);
ac->ac_bh = NULL;
}
static int ocfs2_info_scan_inode_alloc(struct ocfs2_super *osb,
struct inode *inode_alloc, u64 blkno,
struct ocfs2_info_freeinode *fi,
u32 slot)
{
int status = 0, unlock = 0;
struct buffer_head *bh = NULL;
struct ocfs2_dinode *dinode_alloc = NULL;
if (inode_alloc)
mutex_lock(&inode_alloc->i_mutex);
if (o2info_coherent(&fi->ifi_req)) {
status = ocfs2_inode_lock(inode_alloc, &bh, 0);
if (status < 0) {
mlog_errno(status);
goto bail;
}
unlock = 1;
} else {
status = ocfs2_read_blocks_sync(osb, blkno, 1, &bh);
if (status < 0) {
mlog_errno(status);
goto bail;
}
}
dinode_alloc = (struct ocfs2_dinode *)bh->b_data;
fi->ifi_stat[slot].lfi_total =
le32_to_cpu(dinode_alloc->id1.bitmap1.i_total);
fi->ifi_stat[slot].lfi_free =
le32_to_cpu(dinode_alloc->id1.bitmap1.i_total) -
le32_to_cpu(dinode_alloc->id1.bitmap1.i_used);
bail:
if (unlock)
ocfs2_inode_unlock(inode_alloc, 0);
if (inode_alloc)
mutex_unlock(&inode_alloc->i_mutex);
brelse(bh);
return status;
}
static int ocfs2_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
{
struct page *page = vmf->page;
struct inode *inode = vma->vm_file->f_path.dentry->d_inode;
struct buffer_head *di_bh = NULL;
sigset_t blocked, oldset;
int ret, ret2;
ret = ocfs2_vm_op_block_sigs(&blocked, &oldset);
if (ret < 0) {
mlog_errno(ret);
return ret;
}
/*
* The cluster locks taken will block a truncate from another
* node. Taking the data lock will also ensure that we don't
* attempt page truncation as part of a downconvert.
*/
ret = ocfs2_inode_lock(inode, &di_bh, 1);
if (ret < 0) {
mlog_errno(ret);
goto out;
}
/*
* The alloc sem should be enough to serialize with
* ocfs2_truncate_file() changing i_size as well as any thread
* modifying the inode btree.
*/
down_write(&OCFS2_I(inode)->ip_alloc_sem);
ret = __ocfs2_page_mkwrite(inode, di_bh, page);
up_write(&OCFS2_I(inode)->ip_alloc_sem);
brelse(di_bh);
ocfs2_inode_unlock(inode, 1);
out:
ret2 = ocfs2_vm_op_unblock_sigs(&oldset);
if (ret2 < 0)
mlog_errno(ret2);
if (ret)
ret = VM_FAULT_SIGBUS;
return ret;
}
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 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;
trace_ocfs2_bmap((unsigned long long)OCFS2_I(inode)->ip_blkno,
(unsigned long long)block);
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;
return status;
}
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;
}
/*
* This is called from our getattr.
*/
int ocfs2_inode_revalidate(struct dentry *dentry)
{
struct inode *inode = dentry->d_inode;
int status = 0;
mlog_entry("(inode = 0x%p, ino = %llu)\n", inode,
inode ? (unsigned long long)OCFS2_I(inode)->ip_blkno : 0ULL);
if (!inode) {
mlog(0, "eep, no inode!\n");
status = -ENOENT;
goto bail;
}
spin_lock(&OCFS2_I(inode)->ip_lock);
if (OCFS2_I(inode)->ip_flags & OCFS2_INODE_DELETED) {
spin_unlock(&OCFS2_I(inode)->ip_lock);
mlog(0, "inode deleted!\n");
status = -ENOENT;
goto bail;
}
spin_unlock(&OCFS2_I(inode)->ip_lock);
/* Let ocfs2_inode_lock do the work of updating our struct
* inode for us. */
status = ocfs2_inode_lock(inode, NULL, 0);
if (status < 0) {
if (status != -ENOENT)
mlog_errno(status);
goto bail;
}
ocfs2_inode_unlock(inode, 0);
bail:
mlog_exit(status);
return status;
}
static int ocfs2_move_extent(struct ocfs2_move_extents_context *context,
u32 cpos, u32 phys_cpos, u32 *new_phys_cpos,
u32 len, int ext_flags)
{
int ret, credits = 0, extra_blocks = 0, goal_bit = 0;
handle_t *handle;
struct inode *inode = context->inode;
struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
struct inode *tl_inode = osb->osb_tl_inode;
struct inode *gb_inode = NULL;
struct buffer_head *gb_bh = NULL;
struct buffer_head *gd_bh = NULL;
struct ocfs2_group_desc *gd;
struct ocfs2_refcount_tree *ref_tree = NULL;
u32 move_max_hop = ocfs2_blocks_to_clusters(inode->i_sb,
context->range->me_threshold);
u64 phys_blkno, new_phys_blkno;
phys_blkno = ocfs2_clusters_to_blocks(inode->i_sb, phys_cpos);
if ((ext_flags & OCFS2_EXT_REFCOUNTED) && len) {
BUG_ON(!(OCFS2_I(inode)->ip_dyn_features &
OCFS2_HAS_REFCOUNT_FL));
BUG_ON(!context->refcount_loc);
ret = ocfs2_lock_refcount_tree(osb, context->refcount_loc, 1,
&ref_tree, NULL);
if (ret) {
mlog_errno(ret);
return ret;
}
ret = ocfs2_prepare_refcount_change_for_del(inode,
context->refcount_loc,
phys_blkno,
len,
&credits,
&extra_blocks);
if (ret) {
mlog_errno(ret);
goto out;
}
}
ret = ocfs2_lock_allocators_move_extents(inode, &context->et, len, 1,
&context->meta_ac,
NULL, extra_blocks, &credits);
if (ret) {
mlog_errno(ret);
goto out;
}
/*
* need to count 2 extra credits for global_bitmap inode and
* group descriptor.
