extent_map.o sysfs.o struct-funcs.o xattr.o ordered-data.o \
extent_io.o volumes.o async-thread.o ioctl.o locking.o orphan.o \
export.o tree-log.o acl.o free-space-cache.o zlib.o lzo.o \
- compression.o delayed-ref.o relocation.o delayed-inode.o
- compression.o delayed-ref.o relocation.o scrub.o
++ compression.o delayed-ref.o relocation.o delayed-inode.o scrub.o
#include <linux/mm.h>
#include <linux/highmem.h>
#include <linux/fs.h>
+ #include <linux/rwsem.h>
#include <linux/completion.h>
#include <linux/backing-dev.h>
#include <linux/wait.h>
#include "extent_io.h"
#include "extent_map.h"
#include "async-thread.h"
+ #include "ioctl.h"
struct btrfs_trans_handle;
struct btrfs_transaction;
/* For storing free space cache */
#define BTRFS_FREE_SPACE_OBJECTID -11ULL
+/*
+ * The inode number assigned to the special inode for sotring
+ * free ino cache
+ */
+#define BTRFS_FREE_INO_OBJECTID -12ULL
+
/* dummy objectid represents multiple objectids */
#define BTRFS_MULTIPLE_OBJECTIDS -255ULL
struct extent_map_tree map_tree;
};
- #define BTRFS_UUID_SIZE 16
struct btrfs_dev_item {
/* the internal btrfs device id */
__le64 devid;
sizeof(struct btrfs_stripe) * (num_stripes - 1);
}
- #define BTRFS_FSID_SIZE 16
#define BTRFS_HEADER_FLAG_WRITTEN (1ULL << 0)
#define BTRFS_HEADER_FLAG_RELOC (1ULL << 1)
/* use full backrefs for extent pointers in the block */
#define BTRFS_BLOCK_FLAG_FULL_BACKREF (1ULL << 8)
+ /*
+ * this flag is only used internally by scrub and may be changed at any time
+ * it is only declared here to avoid collisions
+ */
+ #define BTRFS_EXTENT_FLAG_SUPER (1ULL << 48)
+
struct btrfs_tree_block_info {
struct btrfs_disk_key key;
u8 level;
u64 total_bytes; /* total bytes in the space,
this doesn't take mirrors into account */
u64 bytes_used; /* total bytes used,
- this does't take mirrors into account */
+ this doesn't take mirrors into account */
u64 bytes_pinned; /* total bytes pinned, will be freed when the
transaction finishes */
u64 bytes_reserved; /* total bytes the allocator has reserved for
*/
unsigned long reservation_progress;
- int full:1; /* indicates that we cannot allocate any more
+ unsigned int full:1; /* indicates that we cannot allocate any more
chunks for this space */
- int chunk_alloc:1; /* set if we are allocating a chunk */
+ unsigned int chunk_alloc:1; /* set if we are allocating a chunk */
- int force_alloc; /* set if we need to force a chunk alloc for
- this space */
+ unsigned int force_alloc; /* set if we need to force a chunk
+ alloc for this space */
struct list_head list;
u64 bytes_super;
u64 flags;
u64 sectorsize;
- int extents_thresh;
- int free_extents;
- int total_bitmaps;
unsigned int ro:1;
unsigned int dirty:1;
unsigned int iref:1;
struct btrfs_space_info *space_info;
/* free space cache stuff */
- spinlock_t tree_lock;
- struct rb_root free_space_offset;
- u64 free_space;
+ struct btrfs_free_space_ctl *free_space_ctl;
/* block group cache stuff */
struct rb_node cache_node;
struct reloc_control;
struct btrfs_device;
struct btrfs_fs_devices;
+struct btrfs_delayed_root;
struct btrfs_fs_info {
u8 fsid[BTRFS_FSID_SIZE];
u8 chunk_tree_uuid[BTRFS_UUID_SIZE];
/* logical->physical extent mapping */
struct btrfs_mapping_tree mapping_tree;
- /* block reservation for extent, checksum and root tree */
+ /*
+ * block reservation for extent, checksum, root tree and
+ * delayed dir index item
+ */
struct btrfs_block_rsv global_block_rsv;
/* block reservation for delay allocation */
struct btrfs_block_rsv delalloc_block_rsv;
* for the sys_munmap function call path
*/
struct btrfs_workers fixup_workers;
+ struct btrfs_workers delayed_workers;
struct task_struct *transaction_kthread;
struct task_struct *cleaner_kthread;
int thread_pool_size;
void *bdev_holder;
+ /* private scrub information */
+ struct mutex scrub_lock;
+ atomic_t scrubs_running;
+ atomic_t scrub_pause_req;
+ atomic_t scrubs_paused;
+ atomic_t scrub_cancel_req;
+ wait_queue_head_t scrub_pause_wait;
+ struct rw_semaphore scrub_super_lock;
+ int scrub_workers_refcnt;
+ struct btrfs_workers scrub_workers;
+
/* filesystem state */
u64 fs_state;
+
+ struct btrfs_delayed_root *delayed_root;
};
/*
spinlock_t accounting_lock;
struct btrfs_block_rsv *block_rsv;
+ /* free ino cache stuff */
+ struct mutex fs_commit_mutex;
+ struct btrfs_free_space_ctl *free_ino_ctl;
+ enum btrfs_caching_type cached;
+ spinlock_t cache_lock;
+ wait_queue_head_t cache_wait;
+ struct btrfs_free_space_ctl *free_ino_pinned;
+ u64 cache_progress;
+ struct inode *cache_inode;
+
struct mutex log_mutex;
wait_queue_head_t log_writer_wait;
wait_queue_head_t log_commit_wait[2];
/* red-black tree that keeps track of in-memory inodes */
struct rb_root inode_tree;
+ /*
+ * radix tree that keeps track of delayed nodes of every inode,
+ * protected by inode_lock
+ */
+ struct radix_tree_root delayed_nodes_tree;
/*
* right now this just gets used so that a root has its own devid
* for stat. It may be used for more later
return btrfs_stripe_offset(eb, btrfs_stripe_nr(c, nr));
}
-static inline void btrfs_set_stripe_offset_nr(struct extent_buffer *eb,
- struct btrfs_chunk *c, int nr,
- u64 val)
-{
- btrfs_set_stripe_offset(eb, btrfs_stripe_nr(c, nr), val);
-}
-
static inline u64 btrfs_stripe_devid_nr(struct extent_buffer *eb,
struct btrfs_chunk *c, int nr)
{
return btrfs_stripe_devid(eb, btrfs_stripe_nr(c, nr));
}
-static inline void btrfs_set_stripe_devid_nr(struct extent_buffer *eb,
- struct btrfs_chunk *c, int nr,
- u64 val)
-{
- btrfs_set_stripe_devid(eb, btrfs_stripe_nr(c, nr), val);
-}
-
/* struct btrfs_block_group_item */
BTRFS_SETGET_STACK_FUNCS(block_group_used, struct btrfs_block_group_item,
used, 64);
return (struct btrfs_timespec *)ptr;
}
-static inline struct btrfs_timespec *
-btrfs_inode_otime(struct btrfs_inode_item *inode_item)
-{
- unsigned long ptr = (unsigned long)inode_item;
- ptr += offsetof(struct btrfs_inode_item, otime);
- return (struct btrfs_timespec *)ptr;
-}
-
BTRFS_SETGET_FUNCS(timespec_sec, struct btrfs_timespec, sec, 64);
BTRFS_SETGET_FUNCS(timespec_nsec, struct btrfs_timespec, nsec, 32);
return (u8 *)ptr;
}
-static inline u8 *btrfs_super_fsid(struct extent_buffer *eb)
-{
- unsigned long ptr = offsetof(struct btrfs_super_block, fsid);
- return (u8 *)ptr;
-}
-
-static inline u8 *btrfs_header_csum(struct extent_buffer *eb)
-{
- unsigned long ptr = offsetof(struct btrfs_header, csum);
- return (u8 *)ptr;
-}
-
-static inline struct btrfs_node *btrfs_buffer_node(struct extent_buffer *eb)
-{
- return NULL;
-}
-
-static inline struct btrfs_leaf *btrfs_buffer_leaf(struct extent_buffer *eb)
-{
- return NULL;
-}
-
-static inline struct btrfs_header *btrfs_buffer_header(struct extent_buffer *eb)
-{
- return NULL;
-}
-
static inline int btrfs_is_leaf(struct extent_buffer *eb)
{
return btrfs_header_level(eb) == 0;
return sb->s_fs_info;
}
-static inline int btrfs_set_root_name(struct btrfs_root *root,
- const char *name, int len)
-{
- /* if we already have a name just free it */
- kfree(root->name);
-
- root->name = kmalloc(len+1, GFP_KERNEL);
- if (!root->name)
- return -ENOMEM;
-
- memcpy(root->name, name, len);
- root->name[len] = '\0';
-
- return 0;
-}
-
static inline u32 btrfs_level_size(struct btrfs_root *root, int level)
{
if (level == 0)
}
/* extent-tree.c */
+static inline u64 btrfs_calc_trans_metadata_size(struct btrfs_root *root,
+ int num_items)
+{
+ return (root->leafsize + root->nodesize * (BTRFS_MAX_LEVEL - 1)) *
+ 3 * num_items;
+}
+
void btrfs_put_block_group(struct btrfs_block_group_cache *cache);
int btrfs_run_delayed_refs(struct btrfs_trans_handle *trans,
struct btrfs_root *root, unsigned long count);
u64 num_bytes, u64 *refs, u64 *flags);
int btrfs_pin_extent(struct btrfs_root *root,
u64 bytenr, u64 num, int reserved);
-int btrfs_drop_leaf_ref(struct btrfs_trans_handle *trans,
- struct btrfs_root *root, struct extent_buffer *leaf);
int btrfs_cross_ref_exist(struct btrfs_trans_handle *trans,
struct btrfs_root *root,
u64 objectid, u64 offset, u64 bytenr);
-int btrfs_copy_pinned(struct btrfs_root *root, struct extent_io_tree *copy);
struct btrfs_block_group_cache *btrfs_lookup_block_group(
struct btrfs_fs_info *info,
u64 bytenr);
struct btrfs_root *root, struct extent_buffer *parent,
int start_slot, int cache_only, u64 *last_ret,
struct btrfs_key *progress);
-void btrfs_release_path(struct btrfs_root *root, struct btrfs_path *p);
+void btrfs_release_path(struct btrfs_path *p);
struct btrfs_path *btrfs_alloc_path(void);
void btrfs_free_path(struct btrfs_path *p);
void btrfs_set_path_blocking(struct btrfs_path *p);
+void btrfs_clear_path_blocking(struct btrfs_path *p,
+ struct extent_buffer *held);
void btrfs_unlock_up_safe(struct btrfs_path *p, int level);
int btrfs_del_items(struct btrfs_trans_handle *trans, struct btrfs_root *root,
return btrfs_del_items(trans, root, path, path->slots[0], 1);
}
+int setup_items_for_insert(struct btrfs_trans_handle *trans,
+ struct btrfs_root *root, struct btrfs_path *path,
+ struct btrfs_key *cpu_key, u32 *data_size,
+ u32 total_data, u32 total_size, int nr);
int btrfs_insert_item(struct btrfs_trans_handle *trans, struct btrfs_root
*root, struct btrfs_key *key, void *data, u32 data_size);
-int btrfs_insert_some_items(struct btrfs_trans_handle *trans,
- struct btrfs_root *root,
- struct btrfs_path *path,
- struct btrfs_key *cpu_key, u32 *data_size,
- int nr);
int btrfs_insert_empty_items(struct btrfs_trans_handle *trans,
struct btrfs_root *root,
struct btrfs_path *path,
*item);
int btrfs_find_last_root(struct btrfs_root *root, u64 objectid, struct
btrfs_root_item *item, struct btrfs_key *key);
-int btrfs_search_root(struct btrfs_root *root, u64 search_start,
- u64 *found_objectid);
int btrfs_find_dead_roots(struct btrfs_root *root, u64 objectid);
int btrfs_find_orphan_roots(struct btrfs_root *tree_root);
int btrfs_set_root_node(struct btrfs_root_item *item,
/* dir-item.c */
int btrfs_insert_dir_item(struct btrfs_trans_handle *trans,
struct btrfs_root *root, const char *name,
- int name_len, u64 dir,
+ int name_len, struct inode *dir,
struct btrfs_key *location, u8 type, u64 index);
struct btrfs_dir_item *btrfs_lookup_dir_item(struct btrfs_trans_handle *trans,
struct btrfs_root *root,
struct btrfs_root *root, u64 offset);
int btrfs_find_orphan_item(struct btrfs_root *root, u64 offset);
-/* inode-map.c */
-int btrfs_find_free_objectid(struct btrfs_trans_handle *trans,
- struct btrfs_root *fs_root,
- u64 dirid, u64 *objectid);
-int btrfs_find_highest_inode(struct btrfs_root *fs_root, u64 *objectid);
-
/* inode-item.c */
int btrfs_insert_inode_ref(struct btrfs_trans_handle *trans,
struct btrfs_root *root,
struct btrfs_ordered_sum *sums);
int btrfs_csum_one_bio(struct btrfs_root *root, struct inode *inode,
struct bio *bio, u64 file_start, int contig);
-int btrfs_csum_file_bytes(struct btrfs_root *root, struct inode *inode,
- u64 start, unsigned long len);
struct btrfs_csum_item *btrfs_lookup_csum(struct btrfs_trans_handle *trans,
struct btrfs_root *root,
struct btrfs_path *path,
u64 bytenr, int cow);
- int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start,
- u64 end, struct list_head *list);
+ int btrfs_csum_truncate(struct btrfs_trans_handle *trans,
+ struct btrfs_root *root, struct btrfs_path *path,
+ u64 isize);
+ int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end,
+ struct list_head *list, int search_commit);
/* inode.c */
/* RHEL and EL kernels have a patch that renames PG_checked to FsMisc */
u32 min_type);
int btrfs_start_delalloc_inodes(struct btrfs_root *root, int delay_iput);
-int btrfs_start_one_delalloc_inode(struct btrfs_root *root, int delay_iput,
- int sync);
int btrfs_set_extent_delalloc(struct inode *inode, u64 start, u64 end,
struct extent_state **cached_state);
int btrfs_writepages(struct address_space *mapping,
int btrfs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf);
int btrfs_readpage(struct file *file, struct page *page);
void btrfs_evict_inode(struct inode *inode);
-void btrfs_put_inode(struct inode *inode);
int btrfs_write_inode(struct inode *inode, struct writeback_control *wbc);
void btrfs_dirty_inode(struct inode *inode);
struct inode *btrfs_alloc_inode(struct super_block *sb);
long btrfs_ioctl_trans_end(struct file *file);
struct inode *btrfs_iget(struct super_block *s, struct btrfs_key *location,
struct btrfs_root *root, int *was_new);
-int btrfs_commit_write(struct file *file, struct page *page,
- unsigned from, unsigned to);
struct extent_map *btrfs_get_extent(struct inode *inode, struct page *page,
- size_t page_offset, u64 start, u64 end,
+ size_t pg_offset, u64 start, u64 end,
int create);
int btrfs_update_inode(struct btrfs_trans_handle *trans,
struct btrfs_root *root,
int btrfs_sync_file(struct file *file, int datasync);
int btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
int skip_pinned);
-int btrfs_check_file(struct btrfs_root *root, struct inode *inode);
extern const struct file_operations btrfs_file_operations;
int btrfs_drop_extents(struct btrfs_trans_handle *trans, struct inode *inode,
u64 start, u64 end, u64 *hint_byte, int drop_cache);
/* sysfs.c */
int btrfs_init_sysfs(void);
void btrfs_exit_sysfs(void);
-int btrfs_sysfs_add_super(struct btrfs_fs_info *fs);
-int btrfs_sysfs_add_root(struct btrfs_root *root);
-void btrfs_sysfs_del_root(struct btrfs_root *root);
-void btrfs_sysfs_del_super(struct btrfs_fs_info *root);
/* xattr.c */
ssize_t btrfs_listxattr(struct dentry *dentry, char *buffer, size_t size);
u64 *bytes_to_reserve);
void btrfs_reloc_post_snapshot(struct btrfs_trans_handle *trans,
struct btrfs_pending_snapshot *pending);
+
+ /* scrub.c */
+ int btrfs_scrub_dev(struct btrfs_root *root, u64 devid, u64 start, u64 end,
+ struct btrfs_scrub_progress *progress, int readonly);
+ int btrfs_scrub_pause(struct btrfs_root *root);
+ int btrfs_scrub_pause_super(struct btrfs_root *root);
+ int btrfs_scrub_continue(struct btrfs_root *root);
+ int btrfs_scrub_continue_super(struct btrfs_root *root);
+ int btrfs_scrub_cancel(struct btrfs_root *root);
+ int btrfs_scrub_cancel_dev(struct btrfs_root *root, struct btrfs_device *dev);
+ int btrfs_scrub_cancel_devid(struct btrfs_root *root, u64 devid);
+ int btrfs_scrub_progress(struct btrfs_root *root, u64 devid,
+ struct btrfs_scrub_progress *progress);
+
#endif
#include <linux/crc32c.h>
#include <linux/slab.h>
#include <linux/migrate.h>
+#include <linux/ratelimit.h>
#include <asm/unaligned.h>
#include "compat.h"
#include "ctree.h"
#include "locking.h"
#include "tree-log.h"
#include "free-space-cache.h"
+#include "inode-map.h"
static struct extent_io_ops btree_extent_io_ops;
static void end_workqueue_fn(struct btrfs_work *work);
* that covers the entire device
*/
static struct extent_map *btree_get_extent(struct inode *inode,
- struct page *page, size_t page_offset, u64 start, u64 len,
+ struct page *page, size_t pg_offset, u64 start, u64 len,
int create)
{
struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
}
read_unlock(&em_tree->lock);
- em = alloc_extent_map(GFP_NOFS);
+ em = alloc_extent_map();
if (!em) {
em = ERR_PTR(-ENOMEM);
goto out;
memcpy(&found, result, csum_size);
read_extent_buffer(buf, &val, 0, csum_size);
- if (printk_ratelimit()) {
- printk(KERN_INFO "btrfs: %s checksum verify "
+ printk_ratelimited(KERN_INFO "btrfs: %s checksum verify "
"failed on %llu wanted %X found %X "
"level %d\n",
root->fs_info->sb->s_id,
(unsigned long long)buf->start, val, found,
btrfs_header_level(buf));
- }
if (result != (char *)&inline_result)
kfree(result);
return 1;
ret = 0;
goto out;
}
- if (printk_ratelimit()) {
- printk("parent transid verify failed on %llu wanted %llu "
+ printk_ratelimited("parent transid verify failed on %llu wanted %llu "
"found %llu\n",
(unsigned long long)eb->start,
(unsigned long long)parent_transid,
(unsigned long long)btrfs_header_generation(eb));
- }
ret = 1;
clear_extent_buffer_uptodate(io_tree, eb, &cached_state);
out:
len = page->private >> 2;
WARN_ON(len == 0);
- eb = alloc_extent_buffer(tree, start, len, page, GFP_NOFS);
+ eb = alloc_extent_buffer(tree, start, len, page);
if (eb == NULL) {
WARN_ON(1);
goto out;
len = page->private >> 2;
WARN_ON(len == 0);
- eb = alloc_extent_buffer(tree, start, len, page, GFP_NOFS);
+ eb = alloc_extent_buffer(tree, start, len, page);
if (eb == NULL) {
ret = -EIO;
goto out;
found_start = btrfs_header_bytenr(eb);
if (found_start != start) {
- if (printk_ratelimit()) {
- printk(KERN_INFO "btrfs bad tree block start "
+ printk_ratelimited(KERN_INFO "btrfs bad tree block start "
"%llu %llu\n",
(unsigned long long)found_start,
(unsigned long long)eb->start);
- }
ret = -EIO;
goto err;
}
goto err;
}
if (check_tree_block_fsid(root, eb)) {
- if (printk_ratelimit()) {
- printk(KERN_INFO "btrfs bad fsid on block %llu\n",
+ printk_ratelimited(KERN_INFO "btrfs bad fsid on block %llu\n",
(unsigned long long)eb->start);
- }
ret = -EIO;
goto err;
}
return 256 * limit;
}
-int btrfs_congested_async(struct btrfs_fs_info *info, int iodone)
-{
- return atomic_read(&info->nr_async_bios) >
- btrfs_async_submit_limit(info);
-}
-
static void run_one_async_start(struct btrfs_work *work)
{
struct async_submit_bio *async;
.writepages = btree_writepages,
.releasepage = btree_releasepage,
.invalidatepage = btree_invalidatepage,
- .sync_page = block_sync_page,
#ifdef CONFIG_MIGRATION
.migratepage = btree_migratepage,
#endif
struct inode *btree_inode = root->fs_info->btree_inode;
struct extent_buffer *eb;
eb = find_extent_buffer(&BTRFS_I(btree_inode)->io_tree,
- bytenr, blocksize, GFP_NOFS);
+ bytenr, blocksize);
return eb;
}
struct extent_buffer *eb;
eb = alloc_extent_buffer(&BTRFS_I(btree_inode)->io_tree,
- bytenr, blocksize, NULL, GFP_NOFS);
+ bytenr, blocksize, NULL);
return eb;
}
root->name = NULL;
root->in_sysfs = 0;
root->inode_tree = RB_ROOT;
+ INIT_RADIX_TREE(&root->delayed_nodes_tree, GFP_ATOMIC);
root->block_rsv = NULL;
root->orphan_block_rsv = NULL;
root->log_transid = 0;
root->last_log_commit = 0;
extent_io_tree_init(&root->dirty_log_pages,
- fs_info->btree_inode->i_mapping, GFP_NOFS);
+ fs_info->btree_inode->i_mapping);
memset(&root->root_key, 0, sizeof(root->root_key));
memset(&root->root_item, 0, sizeof(root->root_item));
return root;
}
-struct btrfs_root *btrfs_lookup_fs_root(struct btrfs_fs_info *fs_info,
- u64 root_objectid)
-{
- struct btrfs_root *root;
-
- if (root_objectid == BTRFS_ROOT_TREE_OBJECTID)
- return fs_info->tree_root;
- if (root_objectid == BTRFS_EXTENT_TREE_OBJECTID)
- return fs_info->extent_root;
-
- root = radix_tree_lookup(&fs_info->fs_roots_radix,
- (unsigned long)root_objectid);
- return root;
-}
-
struct btrfs_root *btrfs_read_fs_root_no_name(struct btrfs_fs_info *fs_info,
struct btrfs_key *location)
{
if (IS_ERR(root))
return root;
+ root->free_ino_ctl = kzalloc(sizeof(*root->free_ino_ctl), GFP_NOFS);
+ if (!root->free_ino_ctl)
+ goto fail;
+ root->free_ino_pinned = kzalloc(sizeof(*root->free_ino_pinned),
+ GFP_NOFS);
+ if (!root->free_ino_pinned)
+ goto fail;
+
+ btrfs_init_free_ino_ctl(root);
+ mutex_init(&root->fs_commit_mutex);
+ spin_lock_init(&root->cache_lock);
+ init_waitqueue_head(&root->cache_wait);
+
set_anon_super(&root->anon_super, NULL);
if (btrfs_root_refs(&root->root_item) == 0) {
return ERR_PTR(ret);
}
-struct btrfs_root *btrfs_read_fs_root(struct btrfs_fs_info *fs_info,
- struct btrfs_key *location,
- const char *name, int namelen)
-{
- return btrfs_read_fs_root_no_name(fs_info, location);
-#if 0
- struct btrfs_root *root;
- int ret;
-
- root = btrfs_read_fs_root_no_name(fs_info, location);
- if (!root)
- return NULL;
-
- if (root->in_sysfs)
- return root;
-
- ret = btrfs_set_root_name(root, name, namelen);
- if (ret) {
- free_extent_buffer(root->node);
- kfree(root);
- return ERR_PTR(ret);
- }
-
- ret = btrfs_sysfs_add_root(root);
- if (ret) {
- free_extent_buffer(root->node);
- kfree(root->name);
- kfree(root);
- return ERR_PTR(ret);
- }
- root->in_sysfs = 1;
- return root;
-#endif
-}
-
static int btrfs_congested_fn(void *congested_data, int bdi_bits)
{
struct btrfs_fs_info *info = (struct btrfs_fs_info *)congested_data;
return ret;
}
-/*
- * this unplugs every device on the box, and it is only used when page
- * is null
- */
-static void __unplug_io_fn(struct backing_dev_info *bdi, struct page *page)
-{
- struct btrfs_device *device;
- struct btrfs_fs_info *info;
-
- info = (struct btrfs_fs_info *)bdi->unplug_io_data;
- list_for_each_entry(device, &info->fs_devices->devices, dev_list) {
- if (!device->bdev)
- continue;
-
- bdi = blk_get_backing_dev_info(device->bdev);
- if (bdi->unplug_io_fn)
- bdi->unplug_io_fn(bdi, page);
- }
-}
-
-static void btrfs_unplug_io_fn(struct backing_dev_info *bdi, struct page *page)
-{
- struct inode *inode;
- struct extent_map_tree *em_tree;
- struct extent_map *em;
- struct address_space *mapping;
- u64 offset;
-
- /* the generic O_DIRECT read code does this */
- if (1 || !page) {
- __unplug_io_fn(bdi, page);
- return;
- }
-
- /*
- * page->mapping may change at any time. Get a consistent copy
- * and use that for everything below
- */
- smp_mb();
- mapping = page->mapping;
- if (!mapping)
- return;
-
- inode = mapping->host;
-
- /*
- * don't do the expensive searching for a small number of
- * devices
- */
- if (BTRFS_I(inode)->root->fs_info->fs_devices->open_devices <= 2) {
- __unplug_io_fn(bdi, page);
- return;
- }
-
- offset = page_offset(page);
-
- em_tree = &BTRFS_I(inode)->extent_tree;
- read_lock(&em_tree->lock);
- em = lookup_extent_mapping(em_tree, offset, PAGE_CACHE_SIZE);
- read_unlock(&em_tree->lock);
- if (!em) {
- __unplug_io_fn(bdi, page);
- return;
- }
-
- if (em->block_start >= EXTENT_MAP_LAST_BYTE) {
- free_extent_map(em);
- __unplug_io_fn(bdi, page);
- return;
- }
- offset = offset - em->start;
- btrfs_unplug_page(&BTRFS_I(inode)->root->fs_info->mapping_tree,
- em->block_start + offset, page);
- free_extent_map(em);
-}
-
/*
* If this fails, caller must call bdi_destroy() to get rid of the
* bdi again.