*/
credits += OCFS2_INODE_UPDATE_CREDITS + 1;
/*
* ocfs2_move_extent() didn't reserve any clusters in lock_allocators()
* logic, while we still need to lock the global_bitmap.
*/
gb_inode = ocfs2_get_system_file_inode(osb, GLOBAL_BITMAP_SYSTEM_INODE,
OCFS2_INVALID_SLOT);
if (!gb_inode) {
mlog(ML_ERROR, "unable to get global_bitmap inode\n");
ret = -EIO;
goto out;
}
mutex_lock(&gb_inode->i_mutex);
ret = ocfs2_inode_lock(gb_inode, &gb_bh, 1);
if (ret) {
mlog_errno(ret);
goto out_unlock_gb_mutex;
}
mutex_lock(&tl_inode->i_mutex);
handle = ocfs2_start_trans(osb, credits);
if (IS_ERR(handle)) {
ret = PTR_ERR(handle);
mlog_errno(ret);
goto out_unlock_tl_inode;
}
new_phys_blkno = ocfs2_clusters_to_blocks(inode->i_sb, *new_phys_cpos);
ret = ocfs2_find_victim_alloc_group(inode, new_phys_blkno,
GLOBAL_BITMAP_SYSTEM_INODE,
OCFS2_INVALID_SLOT,
&goal_bit, &gd_bh);
if (ret) {
mlog_errno(ret);
goto out_commit;
}
/*
* probe the victim cluster group to find a proper
* region to fit wanted movement, it even will perfrom
* a best-effort attempt by compromising to a threshold
* around the goal.
*/
ocfs2_probe_alloc_group(inode, gd_bh, &goal_bit, len, move_max_hop,
new_phys_cpos);
if (!*new_phys_cpos) {
ret = -ENOSPC;
goto out_commit;
}
ret = __ocfs2_move_extent(handle, context, cpos, len, phys_cpos,
*new_phys_cpos, ext_flags);
if (ret) {
mlog_errno(ret);
goto out_commit;
}
gd = (struct ocfs2_group_desc *)gd_bh->b_data;
ret = ocfs2_alloc_dinode_update_counts(gb_inode, handle, gb_bh, len,
le16_to_cpu(gd->bg_chain));
if (ret) {
mlog_errno(ret);
goto out_commit;
}
ret = ocfs2_block_group_set_bits(handle, gb_inode, gd, gd_bh,
goal_bit, len);
if (ret)
mlog_errno(ret);
/*
* Here we should write the new page out first if we are
* in write-back mode.
*/
ret = ocfs2_cow_sync_writeback(inode->i_sb, context->inode, cpos, len);
if (ret)
mlog_errno(ret);
out_commit:
ocfs2_commit_trans(osb, handle);
brelse(gd_bh);
out_unlock_tl_inode:
mutex_unlock(&tl_inode->i_mutex);
ocfs2_inode_unlock(gb_inode, 1);
out_unlock_gb_mutex:
mutex_unlock(&gb_inode->i_mutex);
brelse(gb_bh);
iput(gb_inode);
out:
if (context->meta_ac) {
ocfs2_free_alloc_context(context->meta_ac);
context->meta_ac = NULL;
}
if (ref_tree)
ocfs2_unlock_refcount_tree(osb, ref_tree, 1);
return ret;
}
/*
* return any unused bits to the bitmap and write out a clean
* local_alloc.