return err;
bdi->ra_pages = default_backing_dev_info.ra_pages;
- bdi->unplug_io_fn = btrfs_unplug_io_fn;
- bdi->unplug_io_data = info;
bdi->congested_fn = btrfs_congested_fn;
bdi->congested_data = info;
return 0;
struct btrfs_root *csum_root = kzalloc(sizeof(struct btrfs_root),
GFP_NOFS);
struct btrfs_root *tree_root = btrfs_sb(sb);
- struct btrfs_fs_info *fs_info = tree_root->fs_info;
+ struct btrfs_fs_info *fs_info = NULL;
struct btrfs_root *chunk_root = kzalloc(sizeof(struct btrfs_root),
GFP_NOFS);
struct btrfs_root *dev_root = kzalloc(sizeof(struct btrfs_root),
struct btrfs_super_block *disk_super;
- if (!extent_root || !tree_root || !fs_info ||
+ if (!extent_root || !tree_root || !tree_root->fs_info ||
!chunk_root || !dev_root || !csum_root) {
err = -ENOMEM;
goto fail;
}
+ fs_info = tree_root->fs_info;
ret = init_srcu_struct(&fs_info->subvol_srcu);
if (ret) {
INIT_LIST_HEAD(&fs_info->ordered_extents);
spin_lock_init(&fs_info->ordered_extent_lock);
+ fs_info->delayed_root = kmalloc(sizeof(struct btrfs_delayed_root),
+ GFP_NOFS);
+ if (!fs_info->delayed_root) {
+ err = -ENOMEM;
+ goto fail_iput;
+ }
+ btrfs_init_delayed_root(fs_info->delayed_root);
+ mutex_init(&fs_info->scrub_lock);
+ atomic_set(&fs_info->scrubs_running, 0);
+ atomic_set(&fs_info->scrub_pause_req, 0);
+ atomic_set(&fs_info->scrubs_paused, 0);
+ atomic_set(&fs_info->scrub_cancel_req, 0);
+ init_waitqueue_head(&fs_info->scrub_pause_wait);
+ init_rwsem(&fs_info->scrub_super_lock);
+ fs_info->scrub_workers_refcnt = 0;
+ btrfs_init_workers(&fs_info->scrub_workers, "scrub",
+ fs_info->thread_pool_size, &fs_info->generic_worker);
+
sb->s_blocksize = 4096;
sb->s_blocksize_bits = blksize_bits(4096);
sb->s_bdi = &fs_info->bdi;
RB_CLEAR_NODE(&BTRFS_I(fs_info->btree_inode)->rb_node);
extent_io_tree_init(&BTRFS_I(fs_info->btree_inode)->io_tree,
- fs_info->btree_inode->i_mapping,
- GFP_NOFS);
- extent_map_tree_init(&BTRFS_I(fs_info->btree_inode)->extent_tree,
- GFP_NOFS);
+ fs_info->btree_inode->i_mapping);
+ extent_map_tree_init(&BTRFS_I(fs_info->btree_inode)->extent_tree);
BTRFS_I(fs_info->btree_inode)->io_tree.ops = &btree_extent_io_ops;
fs_info->block_group_cache_tree = RB_ROOT;
extent_io_tree_init(&fs_info->freed_extents[0],
- fs_info->btree_inode->i_mapping, GFP_NOFS);
+ fs_info->btree_inode->i_mapping);
extent_io_tree_init(&fs_info->freed_extents[1],
- fs_info->btree_inode->i_mapping, GFP_NOFS);
+ fs_info->btree_inode->i_mapping);
fs_info->pinned_extents = &fs_info->freed_extents[0];
fs_info->do_barriers = 1;
bh = btrfs_read_dev_super(fs_devices->latest_bdev);
if (!bh) {
err = -EINVAL;
- goto fail_iput;
+ goto fail_alloc;
}
memcpy(&fs_info->super_copy, bh->b_data, sizeof(fs_info->super_copy));
disk_super = &fs_info->super_copy;
if (!btrfs_super_root(disk_super))
- goto fail_iput;
+ goto fail_alloc;
/* check FS state, whether FS is broken. */
fs_info->fs_state |= btrfs_super_flags(disk_super);
ret = btrfs_parse_options(tree_root, options);
if (ret) {
err = ret;
- goto fail_iput;
+ goto fail_alloc;
}
features = btrfs_super_incompat_flags(disk_super) &
"unsupported optional features (%Lx).\n",
(unsigned long long)features);
err = -EINVAL;
- goto fail_iput;
+ goto fail_alloc;
}
features = btrfs_super_incompat_flags(disk_super);
"unsupported option features (%Lx).\n",
(unsigned long long)features);
err = -EINVAL;
- goto fail_iput;
+ goto fail_alloc;
}
btrfs_init_workers(&fs_info->generic_worker,
&fs_info->generic_worker);
btrfs_init_workers(&fs_info->endio_freespace_worker, "freespace-write",
1, &fs_info->generic_worker);
+ btrfs_init_workers(&fs_info->delayed_workers, "delayed-meta",
+ fs_info->thread_pool_size,
+ &fs_info->generic_worker);
/*
* endios are largely parallel and should have a very
btrfs_start_workers(&fs_info->endio_meta_write_workers, 1);
btrfs_start_workers(&fs_info->endio_write_workers, 1);
btrfs_start_workers(&fs_info->endio_freespace_worker, 1);
+ btrfs_start_workers(&fs_info->delayed_workers, 1);
fs_info->bdi.ra_pages *= btrfs_super_num_devices(disk_super);
fs_info->bdi.ra_pages = max(fs_info->bdi.ra_pages,
btrfs_stop_workers(&fs_info->endio_write_workers);
btrfs_stop_workers(&fs_info->endio_freespace_worker);
btrfs_stop_workers(&fs_info->submit_workers);
+ btrfs_stop_workers(&fs_info->delayed_workers);
+fail_alloc:
+ kfree(fs_info->delayed_root);
fail_iput:
invalidate_inode_pages2(fs_info->btree_inode->i_mapping);
iput(fs_info->btree_inode);
if (uptodate) {
set_buffer_uptodate(bh);
} else {
- if (printk_ratelimit()) {
- printk(KERN_WARNING "lost page write due to "
+ printk_ratelimited(KERN_WARNING "lost page write due to "
"I/O error on %s\n",
bdevname(bh->b_bdev, b));
- }
/* note, we dont' set_buffer_write_io_error because we have
* our own ways of dealing with the IO errors
*/
if (btrfs_root_refs(&root->root_item) == 0)
synchronize_srcu(&fs_info->subvol_srcu);
+ __btrfs_remove_free_space_cache(root->free_ino_pinned);
+ __btrfs_remove_free_space_cache(root->free_ino_ctl);
free_fs_root(root);
return 0;
}
static void free_fs_root(struct btrfs_root *root)
{
+ iput(root->cache_inode);
WARN_ON(!RB_EMPTY_ROOT(&root->inode_tree));
if (root->anon_super.s_dev) {
down_write(&root->anon_super.s_umount);
}
free_extent_buffer(root->node);
free_extent_buffer(root->commit_root);
+ kfree(root->free_ino_ctl);
+ kfree(root->free_ino_pinned);
kfree(root->name);
kfree(root);
}
fs_info->closing = 1;
smp_mb();
+ btrfs_scrub_cancel(root);
btrfs_put_block_group_cache(fs_info);
/*
* ERROR state on disk.
*
* 2. when btrfs flips readonly just in btrfs_commit_super,
- * and in such case, btrfs cannnot write sb via btrfs_commit_super,
+ * and in such case, btrfs cannot write sb via btrfs_commit_super,
* and since fs_state has been set BTRFS_SUPER_FLAG_ERROR flag,
* btrfs will cleanup all FS resources first and write sb then.
*/
del_fs_roots(fs_info);
iput(fs_info->btree_inode);
+ kfree(fs_info->delayed_root);
btrfs_stop_workers(&fs_info->generic_worker);
btrfs_stop_workers(&fs_info->fixup_workers);
btrfs_stop_workers(&fs_info->endio_write_workers);
btrfs_stop_workers(&fs_info->endio_freespace_worker);
btrfs_stop_workers(&fs_info->submit_workers);
+ btrfs_stop_workers(&fs_info->delayed_workers);
btrfs_close_devices(fs_info->fs_devices);
btrfs_mapping_tree_free(&fs_info->mapping_tree);
u64 num_dirty;
unsigned long thresh = 32 * 1024 * 1024;
+ if (current->flags & PF_MEMALLOC)
+ return;
+
+ btrfs_balance_delayed_items(root);
+
+ num_dirty = root->fs_info->dirty_metadata_bytes;
+
+ if (num_dirty > thresh) {
+ balance_dirty_pages_ratelimited_nr(
+ root->fs_info->btree_inode->i_mapping, 1);
+ }
+ return;
+}
+
+void __btrfs_btree_balance_dirty(struct btrfs_root *root, unsigned long nr)
+{
+ /*
+ * looks as though older kernels can get into trouble with
+ * this code, they end up stuck in balance_dirty_pages forever
+ */
+ u64 num_dirty;
+ unsigned long thresh = 32 * 1024 * 1024;
+
if (current->flags & PF_MEMALLOC)
return;
goto out;
len = page->private >> 2;
- eb = find_extent_buffer(io_tree, bytenr, len, GFP_NOFS);
+ eb = find_extent_buffer(io_tree, bytenr, len);
if (!eb)
goto out;
u32 item_size;
if (item)
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
item = btrfs_lookup_csum(NULL, root->fs_info->csum_root,
path, disk_bytenr, 0);
if (IS_ERR(item)) {
EXTENT_NODATASUM, GFP_NOFS);
} else {
printk(KERN_INFO "btrfs no csum found "
- "for inode %lu start %llu\n",
- inode->i_ino,
+ "for inode %llu start %llu\n",
+ (unsigned long long)
+ btrfs_ino(inode),
(unsigned long long)offset);
}
item = NULL;
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
goto found;
}
btrfs_item_key_to_cpu(path->nodes[0], &found_key,
}
int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end,
- struct list_head *list)
+ struct list_head *list, int search_commit)
{
struct btrfs_key key;
struct btrfs_path *path;
path = btrfs_alloc_path();
BUG_ON(!path);
+ if (search_commit) {
+ path->skip_locking = 1;
+ path->reada = 2;
+ path->search_commit_root = 1;
+ }
+
key.objectid = BTRFS_EXTENT_CSUM_OBJECTID;
key.offset = start;
key.type = BTRFS_EXTENT_CSUM_KEY;
if (key.offset < bytenr)
break;
}
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
}
out:
btrfs_free_path(path);
* at this point, we know the tree has an item, but it isn't big
* enough yet to put our csum in. Grow it
*/
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
ret = btrfs_search_slot(trans, root, &file_key, path,
csum_size, 1);
if (ret < 0)
}
insert:
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
csum_offset = 0;
if (found_next) {
u64 tmp = total_bytes + root->sectorsize;
}
btrfs_mark_buffer_dirty(path->nodes[0]);
if (total_bytes < sums->len) {
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
cond_resched();
goto again;
}
#include <linux/posix_acl.h>
#include <linux/falloc.h>
#include <linux/slab.h>
+#include <linux/ratelimit.h>
#include "compat.h"
#include "ctree.h"
#include "disk-io.h"
#include "compression.h"
#include "locking.h"
#include "free-space-cache.h"
+#include "inode-map.h"
struct btrfs_iget_args {
u64 ino;
unsigned long *nr_written, int unlock);
static int btrfs_init_inode_security(struct btrfs_trans_handle *trans,
- struct inode *inode, struct inode *dir)
+ struct inode *inode, struct inode *dir,
+ const struct qstr *qstr)
{
int err;
err = btrfs_init_acl(trans, inode, dir);
if (!err)
- err = btrfs_xattr_security_init(trans, inode, dir);
+ err = btrfs_xattr_security_init(trans, inode, dir, qstr);
return err;
}
path->leave_spinning = 1;
btrfs_set_trans_block_group(trans, inode);
- key.objectid = inode->i_ino;
+ key.objectid = btrfs_ino(inode);
key.offset = start;
btrfs_set_key_type(&key, BTRFS_EXTENT_DATA_KEY);
datasize = btrfs_file_extent_calc_inline_size(cur_size);
async_extent->start +
async_extent->ram_size - 1, 0);
- em = alloc_extent_map(GFP_NOFS);
+ em = alloc_extent_map();
BUG_ON(!em);
em->start = async_extent->start;
em->len = async_extent->ram_size;
return alloc_hint;
}
+static inline bool is_free_space_inode(struct btrfs_root *root,
+ struct inode *inode)
+{
+ if (root == root->fs_info->tree_root ||
+ BTRFS_I(inode)->location.objectid == BTRFS_FREE_INO_OBJECTID)
+ return true;
+ return false;
+}
+
/*
* when extent_io.c finds a delayed allocation range in the file,
* the call backs end up in this code. The basic idea is to
struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
int ret = 0;
- BUG_ON(root == root->fs_info->tree_root);
+ BUG_ON(is_free_space_inode(root, inode));
trans = btrfs_join_transaction(root, 1);
BUG_ON(IS_ERR(trans));
btrfs_set_trans_block_group(trans, inode);
(u64)-1, &ins, 1);
BUG_ON(ret);
- em = alloc_extent_map(GFP_NOFS);
+ em = alloc_extent_map();
BUG_ON(!em);
em->start = start;
em->orig_start = em->start;
LIST_HEAD(list);
ret = btrfs_lookup_csums_range(root->fs_info->csum_root, bytenr,
- bytenr + num_bytes - 1, &list);
+ bytenr + num_bytes - 1, &list, 0);
if (ret == 0 && list_empty(&list))
return 0;
int type;
int nocow;
int check_prev = 1;
- bool nolock = false;
+ bool nolock;
+ u64 ino = btrfs_ino(inode);
path = btrfs_alloc_path();
BUG_ON(!path);
- if (root == root->fs_info->tree_root) {
- nolock = true;
+
+ nolock = is_free_space_inode(root, inode);
+
+ if (nolock)
trans = btrfs_join_transaction_nolock(root, 1);
- } else {
+ else
trans = btrfs_join_transaction(root, 1);
- }
BUG_ON(IS_ERR(trans));
cow_start = (u64)-1;
cur_offset = start;
while (1) {
- ret = btrfs_lookup_file_extent(trans, root, path, inode->i_ino,
+ ret = btrfs_lookup_file_extent(trans, root, path, ino,
cur_offset, 0);
BUG_ON(ret < 0);
if (ret > 0 && path->slots[0] > 0 && check_prev) {
leaf = path->nodes[0];
btrfs_item_key_to_cpu(leaf, &found_key,
path->slots[0] - 1);
- if (found_key.objectid == inode->i_ino &&
+ if (found_key.objectid == ino &&
found_key.type == BTRFS_EXTENT_DATA_KEY)
path->slots[0]--;
}
num_bytes = 0;
btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
- if (found_key.objectid > inode->i_ino ||
+ if (found_key.objectid > ino ||
found_key.type > BTRFS_EXTENT_DATA_KEY ||
found_key.offset > end)
break;
goto out_check;
if (btrfs_extent_readonly(root, disk_bytenr))
goto out_check;
- if (btrfs_cross_ref_exist(trans, root, inode->i_ino,
+ if (btrfs_cross_ref_exist(trans, root, ino,
found_key.offset -
extent_offset, disk_bytenr))
goto out_check;
goto next_slot;
}
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
if (cow_start != (u64)-1) {
ret = cow_file_range(inode, locked_page, cow_start,
found_key.offset - 1, page_started,
struct extent_map *em;
struct extent_map_tree *em_tree;
em_tree = &BTRFS_I(inode)->extent_tree;
- em = alloc_extent_map(GFP_NOFS);
+ em = alloc_extent_map();
BUG_ON(!em);
em->start = cur_offset;
em->orig_start = em->start;
if (cur_offset > end)
break;
}
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
if (cur_offset <= end && cow_start == (u64)-1)
cow_start = cur_offset;
if (!(state->state & EXTENT_DELALLOC) && (*bits & EXTENT_DELALLOC)) {
struct btrfs_root *root = BTRFS_I(inode)->root;
u64 len = state->end + 1 - state->start;
- int do_list = (root->root_key.objectid !=
- BTRFS_ROOT_TREE_OBJECTID);
+ bool do_list = !is_free_space_inode(root, inode);
if (*bits & EXTENT_FIRST_DELALLOC)
*bits &= ~EXTENT_FIRST_DELALLOC;
if ((state->state & EXTENT_DELALLOC) && (*bits & EXTENT_DELALLOC)) {
struct btrfs_root *root = BTRFS_I(inode)->root;
u64 len = state->end + 1 - state->start;
- int do_list = (root->root_key.objectid !=
- BTRFS_ROOT_TREE_OBJECTID);
+ bool do_list = !is_free_space_inode(root, inode);
if (*bits & EXTENT_FIRST_DELALLOC)
*bits &= ~EXTENT_FIRST_DELALLOC;
skip_sum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM;
- if (root == root->fs_info->tree_root)
+ if (is_free_space_inode(root, inode))
ret = btrfs_bio_wq_end_io(root->fs_info, bio, 2);
else
ret = btrfs_bio_wq_end_io(root->fs_info, bio, 0);
&hint, 0);
BUG_ON(ret);
- ins.objectid = inode->i_ino;
+ ins.objectid = btrfs_ino(inode);
ins.offset = file_pos;
ins.type = BTRFS_EXTENT_DATA_KEY;
ret = btrfs_insert_empty_item(trans, root, path, &ins, sizeof(*fi));
ins.type = BTRFS_EXTENT_ITEM_KEY;
ret = btrfs_alloc_reserved_file_extent(trans, root,
root->root_key.objectid,
- inode->i_ino, file_pos, &ins);
+ btrfs_ino(inode), file_pos, &ins);
BUG_ON(ret);
btrfs_free_path(path);
struct extent_state *cached_state = NULL;
int compress_type = 0;
int ret;
- bool nolock = false;
+ bool nolock;
ret = btrfs_dec_test_ordered_pending(inode, &ordered_extent, start,
end - start + 1);
return 0;
BUG_ON(!ordered_extent);
- nolock = (root == root->fs_info->tree_root);
+ nolock = is_free_space_inode(root, inode);
if (test_bit(BTRFS_ORDERED_NOCOW, &ordered_extent->flags)) {
BUG_ON(!list_empty(&ordered_extent->list));
}
read_unlock(&em_tree->lock);
- if (!em || IS_ERR(em)) {
+ if (IS_ERR_OR_NULL(em)) {
kfree(failrec);
return -EIO;
}
return 0;
zeroit:
- if (printk_ratelimit()) {
- printk(KERN_INFO "btrfs csum failed ino %lu off %llu csum %u "
- "private %llu\n", page->mapping->host->i_ino,
+ printk_ratelimited(KERN_INFO "btrfs csum failed ino %llu off %llu csum %u "
+ "private %llu\n",
+ (unsigned long long)btrfs_ino(page->mapping->host),
(unsigned long long)start, csum,
(unsigned long long)private);
- }
memset(kaddr + offset, 1, end - start + 1);
flush_dcache_page(page);
kunmap_atomic(kaddr, KM_USER0);
/* insert an orphan item to track this unlinked/truncated file */
if (insert >= 1) {
- ret = btrfs_insert_orphan_item(trans, root, inode->i_ino);
+ ret = btrfs_insert_orphan_item(trans, root, btrfs_ino(inode));
BUG_ON(ret);
}
spin_unlock(&root->orphan_lock);
if (trans && delete_item) {
- ret = btrfs_del_orphan_item(trans, root, inode->i_ino);
+ ret = btrfs_del_orphan_item(trans, root, btrfs_ino(inode));
BUG_ON(ret);
}
/*
* if ret == 0 means we found what we were searching for, which
- * is weird, but possible, so only screw with path if we didnt
+ * is weird, but possible, so only screw with path if we didn't
* find the key and see if we have stuff that matches
*/
if (ret > 0) {
break;
/* release the path since we're done with it */
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
/*
* this is where we are basically btrfs_lookup, without the
* try to precache a NULL acl entry for files that don't have
* any xattrs or acls
*/
- maybe_acls = acls_after_inode_item(leaf, path->slots[0], inode->i_ino);
+ maybe_acls = acls_after_inode_item(leaf, path->slots[0],
+ btrfs_ino(inode));
if (!maybe_acls)
cache_no_acl(inode);
struct extent_buffer *leaf;
int ret;
+ /*
+ * If root is tree root, it means this inode is used to
+ * store free space information. And these inodes are updated
+ * when committing the transaction, so they needn't delaye to
+ * be updated, or deadlock will occured.
+ */
+ if (!is_free_space_inode(root, inode)) {
+ ret = btrfs_delayed_update_inode(trans, root, inode);
+ if (!ret)
+ btrfs_set_inode_last_trans(trans, inode);
+ return ret;
+ }
+
path = btrfs_alloc_path();
- BUG_ON(!path);
+ if (!path)
+ return -ENOMEM;
+
path->leave_spinning = 1;
- ret = btrfs_lookup_inode(trans, root, path,
- &BTRFS_I(inode)->location, 1);
+ ret = btrfs_lookup_inode(trans, root, path, &BTRFS_I(inode)->location,
+ 1);
if (ret) {
if (ret > 0)
ret = -ENOENT;
btrfs_unlock_up_safe(path, 1);
leaf = path->nodes[0];
inode_item = btrfs_item_ptr(leaf, path->slots[0],
- struct btrfs_inode_item);
+ struct btrfs_inode_item);
fill_inode_item(trans, leaf, inode_item, inode);
btrfs_mark_buffer_dirty(leaf);
return ret;
}
-
/*
* unlink helper that gets used here in inode.c and in the tree logging
* recovery code. It remove a link in a directory with a given name, and
struct btrfs_dir_item *di;
struct btrfs_key key;
u64 index;
+ u64 ino = btrfs_ino(inode);
+ u64 dir_ino = btrfs_ino(dir);
path = btrfs_alloc_path();
if (!path) {
}
path->leave_spinning = 1;
- di = btrfs_lookup_dir_item(trans, root, path, dir->i_ino,
+ di = btrfs_lookup_dir_item(trans, root, path, dir_ino,
name, name_len, -1);
if (IS_ERR(di)) {
ret = PTR_ERR(di);
ret = btrfs_delete_one_dir_name(trans, root, path, di);
if (ret)
goto err;
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
- ret = btrfs_del_inode_ref(trans, root, name, name_len,
- inode->i_ino,
- dir->i_ino, &index);
+ ret = btrfs_del_inode_ref(trans, root, name, name_len, ino,
+ dir_ino, &index);
if (ret) {
printk(KERN_INFO "btrfs failed to delete reference to %.*s, "
- "inode %lu parent %lu\n", name_len, name,
- inode->i_ino, dir->i_ino);
+ "inode %llu parent %llu\n", name_len, name,
+ (unsigned long long)ino, (unsigned long long)dir_ino);
goto err;
}
- di = btrfs_lookup_dir_index_item(trans, root, path, dir->i_ino,
- index, name, name_len, -1);
- if (IS_ERR(di)) {
- ret = PTR_ERR(di);
- goto err;
- }
- if (!di) {
- ret = -ENOENT;
+ ret = btrfs_delete_delayed_dir_index(trans, root, dir, index);
+ if (ret)
goto err;
- }
- ret = btrfs_delete_one_dir_name(trans, root, path, di);
- btrfs_release_path(root, path);
ret = btrfs_del_inode_ref_in_log(trans, root, name, name_len,
- inode, dir->i_ino);
+ inode, dir_ino);
BUG_ON(ret != 0 && ret != -ENOENT);
ret = btrfs_del_dir_entries_in_log(trans, root, name, name_len,
int check_link = 1;
int err = -ENOSPC;
int ret;
+ u64 ino = btrfs_ino(inode);
+ u64 dir_ino = btrfs_ino(dir);
trans = btrfs_start_transaction(root, 10);
if (!IS_ERR(trans) || PTR_ERR(trans) != -ENOSPC)
return trans;
- if (inode->i_ino == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID)
+ if (ino == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID)
return ERR_PTR(-ENOSPC);
/* check if there is someone else holds reference */
} else {
check_link = 0;
}
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
ret = btrfs_lookup_inode(trans, root, path,
&BTRFS_I(inode)->location, 0);
} else {
check_link = 0;
}
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
if (ret == 0 && S_ISREG(inode->i_mode)) {
ret = btrfs_lookup_file_extent(trans, root, path,
- inode->i_ino, (u64)-1, 0);
+ ino, (u64)-1, 0);
if (ret < 0) {
err = ret;
goto out;
BUG_ON(ret == 0);
if (check_path_shared(root, path))
goto out;
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
}
if (!check_link) {
goto out;
}
- di = btrfs_lookup_dir_item(trans, root, path, dir->i_ino,
+ di = btrfs_lookup_dir_item(trans, root, path, dir_ino,
dentry->d_name.name, dentry->d_name.len, 0);
if (IS_ERR(di)) {
err = PTR_ERR(di);
err = 0;
goto out;
}
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
ref = btrfs_lookup_inode_ref(trans, root, path,
dentry->d_name.name, dentry->d_name.len,
- inode->i_ino, dir->i_ino, 0);
+ ino, dir_ino, 0);
if (IS_ERR(ref)) {
err = PTR_ERR(ref);
goto out;
if (check_path_shared(root, path))
goto out;
index = btrfs_inode_ref_index(path->nodes[0], ref);
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
- di = btrfs_lookup_dir_index_item(trans, root, path, dir->i_ino, index,
+ /*
+ * This is a commit root search, if we can lookup inode item and other
+ * relative items in the commit root, it means the transaction of
+ * dir/file creation has been committed, and the dir index item that we
+ * delay to insert has also been inserted into the commit root. So
+ * we needn't worry about the delayed insertion of the dir index item
+ * here.
+ */
+ di = btrfs_lookup_dir_index_item(trans, root, path, dir_ino, index,
dentry->d_name.name, dentry->d_name.len, 0);
if (IS_ERR(di)) {
err = PTR_ERR(di);
struct btrfs_key key;
u64 index;
int ret;
+ u64 dir_ino = btrfs_ino(dir);
path = btrfs_alloc_path();
if (!path)
return -ENOMEM;
- di = btrfs_lookup_dir_item(trans, root, path, dir->i_ino,
+ di = btrfs_lookup_dir_item(trans, root, path, dir_ino,
name, name_len, -1);
- BUG_ON(!di || IS_ERR(di));
+ BUG_ON(IS_ERR_OR_NULL(di));
leaf = path->nodes[0];
btrfs_dir_item_key_to_cpu(leaf, di, &key);
WARN_ON(key.type != BTRFS_ROOT_ITEM_KEY || key.objectid != objectid);
ret = btrfs_delete_one_dir_name(trans, root, path, di);
BUG_ON(ret);
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
ret = btrfs_del_root_ref(trans, root->fs_info->tree_root,
objectid, root->root_key.objectid,
- dir->i_ino, &index, name, name_len);
+ dir_ino, &index, name, name_len);
if (ret < 0) {
BUG_ON(ret != -ENOENT);
- di = btrfs_search_dir_index_item(root, path, dir->i_ino,
+ di = btrfs_search_dir_index_item(root, path, dir_ino,
name, name_len);
- BUG_ON(!di || IS_ERR(di));
+ BUG_ON(IS_ERR_OR_NULL(di));
leaf = path->nodes[0];
btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
index = key.offset;
}
+ btrfs_release_path(path);
- di = btrfs_lookup_dir_index_item(trans, root, path, dir->i_ino,
- index, name, name_len, -1);
- BUG_ON(!di || IS_ERR(di));
-
- leaf = path->nodes[0];
- btrfs_dir_item_key_to_cpu(leaf, di, &key);
- WARN_ON(key.type != BTRFS_ROOT_ITEM_KEY || key.objectid != objectid);
- ret = btrfs_delete_one_dir_name(trans, root, path, di);
+ ret = btrfs_delete_delayed_dir_index(trans, root, dir, index);
BUG_ON(ret);
- btrfs_release_path(root, path);
btrfs_i_size_write(dir, dir->i_size - name_len * 2);
dir->i_mtime = dir->i_ctime = CURRENT_TIME;
ret = btrfs_update_inode(trans, root, dir);
BUG_ON(ret);
- btrfs_free_path(path);
return 0;
}
unsigned long nr = 0;
if (inode->i_size > BTRFS_EMPTY_DIR_SIZE ||
- inode->i_ino == BTRFS_FIRST_FREE_OBJECTID)
+ btrfs_ino(inode) == BTRFS_FIRST_FREE_OBJECTID)
return -ENOTEMPTY;
trans = __unlink_start_trans(dir, dentry);
btrfs_set_trans_block_group(trans, dir);
- if (unlikely(inode->i_ino == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID)) {
+ if (unlikely(btrfs_ino(inode) == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID)) {
err = btrfs_unlink_subvol(trans, root, dir,
BTRFS_I(inode)->location.objectid,
dentry->d_name.name,
return err;
}
-#if 0
-/*
- * when truncating bytes in a file, it is possible to avoid reading
- * the leaves that contain only checksum items. This can be the
- * majority of the IO required to delete a large file, but it must
- * be done carefully.
- *
- * The keys in the level just above the leaves are checked to make sure
- * the lowest key in a given leaf is a csum key, and starts at an offset
- * after the new size.
- *
- * Then the key for the next leaf is checked to make sure it also has
- * a checksum item for the same file. If it does, we know our target leaf
- * contains only checksum items, and it can be safely freed without reading
- * it.
- *
- * This is just an optimization targeted at large files. It may do
- * nothing. It will return 0 unless things went badly.
- */
-static noinline int drop_csum_leaves(struct btrfs_trans_handle *trans,
- struct btrfs_root *root,
- struct btrfs_path *path,
- struct inode *inode, u64 new_size)
-{
- struct btrfs_key key;
- int ret;
- int nritems;
- struct btrfs_key found_key;
- struct btrfs_key other_key;
- struct btrfs_leaf_ref *ref;
- u64 leaf_gen;
- u64 leaf_start;
-
- path->lowest_level = 1;
- key.objectid = inode->i_ino;
- key.type = BTRFS_CSUM_ITEM_KEY;
- key.offset = new_size;
-again:
- ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
- if (ret < 0)
- goto out;
-
- if (path->nodes[1] == NULL) {
- ret = 0;
- goto out;
- }
- ret = 0;
- btrfs_node_key_to_cpu(path->nodes[1], &found_key, path->slots[1]);
- nritems = btrfs_header_nritems(path->nodes[1]);
-
- if (!nritems)
- goto out;
-
- if (path->slots[1] >= nritems)
- goto next_node;
-
- /* did we find a key greater than anything we want to delete? */
- if (found_key.objectid > inode->i_ino ||
- (found_key.objectid == inode->i_ino && found_key.type > key.type))
- goto out;
-
- /* we check the next key in the node to make sure the leave contains
- * only checksum items. This comparison doesn't work if our
- * leaf is the last one in the node
- */
- if (path->slots[1] + 1 >= nritems) {
-next_node:
- /* search forward from the last key in the node, this
- * will bring us into the next node in the tree
- */
- btrfs_node_key_to_cpu(path->nodes[1], &found_key, nritems - 1);
-
- /* unlikely, but we inc below, so check to be safe */
- if (found_key.offset == (u64)-1)
- goto out;
-
- /* search_forward needs a path with locks held, do the
- * search again for the original key. It is possible
- * this will race with a balance and return a path that
- * we could modify, but this drop is just an optimization
- * and is allowed to miss some leaves.
- */
- btrfs_release_path(root, path);
- found_key.offset++;
-
- /* setup a max key for search_forward */
- other_key.offset = (u64)-1;
- other_key.type = key.type;
- other_key.objectid = key.objectid;
-
- path->keep_locks = 1;
- ret = btrfs_search_forward(root, &found_key, &other_key,
- path, 0, 0);
- path->keep_locks = 0;
- if (ret || found_key.objectid != key.objectid ||
- found_key.type != key.type) {
- ret = 0;
- goto out;
- }
-
- key.offset = found_key.offset;
- btrfs_release_path(root, path);
- cond_resched();
- goto again;
- }
-
- /* we know there's one more slot after us in the tree,
- * read that key so we can verify it is also a checksum item
- */
- btrfs_node_key_to_cpu(path->nodes[1], &other_key, path->slots[1] + 1);
-
- if (found_key.objectid < inode->i_ino)
- goto next_key;
-
- if (found_key.type != key.type || found_key.offset < new_size)
- goto next_key;
-
- /*
- * if the key for the next leaf isn't a csum key from this objectid,
- * we can't be sure there aren't good items inside this leaf.
- * Bail out
- */
- if (other_key.objectid != inode->i_ino || other_key.type != key.type)
- goto out;
-
- leaf_start = btrfs_node_blockptr(path->nodes[1], path->slots[1]);
- leaf_gen = btrfs_node_ptr_generation(path->nodes[1], path->slots[1]);
- /*
- * it is safe to delete this leaf, it contains only
- * csum items from this inode at an offset >= new_size
- */
- ret = btrfs_del_leaf(trans, root, path, leaf_start);
- BUG_ON(ret);
-
- if (root->ref_cows && leaf_gen < trans->transid) {
- ref = btrfs_alloc_leaf_ref(root, 0);
- if (ref) {
- ref->root_gen = root->root_key.offset;
- ref->bytenr = leaf_start;
- ref->owner = 0;
- ref->generation = leaf_gen;
- ref->nritems = 0;
-
- btrfs_sort_leaf_ref(ref);
-
- ret = btrfs_add_leaf_ref(root, ref, 0);
- WARN_ON(ret);
- btrfs_free_leaf_ref(root, ref);
- } else {
- WARN_ON(1);
- }
- }
-next_key:
- btrfs_release_path(root, path);
-
- if (other_key.objectid == inode->i_ino &&
- other_key.type == key.type && other_key.offset > key.offset) {
- key.offset = other_key.offset;
- cond_resched();
- goto again;
- }
- ret = 0;
-out:
- /* fixup any changes we've made to the path */
- path->lowest_level = 0;
- path->keep_locks = 0;
- btrfs_release_path(root, path);
- return ret;
-}
-
-#endif
-
/*
* this can truncate away extent items, csum items and directory items.
* It starts at a high offset and removes keys until it can't find
int encoding;
int ret;
int err = 0;
+ u64 ino = btrfs_ino(inode);
BUG_ON(new_size > 0 && min_type != BTRFS_EXTENT_DATA_KEY);
if (root->ref_cows || root == root->fs_info->tree_root)
btrfs_drop_extent_cache(inode, new_size & (~mask), (u64)-1, 0);
+ /*
+ * This function is also used to drop the items in the log tree before
+ * we relog the inode, so if root != BTRFS_I(inode)->root, it means
+ * it is used to drop the loged items. So we shouldn't kill the delayed
+ * items.