*
* local_alloc_bh is optional. If not passed, we will simply use the
* one off osb. If you do pass it however, be warned that it *will* be
* returned brelse'd and NULL'd out.*/
void ocfs2_shutdown_local_alloc(struct ocfs2_super *osb)
{
int status;
handle_t *handle;
struct inode *local_alloc_inode = NULL;
struct buffer_head *bh = NULL;
struct buffer_head *main_bm_bh = NULL;
struct inode *main_bm_inode = NULL;
struct ocfs2_dinode *alloc_copy = NULL;
struct ocfs2_dinode *alloc = NULL;
cancel_delayed_work(&osb->la_enable_wq);
flush_workqueue(osb->ocfs2_wq);
if (osb->local_alloc_state == OCFS2_LA_UNUSED)
goto out;
local_alloc_inode =
ocfs2_get_system_file_inode(osb,
LOCAL_ALLOC_SYSTEM_INODE,
osb->slot_num);
if (!local_alloc_inode) {
status = -ENOENT;
mlog_errno(status);
goto out;
}
osb->local_alloc_state = OCFS2_LA_DISABLED;
ocfs2_resmap_uninit(&osb->osb_la_resmap);
main_bm_inode = ocfs2_get_system_file_inode(osb,
GLOBAL_BITMAP_SYSTEM_INODE,
OCFS2_INVALID_SLOT);
if (!main_bm_inode) {
status = -EINVAL;
mlog_errno(status);
goto out;
}
inode_lock(main_bm_inode);
status = ocfs2_inode_lock(main_bm_inode, &main_bm_bh, 1);
if (status < 0) {
mlog_errno(status);
goto out_mutex;
}
/* WINDOW_MOVE_CREDITS is a bit heavy... */
handle = ocfs2_start_trans(osb, OCFS2_WINDOW_MOVE_CREDITS);
if (IS_ERR(handle)) {
mlog_errno(PTR_ERR(handle));
handle = NULL;
goto out_unlock;
}
bh = osb->local_alloc_bh;
alloc = (struct ocfs2_dinode *) bh->b_data;
alloc_copy = kmalloc(bh->b_size, GFP_NOFS);
if (!alloc_copy) {
status = -ENOMEM;
goto out_commit;
}
memcpy(alloc_copy, alloc, bh->b_size);
status = ocfs2_journal_access_di(handle, INODE_CACHE(local_alloc_inode),
bh, OCFS2_JOURNAL_ACCESS_WRITE);
if (status < 0) {
mlog_errno(status);
goto out_commit;
}
ocfs2_clear_local_alloc(alloc);
ocfs2_journal_dirty(handle, bh);
brelse(bh);
osb->local_alloc_bh = NULL;
osb->local_alloc_state = OCFS2_LA_UNUSED;
status = ocfs2_sync_local_to_main(osb, handle, alloc_copy,
main_bm_inode, main_bm_bh);
if (status < 0)
mlog_errno(status);
out_commit:
ocfs2_commit_trans(osb, handle);
out_unlock:
brelse(main_bm_bh);
ocfs2_inode_unlock(main_bm_inode, 1);
out_mutex:
inode_unlock(main_bm_inode);
iput(main_bm_inode);
out:
iput(local_alloc_inode);
kfree(alloc_copy);
}
static int ocfs2_set_inode_attr(struct inode *inode, unsigned flags,
unsigned mask)
{
struct ocfs2_inode_info *ocfs2_inode = OCFS2_I(inode);
struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
handle_t *handle = NULL;
struct buffer_head *bh = NULL;
unsigned oldflags;
int status;
mutex_lock(&inode->i_mutex);
status = ocfs2_inode_lock(inode, &bh, 1);
if (status < 0) {
mlog_errno(status);
goto bail;
}
status = -EACCES;
if (!inode_owner_or_capable(inode))
goto bail_unlock;
if (!S_ISDIR(inode->i_mode))
flags &= ~OCFS2_DIRSYNC_FL;
oldflags = ocfs2_inode->ip_attr;
flags = flags & mask;
flags |= oldflags & ~mask;
/*
* The IMMUTABLE and APPEND_ONLY flags can only be changed by
* the relevant capability.
*/
status = -EPERM;
if ((oldflags & OCFS2_IMMUTABLE_FL) || ((flags ^ oldflags) &
(OCFS2_APPEND_FL | OCFS2_IMMUTABLE_FL))) {
if (!capable(CAP_LINUX_IMMUTABLE))
goto bail_unlock;
}
handle = ocfs2_start_trans(osb, OCFS2_INODE_UPDATE_CREDITS);
if (IS_ERR(handle)) {
status = PTR_ERR(handle);
mlog_errno(status);
goto bail_unlock;
}
ocfs2_inode->ip_attr = flags;
ocfs2_set_inode_flags(inode);
status = ocfs2_mark_inode_dirty(handle, inode, bh);
if (status < 0)
mlog_errno(status);
ocfs2_commit_trans(osb, handle);
bail_unlock:
ocfs2_inode_unlock(inode, 1);
bail:
mutex_unlock(&inode->i_mutex);
brelse(bh);
return status;
}
static int ocfs2_info_freefrag_scan_bitmap(struct ocfs2_super *osb,
struct inode *gb_inode, u64 blkno,
struct ocfs2_info_freefrag *ffg)
{
u32 chunks_in_group;
int status = 0, unlock = 0, i;
struct buffer_head *bh = NULL;
struct ocfs2_chain_list *cl = NULL;
struct ocfs2_chain_rec *rec = NULL;
struct ocfs2_dinode *gb_dinode = NULL;
if (gb_inode)
mutex_lock(&gb_inode->i_mutex);
if (o2info_coherent(&ffg->iff_req)) {
status = ocfs2_inode_lock(gb_inode, &bh, 0);
if (status < 0) {
mlog_errno(status);
goto bail;
}
unlock = 1;
} else {
status = ocfs2_read_blocks_sync(osb, blkno, 1, &bh);
if (status < 0) {
mlog_errno(status);
goto bail;
}
}
gb_dinode = (struct ocfs2_dinode *)bh->b_data;
cl = &(gb_dinode->id2.i_chain);
/*
* Chunksize(in) clusters from userspace should be
* less than clusters in a group.