+ */
+ if (min_type == 0 && root == BTRFS_I(inode)->root)
+ btrfs_kill_delayed_inode_items(inode);
+
path = btrfs_alloc_path();
BUG_ON(!path);
path->reada = -1;
- key.objectid = inode->i_ino;
+ key.objectid = ino;
key.offset = (u64)-1;
key.type = (u8)-1;
found_type = btrfs_key_type(&found_key);
encoding = 0;
- if (found_key.objectid != inode->i_ino)
+ if (found_key.objectid != ino)
break;
if (found_type < min_type)
ret = btrfs_free_extent(trans, root, extent_start,
extent_num_bytes, 0,
btrfs_header_owner(leaf),
- inode->i_ino, extent_offset);
+ ino, extent_offset);
BUG_ON(ret);
}
if (path->slots[0] == 0 ||
path->slots[0] != pending_del_slot) {
- if (root->ref_cows) {
+ if (root->ref_cows &&
+ BTRFS_I(inode)->location.objectid !=
+ BTRFS_FREE_INO_OBJECTID) {
err = -EAGAIN;
goto out;
}
BUG_ON(ret);
pending_del_nr = 0;
}
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
goto search_again;
} else {
path->slots[0]--;
while (1) {
em = btrfs_get_extent(inode, NULL, 0, cur_offset,
block_end - cur_offset, 0);
- BUG_ON(IS_ERR(em) || !em);
+ BUG_ON(IS_ERR_OR_NULL(em));
last_byte = min(extent_map_end(em), block_end);
last_byte = (last_byte + mask) & ~mask;
if (!test_bit(EXTENT_FLAG_PREALLOC, &em->flags)) {
break;
err = btrfs_insert_file_extent(trans, root,
- inode->i_ino, cur_offset, 0,
+ btrfs_ino(inode), cur_offset, 0,
0, hole_size, 0, hole_size,
0, 0, 0);
if (err)
truncate_inode_pages(&inode->i_data, 0);
if (inode->i_nlink && (btrfs_root_refs(&root->root_item) != 0 ||
- root == root->fs_info->tree_root))
+ is_free_space_inode(root, inode)))
goto no_delete;
if (is_bad_inode(inode)) {
BUG_ON(ret);
}
+ if (!(root == root->fs_info->tree_root ||
+ root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID))
+ btrfs_return_ino(root, btrfs_ino(inode));
+
nr = trans->blocks_used;
btrfs_end_transaction(trans, root);
btrfs_btree_balance_dirty(root, nr);
path = btrfs_alloc_path();
BUG_ON(!path);
- di = btrfs_lookup_dir_item(NULL, root, path, dir->i_ino, name,
+ di = btrfs_lookup_dir_item(NULL, root, path, btrfs_ino(dir), name,
namelen, 0);
if (IS_ERR(di))
ret = PTR_ERR(di);
- if (!di || IS_ERR(di))
+ if (IS_ERR_OR_NULL(di))
goto out_err;
btrfs_dir_item_key_to_cpu(path->nodes[0], di, location);
leaf = path->nodes[0];
ref = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_root_ref);
- if (btrfs_root_ref_dirid(leaf, ref) != dir->i_ino ||
+ if (btrfs_root_ref_dirid(leaf, ref) != btrfs_ino(dir) ||
btrfs_root_ref_name_len(leaf, ref) != dentry->d_name.len)
goto out;
if (ret)
goto out;
- btrfs_release_path(root->fs_info->tree_root, path);
+ btrfs_release_path(path);
new_root = btrfs_read_fs_root_no_name(root->fs_info, location);
if (IS_ERR(new_root)) {
struct btrfs_inode *entry;
struct rb_node **p;
struct rb_node *parent;
+ u64 ino = btrfs_ino(inode);
again:
p = &root->inode_tree.rb_node;
parent = NULL;
parent = *p;
entry = rb_entry(parent, struct btrfs_inode, rb_node);
- if (inode->i_ino < entry->vfs_inode.i_ino)
+ if (ino < btrfs_ino(&entry->vfs_inode))
p = &parent->rb_left;
- else if (inode->i_ino > entry->vfs_inode.i_ino)
+ else if (ino > btrfs_ino(&entry->vfs_inode))
p = &parent->rb_right;
else {
WARN_ON(!(entry->vfs_inode.i_state &
prev = node;
entry = rb_entry(node, struct btrfs_inode, rb_node);
- if (objectid < entry->vfs_inode.i_ino)
+ if (objectid < btrfs_ino(&entry->vfs_inode))
node = node->rb_left;
- else if (objectid > entry->vfs_inode.i_ino)
+ else if (objectid > btrfs_ino(&entry->vfs_inode))
node = node->rb_right;
else
break;
if (!node) {
while (prev) {
entry = rb_entry(prev, struct btrfs_inode, rb_node);
- if (objectid <= entry->vfs_inode.i_ino) {
+ if (objectid <= btrfs_ino(&entry->vfs_inode)) {
node = prev;
break;
}
}
while (node) {
entry = rb_entry(node, struct btrfs_inode, rb_node);
- objectid = entry->vfs_inode.i_ino + 1;
+ objectid = btrfs_ino(&entry->vfs_inode) + 1;
inode = igrab(&entry->vfs_inode);
if (inode) {
spin_unlock(&root->inode_lock);
static int btrfs_find_actor(struct inode *inode, void *opaque)
{
struct btrfs_iget_args *args = opaque;
- return args->ino == inode->i_ino &&
+ return args->ino == btrfs_ino(inode) &&
args->root == BTRFS_I(inode)->root;
}
return d_splice_alias(inode, dentry);
}
-static unsigned char btrfs_filetype_table[] = {
+unsigned char btrfs_filetype_table[] = {
DT_UNKNOWN, DT_REG, DT_DIR, DT_CHR, DT_BLK, DT_FIFO, DT_SOCK, DT_LNK
};
struct btrfs_key key;
struct btrfs_key found_key;
struct btrfs_path *path;
+ struct list_head ins_list;
+ struct list_head del_list;
int ret;
struct extent_buffer *leaf;
int slot;
char tmp_name[32];
char *name_ptr;
int name_len;
+ int is_curr = 0; /* filp->f_pos points to the current index? */
/* FIXME, use a real flag for deciding about the key type */
if (root->fs_info->tree_root == root)
/* special case for "." */
if (filp->f_pos == 0) {
- over = filldir(dirent, ".", 1,
- 1, inode->i_ino,
- DT_DIR);
+ over = filldir(dirent, ".", 1, 1, btrfs_ino(inode), DT_DIR);
if (over)
return 0;
filp->f_pos = 1;
filp->f_pos = 2;
}
path = btrfs_alloc_path();
+ if (!path)
+ return -ENOMEM;
path->reada = 2;
+ if (key_type == BTRFS_DIR_INDEX_KEY) {
+ INIT_LIST_HEAD(&ins_list);
+ INIT_LIST_HEAD(&del_list);
+ btrfs_get_delayed_items(inode, &ins_list, &del_list);
+ }
+
btrfs_set_key_type(&key, key_type);
key.offset = filp->f_pos;
- key.objectid = inode->i_ino;
+ key.objectid = btrfs_ino(inode);
ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
if (ret < 0)
break;
if (found_key.offset < filp->f_pos)
goto next;
+ if (key_type == BTRFS_DIR_INDEX_KEY &&
+ btrfs_should_delete_dir_index(&del_list,
+ found_key.offset))
+ goto next;
filp->f_pos = found_key.offset;
+ is_curr = 1;
di = btrfs_item_ptr(leaf, slot, struct btrfs_dir_item);
di_cur = 0;
path->slots[0]++;
}
+ if (key_type == BTRFS_DIR_INDEX_KEY) {
+ if (is_curr)
+ filp->f_pos++;
+ ret = btrfs_readdir_delayed_dir_index(filp, dirent, filldir,
+ &ins_list);
+ if (ret)
+ goto nopos;
+ }
+
/* Reached end of directory/root. Bump pos past the last item. */
if (key_type == BTRFS_DIR_INDEX_KEY)
/*
nopos:
ret = 0;
err:
+ if (key_type == BTRFS_DIR_INDEX_KEY)
+ btrfs_put_delayed_items(&ins_list, &del_list);
btrfs_free_path(path);
return ret;
}
return 0;
smp_mb();
- nolock = (root->fs_info->closing && root == root->fs_info->tree_root);
+ if (root->fs_info->closing && is_free_space_inode(root, inode))
+ nolock = true;
if (wbc->sync_mode == WB_SYNC_ALL) {
if (nolock)
btrfs_end_transaction(trans, root);
trans = btrfs_start_transaction(root, 1);
if (IS_ERR(trans)) {
- if (printk_ratelimit()) {
- printk(KERN_ERR "btrfs: fail to "
- "dirty inode %lu error %ld\n",
- inode->i_ino, PTR_ERR(trans));
- }
+ printk_ratelimited(KERN_ERR "btrfs: fail to "
+ "dirty inode %llu error %ld\n",
+ (unsigned long long)btrfs_ino(inode),
+ PTR_ERR(trans));
return;
}
btrfs_set_trans_block_group(trans, inode);
ret = btrfs_update_inode(trans, root, inode);
if (ret) {
- if (printk_ratelimit()) {
- printk(KERN_ERR "btrfs: fail to "
- "dirty inode %lu error %d\n",
- inode->i_ino, ret);
- }
+ printk_ratelimited(KERN_ERR "btrfs: fail to "
+ "dirty inode %llu error %d\n",
+ (unsigned long long)btrfs_ino(inode),
+ ret);
}
}
btrfs_end_transaction(trans, root);
+ if (BTRFS_I(inode)->delayed_node)
+ btrfs_balance_delayed_items(root);
}
/*
struct extent_buffer *leaf;
int ret;
- key.objectid = inode->i_ino;
+ key.objectid = btrfs_ino(inode);
btrfs_set_key_type(&key, BTRFS_DIR_INDEX_KEY);
key.offset = (u64)-1;
leaf = path->nodes[0];
btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
- if (found_key.objectid != inode->i_ino ||
+ if (found_key.objectid != btrfs_ino(inode) ||
btrfs_key_type(&found_key) != BTRFS_DIR_INDEX_KEY) {
BTRFS_I(inode)->index_cnt = 2;
goto out;
int ret = 0;
if (BTRFS_I(dir)->index_cnt == (u64)-1) {
- ret = btrfs_set_inode_index_count(dir);
- if (ret)
- return ret;
+ ret = btrfs_inode_delayed_dir_index_count(dir);
+ if (ret) {
+ ret = btrfs_set_inode_index_count(dir);
+ if (ret)
+ return ret;
+ }
}
*index = BTRFS_I(dir)->index_cnt;
return ERR_PTR(-ENOMEM);
}
+ /*
+ * we have to initialize this early, so we can reclaim the inode
+ * number if we fail afterwards in this function.
+ */
+ inode->i_ino = objectid;
+
if (dir) {
trace_btrfs_inode_request(dir);
goto fail;
inode_init_owner(inode, dir, mode);
- inode->i_ino = objectid;
inode_set_bytes(inode, 0);
inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME;
inode_item = btrfs_item_ptr(path->nodes[0], path->slots[0],
int ret = 0;
struct btrfs_key key;
struct btrfs_root *root = BTRFS_I(parent_inode)->root;
+ u64 ino = btrfs_ino(inode);
+ u64 parent_ino = btrfs_ino(parent_inode);
- if (unlikely(inode->i_ino == BTRFS_FIRST_FREE_OBJECTID)) {
+ if (unlikely(ino == BTRFS_FIRST_FREE_OBJECTID)) {
memcpy(&key, &BTRFS_I(inode)->root->root_key, sizeof(key));
} else {
- key.objectid = inode->i_ino;
+ key.objectid = ino;
btrfs_set_key_type(&key, BTRFS_INODE_ITEM_KEY);
key.offset = 0;
}
- if (unlikely(inode->i_ino == BTRFS_FIRST_FREE_OBJECTID)) {
+ if (unlikely(ino == BTRFS_FIRST_FREE_OBJECTID)) {
ret = btrfs_add_root_ref(trans, root->fs_info->tree_root,
key.objectid, root->root_key.objectid,
- parent_inode->i_ino,
- index, name, name_len);
+ parent_ino, index, name, name_len);
} else if (add_backref) {
- ret = btrfs_insert_inode_ref(trans, root,
- name, name_len, inode->i_ino,
- parent_inode->i_ino, index);
+ ret = btrfs_insert_inode_ref(trans, root, name, name_len, ino,
+ parent_ino, index);
}
if (ret == 0) {
ret = btrfs_insert_dir_item(trans, root, name, name_len,
- parent_inode->i_ino, &key,
+ parent_inode, &key,
btrfs_inode_type(inode), index);
BUG_ON(ret);
if (!new_valid_dev(rdev))
return -EINVAL;
- err = btrfs_find_free_objectid(NULL, root, dir->i_ino, &objectid);
- if (err)
- return err;
-
/*
* 2 for inode item and ref
* 2 for dir items
btrfs_set_trans_block_group(trans, dir);
+ err = btrfs_find_free_ino(root, &objectid);
+ if (err)
+ goto out_unlock;
+
inode = btrfs_new_inode(trans, root, dir, dentry->d_name.name,
- dentry->d_name.len, dir->i_ino, objectid,
+ dentry->d_name.len, btrfs_ino(dir), objectid,
BTRFS_I(dir)->block_group, mode, &index);
if (IS_ERR(inode)) {
err = PTR_ERR(inode);
goto out_unlock;
}
- err = btrfs_init_inode_security(trans, inode, dir);
+ err = btrfs_init_inode_security(trans, inode, dir, &dentry->d_name);
if (err) {
drop_inode = 1;
goto out_unlock;
u64 objectid;
u64 index = 0;
- err = btrfs_find_free_objectid(NULL, root, dir->i_ino, &objectid);
- if (err)
- return err;
/*
* 2 for inode item and ref
* 2 for dir items
btrfs_set_trans_block_group(trans, dir);
+ err = btrfs_find_free_ino(root, &objectid);
+ if (err)
+ goto out_unlock;
+
inode = btrfs_new_inode(trans, root, dir, dentry->d_name.name,
- dentry->d_name.len, dir->i_ino, objectid,
+ dentry->d_name.len, btrfs_ino(dir), objectid,
BTRFS_I(dir)->block_group, mode, &index);
if (IS_ERR(inode)) {
err = PTR_ERR(inode);
goto out_unlock;
}
- err = btrfs_init_inode_security(trans, inode, dir);
+ err = btrfs_init_inode_security(trans, inode, dir, &dentry->d_name);
if (err) {
drop_inode = 1;
goto out_unlock;
int err;
int drop_inode = 0;
- if (inode->i_nlink == 0)
- return -ENOENT;
-
/* do not allow sys_link's with other subvols of the same device */
if (root->objectid != BTRFS_I(inode)->root->objectid)
return -EXDEV;
u64 index = 0;
unsigned long nr = 1;
- err = btrfs_find_free_objectid(NULL, root, dir->i_ino, &objectid);
- if (err)
- return err;
-
/*
* 2 items for inode and ref
* 2 items for dir items
return PTR_ERR(trans);
btrfs_set_trans_block_group(trans, dir);
+ err = btrfs_find_free_ino(root, &objectid);
+ if (err)
+ goto out_fail;
+
inode = btrfs_new_inode(trans, root, dir, dentry->d_name.name,
- dentry->d_name.len, dir->i_ino, objectid,
+ dentry->d_name.len, btrfs_ino(dir), objectid,
BTRFS_I(dir)->block_group, S_IFDIR | mode,
&index);
if (IS_ERR(inode)) {
drop_on_err = 1;
- err = btrfs_init_inode_security(trans, inode, dir);
+ err = btrfs_init_inode_security(trans, inode, dir, &dentry->d_name);
if (err)
goto out_fail;
u64 bytenr;
u64 extent_start = 0;
u64 extent_end = 0;
- u64 objectid = inode->i_ino;
+ u64 objectid = btrfs_ino(inode);
u32 found_type;
struct btrfs_path *path = NULL;
struct btrfs_root *root = BTRFS_I(inode)->root;
else
goto out;
}
- em = alloc_extent_map(GFP_NOFS);
+ em = alloc_extent_map();
if (!em) {
err = -ENOMEM;
goto out;
kunmap(page);
free_extent_map(em);
em = NULL;
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
trans = btrfs_join_transaction(root, 1);
if (IS_ERR(trans))
return ERR_CAST(trans);
em->block_start = EXTENT_MAP_HOLE;
set_bit(EXTENT_FLAG_VACANCY, &em->flags);
insert:
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
if (em->start > start || extent_map_end(em) <= start) {
printk(KERN_ERR "Btrfs: bad extent! em: [%llu %llu] passed "
"[%llu %llu]\n", (unsigned long long)em->start,
u64 hole_start = start;
u64 hole_len = len;
- em = alloc_extent_map(GFP_NOFS);
+ em = alloc_extent_map();
if (!em) {
err = -ENOMEM;
goto out;
}
if (!em) {
- em = alloc_extent_map(GFP_NOFS);
+ em = alloc_extent_map();
if (!em) {
em = ERR_PTR(-ENOMEM);
goto out;
if (!path)
return -ENOMEM;
- ret = btrfs_lookup_file_extent(trans, root, path, inode->i_ino,
+ ret = btrfs_lookup_file_extent(trans, root, path, btrfs_ino(inode),
offset, 0);
if (ret < 0)
goto out;
ret = 0;
leaf = path->nodes[0];
btrfs_item_key_to_cpu(leaf, &key, slot);
- if (key.objectid != inode->i_ino ||
+ if (key.objectid != btrfs_ino(inode) ||
key.type != BTRFS_EXTENT_DATA_KEY) {
/* not our file or wrong item type, must cow */
goto out;
* look for other files referencing this extent, if we
* find any we must cow
*/
- if (btrfs_cross_ref_exist(trans, root, inode->i_ino,
+ if (btrfs_cross_ref_exist(trans, root, btrfs_ino(inode),
key.offset - backref_offset, disk_bytenr))
goto out;
flush_dcache_page(bvec->bv_page);
if (csum != *private) {
- printk(KERN_ERR "btrfs csum failed ino %lu off"
+ printk(KERN_ERR "btrfs csum failed ino %llu off"
" %llu csum %u private %u\n",
- inode->i_ino, (unsigned long long)start,
+ (unsigned long long)btrfs_ino(inode),
+ (unsigned long long)start,
csum, *private);
err = -EIO;
}
struct btrfs_dio_private *dip = bio->bi_private;
if (err) {
- printk(KERN_ERR "btrfs direct IO failed ino %lu rw %lu "
+ printk(KERN_ERR "btrfs direct IO failed ino %llu rw %lu "
"sector %#Lx len %u err no %d\n",
- dip->inode->i_ino, bio->bi_rw,
+ (unsigned long long)btrfs_ino(dip->inode), bio->bi_rw,
(unsigned long long)bio->bi_sector, bio->bi_size, err);
dip->errors = 1;
ei->dummy_inode = 0;
ei->force_compress = BTRFS_COMPRESS_NONE;
+ ei->delayed_node = NULL;
+
inode = &ei->vfs_inode;
- extent_map_tree_init(&ei->extent_tree, GFP_NOFS);
- extent_io_tree_init(&ei->io_tree, &inode->i_data, GFP_NOFS);
- extent_io_tree_init(&ei->io_failure_tree, &inode->i_data, GFP_NOFS);
+ extent_map_tree_init(&ei->extent_tree);
+ extent_io_tree_init(&ei->io_tree, &inode->i_data);
+ extent_io_tree_init(&ei->io_failure_tree, &inode->i_data);
mutex_init(&ei->log_mutex);
btrfs_ordered_inode_tree_init(&ei->ordered_tree);
INIT_LIST_HEAD(&ei->i_orphan);
spin_lock(&root->orphan_lock);
if (!list_empty(&BTRFS_I(inode)->i_orphan)) {
- printk(KERN_INFO "BTRFS: inode %lu still on the orphan list\n",
- inode->i_ino);
+ printk(KERN_INFO "BTRFS: inode %llu still on the orphan list\n",
+ (unsigned long long)btrfs_ino(inode));
list_del_init(&BTRFS_I(inode)->i_orphan);
}
spin_unlock(&root->orphan_lock);
inode_tree_del(inode);
btrfs_drop_extent_cache(inode, 0, (u64)-1, 0);
free:
+ btrfs_remove_delayed_node(inode);
call_rcu(&inode->i_rcu, btrfs_i_callback);
}
struct btrfs_root *root = BTRFS_I(inode)->root;
if (btrfs_root_refs(&root->root_item) == 0 &&
- root != root->fs_info->tree_root)
+ !is_free_space_inode(root, inode))
return 1;
else
return generic_drop_inode(inode);
u64 index = 0;
u64 root_objectid;
int ret;
+ u64 old_ino = btrfs_ino(old_inode);
- if (new_dir->i_ino == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID)
+ if (btrfs_ino(new_dir) == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID)
return -EPERM;
/* we only allow rename subvolume link between subvolumes */
- if (old_inode->i_ino != BTRFS_FIRST_FREE_OBJECTID && root != dest)
+ if (old_ino != BTRFS_FIRST_FREE_OBJECTID && root != dest)
return -EXDEV;
- if (old_inode->i_ino == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID ||
- (new_inode && new_inode->i_ino == BTRFS_FIRST_FREE_OBJECTID))
+ if (old_ino == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID ||
+ (new_inode && btrfs_ino(new_inode) == BTRFS_FIRST_FREE_OBJECTID))
return -ENOTEMPTY;
if (S_ISDIR(old_inode->i_mode) && new_inode &&
filemap_flush(old_inode->i_mapping);
/* close the racy window with snapshot create/destroy ioctl */
- if (old_inode->i_ino == BTRFS_FIRST_FREE_OBJECTID)
+ if (old_ino == BTRFS_FIRST_FREE_OBJECTID)
down_read(&root->fs_info->subvol_sem);
/*
* We want to reserve the absolute worst case amount of items. So if
if (ret)
goto out_fail;
- if (unlikely(old_inode->i_ino == BTRFS_FIRST_FREE_OBJECTID)) {
+ if (unlikely(old_ino == BTRFS_FIRST_FREE_OBJECTID)) {
/* force full log commit if subvolume involved. */
root->fs_info->last_trans_log_full_commit = trans->transid;
} else {
ret = btrfs_insert_inode_ref(trans, dest,
new_dentry->d_name.name,
new_dentry->d_name.len,
- old_inode->i_ino,
- new_dir->i_ino, index);
+ old_ino,
+ btrfs_ino(new_dir), index);
if (ret)
goto out_fail;
/*
* make sure the inode gets flushed if it is replacing
* something.
*/
- if (new_inode && new_inode->i_size &&
- old_inode && S_ISREG(old_inode->i_mode)) {
+ if (new_inode && new_inode->i_size && S_ISREG(old_inode->i_mode))
btrfs_add_ordered_operation(trans, root, old_inode);
- }
old_dir->i_ctime = old_dir->i_mtime = ctime;
new_dir->i_ctime = new_dir->i_mtime = ctime;
if (old_dentry->d_parent != new_dentry->d_parent)
btrfs_record_unlink_dir(trans, old_dir, old_inode, 1);
- if (unlikely(old_inode->i_ino == BTRFS_FIRST_FREE_OBJECTID)) {
+ if (unlikely(old_ino == BTRFS_FIRST_FREE_OBJECTID)) {
root_objectid = BTRFS_I(old_inode)->root->root_key.objectid;
ret = btrfs_unlink_subvol(trans, root, old_dir, root_objectid,
old_dentry->d_name.name,
if (new_inode) {
new_inode->i_ctime = CURRENT_TIME;
- if (unlikely(new_inode->i_ino ==
+ if (unlikely(btrfs_ino(new_inode) ==
BTRFS_EMPTY_SUBVOL_DIR_OBJECTID)) {
root_objectid = BTRFS_I(new_inode)->location.objectid;
ret = btrfs_unlink_subvol(trans, dest, new_dir,
new_dentry->d_name.len, 0, index);
BUG_ON(ret);
- if (old_inode->i_ino != BTRFS_FIRST_FREE_OBJECTID) {
+ if (old_ino != BTRFS_FIRST_FREE_OBJECTID) {
struct dentry *parent = dget_parent(new_dentry);
btrfs_log_new_name(trans, old_inode, old_dir, parent);
dput(parent);
out_fail:
btrfs_end_transaction_throttle(trans, root);
out_notrans:
- if (old_inode->i_ino == BTRFS_FIRST_FREE_OBJECTID)
+ if (old_ino == BTRFS_FIRST_FREE_OBJECTID)
up_read(&root->fs_info->subvol_sem);
return ret;
return 0;
}
-int btrfs_start_one_delalloc_inode(struct btrfs_root *root, int delay_iput,
- int sync)
-{
- struct btrfs_inode *binode;
- struct inode *inode = NULL;
-
- spin_lock(&root->fs_info->delalloc_lock);
- while (!list_empty(&root->fs_info->delalloc_inodes)) {
- binode = list_entry(root->fs_info->delalloc_inodes.next,
- struct btrfs_inode, delalloc_inodes);
- inode = igrab(&binode->vfs_inode);
- if (inode) {
- list_move_tail(&binode->delalloc_inodes,
- &root->fs_info->delalloc_inodes);
- break;
- }
-
- list_del_init(&binode->delalloc_inodes);
- cond_resched_lock(&root->fs_info->delalloc_lock);
- }
- spin_unlock(&root->fs_info->delalloc_lock);
-
- if (inode) {
- if (sync) {
- filemap_write_and_wait(inode->i_mapping);
- /*
- * We have to do this because compression doesn't
- * actually set PG_writeback until it submits the pages
- * for IO, which happens in an async thread, so we could
- * race and not actually wait for any writeback pages
- * because they've not been submitted yet. Technically
- * this could still be the case for the ordered stuff
- * since the async thread may not have started to do its
- * work yet. If this becomes the case then we need to
- * figure out a way to make sure that in writepage we
- * wait for any async pages to be submitted before
- * returning so that fdatawait does what its supposed to
- * do.
- */
- btrfs_wait_ordered_range(inode, 0, (u64)-1);
- } else {
- filemap_flush(inode->i_mapping);
- }
- if (delay_iput)
- btrfs_add_delayed_iput(inode);
- else
- iput(inode);
- return 1;
- }
- return 0;
-}
-
static int btrfs_symlink(struct inode *dir, struct dentry *dentry,
const char *symname)
{
if (name_len > BTRFS_MAX_INLINE_DATA_SIZE(root))
return -ENAMETOOLONG;
- err = btrfs_find_free_objectid(NULL, root, dir->i_ino, &objectid);
- if (err)
- return err;
/*
* 2 items for inode item and ref
* 2 items for dir items
btrfs_set_trans_block_group(trans, dir);
+ err = btrfs_find_free_ino(root, &objectid);
+ if (err)
+ goto out_unlock;
+
inode = btrfs_new_inode(trans, root, dir, dentry->d_name.name,
- dentry->d_name.len, dir->i_ino, objectid,
+ dentry->d_name.len, btrfs_ino(dir), objectid,
BTRFS_I(dir)->block_group, S_IFLNK|S_IRWXUGO,
&index);
if (IS_ERR(inode)) {
goto out_unlock;
}
- err = btrfs_init_inode_security(trans, inode, dir);
+ err = btrfs_init_inode_security(trans, inode, dir, &dentry->d_name);
if (err) {
drop_inode = 1;
goto out_unlock;
path = btrfs_alloc_path();
BUG_ON(!path);
- key.objectid = inode->i_ino;
+ key.objectid = btrfs_ino(inode);
key.offset = 0;
btrfs_set_key_type(&key, BTRFS_EXTENT_DATA_KEY);
datasize = btrfs_file_extent_calc_inline_size(name_len);
.writepage = btrfs_writepage,
.writepages = btrfs_writepages,
.readpages = btrfs_readpages,
- .sync_page = block_sync_page,
.direct_IO = btrfs_direct_IO,
.invalidatepage = btrfs_invalidatepage,
.releasepage = btrfs_releasepage,
#include "print-tree.h"
#include "volumes.h"
#include "locking.h"
+#include "inode-map.h"
/* Mask out flags that are inappropriate for the given type of inode. */
static inline __u32 btrfs_mask_flags(umode_t mode, __u32 flags)
iflags |= FS_NOATIME_FL;
if (flags & BTRFS_INODE_DIRSYNC)
iflags |= FS_DIRSYNC_FL;
+ if (flags & BTRFS_INODE_NODATACOW)
+ iflags |= FS_NOCOW_FL;
+
+ if ((flags & BTRFS_INODE_COMPRESS) && !(flags & BTRFS_INODE_NOCOMPRESS))
+ iflags |= FS_COMPR_FL;
+ else if (flags & BTRFS_INODE_NOCOMPRESS)
+ iflags |= FS_NOCOMP_FL;
return iflags;
}
if (flags & ~(FS_IMMUTABLE_FL | FS_APPEND_FL | \
FS_NOATIME_FL | FS_NODUMP_FL | \
FS_SYNC_FL | FS_DIRSYNC_FL | \
- FS_NOCOMP_FL | FS_COMPR_FL | \
- FS_NOCOW_FL | FS_COW_FL))
+ FS_NOCOMP_FL | FS_COMPR_FL |
+ FS_NOCOW_FL))
return -EOPNOTSUPP;
if ((flags & FS_NOCOMP_FL) && (flags & FS_COMPR_FL))
return -EINVAL;
- if ((flags & FS_NOCOW_FL) && (flags & FS_COW_FL))
- return -EINVAL;
-
return 0;
}
if (ret)
return ret;
- if (!is_owner_or_cap(inode))
+ if (!inode_owner_or_capable(inode))
return -EACCES;
mutex_lock(&inode->i_mutex);
ip->flags |= BTRFS_INODE_DIRSYNC;
else
ip->flags &= ~BTRFS_INODE_DIRSYNC;
+ if (flags & FS_NOCOW_FL)
+ ip->flags |= BTRFS_INODE_NODATACOW;
+ else
+ ip->flags &= ~BTRFS_INODE_NODATACOW;
/*
* The COMPRESS flag can only be changed by users, while the NOCOMPRESS
} else if (flags & FS_COMPR_FL) {
ip->flags |= BTRFS_INODE_COMPRESS;
ip->flags &= ~BTRFS_INODE_NOCOMPRESS;
+ } else {
+ ip->flags &= ~(BTRFS_INODE_COMPRESS | BTRFS_INODE_NOCOMPRESS);
}
- if (flags & FS_NOCOW_FL)
- ip->flags |= BTRFS_INODE_NODATACOW;
- else if (flags & FS_COW_FL)
- ip->flags &= ~BTRFS_INODE_NODATACOW;
trans = btrfs_join_transaction(root, 1);
BUG_ON(IS_ERR(trans));
u64 new_dirid = BTRFS_FIRST_FREE_OBJECTID;
u64 index = 0;
- ret = btrfs_find_free_objectid(NULL, root->fs_info->tree_root,
- 0, &objectid);
+ ret = btrfs_find_free_objectid(root->fs_info->tree_root, &objectid);
if (ret) {
dput(parent);
return ret;
BUG_ON(ret);
ret = btrfs_insert_dir_item(trans, root,
- name, namelen, dir->i_ino, &key,
+ name, namelen, dir, &key,
BTRFS_FT_DIR, index);
if (ret)
goto fail;
ret = btrfs_add_root_ref(trans, root->fs_info->tree_root,
objectid, root->root_key.objectid,
- dir->i_ino, index, name, namelen);
+ btrfs_ino(dir), index, name, namelen);
BUG_ON(ret);
int ret = 0;
u64 flags = 0;
- if (inode->i_ino != BTRFS_FIRST_FREE_OBJECTID)
+ if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID)
return -EINVAL;
down_read(&root->fs_info->subvol_sem);
if (root->fs_info->sb->s_flags & MS_RDONLY)
return -EROFS;
- if (inode->i_ino != BTRFS_FIRST_FREE_OBJECTID)
+ if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID)
return -EINVAL;
if (copy_from_user(&flags, arg, sizeof(flags)))
if (flags & ~BTRFS_SUBVOL_RDONLY)
return -EOPNOTSUPP;
- if (!is_owner_or_cap(inode))
+ if (!inode_owner_or_capable(inode))
return -EACCES;
down_write(&root->fs_info->subvol_sem);
}
ret = copy_to_sk(root, path, &key, sk, args->buf,
&sk_offset, &num_found);
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
if (ret || num_found >= sk->nr_items)
break;
if (key.offset == BTRFS_FIRST_FREE_OBJECTID)
break;
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
key.objectid = key.offset;
key.offset = (u64)-1;
dirid = key.objectid;
goto out_dput;
}
- if (inode->i_ino != BTRFS_FIRST_FREE_OBJECTID) {
+ if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID) {
err = -EINVAL;
goto out_dput;
}
return ret;
}
+ static long btrfs_ioctl_fs_info(struct btrfs_root *root, void __user *arg)
+ {
+ struct btrfs_ioctl_fs_info_args fi_args;
+ struct btrfs_device *device;
+ struct btrfs_device *next;
+ struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
+
+ if (!capable(CAP_SYS_ADMIN))
+ return -EPERM;
+
+ fi_args.num_devices = fs_devices->num_devices;
+ fi_args.max_id = 0;
+ memcpy(&fi_args.fsid, root->fs_info->fsid, sizeof(fi_args.fsid));
+
+ mutex_lock(&fs_devices->device_list_mutex);
+ list_for_each_entry_safe(device, next, &fs_devices->devices, dev_list) {
+ if (device->devid > fi_args.max_id)
+ fi_args.max_id = device->devid;
+ }
+ mutex_unlock(&fs_devices->device_list_mutex);
+
+ if (copy_to_user(arg, &fi_args, sizeof(fi_args)))
+ return -EFAULT;
+
+ return 0;
+ }
+
+ static long btrfs_ioctl_dev_info(struct btrfs_root *root, void __user *arg)
+ {
+ struct btrfs_ioctl_dev_info_args *di_args;
+ struct btrfs_device *dev;
+ struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
+ int ret = 0;
+ char *s_uuid = NULL;
+ char empty_uuid[BTRFS_UUID_SIZE] = {0};
+
+ if (!capable(CAP_SYS_ADMIN))
+ return -EPERM;
+
+ di_args = memdup_user(arg, sizeof(*di_args));
+ if (IS_ERR(di_args))
+ return PTR_ERR(di_args);
+
+ if (memcmp(empty_uuid, di_args->uuid, BTRFS_UUID_SIZE) != 0)
+ s_uuid = di_args->uuid;
+
+ mutex_lock(&fs_devices->device_list_mutex);
+ dev = btrfs_find_device(root, di_args->devid, s_uuid, NULL);
+ mutex_unlock(&fs_devices->device_list_mutex);
+
+ if (!dev) {
+ ret = -ENODEV;
+ goto out;
+ }
+
+ di_args->devid = dev->devid;
+ di_args->bytes_used = dev->bytes_used;
+ di_args->total_bytes = dev->total_bytes;
+ memcpy(di_args->uuid, dev->uuid, sizeof(di_args->uuid));
+ strncpy(di_args->path, dev->name, sizeof(di_args->path));
+
+ out:
+ if (ret == 0 && copy_to_user(arg, di_args, sizeof(*di_args)))
+ ret = -EFAULT;
+
+ kfree(di_args);
+ return ret;
+ }
+
static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd,
u64 off, u64 olen, u64 destoff)
{
}
/* clone data */
- key.objectid = src->i_ino;
+ key.objectid = btrfs_ino(src);
key.type = BTRFS_EXTENT_DATA_KEY;
key.offset = 0;
btrfs_item_key_to_cpu(leaf, &key, slot);
if (btrfs_key_type(&key) > BTRFS_EXTENT_DATA_KEY ||
- key.objectid != src->i_ino)
+ key.objectid != btrfs_ino(src))
break;
if (btrfs_key_type(&key) == BTRFS_EXTENT_DATA_KEY) {
datal = btrfs_file_extent_ram_bytes(leaf,
extent);
}
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
if (key.offset + datal <= off ||
key.offset >= off+len)
goto next;
memcpy(&new_key, &key, sizeof(new_key));
- new_key.objectid = inode->i_ino;
+ new_key.objectid = btrfs_ino(inode);
if (off <= key.offset)
new_key.offset = key.offset + destoff - off;
else
ret = btrfs_inc_extent_ref(trans, root,
disko, diskl, 0,
root->root_key.objectid,
- inode->i_ino,
+ btrfs_ino(inode),
new_key.offset - datao);
BUG_ON(ret);
}
}
btrfs_mark_buffer_dirty(leaf);
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
inode->i_mtime = inode->i_ctime = CURRENT_TIME;
btrfs_end_transaction(trans, root);
}
next:
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
key.offset++;
}
ret = 0;
out:
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
unlock_extent(&BTRFS_I(src)->io_tree, off, off+len, GFP_NOFS);
out_unlock:
mutex_unlock(&src->i_mutex);
return btrfs_wait_for_commit(root, transid);
}
+ static long btrfs_ioctl_scrub(struct btrfs_root *root, void __user *arg)
+ {
+ int ret;
+ struct btrfs_ioctl_scrub_args *sa;
+
+ if (!capable(CAP_SYS_ADMIN))
+ return -EPERM;
+
+ sa = memdup_user(arg, sizeof(*sa));
+ if (IS_ERR(sa))
+ return PTR_ERR(sa);
+
+ ret = btrfs_scrub_dev(root, sa->devid, sa->start, sa->end,
+ &sa->progress, sa->flags & BTRFS_SCRUB_READONLY);
+
+ if (copy_to_user(arg, sa, sizeof(*sa)))
+ ret = -EFAULT;
+
+ kfree(sa);
+ return ret;
+ }
+
+ static long btrfs_ioctl_scrub_cancel(struct btrfs_root *root, void __user *arg)
+ {
+ if (!capable(CAP_SYS_ADMIN))
+ return -EPERM;
+
+ return btrfs_scrub_cancel(root);
+ }
+
+ static long btrfs_ioctl_scrub_progress(struct btrfs_root *root,
+ void __user *arg)
+ {
+ struct btrfs_ioctl_scrub_args *sa;
+ int ret;
+
+ if (!capable(CAP_SYS_ADMIN))
+ return -EPERM;
+
+ sa = memdup_user(arg, sizeof(*sa));
+ if (IS_ERR(sa))
+ return PTR_ERR(sa);
+
+ ret = btrfs_scrub_progress(root, sa->devid, &sa->progress);
+
+ if (copy_to_user(arg, sa, sizeof(*sa)))
+ ret = -EFAULT;
+
+ kfree(sa);
+ return ret;
+ }
+
long btrfs_ioctl(struct file *file, unsigned int
cmd, unsigned long arg)
{
return btrfs_ioctl_add_dev(root, argp);
case BTRFS_IOC_RM_DEV:
return btrfs_ioctl_rm_dev(root, argp);
+ case BTRFS_IOC_FS_INFO:
+ return btrfs_ioctl_fs_info(root, argp);
+ case BTRFS_IOC_DEV_INFO:
+ return btrfs_ioctl_dev_info(root, argp);
case BTRFS_IOC_BALANCE:
return btrfs_balance(root->fs_info->dev_root);
case BTRFS_IOC_CLONE:
return btrfs_ioctl_start_sync(file, argp);
case BTRFS_IOC_WAIT_SYNC:
return btrfs_ioctl_wait_sync(file, argp);
+ case BTRFS_IOC_SCRUB:
+ return btrfs_ioctl_scrub(root, argp);
+ case BTRFS_IOC_SCRUB_CANCEL:
+ return btrfs_ioctl_scrub_cancel(root, argp);
+ case BTRFS_IOC_SCRUB_PROGRESS:
+ return btrfs_ioctl_scrub_progress(root, argp);
}
return -ENOTTY;
#include "btrfs_inode.h"
#include "async-thread.h"
#include "free-space-cache.h"
+#include "inode-map.h"
/*
* backref_node, mapping_node and tree_block start with this
return 1;
}
+
static int should_ignore_root(struct btrfs_root *root)
{
struct btrfs_root *reloc_root;
*/
return 1;
}
-
/*
* find reloc tree by address of tree root
*/
lower = upper;
upper = NULL;
}
- btrfs_release_path(root, path2);
+ btrfs_release_path(path2);
next:
if (ptr < end) {
ptr += btrfs_extent_inline_ref_size(key.type);
if (ptr >= end)
path1->slots[0]++;
}
- btrfs_release_path(rc->extent_root, path1);
+ btrfs_release_path(path1);
cur->checked = 1;
WARN_ON(exist);
prev = node;
entry = rb_entry(node, struct btrfs_inode, rb_node);
- if (objectid < entry->vfs_inode.i_ino)
+ if (objectid < btrfs_ino(&entry->vfs_inode))
node = node->rb_left;
- else if (objectid > entry->vfs_inode.i_ino)
+ else if (objectid > btrfs_ino(&entry->vfs_inode))
node = node->rb_right;
else
break;
if (!node) {
while (prev) {
entry = rb_entry(prev, struct btrfs_inode, rb_node);
- if (objectid <= entry->vfs_inode.i_ino) {
+ if (objectid <= btrfs_ino(&entry->vfs_inode)) {
node = prev;
break;
}
return inode;
}
- objectid = entry->vfs_inode.i_ino + 1;
+ objectid = btrfs_ino(&entry->vfs_inode) + 1;
if (cond_resched_lock(&root->inode_lock))
goto again;
return -ENOMEM;
bytenr -= BTRFS_I(reloc_inode)->index_cnt;
- ret = btrfs_lookup_file_extent(NULL, root, path, reloc_inode->i_ino,
+ ret = btrfs_lookup_file_extent(NULL, root, path, btrfs_ino(reloc_inode),
bytenr, 0);
if (ret < 0)
goto out;
if (first) {
inode = find_next_inode(root, key.objectid);
first = 0;
- } else if (inode && inode->i_ino < key.objectid) {
+ } else if (inode && btrfs_ino(inode) < key.objectid) {
btrfs_add_delayed_iput(inode);
inode = find_next_inode(root, key.objectid);
}
- if (inode && inode->i_ino == key.objectid) {
+ if (inode && btrfs_ino(inode) == key.objectid) {
end = key.offset +
btrfs_file_extent_num_bytes(leaf, fi);
WARN_ON(!IS_ALIGNED(key.offset,
btrfs_node_key_to_cpu(path->nodes[level], &key,
path->slots[level]);
- btrfs_release_path(src, path);
+ btrfs_release_path(path);
path->lowest_level = level;
ret = btrfs_search_slot(trans, src, &key, path, 0, 1);
struct inode *inode = NULL;
u64 objectid;
u64 start, end;
+ u64 ino;
objectid = min_key->objectid;
while (1) {
inode = find_next_inode(root, objectid);
if (!inode)
break;
+ ino = btrfs_ino(inode);
- if (inode->i_ino > max_key->objectid) {
+ if (ino > max_key->objectid) {
iput(inode);
break;
}
- objectid = inode->i_ino + 1;
+ objectid = ino + 1;
if (!S_ISREG(inode->i_mode))
continue;
- if (unlikely(min_key->objectid == inode->i_ino)) {
+ if (unlikely(min_key->objectid == ino)) {
if (min_key->type > BTRFS_EXTENT_DATA_KEY)
continue;
if (min_key->type < BTRFS_EXTENT_DATA_KEY)
start = 0;
}
- if (unlikely(max_key->objectid == inode->i_ino)) {
+ if (unlikely(max_key->objectid == ino)) {
if (max_key->type < BTRFS_EXTENT_DATA_KEY)
continue;
if (max_key->type > BTRFS_EXTENT_DATA_KEY) {
root = next->root;
BUG_ON(!root);
- /* no other choice for non-refernce counted tree */
+ /* no other choice for non-references counted tree */
if (!root->ref_cows)
return root;
path->locks[upper->level] = 0;
slot = path->slots[upper->level];
- btrfs_release_path(NULL, path);
+ btrfs_release_path(path);
} else {
ret = btrfs_bin_search(upper->eb, key, upper->level,
&slot);
} else {
path->lowest_level = node->level;
ret = btrfs_search_slot(trans, root, key, path, 0, 1);
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
if (ret > 0)
ret = 0;
}
struct extent_map *em;
int ret = 0;
- em = alloc_extent_map(GFP_NOFS);
+ em = alloc_extent_map();
if (!em)
return -ENOMEM;
#endif
}
- btrfs_release_path(rc->extent_root, path);
+ btrfs_release_path(path);
BUG_ON(level == -1);
key.offset = 0;
inode = btrfs_iget(fs_info->sb, &key, root, NULL);
- if (!inode || IS_ERR(inode) || is_bad_inode(inode)) {
+ if (IS_ERR_OR_NULL(inode) || is_bad_inode(inode)) {
if (inode && !IS_ERR(inode))
iput(inode);
return -ENOENT;
}
path->slots[0]++;
}
- btrfs_release_path(rc->extent_root, path);
+ btrfs_release_path(path);
if (err)
free_block_list(blocks);
return err;
EXTENT_DIRTY);
if (ret == 0 && start <= key.objectid) {
- btrfs_release_path(rc->extent_root, path);
+ btrfs_release_path(path);
rc->search_start = end + 1;
} else {
rc->search_start = key.objectid + key.offset;
return 0;
}
}
- btrfs_release_path(rc->extent_root, path);
+ btrfs_release_path(path);
return ret;
}
flags = BTRFS_EXTENT_FLAG_DATA;
if (path_change) {
- btrfs_release_path(rc->extent_root, path);
+ btrfs_release_path(path);
path->search_commit_root = 1;
path->skip_locking = 1;
(flags & BTRFS_EXTENT_FLAG_DATA)) {
ret = add_data_references(rc, &key, path, &blocks);
} else {
- btrfs_release_path(rc->extent_root, path);
+ btrfs_release_path(path);
ret = 0;
}
if (ret < 0) {
}
}
- btrfs_release_path(rc->extent_root, path);
+ btrfs_release_path(path);
clear_extent_bits(&rc->processed_blocks, 0, (u64)-1, EXTENT_DIRTY,
GFP_NOFS);
btrfs_set_inode_flags(leaf, item, BTRFS_INODE_NOCOMPRESS |
BTRFS_INODE_PREALLOC);
btrfs_mark_buffer_dirty(leaf);
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
out:
btrfs_free_path(path);
return ret;
if (IS_ERR(trans))
return ERR_CAST(trans);
- err = btrfs_find_free_objectid(trans, root, objectid, &objectid);
+ err = btrfs_find_free_objectid(root, &objectid);
if (err)
goto out;
INIT_LIST_HEAD(&rc->reloc_roots);
backref_cache_init(&rc->backref_cache);
mapping_tree_init(&rc->reloc_root_tree);
- extent_io_tree_init(&rc->processed_blocks, NULL, GFP_NOFS);
+ extent_io_tree_init(&rc->processed_blocks, NULL);
return rc;
}
}
leaf = path->nodes[0];
btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
- btrfs_release_path(root->fs_info->tree_root, path);
+ btrfs_release_path(path);
if (key.objectid != BTRFS_TREE_RELOC_OBJECTID ||
key.type != BTRFS_ROOT_ITEM_KEY)
key.offset--;
}
- btrfs_release_path(root->fs_info->tree_root, path);
+ btrfs_release_path(path);
if (list_empty(&reloc_roots))
goto out;
disk_bytenr = file_pos + BTRFS_I(inode)->index_cnt;
ret = btrfs_lookup_csums_range(root->fs_info->csum_root, disk_bytenr,
- disk_bytenr + len - 1, &list);
+ disk_bytenr + len - 1, &list, 0);
while (!list_empty(&list)) {
sums = list_entry(list.next, struct btrfs_ordered_sum, list);
--- /dev/null
- rw |= REQ_SYNC | REQ_UNPLUG;
+ /*
+ * Copyright (C) 2011 STRATO. All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public
+ * License v2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public
+ * License along with this program; if not, write to the
+ * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+ * Boston, MA 021110-1307, USA.
+ */
+
+ #include <linux/sched.h>
+ #include <linux/pagemap.h>
+ #include <linux/writeback.h>
+ #include <linux/blkdev.h>
+ #include <linux/rbtree.h>
+ #include <linux/slab.h>
+ #include <linux/workqueue.h>
+ #include "ctree.h"
+ #include "volumes.h"
+ #include "disk-io.h"
+ #include "ordered-data.h"
+
+ /*
+ * This is only the first step towards a full-features scrub. It reads all
+ * extent and super block and verifies the checksums. In case a bad checksum
+ * is found or the extent cannot be read, good data will be written back if
+ * any can be found.