*/
if (ffg->iff_chunksize > le16_to_cpu(cl->cl_cpg)) {
status = -EINVAL;
goto bail;
}
memset(&ffg->iff_ffs, 0, sizeof(struct ocfs2_info_freefrag_stats));
ffg->iff_ffs.ffs_min = ~0U;
ffg->iff_ffs.ffs_clusters =
le32_to_cpu(gb_dinode->id1.bitmap1.i_total);
ffg->iff_ffs.ffs_free_clusters = ffg->iff_ffs.ffs_clusters -
le32_to_cpu(gb_dinode->id1.bitmap1.i_used);
chunks_in_group = le16_to_cpu(cl->cl_cpg) / ffg->iff_chunksize + 1;
for (i = 0; i < le16_to_cpu(cl->cl_next_free_rec); i++) {
rec = &(cl->cl_recs[i]);
status = ocfs2_info_freefrag_scan_chain(osb, gb_inode,
gb_dinode,
rec, ffg,
chunks_in_group);
if (status)
goto bail;
}
if (ffg->iff_ffs.ffs_free_chunks_real)
ffg->iff_ffs.ffs_avg = (ffg->iff_ffs.ffs_avg /
ffg->iff_ffs.ffs_free_chunks_real);
bail:
if (unlock)
ocfs2_inode_unlock(gb_inode, 0);
if (gb_inode)
mutex_unlock(&gb_inode->i_mutex);
if (gb_inode)
iput(gb_inode);
brelse(bh);
return status;
}
static void ocfs2_delete_inode(struct inode *inode)
{
int wipe, status;
sigset_t oldset;
struct buffer_head *di_bh = NULL;
trace_ocfs2_delete_inode(inode->i_ino,
(unsigned long long)OCFS2_I(inode)->ip_blkno,
is_bad_inode(inode));
/* When we fail in read_inode() we mark inode as bad. The second test
* catches the case when inode allocation fails before allocating
* a block for inode. */
if (is_bad_inode(inode) || !OCFS2_I(inode)->ip_blkno)
goto bail;
dquot_initialize(inode);
if (!ocfs2_inode_is_valid_to_delete(inode)) {
/* It's probably not necessary to truncate_inode_pages
* here but we do it for safety anyway (it will most
* likely be a no-op anyway) */
ocfs2_cleanup_delete_inode(inode, 0);
goto bail;
}
/* We want to block signals in delete_inode as the lock and
* messaging paths may return us -ERESTARTSYS. Which would
* cause us to exit early, resulting in inodes being orphaned
* forever. */
ocfs2_block_signals(&oldset);
/*
* Synchronize us against ocfs2_get_dentry. We take this in
* shared mode so that all nodes can still concurrently
* process deletes.
*/
status = ocfs2_nfs_sync_lock(OCFS2_SB(inode->i_sb), 0);
if (status < 0) {
mlog(ML_ERROR, "getting nfs sync lock(PR) failed %d\n", status);
ocfs2_cleanup_delete_inode(inode, 0);
goto bail_unblock;
}
/* Lock down the inode. This gives us an up to date view of
* it's metadata (for verification), and allows us to
* serialize delete_inode on multiple nodes.
*
* Even though we might be doing a truncate, we don't take the
* allocation lock here as it won't be needed - nobody will
* have the file open.
*/
status = ocfs2_inode_lock(inode, &di_bh, 1);
if (status < 0) {
if (status != -ENOENT)
mlog_errno(status);
ocfs2_cleanup_delete_inode(inode, 0);
goto bail_unlock_nfs_sync;
}
/* Query the cluster. This will be the final decision made
* before we go ahead and wipe the inode. */
status = ocfs2_query_inode_wipe(inode, di_bh, &wipe);
if (!wipe || status < 0) {
/* Error and remote inode busy both mean we won't be
* removing the inode, so they take almost the same
* path. */
if (status < 0)
mlog_errno(status);
/* Someone in the cluster has disallowed a wipe of
* this inode, or it was never completely
* orphaned. Write out the pages and exit now. */
ocfs2_cleanup_delete_inode(inode, 1);
goto bail_unlock_inode;
}
ocfs2_cleanup_delete_inode(inode, 0);
status = ocfs2_wipe_inode(inode, di_bh);
if (status < 0) {
if (status != -EDEADLK)
mlog_errno(status);
goto bail_unlock_inode;
}
/*
* Mark the inode as successfully deleted.