+ *
+ * Future enhancements:
+ * - To enhance the performance, better read-ahead strategies for the
+ * extent-tree can be employed.
+ * - In case an unrepairable extent is encountered, track which files are
+ * affected and report them
+ * - In case of a read error on files with nodatasum, map the file and read
+ * the extent to trigger a writeback of the good copy
+ * - track and record media errors, throw out bad devices
+ * - add a mode to also read unallocated space
+ * - make the prefetch cancellable
+ */
+
+ struct scrub_bio;
+ struct scrub_page;
+ struct scrub_dev;
+ static void scrub_bio_end_io(struct bio *bio, int err);
+ static void scrub_checksum(struct btrfs_work *work);
+ static int scrub_checksum_data(struct scrub_dev *sdev,
+ struct scrub_page *spag, void *buffer);
+ static int scrub_checksum_tree_block(struct scrub_dev *sdev,
+ struct scrub_page *spag, u64 logical,
+ void *buffer);
+ static int scrub_checksum_super(struct scrub_bio *sbio, void *buffer);
+ static int scrub_fixup_check(struct scrub_bio *sbio, int ix);
+ static void scrub_fixup_end_io(struct bio *bio, int err);
+ static int scrub_fixup_io(int rw, struct block_device *bdev, sector_t sector,
+ struct page *page);
+ static void scrub_fixup(struct scrub_bio *sbio, int ix);
+
+ #define SCRUB_PAGES_PER_BIO 16 /* 64k per bio */
+ #define SCRUB_BIOS_PER_DEV 16 /* 1 MB per device in flight */
+
+ struct scrub_page {
+ u64 flags; /* extent flags */
+ u64 generation;
+ u64 mirror_num;
+ int have_csum;
+ u8 csum[BTRFS_CSUM_SIZE];
+ };
+
+ struct scrub_bio {
+ int index;
+ struct scrub_dev *sdev;
+ struct bio *bio;
+ int err;
+ u64 logical;
+ u64 physical;
+ struct scrub_page spag[SCRUB_PAGES_PER_BIO];
+ u64 count;
+ int next_free;
+ struct btrfs_work work;
+ };
+
+ struct scrub_dev {
+ struct scrub_bio *bios[SCRUB_BIOS_PER_DEV];
+ struct btrfs_device *dev;
+ int first_free;
+ int curr;
+ atomic_t in_flight;
+ spinlock_t list_lock;
+ wait_queue_head_t list_wait;
+ u16 csum_size;
+ struct list_head csum_list;
+ atomic_t cancel_req;
+ int readonly;
+ /*
+ * statistics
+ */
+ struct btrfs_scrub_progress stat;
+ spinlock_t stat_lock;
+ };
+
+ static void scrub_free_csums(struct scrub_dev *sdev)
+ {
+ while (!list_empty(&sdev->csum_list)) {
+ struct btrfs_ordered_sum *sum;
+ sum = list_first_entry(&sdev->csum_list,
+ struct btrfs_ordered_sum, list);
+ list_del(&sum->list);
+ kfree(sum);
+ }
+ }
+
+ static noinline_for_stack void scrub_free_dev(struct scrub_dev *sdev)
+ {
+ int i;
+ int j;
+ struct page *last_page;
+
+ if (!sdev)
+ return;
+
+ for (i = 0; i < SCRUB_BIOS_PER_DEV; ++i) {
+ struct scrub_bio *sbio = sdev->bios[i];
+ struct bio *bio;
+
+ if (!sbio)
+ break;
+
+ bio = sbio->bio;
+ if (bio) {
+ last_page = NULL;
+ for (j = 0; j < bio->bi_vcnt; ++j) {
+ if (bio->bi_io_vec[j].bv_page == last_page)
+ continue;
+ last_page = bio->bi_io_vec[j].bv_page;
+ __free_page(last_page);
+ }
+ bio_put(bio);
+ }
+ kfree(sbio);
+ }
+
+ scrub_free_csums(sdev);
+ kfree(sdev);
+ }
+
+ static noinline_for_stack
+ struct scrub_dev *scrub_setup_dev(struct btrfs_device *dev)
+ {
+ struct scrub_dev *sdev;
+ int i;
+ int j;
+ int ret;
+ struct btrfs_fs_info *fs_info = dev->dev_root->fs_info;
+
+ sdev = kzalloc(sizeof(*sdev), GFP_NOFS);
+ if (!sdev)
+ goto nomem;
+ sdev->dev = dev;
+ for (i = 0; i < SCRUB_BIOS_PER_DEV; ++i) {
+ struct bio *bio;
+ struct scrub_bio *sbio;
+
+ sbio = kzalloc(sizeof(*sbio), GFP_NOFS);
+ if (!sbio)
+ goto nomem;
+ sdev->bios[i] = sbio;
+
+ bio = bio_kmalloc(GFP_NOFS, SCRUB_PAGES_PER_BIO);
+ if (!bio)
+ goto nomem;
+
+ sbio->index = i;
+ sbio->sdev = sdev;
+ sbio->bio = bio;
+ sbio->count = 0;
+ sbio->work.func = scrub_checksum;
+ bio->bi_private = sdev->bios[i];
+ bio->bi_end_io = scrub_bio_end_io;
+ bio->bi_sector = 0;
+ bio->bi_bdev = dev->bdev;
+ bio->bi_size = 0;
+
+ for (j = 0; j < SCRUB_PAGES_PER_BIO; ++j) {
+ struct page *page;
+ page = alloc_page(GFP_NOFS);
+ if (!page)
+ goto nomem;
+
+ ret = bio_add_page(bio, page, PAGE_SIZE, 0);
+ if (!ret)
+ goto nomem;
+ }
+ WARN_ON(bio->bi_vcnt != SCRUB_PAGES_PER_BIO);
+
+ if (i != SCRUB_BIOS_PER_DEV-1)
+ sdev->bios[i]->next_free = i + 1;
+ else
+ sdev->bios[i]->next_free = -1;
+ }
+ sdev->first_free = 0;
+ sdev->curr = -1;
+ atomic_set(&sdev->in_flight, 0);
+ atomic_set(&sdev->cancel_req, 0);
+ sdev->csum_size = btrfs_super_csum_size(&fs_info->super_copy);
+ INIT_LIST_HEAD(&sdev->csum_list);
+
+ spin_lock_init(&sdev->list_lock);
+ spin_lock_init(&sdev->stat_lock);
+ init_waitqueue_head(&sdev->list_wait);
+ return sdev;
+
+ nomem:
+ scrub_free_dev(sdev);
+ return ERR_PTR(-ENOMEM);
+ }
+
+ /*
+ * scrub_recheck_error gets called when either verification of the page
+ * failed or the bio failed to read, e.g. with EIO. In the latter case,
+ * recheck_error gets called for every page in the bio, even though only
+ * one may be bad
+ */
+ static void scrub_recheck_error(struct scrub_bio *sbio, int ix)
+ {
+ if (sbio->err) {
+ if (scrub_fixup_io(READ, sbio->sdev->dev->bdev,
+ (sbio->physical + ix * PAGE_SIZE) >> 9,
+ sbio->bio->bi_io_vec[ix].bv_page) == 0) {
+ if (scrub_fixup_check(sbio, ix) == 0)
+ return;
+ }
+ }
+
+ scrub_fixup(sbio, ix);
+ }
+
+ static int scrub_fixup_check(struct scrub_bio *sbio, int ix)
+ {
+ int ret = 1;
+ struct page *page;
+ void *buffer;
+ u64 flags = sbio->spag[ix].flags;
+
+ page = sbio->bio->bi_io_vec[ix].bv_page;
+ buffer = kmap_atomic(page, KM_USER0);
+ if (flags & BTRFS_EXTENT_FLAG_DATA) {
+ ret = scrub_checksum_data(sbio->sdev,
+ sbio->spag + ix, buffer);
+ } else if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) {
+ ret = scrub_checksum_tree_block(sbio->sdev,
+ sbio->spag + ix,
+ sbio->logical + ix * PAGE_SIZE,
+ buffer);
+ } else {
+ WARN_ON(1);
+ }
+ kunmap_atomic(buffer, KM_USER0);
+
+ return ret;
+ }
+
+ static void scrub_fixup_end_io(struct bio *bio, int err)
+ {
+ complete((struct completion *)bio->bi_private);
+ }
+
+ static void scrub_fixup(struct scrub_bio *sbio, int ix)
+ {
+ struct scrub_dev *sdev = sbio->sdev;
+ struct btrfs_fs_info *fs_info = sdev->dev->dev_root->fs_info;
+ struct btrfs_mapping_tree *map_tree = &fs_info->mapping_tree;
+ struct btrfs_multi_bio *multi = NULL;
+ u64 logical = sbio->logical + ix * PAGE_SIZE;
+ u64 length;
+ int i;
+ int ret;
+ DECLARE_COMPLETION_ONSTACK(complete);
+
+ if ((sbio->spag[ix].flags & BTRFS_EXTENT_FLAG_DATA) &&
+ (sbio->spag[ix].have_csum == 0)) {
+ /*
+ * nodatasum, don't try to fix anything
+ * FIXME: we can do better, open the inode and trigger a
+ * writeback
+ */
+ goto uncorrectable;
+ }
+
+ length = PAGE_SIZE;
+ ret = btrfs_map_block(map_tree, REQ_WRITE, logical, &length,
+ &multi, 0);
+ if (ret || !multi || length < PAGE_SIZE) {
+ printk(KERN_ERR
+ "scrub_fixup: btrfs_map_block failed us for %llu\n",
+ (unsigned long long)logical);
+ WARN_ON(1);
+ return;
+ }
+
+ if (multi->num_stripes == 1)
+ /* there aren't any replicas */
+ goto uncorrectable;
+
+ /*
+ * first find a good copy
+ */
+ for (i = 0; i < multi->num_stripes; ++i) {
+ if (i == sbio->spag[ix].mirror_num)
+ continue;
+
+ if (scrub_fixup_io(READ, multi->stripes[i].dev->bdev,
+ multi->stripes[i].physical >> 9,
+ sbio->bio->bi_io_vec[ix].bv_page)) {
+ /* I/O-error, this is not a good copy */
+ continue;
+ }
+
+ if (scrub_fixup_check(sbio, ix) == 0)
+ break;
+ }
+ if (i == multi->num_stripes)
+ goto uncorrectable;
+
+ if (!sdev->readonly) {
+ /*
+ * bi_io_vec[ix].bv_page now contains good data, write it back
+ */
+ if (scrub_fixup_io(WRITE, sdev->dev->bdev,
+ (sbio->physical + ix * PAGE_SIZE) >> 9,
+ sbio->bio->bi_io_vec[ix].bv_page)) {
+ /* I/O-error, writeback failed, give up */
+ goto uncorrectable;
+ }
+ }
+
+ kfree(multi);
+ spin_lock(&sdev->stat_lock);
+ ++sdev->stat.corrected_errors;
+ spin_unlock(&sdev->stat_lock);
+
+ if (printk_ratelimit())
+ printk(KERN_ERR "btrfs: fixed up at %llu\n",
+ (unsigned long long)logical);
+ return;
+
+ uncorrectable:
+ kfree(multi);
+ spin_lock(&sdev->stat_lock);
+ ++sdev->stat.uncorrectable_errors;
+ spin_unlock(&sdev->stat_lock);
+
+ if (printk_ratelimit())
+ printk(KERN_ERR "btrfs: unable to fixup at %llu\n",
+ (unsigned long long)logical);
+ }
+
+ static int scrub_fixup_io(int rw, struct block_device *bdev, sector_t sector,
+ struct page *page)
+ {
+ struct bio *bio = NULL;
+ int ret;
+ DECLARE_COMPLETION_ONSTACK(complete);
+
+ /* we are going to wait on this IO */
- btrfs_release_path(root, path);
++ rw |= REQ_SYNC;
+
+ bio = bio_alloc(GFP_NOFS, 1);
+ bio->bi_bdev = bdev;
+ bio->bi_sector = sector;
+ bio_add_page(bio, page, PAGE_SIZE, 0);
+ bio->bi_end_io = scrub_fixup_end_io;
+ bio->bi_private = &complete;
+ submit_bio(rw, bio);
+
+ wait_for_completion(&complete);
+
+ ret = !test_bit(BIO_UPTODATE, &bio->bi_flags);
+ bio_put(bio);
+ return ret;
+ }
+
+ static void scrub_bio_end_io(struct bio *bio, int err)
+ {
+ struct scrub_bio *sbio = bio->bi_private;
+ struct scrub_dev *sdev = sbio->sdev;
+ struct btrfs_fs_info *fs_info = sdev->dev->dev_root->fs_info;
+
+ sbio->err = err;
+
+ btrfs_queue_worker(&fs_info->scrub_workers, &sbio->work);
+ }
+
+ static void scrub_checksum(struct btrfs_work *work)
+ {
+ struct scrub_bio *sbio = container_of(work, struct scrub_bio, work);
+ struct scrub_dev *sdev = sbio->sdev;
+ struct page *page;
+ void *buffer;
+ int i;
+ u64 flags;
+ u64 logical;
+ int ret;
+
+ if (sbio->err) {
+ for (i = 0; i < sbio->count; ++i)
+ scrub_recheck_error(sbio, i);
+
+ sbio->bio->bi_flags &= ~(BIO_POOL_MASK - 1);
+ sbio->bio->bi_flags |= 1 << BIO_UPTODATE;
+ sbio->bio->bi_phys_segments = 0;
+ sbio->bio->bi_idx = 0;
+
+ for (i = 0; i < sbio->count; i++) {
+ struct bio_vec *bi;
+ bi = &sbio->bio->bi_io_vec[i];
+ bi->bv_offset = 0;
+ bi->bv_len = PAGE_SIZE;
+ }
+
+ spin_lock(&sdev->stat_lock);
+ ++sdev->stat.read_errors;
+ spin_unlock(&sdev->stat_lock);
+ goto out;
+ }
+ for (i = 0; i < sbio->count; ++i) {
+ page = sbio->bio->bi_io_vec[i].bv_page;
+ buffer = kmap_atomic(page, KM_USER0);
+ flags = sbio->spag[i].flags;
+ logical = sbio->logical + i * PAGE_SIZE;
+ ret = 0;
+ if (flags & BTRFS_EXTENT_FLAG_DATA) {
+ ret = scrub_checksum_data(sdev, sbio->spag + i, buffer);
+ } else if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) {
+ ret = scrub_checksum_tree_block(sdev, sbio->spag + i,
+ logical, buffer);
+ } else if (flags & BTRFS_EXTENT_FLAG_SUPER) {
+ BUG_ON(i);
+ (void)scrub_checksum_super(sbio, buffer);
+ } else {
+ WARN_ON(1);
+ }
+ kunmap_atomic(buffer, KM_USER0);
+ if (ret)
+ scrub_recheck_error(sbio, i);
+ }
+
+ out:
+ spin_lock(&sdev->list_lock);
+ sbio->next_free = sdev->first_free;
+ sdev->first_free = sbio->index;
+ spin_unlock(&sdev->list_lock);
+ atomic_dec(&sdev->in_flight);
+ wake_up(&sdev->list_wait);
+ }
+
+ static int scrub_checksum_data(struct scrub_dev *sdev,
+ struct scrub_page *spag, void *buffer)
+ {
+ u8 csum[BTRFS_CSUM_SIZE];
+ u32 crc = ~(u32)0;
+ int fail = 0;
+ struct btrfs_root *root = sdev->dev->dev_root;
+
+ if (!spag->have_csum)
+ return 0;
+
+ crc = btrfs_csum_data(root, buffer, crc, PAGE_SIZE);
+ btrfs_csum_final(crc, csum);
+ if (memcmp(csum, spag->csum, sdev->csum_size))
+ fail = 1;
+
+ spin_lock(&sdev->stat_lock);
+ ++sdev->stat.data_extents_scrubbed;
+ sdev->stat.data_bytes_scrubbed += PAGE_SIZE;
+ if (fail)
+ ++sdev->stat.csum_errors;
+ spin_unlock(&sdev->stat_lock);
+
+ return fail;
+ }
+
+ static int scrub_checksum_tree_block(struct scrub_dev *sdev,
+ struct scrub_page *spag, u64 logical,
+ void *buffer)
+ {
+ struct btrfs_header *h;
+ struct btrfs_root *root = sdev->dev->dev_root;
+ struct btrfs_fs_info *fs_info = root->fs_info;
+ u8 csum[BTRFS_CSUM_SIZE];
+ u32 crc = ~(u32)0;
+ int fail = 0;
+ int crc_fail = 0;
+
+ /*
+ * we don't use the getter functions here, as we
+ * a) don't have an extent buffer and
+ * b) the page is already kmapped
+ */
+ h = (struct btrfs_header *)buffer;
+
+ if (logical != le64_to_cpu(h->bytenr))
+ ++fail;
+
+ if (spag->generation != le64_to_cpu(h->generation))
+ ++fail;
+
+ if (memcmp(h->fsid, fs_info->fsid, BTRFS_UUID_SIZE))
+ ++fail;
+
+ if (memcmp(h->chunk_tree_uuid, fs_info->chunk_tree_uuid,
+ BTRFS_UUID_SIZE))
+ ++fail;
+
+ crc = btrfs_csum_data(root, buffer + BTRFS_CSUM_SIZE, crc,
+ PAGE_SIZE - BTRFS_CSUM_SIZE);
+ btrfs_csum_final(crc, csum);
+ if (memcmp(csum, h->csum, sdev->csum_size))
+ ++crc_fail;
+
+ spin_lock(&sdev->stat_lock);
+ ++sdev->stat.tree_extents_scrubbed;
+ sdev->stat.tree_bytes_scrubbed += PAGE_SIZE;
+ if (crc_fail)
+ ++sdev->stat.csum_errors;
+ if (fail)
+ ++sdev->stat.verify_errors;
+ spin_unlock(&sdev->stat_lock);
+
+ return fail || crc_fail;
+ }
+
+ static int scrub_checksum_super(struct scrub_bio *sbio, void *buffer)
+ {
+ struct btrfs_super_block *s;
+ u64 logical;
+ struct scrub_dev *sdev = sbio->sdev;
+ struct btrfs_root *root = sdev->dev->dev_root;
+ struct btrfs_fs_info *fs_info = root->fs_info;
+ u8 csum[BTRFS_CSUM_SIZE];
+ u32 crc = ~(u32)0;
+ int fail = 0;
+
+ s = (struct btrfs_super_block *)buffer;
+ logical = sbio->logical;
+
+ if (logical != le64_to_cpu(s->bytenr))
+ ++fail;
+
+ if (sbio->spag[0].generation != le64_to_cpu(s->generation))
+ ++fail;
+
+ if (memcmp(s->fsid, fs_info->fsid, BTRFS_UUID_SIZE))
+ ++fail;
+
+ crc = btrfs_csum_data(root, buffer + BTRFS_CSUM_SIZE, crc,
+ PAGE_SIZE - BTRFS_CSUM_SIZE);
+ btrfs_csum_final(crc, csum);
+ if (memcmp(csum, s->csum, sbio->sdev->csum_size))
+ ++fail;
+
+ if (fail) {
+ /*
+ * if we find an error in a super block, we just report it.