*
* This is important for ocfs2_clear_inode() as it will check
* this flag and skip any checkpointing work
*
* ocfs2_stuff_meta_lvb() also uses this flag to invalidate
* the LVB for other nodes.
*/
OCFS2_I(inode)->ip_flags |= OCFS2_INODE_DELETED;
bail_unlock_inode:
ocfs2_inode_unlock(inode, 1);
brelse(di_bh);
bail_unlock_nfs_sync:
ocfs2_nfs_sync_unlock(OCFS2_SB(inode->i_sb), 0);
bail_unblock:
ocfs2_unblock_signals(&oldset);
bail:
return;
}
static int ocfs2_wipe_inode(struct inode *inode,
struct buffer_head *di_bh)
{
int status, orphaned_slot = -1;
struct inode *orphan_dir_inode = NULL;
struct buffer_head *orphan_dir_bh = NULL;
struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
struct ocfs2_dinode *di = (struct ocfs2_dinode *) di_bh->b_data;
if (!(OCFS2_I(inode)->ip_flags & OCFS2_INODE_SKIP_ORPHAN_DIR)) {
orphaned_slot = le16_to_cpu(di->i_orphaned_slot);
status = ocfs2_check_orphan_recovery_state(osb, orphaned_slot);
if (status)
return status;
orphan_dir_inode = ocfs2_get_system_file_inode(osb,
ORPHAN_DIR_SYSTEM_INODE,
orphaned_slot);
if (!orphan_dir_inode) {
status = -EEXIST;
mlog_errno(status);
goto bail;
}
/* Lock the orphan dir. The lock will be held for the entire
* delete_inode operation. We do this now to avoid races with
* recovery completion on other nodes. */
mutex_lock(&orphan_dir_inode->i_mutex);
status = ocfs2_inode_lock(orphan_dir_inode, &orphan_dir_bh, 1);
if (status < 0) {
mutex_unlock(&orphan_dir_inode->i_mutex);
mlog_errno(status);
goto bail;
}
}
/* we do this while holding the orphan dir lock because we
* don't want recovery being run from another node to try an
* inode delete underneath us -- this will result in two nodes
* truncating the same file! */
status = ocfs2_truncate_for_delete(osb, inode, di_bh);
if (status < 0) {
mlog_errno(status);
goto bail_unlock_dir;
}
/* Remove any dir index tree */
if (S_ISDIR(inode->i_mode)) {
status = ocfs2_dx_dir_truncate(inode, di_bh);
if (status) {
mlog_errno(status);
goto bail_unlock_dir;
}
}
/*Free extended attribute resources associated with this inode.*/
status = ocfs2_xattr_remove(inode, di_bh);
if (status < 0) {
mlog_errno(status);
goto bail_unlock_dir;
}
status = ocfs2_remove_refcount_tree(inode, di_bh);
if (status < 0) {
mlog_errno(status);
goto bail_unlock_dir;
}
status = ocfs2_remove_inode(inode, di_bh, orphan_dir_inode,
orphan_dir_bh);
if (status < 0)
mlog_errno(status);
bail_unlock_dir:
if (OCFS2_I(inode)->ip_flags & OCFS2_INODE_SKIP_ORPHAN_DIR)
return status;
ocfs2_inode_unlock(orphan_dir_inode, 1);
mutex_unlock(&orphan_dir_inode->i_mutex);
brelse(orphan_dir_bh);
bail:
iput(orphan_dir_inode);
ocfs2_signal_wipe_completion(osb, orphaned_slot);
return status;
}
static int ocfs2_remove_inode(struct inode *inode,
struct buffer_head *di_bh,
struct inode *orphan_dir_inode,
struct buffer_head *orphan_dir_bh)
{
int status;
struct inode *inode_alloc_inode = NULL;
struct buffer_head *inode_alloc_bh = NULL;
handle_t *handle;
struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
struct ocfs2_dinode *di = (struct ocfs2_dinode *) di_bh->b_data;
inode_alloc_inode =
ocfs2_get_system_file_inode(osb, INODE_ALLOC_SYSTEM_INODE,
le16_to_cpu(di->i_suballoc_slot));
if (!inode_alloc_inode) {
status = -EEXIST;
mlog_errno(status);
goto bail;
}
mutex_lock(&inode_alloc_inode->i_mutex);
status = ocfs2_inode_lock(inode_alloc_inode, &inode_alloc_bh, 1);
if (status < 0) {
mutex_unlock(&inode_alloc_inode->i_mutex);
mlog_errno(status);
goto bail;
}
handle = ocfs2_start_trans(osb, OCFS2_DELETE_INODE_CREDITS +
ocfs2_quota_trans_credits(inode->i_sb));
if (IS_ERR(handle)) {
status = PTR_ERR(handle);
mlog_errno(status);
goto bail_unlock;
}
if (!(OCFS2_I(inode)->ip_flags & OCFS2_INODE_SKIP_ORPHAN_DIR)) {
status = ocfs2_orphan_del(osb, handle, orphan_dir_inode, inode,
orphan_dir_bh);
if (status < 0) {
mlog_errno(status);
goto bail_commit;
}
}
/* set the inodes dtime */
status = ocfs2_journal_access_di(handle, INODE_CACHE(inode), di_bh,
OCFS2_JOURNAL_ACCESS_WRITE);
if (status < 0) {
mlog_errno(status);
goto bail_commit;
}
di->i_dtime = cpu_to_le64(CURRENT_TIME.tv_sec);
di->i_flags &= cpu_to_le32(~(OCFS2_VALID_FL | OCFS2_ORPHANED_FL));
ocfs2_journal_dirty(handle, di_bh);
ocfs2_remove_from_cache(INODE_CACHE(inode), di_bh);
dquot_free_inode(inode);
status = ocfs2_free_dinode(handle, inode_alloc_inode,
inode_alloc_bh, di);
if (status < 0)
mlog_errno(status);
bail_commit:
ocfs2_commit_trans(osb, handle);
bail_unlock:
ocfs2_inode_unlock(inode_alloc_inode, 1);
mutex_unlock(&inode_alloc_inode->i_mutex);
brelse(inode_alloc_bh);
bail:
iput(inode_alloc_inode);
return status;
}
/*
* Extend the filesystem to the new number of clusters specified. This entry
* point is only used to extend the current filesystem to the end of the last
* existing group.
*/
int ocfs2_group_extend(struct inode * inode, int new_clusters)
{
int ret;
handle_t *handle;
struct buffer_head *main_bm_bh = NULL;
struct buffer_head *group_bh = NULL;
struct inode *main_bm_inode = NULL;
struct ocfs2_dinode *fe = NULL;
struct ocfs2_group_desc *group = NULL;
struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
u16 cl_bpc;
u32 first_new_cluster;
u64 lgd_blkno;
if (ocfs2_is_hard_readonly(osb) || ocfs2_is_soft_readonly(osb))
return -EROFS;
if (new_clusters < 0)
return -EINVAL;
else if (new_clusters == 0)
return 0;
main_bm_inode = ocfs2_get_system_file_inode(osb,
GLOBAL_BITMAP_SYSTEM_INODE,
OCFS2_INVALID_SLOT);
if (!main_bm_inode) {
ret = -EINVAL;
mlog_errno(ret);
goto out;
}
mutex_lock(&main_bm_inode->i_mutex);
ret = ocfs2_inode_lock(main_bm_inode, &main_bm_bh, 1);
if (ret < 0) {
mlog_errno(ret);
goto out_mutex;
}
fe = (struct ocfs2_dinode *)main_bm_bh->b_data;
/* main_bm_bh is validated by inode read inside ocfs2_inode_lock(),
* so any corruption is a code bug. */
BUG_ON(!OCFS2_IS_VALID_DINODE(fe));
if (le16_to_cpu(fe->id2.i_chain.cl_cpg) !=
ocfs2_group_bitmap_size(osb->sb, 0,
osb->s_feature_incompat) * 8) {
mlog(ML_ERROR, "The disk is too old and small. "
"Force to do offline resize.");
ret = -EINVAL;
goto out_unlock;
}
first_new_cluster = le32_to_cpu(fe->i_clusters);
lgd_blkno = ocfs2_which_cluster_group(main_bm_inode,
first_new_cluster - 1);
ret = ocfs2_read_group_descriptor(main_bm_inode, fe, lgd_blkno,
&group_bh);
if (ret < 0) {
mlog_errno(ret);
goto out_unlock;
}
group = (struct ocfs2_group_desc *)group_bh->b_data;
cl_bpc = le16_to_cpu(fe->id2.i_chain.cl_bpc);
if (le16_to_cpu(group->bg_bits) / cl_bpc + new_clusters >
le16_to_cpu(fe->id2.i_chain.cl_cpg)) {
ret = -EINVAL;
goto out_unlock;
}
trace_ocfs2_group_extend(
(unsigned long long)le64_to_cpu(group->bg_blkno), new_clusters);
handle = ocfs2_start_trans(osb, OCFS2_GROUP_EXTEND_CREDITS);
if (IS_ERR(handle)) {
mlog_errno(PTR_ERR(handle));
ret = -EINVAL;
goto out_unlock;
}
/* update the last group descriptor and inode. */
ret = ocfs2_update_last_group_and_inode(handle, main_bm_inode,
main_bm_bh, group_bh,
first_new_cluster,
new_clusters);
if (ret) {
mlog_errno(ret);
goto out_commit;
}
ocfs2_update_super_and_backups(main_bm_inode, new_clusters);
out_commit:
ocfs2_commit_trans(osb, handle);
out_unlock:
brelse(group_bh);
brelse(main_bm_bh);
ocfs2_inode_unlock(main_bm_inode, 1);
out_mutex:
mutex_unlock(&main_bm_inode->i_mutex);
iput(main_bm_inode);
out:
return ret;
}
/* Add a new group descriptor to global_bitmap. */
int ocfs2_group_add(struct inode *inode, struct ocfs2_new_group_input *input)
{
int ret;
handle_t *handle;
struct buffer_head *main_bm_bh = NULL;
struct inode *main_bm_inode = NULL;