+ * They will get written with the next transaction commit
+ * anyway
+ */
+ spin_lock(&sdev->stat_lock);
+ ++sdev->stat.super_errors;
+ spin_unlock(&sdev->stat_lock);
+ }
+
+ return fail;
+ }
+
+ static int scrub_submit(struct scrub_dev *sdev)
+ {
+ struct scrub_bio *sbio;
+
+ if (sdev->curr == -1)
+ return 0;
+
+ sbio = sdev->bios[sdev->curr];
+
+ sbio->bio->bi_sector = sbio->physical >> 9;
+ sbio->bio->bi_size = sbio->count * PAGE_SIZE;
+ sbio->bio->bi_next = NULL;
+ sbio->bio->bi_flags |= 1 << BIO_UPTODATE;
+ sbio->bio->bi_comp_cpu = -1;
+ sbio->bio->bi_bdev = sdev->dev->bdev;
+ sbio->err = 0;
+ sdev->curr = -1;
+ atomic_inc(&sdev->in_flight);
+
+ submit_bio(0, sbio->bio);
+
+ return 0;
+ }
+
+ static int scrub_page(struct scrub_dev *sdev, u64 logical, u64 len,
+ u64 physical, u64 flags, u64 gen, u64 mirror_num,
+ u8 *csum, int force)
+ {
+ struct scrub_bio *sbio;
+
+ again:
+ /*
+ * grab a fresh bio or wait for one to become available
+ */
+ while (sdev->curr == -1) {
+ spin_lock(&sdev->list_lock);
+ sdev->curr = sdev->first_free;
+ if (sdev->curr != -1) {
+ sdev->first_free = sdev->bios[sdev->curr]->next_free;
+ sdev->bios[sdev->curr]->next_free = -1;
+ sdev->bios[sdev->curr]->count = 0;
+ spin_unlock(&sdev->list_lock);
+ } else {
+ spin_unlock(&sdev->list_lock);
+ wait_event(sdev->list_wait, sdev->first_free != -1);
+ }
+ }
+ sbio = sdev->bios[sdev->curr];
+ if (sbio->count == 0) {
+ sbio->physical = physical;
+ sbio->logical = logical;
+ } else if (sbio->physical + sbio->count * PAGE_SIZE != physical) {
+ scrub_submit(sdev);
+ goto again;
+ }
+ sbio->spag[sbio->count].flags = flags;
+ sbio->spag[sbio->count].generation = gen;
+ sbio->spag[sbio->count].have_csum = 0;
+ sbio->spag[sbio->count].mirror_num = mirror_num;
+ if (csum) {
+ sbio->spag[sbio->count].have_csum = 1;
+ memcpy(sbio->spag[sbio->count].csum, csum, sdev->csum_size);
+ }
+ ++sbio->count;
+ if (sbio->count == SCRUB_PAGES_PER_BIO || force)
+ scrub_submit(sdev);
+
+ return 0;
+ }
+
+ static int scrub_find_csum(struct scrub_dev *sdev, u64 logical, u64 len,
+ u8 *csum)
+ {
+ struct btrfs_ordered_sum *sum = NULL;
+ int ret = 0;
+ unsigned long i;
+ unsigned long num_sectors;
+ u32 sectorsize = sdev->dev->dev_root->sectorsize;
+
+ while (!list_empty(&sdev->csum_list)) {
+ sum = list_first_entry(&sdev->csum_list,
+ struct btrfs_ordered_sum, list);
+ if (sum->bytenr > logical)
+ return 0;
+ if (sum->bytenr + sum->len > logical)
+ break;
+
+ ++sdev->stat.csum_discards;
+ list_del(&sum->list);
+ kfree(sum);
+ sum = NULL;
+ }
+ if (!sum)
+ return 0;
+
+ num_sectors = sum->len / sectorsize;
+ for (i = 0; i < num_sectors; ++i) {
+ if (sum->sums[i].bytenr == logical) {
+ memcpy(csum, &sum->sums[i].sum, sdev->csum_size);
+ ret = 1;
+ break;
+ }
+ }
+ if (ret && i == num_sectors - 1) {
+ list_del(&sum->list);
+ kfree(sum);
+ }
+ return ret;
+ }
+
+ /* scrub extent tries to collect up to 64 kB for each bio */
+ static int scrub_extent(struct scrub_dev *sdev, u64 logical, u64 len,
+ u64 physical, u64 flags, u64 gen, u64 mirror_num)
+ {
+ int ret;
+ u8 csum[BTRFS_CSUM_SIZE];
+
+ while (len) {
+ u64 l = min_t(u64, len, PAGE_SIZE);
+ int have_csum = 0;
+
+ if (flags & BTRFS_EXTENT_FLAG_DATA) {
+ /* push csums to sbio */
+ have_csum = scrub_find_csum(sdev, logical, l, csum);
+ if (have_csum == 0)
+ ++sdev->stat.no_csum;
+ }
+ ret = scrub_page(sdev, logical, l, physical, flags, gen,
+ mirror_num, have_csum ? csum : NULL, 0);
+ if (ret)
+ return ret;
+ len -= l;
+ logical += l;
+ physical += l;
+ }
+ return 0;
+ }
+
+ static noinline_for_stack int scrub_stripe(struct scrub_dev *sdev,
+ struct map_lookup *map, int num, u64 base, u64 length)
+ {
+ struct btrfs_path *path;
+ struct btrfs_fs_info *fs_info = sdev->dev->dev_root->fs_info;
+ struct btrfs_root *root = fs_info->extent_root;
+ struct btrfs_root *csum_root = fs_info->csum_root;
+ struct btrfs_extent_item *extent;
+ u64 flags;
+ int ret;
+ int slot;
+ int i;
+ u64 nstripes;
+ int start_stripe;
+ struct extent_buffer *l;
+ struct btrfs_key key;
+ u64 physical;
+ u64 logical;
+ u64 generation;
+ u64 mirror_num;
+
+ u64 increment = map->stripe_len;
+ u64 offset;
+
+ nstripes = length;
+ offset = 0;
+ do_div(nstripes, map->stripe_len);
+ if (map->type & BTRFS_BLOCK_GROUP_RAID0) {
+ offset = map->stripe_len * num;
+ increment = map->stripe_len * map->num_stripes;
+ mirror_num = 0;
+ } else if (map->type & BTRFS_BLOCK_GROUP_RAID10) {
+ int factor = map->num_stripes / map->sub_stripes;
+ offset = map->stripe_len * (num / map->sub_stripes);
+ increment = map->stripe_len * factor;
+ mirror_num = num % map->sub_stripes;
+ } else if (map->type & BTRFS_BLOCK_GROUP_RAID1) {
+ increment = map->stripe_len;
+ mirror_num = num % map->num_stripes;
+ } else if (map->type & BTRFS_BLOCK_GROUP_DUP) {
+ increment = map->stripe_len;
+ mirror_num = num % map->num_stripes;
+ } else {
+ increment = map->stripe_len;
+ mirror_num = 0;
+ }
+
+ path = btrfs_alloc_path();
+ if (!path)
+ return -ENOMEM;
+
+ path->reada = 2;
+ path->search_commit_root = 1;
+ path->skip_locking = 1;
+
+ /*
+ * find all extents for each stripe and just read them to get
+ * them into the page cache
+ * FIXME: we can do better. build a more intelligent prefetching
+ */
+ logical = base + offset;
+ physical = map->stripes[num].physical;
+ ret = 0;
+ for (i = 0; i < nstripes; ++i) {
+ key.objectid = logical;
+ key.type = BTRFS_EXTENT_ITEM_KEY;
+ key.offset = (u64)0;
+
+ ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
+ if (ret < 0)
+ goto out;
+
+ l = path->nodes[0];
+ slot = path->slots[0];
+ btrfs_item_key_to_cpu(l, &key, slot);
+ if (key.objectid != logical) {
+ ret = btrfs_previous_item(root, path, 0,
+ BTRFS_EXTENT_ITEM_KEY);
+ if (ret < 0)
+ goto out;
+ }
+
+ while (1) {
+ l = path->nodes[0];
+ slot = path->slots[0];
+ if (slot >= btrfs_header_nritems(l)) {
+ ret = btrfs_next_leaf(root, path);
+ if (ret == 0)
+ continue;
+ if (ret < 0)
+ goto out;
+
+ break;
+ }
+ btrfs_item_key_to_cpu(l, &key, slot);
+
+ if (key.objectid >= logical + map->stripe_len)
+ break;
+
+ path->slots[0]++;
+ }
- btrfs_release_path(root, path);
++ btrfs_release_path(path);
+ logical += increment;
+ physical += map->stripe_len;
+ cond_resched();
+ }
+
+ /*
+ * collect all data csums for the stripe to avoid seeking during
+ * the scrub. This might currently (crc32) end up to be about 1MB
+ */
+ start_stripe = 0;
+ again:
+ logical = base + offset + start_stripe * increment;
+ for (i = start_stripe; i < nstripes; ++i) {
+ ret = btrfs_lookup_csums_range(csum_root, logical,
+ logical + map->stripe_len - 1,
+ &sdev->csum_list, 1);
+ if (ret)
+ goto out;
+
+ logical += increment;
+ cond_resched();
+ }
+ /*
+ * now find all extents for each stripe and scrub them
+ */
+ logical = base + offset + start_stripe * increment;
+ physical = map->stripes[num].physical + start_stripe * map->stripe_len;
+ ret = 0;
+ for (i = start_stripe; i < nstripes; ++i) {
+ /*
+ * canceled?
+ */
+ if (atomic_read(&fs_info->scrub_cancel_req) ||
+ atomic_read(&sdev->cancel_req)) {
+ ret = -ECANCELED;
+ goto out;
+ }
+ /*
+ * check to see if we have to pause
+ */
+ if (atomic_read(&fs_info->scrub_pause_req)) {
+ /* push queued extents */
+ scrub_submit(sdev);
+ wait_event(sdev->list_wait,
+ atomic_read(&sdev->in_flight) == 0);
+ atomic_inc(&fs_info->scrubs_paused);
+ wake_up(&fs_info->scrub_pause_wait);
+ mutex_lock(&fs_info->scrub_lock);
+ while (atomic_read(&fs_info->scrub_pause_req)) {
+ mutex_unlock(&fs_info->scrub_lock);
+ wait_event(fs_info->scrub_pause_wait,
+ atomic_read(&fs_info->scrub_pause_req) == 0);
+ mutex_lock(&fs_info->scrub_lock);
+ }
+ atomic_dec(&fs_info->scrubs_paused);
+ mutex_unlock(&fs_info->scrub_lock);
+ wake_up(&fs_info->scrub_pause_wait);
+ scrub_free_csums(sdev);
+ start_stripe = i;
+ goto again;
+ }
+
+ key.objectid = logical;
+ key.type = BTRFS_EXTENT_ITEM_KEY;
+ key.offset = (u64)0;
+
+ ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
+ if (ret < 0)
+ goto out;
+
+ l = path->nodes[0];
+ slot = path->slots[0];
+ btrfs_item_key_to_cpu(l, &key, slot);
+ if (key.objectid != logical) {
+ ret = btrfs_previous_item(root, path, 0,
+ BTRFS_EXTENT_ITEM_KEY);
+ if (ret < 0)
+ goto out;
+ }
+
+ while (1) {
+ l = path->nodes[0];
+ slot = path->slots[0];
+ if (slot >= btrfs_header_nritems(l)) {
+ ret = btrfs_next_leaf(root, path);
+ if (ret == 0)
+ continue;
+ if (ret < 0)
+ goto out;
+
+ break;
+ }
+ btrfs_item_key_to_cpu(l, &key, slot);
+
+ if (key.objectid + key.offset <= logical)
+ goto next;
+
+ if (key.objectid >= logical + map->stripe_len)
+ break;
+
+ if (btrfs_key_type(&key) != BTRFS_EXTENT_ITEM_KEY)
+ goto next;
+
+ extent = btrfs_item_ptr(l, slot,
+ struct btrfs_extent_item);
+ flags = btrfs_extent_flags(l, extent);
+ generation = btrfs_extent_generation(l, extent);
+
+ if (key.objectid < logical &&
+ (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK)) {
+ printk(KERN_ERR
+ "btrfs scrub: tree block %llu spanning "
+ "stripes, ignored. logical=%llu\n",
+ (unsigned long long)key.objectid,
+ (unsigned long long)logical);
+ goto next;
+ }
+
+ /*
+ * trim extent to this stripe
+ */
+ if (key.objectid < logical) {
+ key.offset -= logical - key.objectid;
+ key.objectid = logical;
+ }
+ if (key.objectid + key.offset >
+ logical + map->stripe_len) {
+ key.offset = logical + map->stripe_len -
+ key.objectid;
+ }
+
+ ret = scrub_extent(sdev, key.objectid, key.offset,
+ key.objectid - logical + physical,
+ flags, generation, mirror_num);
+ if (ret)
+ goto out;
+
+ next:
+ path->slots[0]++;
+ }
- btrfs_release_path(root, path);
++ btrfs_release_path(path);
+ logical += increment;
+ physical += map->stripe_len;
+ spin_lock(&sdev->stat_lock);
+ sdev->stat.last_physical = physical;
+ spin_unlock(&sdev->stat_lock);
+ }
+ /* push queued extents */
+ scrub_submit(sdev);
+
+ out:
+ btrfs_free_path(path);
+ return ret < 0 ? ret : 0;
+ }
+
+ static noinline_for_stack int scrub_chunk(struct scrub_dev *sdev,
+ u64 chunk_tree, u64 chunk_objectid, u64 chunk_offset, u64 length)
+ {
+ struct btrfs_mapping_tree *map_tree =
+ &sdev->dev->dev_root->fs_info->mapping_tree;
+ struct map_lookup *map;
+ struct extent_map *em;
+ int i;
+ int ret = -EINVAL;
+
+ read_lock(&map_tree->map_tree.lock);
+ em = lookup_extent_mapping(&map_tree->map_tree, chunk_offset, 1);
+ read_unlock(&map_tree->map_tree.lock);
+
+ if (!em)
+ return -EINVAL;
+
+ map = (struct map_lookup *)em->bdev;
+ if (em->start != chunk_offset)
+ goto out;
+
+ if (em->len < length)
+ goto out;
+
+ for (i = 0; i < map->num_stripes; ++i) {
+ if (map->stripes[i].dev == sdev->dev) {
+ ret = scrub_stripe(sdev, map, i, chunk_offset, length);
+ if (ret)
+ goto out;
+ }
+ }
+ out:
+ free_extent_map(em);
+
+ return ret;
+ }
+
+ static noinline_for_stack
+ int scrub_enumerate_chunks(struct scrub_dev *sdev, u64 start, u64 end)
+ {
+ struct btrfs_dev_extent *dev_extent = NULL;
+ struct btrfs_path *path;
+ struct btrfs_root *root = sdev->dev->dev_root;
+ struct btrfs_fs_info *fs_info = root->fs_info;
+ u64 length;
+ u64 chunk_tree;
+ u64 chunk_objectid;
+ u64 chunk_offset;
+ int ret;
+ int slot;
+ struct extent_buffer *l;
+ struct btrfs_key key;
+ struct btrfs_key found_key;
+ struct btrfs_block_group_cache *cache;
+
+ path = btrfs_alloc_path();
+ if (!path)
+ return -ENOMEM;
+
+ path->reada = 2;
+ path->search_commit_root = 1;
+ path->skip_locking = 1;
+
+ key.objectid = sdev->dev->devid;
+ key.offset = 0ull;
+ key.type = BTRFS_DEV_EXTENT_KEY;
+
+
+ while (1) {
+ ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
+ if (ret < 0)
+ goto out;
+ ret = 0;
+
+ l = path->nodes[0];
+ slot = path->slots[0];
+
+ btrfs_item_key_to_cpu(l, &found_key, slot);
+
+ if (found_key.objectid != sdev->dev->devid)
+ break;
+
+ if (btrfs_key_type(&key) != BTRFS_DEV_EXTENT_KEY)
+ break;
+
+ if (found_key.offset >= end)
+ break;
+
+ if (found_key.offset < key.offset)
+ break;
+
+ dev_extent = btrfs_item_ptr(l, slot, struct btrfs_dev_extent);
+ length = btrfs_dev_extent_length(l, dev_extent);
+
+ if (found_key.offset + length <= start) {
+ key.offset = found_key.offset + length;
- btrfs_release_path(root, path);
++ btrfs_release_path(path);
+ continue;
+ }
+
+ chunk_tree = btrfs_dev_extent_chunk_tree(l, dev_extent);
+ chunk_objectid = btrfs_dev_extent_chunk_objectid(l, dev_extent);
+ chunk_offset = btrfs_dev_extent_chunk_offset(l, dev_extent);
+
+ /*
+ * get a reference on the corresponding block group to prevent
+ * the chunk from going away while we scrub it
+ */
+ cache = btrfs_lookup_block_group(fs_info, chunk_offset);
+ if (!cache) {
+ ret = -ENOENT;
+ goto out;
+ }
+ ret = scrub_chunk(sdev, chunk_tree, chunk_objectid,
+ chunk_offset, length);
+ btrfs_put_block_group(cache);
+ if (ret)
+ break;
+
+ key.offset = found_key.offset + length;
++ btrfs_release_path(path);
+ }
+
+ out:
+ btrfs_free_path(path);
+ return ret;
+ }
+
+ static noinline_for_stack int scrub_supers(struct scrub_dev *sdev)
+ {
+ int i;
+ u64 bytenr;
+ u64 gen;
+ int ret;
+ struct btrfs_device *device = sdev->dev;
+ struct btrfs_root *root = device->dev_root;
+
+ gen = root->fs_info->last_trans_committed;
+
+ for (i = 0; i < BTRFS_SUPER_MIRROR_MAX; i++) {
+ bytenr = btrfs_sb_offset(i);
+ if (bytenr + BTRFS_SUPER_INFO_SIZE >= device->total_bytes)
+ break;
+
+ ret = scrub_page(sdev, bytenr, PAGE_SIZE, bytenr,
+ BTRFS_EXTENT_FLAG_SUPER, gen, i, NULL, 1);
+ if (ret)
+ return ret;
+ }
+ wait_event(sdev->list_wait, atomic_read(&sdev->in_flight) == 0);
+
+ return 0;
+ }
+
+ /*
+ * get a reference count on fs_info->scrub_workers. start worker if necessary
+ */
+ static noinline_for_stack int scrub_workers_get(struct btrfs_root *root)
+ {
+ struct btrfs_fs_info *fs_info = root->fs_info;
+
+ mutex_lock(&fs_info->scrub_lock);
+ if (fs_info->scrub_workers_refcnt == 0)
+ btrfs_start_workers(&fs_info->scrub_workers, 1);
+ ++fs_info->scrub_workers_refcnt;
+ mutex_unlock(&fs_info->scrub_lock);
+
+ return 0;
+ }
+
+ static noinline_for_stack void scrub_workers_put(struct btrfs_root *root)
+ {
+ struct btrfs_fs_info *fs_info = root->fs_info;
+
+ mutex_lock(&fs_info->scrub_lock);
+ if (--fs_info->scrub_workers_refcnt == 0)
+ btrfs_stop_workers(&fs_info->scrub_workers);
+ WARN_ON(fs_info->scrub_workers_refcnt < 0);
+ mutex_unlock(&fs_info->scrub_lock);
+ }
+
+
+ int btrfs_scrub_dev(struct btrfs_root *root, u64 devid, u64 start, u64 end,
+ struct btrfs_scrub_progress *progress, int readonly)
+ {
+ struct scrub_dev *sdev;
+ struct btrfs_fs_info *fs_info = root->fs_info;
+ int ret;
+ struct btrfs_device *dev;
+
+ if (root->fs_info->closing)
+ return -EINVAL;
+
+ /*
+ * check some assumptions
+ */
+ if (root->sectorsize != PAGE_SIZE ||
+ root->sectorsize != root->leafsize ||
+ root->sectorsize != root->nodesize) {
+ printk(KERN_ERR "btrfs_scrub: size assumptions fail\n");
+ return -EINVAL;
+ }
+
+ ret = scrub_workers_get(root);
+ if (ret)
+ return ret;
+
+ mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
+ dev = btrfs_find_device(root, devid, NULL, NULL);
+ if (!dev || dev->missing) {
+ mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
+ scrub_workers_put(root);
+ return -ENODEV;
+ }
+ mutex_lock(&fs_info->scrub_lock);
+
+ if (!dev->in_fs_metadata) {
+ mutex_unlock(&fs_info->scrub_lock);
+ mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
+ scrub_workers_put(root);
+ return -ENODEV;
+ }
+
+ if (dev->scrub_device) {
+ mutex_unlock(&fs_info->scrub_lock);
+ mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
+ scrub_workers_put(root);
+ return -EINPROGRESS;
+ }
+ sdev = scrub_setup_dev(dev);
+ if (IS_ERR(sdev)) {
+ mutex_unlock(&fs_info->scrub_lock);
+ mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
+ scrub_workers_put(root);
+ return PTR_ERR(sdev);
+ }
+ sdev->readonly = readonly;
+ dev->scrub_device = sdev;
+
+ atomic_inc(&fs_info->scrubs_running);
+ mutex_unlock(&fs_info->scrub_lock);
+ mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
+
+ down_read(&fs_info->scrub_super_lock);
+ ret = scrub_supers(sdev);
+ up_read(&fs_info->scrub_super_lock);
+
+ if (!ret)
+ ret = scrub_enumerate_chunks(sdev, start, end);
+
+ wait_event(sdev->list_wait, atomic_read(&sdev->in_flight) == 0);
+
+ atomic_dec(&fs_info->scrubs_running);
+ wake_up(&fs_info->scrub_pause_wait);
+
+ if (progress)
+ memcpy(progress, &sdev->stat, sizeof(*progress));
+
+ mutex_lock(&fs_info->scrub_lock);
+ dev->scrub_device = NULL;
+ mutex_unlock(&fs_info->scrub_lock);
+
+ scrub_free_dev(sdev);
+ scrub_workers_put(root);
+
+ return ret;
+ }
+
+ int btrfs_scrub_pause(struct btrfs_root *root)
+ {
+ struct btrfs_fs_info *fs_info = root->fs_info;
+
+ mutex_lock(&fs_info->scrub_lock);
+ atomic_inc(&fs_info->scrub_pause_req);
+ while (atomic_read(&fs_info->scrubs_paused) !=
+ atomic_read(&fs_info->scrubs_running)) {
+ mutex_unlock(&fs_info->scrub_lock);
+ wait_event(fs_info->scrub_pause_wait,
+ atomic_read(&fs_info->scrubs_paused) ==
+ atomic_read(&fs_info->scrubs_running));
+ mutex_lock(&fs_info->scrub_lock);
+ }
+ mutex_unlock(&fs_info->scrub_lock);
+
+ return 0;
+ }
+
+ int btrfs_scrub_continue(struct btrfs_root *root)
+ {
+ struct btrfs_fs_info *fs_info = root->fs_info;
+
+ atomic_dec(&fs_info->scrub_pause_req);
+ wake_up(&fs_info->scrub_pause_wait);
+ return 0;
+ }
+
+ int btrfs_scrub_pause_super(struct btrfs_root *root)
+ {
+ down_write(&root->fs_info->scrub_super_lock);
+ return 0;
+ }
+
+ int btrfs_scrub_continue_super(struct btrfs_root *root)
+ {
+ up_write(&root->fs_info->scrub_super_lock);
+ return 0;
+ }
+
+ int btrfs_scrub_cancel(struct btrfs_root *root)
+ {
+ struct btrfs_fs_info *fs_info = root->fs_info;
+
+ mutex_lock(&fs_info->scrub_lock);
+ if (!atomic_read(&fs_info->scrubs_running)) {
+ mutex_unlock(&fs_info->scrub_lock);
+ return -ENOTCONN;
+ }
+
+ atomic_inc(&fs_info->scrub_cancel_req);
+ while (atomic_read(&fs_info->scrubs_running)) {
+ mutex_unlock(&fs_info->scrub_lock);
+ wait_event(fs_info->scrub_pause_wait,
+ atomic_read(&fs_info->scrubs_running) == 0);
+ mutex_lock(&fs_info->scrub_lock);
+ }
+ atomic_dec(&fs_info->scrub_cancel_req);
+ mutex_unlock(&fs_info->scrub_lock);
+
+ return 0;
+ }
+
+ int btrfs_scrub_cancel_dev(struct btrfs_root *root, struct btrfs_device *dev)
+ {
+ struct btrfs_fs_info *fs_info = root->fs_info;
+ struct scrub_dev *sdev;
+
+ mutex_lock(&fs_info->scrub_lock);
+ sdev = dev->scrub_device;
+ if (!sdev) {
+ mutex_unlock(&fs_info->scrub_lock);
+ return -ENOTCONN;
+ }
+ atomic_inc(&sdev->cancel_req);
+ while (dev->scrub_device) {
+ mutex_unlock(&fs_info->scrub_lock);
+ wait_event(fs_info->scrub_pause_wait,
+ dev->scrub_device == NULL);
+ mutex_lock(&fs_info->scrub_lock);
+ }
+ mutex_unlock(&fs_info->scrub_lock);
+
+ return 0;
+ }
+ int btrfs_scrub_cancel_devid(struct btrfs_root *root, u64 devid)
+ {
+ struct btrfs_fs_info *fs_info = root->fs_info;
+ struct btrfs_device *dev;
+ int ret;
+
+ /*
+ * we have to hold the device_list_mutex here so the device
+ * does not go away in cancel_dev. FIXME: find a better solution
+ */
+ mutex_lock(&fs_info->fs_devices->device_list_mutex);
+ dev = btrfs_find_device(root, devid, NULL, NULL);
+ if (!dev) {
+ mutex_unlock(&fs_info->fs_devices->device_list_mutex);
+ return -ENODEV;
+ }
+ ret = btrfs_scrub_cancel_dev(root, dev);
+ mutex_unlock(&fs_info->fs_devices->device_list_mutex);
+
+ return ret;
+ }
+
+ int btrfs_scrub_progress(struct btrfs_root *root, u64 devid,
+ struct btrfs_scrub_progress *progress)
+ {
+ struct btrfs_device *dev;
+ struct scrub_dev *sdev = NULL;
+
+ mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
+ dev = btrfs_find_device(root, devid, NULL, NULL);
+ if (dev)
+ sdev = dev->scrub_device;
+ if (sdev)
+ memcpy(progress, &sdev->stat, sizeof(*progress));
+ mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
+
+ return dev ? (sdev ? 0 : -ENOTCONN) : -ENODEV;
+ }
#include "transaction.h"
#include "locking.h"
#include "tree-log.h"
+#include "inode-map.h"
#define BTRFS_ROOT_TRANS_TAG 0
INIT_LIST_HEAD(&cur_trans->pending_snapshots);
list_add_tail(&cur_trans->list, &root->fs_info->trans_list);
extent_io_tree_init(&cur_trans->dirty_pages,
- root->fs_info->btree_inode->i_mapping,
- GFP_NOFS);
+ root->fs_info->btree_inode->i_mapping);
spin_lock(&root->fs_info->new_trans_lock);
root->fs_info->running_transaction = cur_trans;
spin_unlock(&root->fs_info->new_trans_lock);
return ret;
}
-#if 0
-/*
- * rate limit against the drop_snapshot code. This helps to slow down new
- * operations if the drop_snapshot code isn't able to keep up.