struct ocfs2_dinode *fe = NULL;
struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
struct buffer_head *group_bh = NULL;
struct ocfs2_group_desc *group = NULL;
struct ocfs2_chain_list *cl;
struct ocfs2_chain_rec *cr;
u16 cl_bpc;
if (ocfs2_is_hard_readonly(osb) || ocfs2_is_soft_readonly(osb))
return -EROFS;
main_bm_inode = ocfs2_get_system_file_inode(osb,
GLOBAL_BITMAP_SYSTEM_INODE,
OCFS2_INVALID_SLOT);
if (!main_bm_inode) {
ret = -EINVAL;
mlog_errno(ret);
goto out;
}
mutex_lock(&main_bm_inode->i_mutex);
ret = ocfs2_inode_lock(main_bm_inode, &main_bm_bh, 1);
if (ret < 0) {
mlog_errno(ret);
goto out_mutex;
}
fe = (struct ocfs2_dinode *)main_bm_bh->b_data;
if (le16_to_cpu(fe->id2.i_chain.cl_cpg) !=
ocfs2_group_bitmap_size(osb->sb, 0,
osb->s_feature_incompat) * 8) {
mlog(ML_ERROR, "The disk is too old and small."
" Force to do offline resize.");
ret = -EINVAL;
goto out_unlock;
}
ret = ocfs2_read_blocks_sync(osb, input->group, 1, &group_bh);
if (ret < 0) {
mlog(ML_ERROR, "Can't read the group descriptor # %llu "
"from the device.", (unsigned long long)input->group);
goto out_unlock;
}
ocfs2_set_new_buffer_uptodate(INODE_CACHE(inode), group_bh);
ret = ocfs2_verify_group_and_input(main_bm_inode, fe, input, group_bh);
if (ret) {
mlog_errno(ret);
goto out_unlock;
}
trace_ocfs2_group_add((unsigned long long)input->group,
input->chain, input->clusters, input->frees);
handle = ocfs2_start_trans(osb, OCFS2_GROUP_ADD_CREDITS);
if (IS_ERR(handle)) {
mlog_errno(PTR_ERR(handle));
ret = -EINVAL;
goto out_unlock;
}
cl_bpc = le16_to_cpu(fe->id2.i_chain.cl_bpc);
cl = &fe->id2.i_chain;
cr = &cl->cl_recs[input->chain];
ret = ocfs2_journal_access_gd(handle, INODE_CACHE(main_bm_inode),
group_bh, OCFS2_JOURNAL_ACCESS_WRITE);
if (ret < 0) {
mlog_errno(ret);
goto out_commit;
}
group = (struct ocfs2_group_desc *)group_bh->b_data;
group->bg_next_group = cr->c_blkno;
ocfs2_journal_dirty(handle, group_bh);
ret = ocfs2_journal_access_di(handle, INODE_CACHE(main_bm_inode),
main_bm_bh, OCFS2_JOURNAL_ACCESS_WRITE);
if (ret < 0) {
mlog_errno(ret);
goto out_commit;
}
if (input->chain == le16_to_cpu(cl->cl_next_free_rec)) {
le16_add_cpu(&cl->cl_next_free_rec, 1);
memset(cr, 0, sizeof(struct ocfs2_chain_rec));
}
cr->c_blkno = cpu_to_le64(input->group);
le32_add_cpu(&cr->c_total, input->clusters * cl_bpc);
le32_add_cpu(&cr->c_free, input->frees * cl_bpc);
le32_add_cpu(&fe->id1.bitmap1.i_total, input->clusters *cl_bpc);
le32_add_cpu(&fe->id1.bitmap1.i_used,
(input->clusters - input->frees) * cl_bpc);
le32_add_cpu(&fe->i_clusters, input->clusters);
ocfs2_journal_dirty(handle, main_bm_bh);
spin_lock(&OCFS2_I(main_bm_inode)->ip_lock);
OCFS2_I(main_bm_inode)->ip_clusters = le32_to_cpu(fe->i_clusters);
le64_add_cpu(&fe->i_size, input->clusters << osb->s_clustersize_bits);
spin_unlock(&OCFS2_I(main_bm_inode)->ip_lock);
i_size_write(main_bm_inode, le64_to_cpu(fe->i_size));
ocfs2_update_super_and_backups(main_bm_inode, input->clusters);
out_commit:
ocfs2_commit_trans(osb, handle);
out_unlock:
brelse(group_bh);
brelse(main_bm_bh);
ocfs2_inode_unlock(main_bm_inode, 1);
out_mutex:
mutex_unlock(&main_bm_inode->i_mutex);
iput(main_bm_inode);
out:
return ret;
}
static int ocfs2_read_locked_inode(struct inode *inode,
struct ocfs2_find_inode_args *args)
{
struct super_block *sb;
struct ocfs2_super *osb;
struct ocfs2_dinode *fe;
struct buffer_head *bh = NULL;
int status, can_lock;
u32 generation = 0;
status = -EINVAL;
if (inode == NULL || inode->i_sb == NULL) {
mlog(ML_ERROR, "bad inode\n");
return status;
}
sb = inode->i_sb;
osb = OCFS2_SB(sb);
if (!args) {
mlog(ML_ERROR, "bad inode args\n");
make_bad_inode(inode);
return status;
}
/*
* To improve performance of cold-cache inode stats, we take
* the cluster lock here if possible.
*
* Generally, OCFS2 never trusts the contents of an inode
* unless it's holding a cluster lock, so taking it here isn't
* a correctness issue as much as it is a performance
* improvement.
*
* There are three times when taking the lock is not a good idea:
*
* 1) During startup, before we have initialized the DLM.
*
* 2) If we are reading certain system files which never get
* cluster locks (local alloc, truncate log).
*
* 3) If the process doing the iget() is responsible for
* orphan dir recovery. We're holding the orphan dir lock and
* can get into a deadlock with another process on another
* node in ->delete_inode().
*
* #1 and #2 can be simply solved by never taking the lock
* here for system files (which are the only type we read
* during mount). It's a heavier approach, but our main
* concern is user-accessible files anyway.
*
* #3 works itself out because we'll eventually take the
* cluster lock before trusting anything anyway.
*/
can_lock = !(args->fi_flags & OCFS2_FI_FLAG_SYSFILE)
&& !(args->fi_flags & OCFS2_FI_FLAG_ORPHAN_RECOVERY)
&& !ocfs2_mount_local(osb);
trace_ocfs2_read_locked_inode(
(unsigned long long)OCFS2_I(inode)->ip_blkno, can_lock);
/*
* To maintain backwards compatibility with older versions of
* ocfs2-tools, we still store the generation value for system
* files. The only ones that actually matter to userspace are
* the journals, but it's easier and inexpensive to just flag
* all system files similarly.
*/
if (args->fi_flags & OCFS2_FI_FLAG_SYSFILE)
generation = osb->fs_generation;
ocfs2_inode_lock_res_init(&OCFS2_I(inode)->ip_inode_lockres,
OCFS2_LOCK_TYPE_META,
generation, inode);
ocfs2_inode_lock_res_init(&OCFS2_I(inode)->ip_open_lockres,
OCFS2_LOCK_TYPE_OPEN,
0, inode);
if (can_lock) {
status = ocfs2_open_lock(inode);
if (status) {
make_bad_inode(inode);
mlog_errno(status);
return status;
}
status = ocfs2_inode_lock(inode, NULL, 0);
if (status) {
make_bad_inode(inode);
mlog_errno(status);
return status;
}
}
if (args->fi_flags & OCFS2_FI_FLAG_ORPHAN_RECOVERY) {
status = ocfs2_try_open_lock(inode, 0);
if (status) {
make_bad_inode(inode);
return status;
}
}
if (can_lock) {
status = ocfs2_read_inode_block_full(inode, &bh,
OCFS2_BH_IGNORE_CACHE);
} else {
status = ocfs2_read_blocks_sync(osb, args->fi_blkno, 1, &bh);
/*
* If buffer is in jbd, then its checksum may not have been
* computed as yet.
*/
if (!status && !buffer_jbd(bh))
status = ocfs2_validate_inode_block(osb->sb, bh);
}
if (status < 0) {
mlog_errno(status);
goto bail;
}
status = -EINVAL;
fe = (struct ocfs2_dinode *) bh->b_data;
/*
* This is a code bug. Right now the caller needs to
* understand whether it is asking for a system file inode or
* not so the proper lock names can be built.
*/
mlog_bug_on_msg(!!(fe->i_flags & cpu_to_le32(OCFS2_SYSTEM_FL)) !=
!!(args->fi_flags & OCFS2_FI_FLAG_SYSFILE),
"Inode %llu: system file state is ambigous\n",
(unsigned long long)args->fi_blkno);
if (S_ISCHR(le16_to_cpu(fe->i_mode)) ||
S_ISBLK(le16_to_cpu(fe->i_mode)))
inode->i_rdev = huge_decode_dev(le64_to_cpu(fe->id1.dev1.i_rdev));
ocfs2_populate_inode(inode, fe, 0);
BUG_ON(args->fi_blkno != le64_to_cpu(fe->i_blkno));
status = 0;
bail:
if (can_lock)
ocfs2_inode_unlock(inode, 0);
if (status < 0)
make_bad_inode(inode);
if (args && bh)
brelse(bh);
return status;
}