- */
-static void throttle_on_drops(struct btrfs_root *root)
-{
- struct btrfs_fs_info *info = root->fs_info;
- int harder_count = 0;
-
-harder:
- if (atomic_read(&info->throttles)) {
- DEFINE_WAIT(wait);
- int thr;
- thr = atomic_read(&info->throttle_gen);
-
- do {
- prepare_to_wait(&info->transaction_throttle,
- &wait, TASK_UNINTERRUPTIBLE);
- if (!atomic_read(&info->throttles)) {
- finish_wait(&info->transaction_throttle, &wait);
- break;
- }
- schedule();
- finish_wait(&info->transaction_throttle, &wait);
- } while (thr == atomic_read(&info->throttle_gen));
- harder_count++;
-
- if (root->fs_info->total_ref_cache_size > 1 * 1024 * 1024 &&
- harder_count < 2)
- goto harder;
-
- if (root->fs_info->total_ref_cache_size > 5 * 1024 * 1024 &&
- harder_count < 10)
- goto harder;
-
- if (root->fs_info->total_ref_cache_size > 10 * 1024 * 1024 &&
- harder_count < 20)
- goto harder;
- }
-}
-#endif
-
void btrfs_throttle(struct btrfs_root *root)
{
mutex_lock(&root->fs_info->trans_mutex);
int btrfs_end_transaction(struct btrfs_trans_handle *trans,
struct btrfs_root *root)
{
- return __btrfs_end_transaction(trans, root, 0, 1);
+ int ret;
+
+ ret = __btrfs_end_transaction(trans, root, 0, 1);
+ if (ret)
+ return ret;
+ return 0;
}
int btrfs_end_transaction_throttle(struct btrfs_trans_handle *trans,
struct btrfs_root *root)
{
- return __btrfs_end_transaction(trans, root, 1, 1);
+ int ret;
+
+ ret = __btrfs_end_transaction(trans, root, 1, 1);
+ if (ret)
+ return ret;
+ return 0;
}
int btrfs_end_transaction_nolock(struct btrfs_trans_handle *trans,
struct btrfs_root *root)
{
- return __btrfs_end_transaction(trans, root, 0, 0);
+ int ret;
+
+ ret = __btrfs_end_transaction(trans, root, 0, 0);
+ if (ret)
+ return ret;
+ return 0;
+}
+
+int btrfs_end_transaction_dmeta(struct btrfs_trans_handle *trans,
+ struct btrfs_root *root)
+{
+ return __btrfs_end_transaction(trans, root, 1, 1);
}
/*
btrfs_update_reloc_root(trans, root);
btrfs_orphan_commit_root(trans, root);
+ btrfs_save_ino_cache(root, trans);
+
if (root->commit_root != root->node) {
+ mutex_lock(&root->fs_commit_mutex);
switch_commit_root(root);
+ btrfs_unpin_free_ino(root);
+ mutex_unlock(&root->fs_commit_mutex);
+
btrfs_set_root_node(&root->root_item,
root->node);
}
return ret;
}
-#if 0
-/*
- * when dropping snapshots, we generate a ton of delayed refs, and it makes
- * sense not to join the transaction while it is trying to flush the current
- * queue of delayed refs out.
- *
- * This is used by the drop snapshot code only
- */
-static noinline int wait_transaction_pre_flush(struct btrfs_fs_info *info)
-{
- DEFINE_WAIT(wait);
-
- mutex_lock(&info->trans_mutex);
- while (info->running_transaction &&
- info->running_transaction->delayed_refs.flushing) {
- prepare_to_wait(&info->transaction_wait, &wait,
- TASK_UNINTERRUPTIBLE);
- mutex_unlock(&info->trans_mutex);
-
- schedule();
-
- mutex_lock(&info->trans_mutex);
- finish_wait(&info->transaction_wait, &wait);
- }
- mutex_unlock(&info->trans_mutex);
- return 0;
-}
-
-/*
- * Given a list of roots that need to be deleted, call btrfs_drop_snapshot on
- * all of them
- */
-int btrfs_drop_dead_root(struct btrfs_root *root)
-{
- struct btrfs_trans_handle *trans;
- struct btrfs_root *tree_root = root->fs_info->tree_root;
- unsigned long nr;
- int ret;
-
- while (1) {
- /*
- * we don't want to jump in and create a bunch of
- * delayed refs if the transaction is starting to close
- */
- wait_transaction_pre_flush(tree_root->fs_info);
- trans = btrfs_start_transaction(tree_root, 1);
-
- /*
- * we've joined a transaction, make sure it isn't
- * closing right now
- */
- if (trans->transaction->delayed_refs.flushing) {
- btrfs_end_transaction(trans, tree_root);
- continue;
- }
-
- ret = btrfs_drop_snapshot(trans, root);
- if (ret != -EAGAIN)
- break;
-
- ret = btrfs_update_root(trans, tree_root,
- &root->root_key,
- &root->root_item);
- if (ret)
- break;
-
- nr = trans->blocks_used;
- ret = btrfs_end_transaction(trans, tree_root);
- BUG_ON(ret);
-
- btrfs_btree_balance_dirty(tree_root, nr);
- cond_resched();
- }
- BUG_ON(ret);
-
- ret = btrfs_del_root(trans, tree_root, &root->root_key);
- BUG_ON(ret);
-
- nr = trans->blocks_used;
- ret = btrfs_end_transaction(trans, tree_root);
- BUG_ON(ret);
-
- free_extent_buffer(root->node);
- free_extent_buffer(root->commit_root);
- kfree(root);
-
- btrfs_btree_balance_dirty(tree_root, nr);
- return ret;
-}
-#endif
-
/*
* new snapshots need to be created at a very specific time in the
* transaction commit. This does the actual creation
goto fail;
}
- ret = btrfs_find_free_objectid(trans, tree_root, 0, &objectid);
+ ret = btrfs_find_free_objectid(tree_root, &objectid);
if (ret) {
pending->error = ret;
goto fail;
BUG_ON(ret);
ret = btrfs_insert_dir_item(trans, parent_root,
dentry->d_name.name, dentry->d_name.len,
- parent_inode->i_ino, &key,
+ parent_inode, &key,
BTRFS_FT_DIR, index);
BUG_ON(ret);
*/
ret = btrfs_add_root_ref(trans, tree_root, objectid,
parent_root->root_key.objectid,
- parent_inode->i_ino, index,
+ btrfs_ino(parent_inode), index,
dentry->d_name.name, dentry->d_name.len);
BUG_ON(ret);
dput(parent);
int ret;
list_for_each_entry(pending, head, list) {
+ /*
+ * We must deal with the delayed items before creating
+ * snapshots, or we will create a snapthot with inconsistent
+ * information.
+ */
+ ret = btrfs_run_delayed_items(trans, fs_info->fs_root);
+ BUG_ON(ret);
+
ret = create_pending_snapshot(trans, fs_info, pending);
BUG_ON(ret);
}
BUG_ON(ret);
}
+ ret = btrfs_run_delayed_items(trans, root);
+ BUG_ON(ret);
+
/*
* rename don't use btrfs_join_transaction, so, once we
* set the transaction to blocked above, we aren't going
ret = create_pending_snapshots(trans, root->fs_info);
BUG_ON(ret);
+ ret = btrfs_run_delayed_items(trans, root);
+ BUG_ON(ret);
+
ret = btrfs_run_delayed_refs(trans, root, (unsigned long)-1);
BUG_ON(ret);
WARN_ON(cur_trans != trans->transaction);
+ btrfs_scrub_pause(root);
/* btrfs_commit_tree_roots is responsible for getting the
* various roots consistent with each other. Every pointer
* in the tree of tree roots has to point to the most up to date
mutex_unlock(&root->fs_info->trans_mutex);
+ btrfs_scrub_continue(root);
+
if (current->journal_info == trans)
current->journal_info = NULL;
root = list_entry(list.next, struct btrfs_root, root_list);
list_del(&root->root_list);
+ btrfs_kill_all_delayed_nodes(root);
+
if (btrfs_header_backref_rev(root->node) <
BTRFS_MIXED_BACKREF_REV)
btrfs_drop_snapshot(root, NULL, 0);
goto insert;
if (item_size == 0) {
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
return 0;
}
dst_copy = kmalloc(item_size, GFP_NOFS);
src_copy = kmalloc(item_size, GFP_NOFS);
if (!dst_copy || !src_copy) {
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
kfree(dst_copy);
kfree(src_copy);
return -ENOMEM;
* sync
*/
if (ret == 0) {
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
return 0;
}
}
insert:
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
/* try to insert the key into the destination tree */
ret = btrfs_insert_empty_item(trans, root, path,
key, item_size);
}
no_copy:
btrfs_mark_buffer_dirty(path->nodes[0]);
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
return 0;
}
* file. This must be done before the btrfs_drop_extents run
* so we don't try to drop this extent.
*/
- ret = btrfs_lookup_file_extent(trans, root, path, inode->i_ino,
+ ret = btrfs_lookup_file_extent(trans, root, path, btrfs_ino(inode),
start, 0);
if (ret == 0 &&
* we don't have to do anything
*/
if (memcmp(&cmp1, &cmp2, sizeof(cmp1)) == 0) {
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
goto out;
}
}
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
saved_nbytes = inode_get_bytes(inode);
/* drop any overlapping extents */
key->objectid, offset, &ins);
BUG_ON(ret);
}
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
if (btrfs_file_extent_compression(eb, item)) {
csum_start = ins.objectid;
ret = btrfs_lookup_csums_range(root->log_root,
csum_start, csum_end - 1,
- &ordered_sums);
+ &ordered_sums, 0);
BUG_ON(ret);
while (!list_empty(&ordered_sums)) {
struct btrfs_ordered_sum *sums;
kfree(sums);
}
} else {
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
}
} else if (found_type == BTRFS_FILE_EXTENT_INLINE) {
/* inline extents are easy, we just overwrite them */
return -ENOMEM;
read_extent_buffer(leaf, name, (unsigned long)(di + 1), name_len);
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
inode = read_one_inode(root, location.objectid);
BUG_ON(!inode);
goto out;
} else
goto out;
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
di = btrfs_lookup_dir_item(NULL, root, path, dirid, name, name_len, 0);
if (di && !IS_ERR(di)) {
goto out;
match = 1;
out:
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
return match;
}
read_extent_buffer(eb, name, (unsigned long)(ref + 1), namelen);
/* if we already have a perfect match, we're done */
- if (inode_in_dir(root, path, dir->i_ino, inode->i_ino,
+ if (inode_in_dir(root, path, btrfs_ino(dir), btrfs_ino(inode),
btrfs_inode_ref_index(eb, ref),
name, namelen)) {
goto out;
if (!backref_in_log(log, key, victim_name,
victim_name_len)) {
btrfs_inc_nlink(inode);
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
ret = btrfs_unlink_inode(trans, root, dir,
inode, victim_name,
*/
search_done = 1;
}
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
insert:
/* insert our name */
BUG_ON(ret);
out_nowrite:
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
iput(dir);
iput(inode);
return 0;
unsigned long ptr;
unsigned long ptr_end;
int name_len;
+ u64 ino = btrfs_ino(inode);
- key.objectid = inode->i_ino;
+ key.objectid = ino;
key.type = BTRFS_INODE_REF_KEY;
key.offset = (u64)-1;
}
btrfs_item_key_to_cpu(path->nodes[0], &key,
path->slots[0]);
- if (key.objectid != inode->i_ino ||
+ if (key.objectid != ino ||
key.type != BTRFS_INODE_REF_KEY)
break;
ptr = btrfs_item_ptr_offset(path->nodes[0], path->slots[0]);
if (key.offset == 0)
break;
key.offset--;
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
}
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
if (nlink != inode->i_nlink) {
inode->i_nlink = nlink;
btrfs_update_inode(trans, root, inode);
if (inode->i_nlink == 0) {
if (S_ISDIR(inode->i_mode)) {
ret = replay_dir_deletes(trans, root, NULL, path,
- inode->i_ino, 1);
+ ino, 1);
BUG_ON(ret);
}
- ret = insert_orphan_item(trans, root, inode->i_ino);
+ ret = insert_orphan_item(trans, root, ino);
BUG_ON(ret);
}
btrfs_free_path(path);
ret = btrfs_del_item(trans, root, path);
BUG_ON(ret);
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
inode = read_one_inode(root, key.offset);
BUG_ON(!inode);
*/
key.offset = (u64)-1;
}
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
return 0;
}
ret = btrfs_insert_empty_item(trans, root, path, &key, 0);
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
if (ret == 0) {
btrfs_inc_nlink(inode);
btrfs_update_inode(trans, root, inode);
exists = 1;
else
exists = 0;
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
if (key->type == BTRFS_DIR_ITEM_KEY) {
dst_di = btrfs_lookup_dir_item(trans, root, path, key->objectid,
} else {
BUG();
}
- if (!dst_di || IS_ERR(dst_di)) {
+ if (IS_ERR_OR_NULL(dst_di)) {
/* we need a sequence number to insert, so we only
* do inserts for the BTRFS_DIR_INDEX_KEY types
*/
if (key->type == BTRFS_DIR_INDEX_KEY)
goto insert;
out:
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
kfree(name);
iput(dir);
return 0;
insert:
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
ret = insert_one_name(trans, root, path, key->objectid, key->offset,
name, name_len, log_type, &log_key);
*end_ret = found_end;
ret = 0;
out:
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
return ret;
}
dir_key->offset,
name, name_len, 0);
}
- if (!log_di || IS_ERR(log_di)) {
+ if (IS_ERR_OR_NULL(log_di)) {
btrfs_dir_item_key_to_cpu(eb, di, &location);
- btrfs_release_path(root, path);
- btrfs_release_path(log, log_path);
+ btrfs_release_path(path);
+ btrfs_release_path(log_path);
inode = read_one_inode(root, location.objectid);
BUG_ON(!inode);
ret = 0;
goto out;
}
- btrfs_release_path(log, log_path);
+ btrfs_release_path(log_path);
kfree(name);
ptr = (unsigned long)(di + 1);
}
ret = 0;
out:
- btrfs_release_path(root, path);
- btrfs_release_path(log, log_path);
+ btrfs_release_path(path);
+ btrfs_release_path(log_path);
return ret;
}
break;
dir_key.offset = found_key.offset + 1;
}
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
if (range_end == (u64)-1)
break;
range_start = range_end + 1;
if (key_type == BTRFS_DIR_LOG_ITEM_KEY) {
key_type = BTRFS_DIR_LOG_INDEX_KEY;
dir_key.type = BTRFS_DIR_INDEX_KEY;
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
goto again;
}
out:
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
btrfs_free_path(log_path);
iput(dir);
return ret;
* the running transaction open, so a full commit can't hop
* in and cause problems either.
*/
+ btrfs_scrub_pause_super(root);
write_ctree_super(trans, root->fs_info->tree_root, 1);
+ btrfs_scrub_continue_super(root);
ret = 0;
mutex_lock(&root->log_mutex);
int ret;
int err = 0;
int bytes_del = 0;
+ u64 dir_ino = btrfs_ino(dir);
if (BTRFS_I(dir)->logged_trans < trans->transid)
return 0;
goto out_unlock;
}
- di = btrfs_lookup_dir_item(trans, log, path, dir->i_ino,
+ di = btrfs_lookup_dir_item(trans, log, path, dir_ino,
name, name_len, -1);
if (IS_ERR(di)) {
err = PTR_ERR(di);
bytes_del += name_len;
BUG_ON(ret);
}
- btrfs_release_path(log, path);
- di = btrfs_lookup_dir_index_item(trans, log, path, dir->i_ino,
+ btrfs_release_path(path);
+ di = btrfs_lookup_dir_index_item(trans, log, path, dir_ino,
index, name, name_len, -1);
if (IS_ERR(di)) {
err = PTR_ERR(di);
if (bytes_del) {
struct btrfs_key key;
- key.objectid = dir->i_ino;
+ key.objectid = dir_ino;
key.offset = 0;
key.type = BTRFS_INODE_ITEM_KEY;
- btrfs_release_path(log, path);
+ btrfs_release_path(path);
ret = btrfs_search_slot(trans, log, &key, path, 0, 1);
if (ret < 0) {
btrfs_mark_buffer_dirty(path->nodes[0]);
} else
ret = 0;
- btrfs_release_path(log, path);
+ btrfs_release_path(path);
}
fail:
btrfs_free_path(path);
log = root->log_root;
mutex_lock(&BTRFS_I(inode)->log_mutex);
- ret = btrfs_del_inode_ref(trans, log, name, name_len, inode->i_ino,
+ ret = btrfs_del_inode_ref(trans, log, name, name_len, btrfs_ino(inode),
dirid, &index);
mutex_unlock(&BTRFS_I(inode)->log_mutex);
if (ret == -ENOSPC) {
struct btrfs_dir_log_item);
btrfs_set_dir_log_end(path->nodes[0], item, last_offset);
btrfs_mark_buffer_dirty(path->nodes[0]);
- btrfs_release_path(log, path);
+ btrfs_release_path(path);
return 0;
}
int nritems;
u64 first_offset = min_offset;
u64 last_offset = (u64)-1;
+ u64 ino = btrfs_ino(inode);
log = root->log_root;
- max_key.objectid = inode->i_ino;
+ max_key.objectid = ino;
max_key.offset = (u64)-1;
max_key.type = key_type;
- min_key.objectid = inode->i_ino;
+ min_key.objectid = ino;
min_key.type = key_type;
min_key.offset = min_offset;
* we didn't find anything from this transaction, see if there
* is anything at all
*/
- if (ret != 0 || min_key.objectid != inode->i_ino ||
- min_key.type != key_type) {
- min_key.objectid = inode->i_ino;
+ if (ret != 0 || min_key.objectid != ino || min_key.type != key_type) {
+ min_key.objectid = ino;
min_key.type = key_type;
min_key.offset = (u64)-1;
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0);
if (ret < 0) {
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
return ret;
}
- ret = btrfs_previous_item(root, path, inode->i_ino, key_type);
+ ret = btrfs_previous_item(root, path, ino, key_type);
/* if ret == 0 there are items for this type,
* create a range to tell us the last key of this type.
}
/* go backward to find any previous key */
- ret = btrfs_previous_item(root, path, inode->i_ino, key_type);
+ ret = btrfs_previous_item(root, path, ino, key_type);
if (ret == 0) {
struct btrfs_key tmp;
btrfs_item_key_to_cpu(path->nodes[0], &tmp, path->slots[0]);
}
}
}
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
/* find the first key from this transaction again */
ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0);
for (i = path->slots[0]; i < nritems; i++) {
btrfs_item_key_to_cpu(src, &min_key, i);
- if (min_key.objectid != inode->i_ino ||
- min_key.type != key_type)
+ if (min_key.objectid != ino || min_key.type != key_type)
goto done;
ret = overwrite_item(trans, log, dst_path, src, i,
&min_key);
goto done;
}
btrfs_item_key_to_cpu(path->nodes[0], &tmp, path->slots[0]);
- if (tmp.objectid != inode->i_ino || tmp.type != key_type) {
+ if (tmp.objectid != ino || tmp.type != key_type) {
last_offset = (u64)-1;
goto done;
}
}
}
done:
- btrfs_release_path(root, path);
- btrfs_release_path(log, dst_path);
+ btrfs_release_path(path);
+ btrfs_release_path(dst_path);
if (err == 0) {
*last_offset_ret = last_offset;
* is valid
*/
ret = insert_dir_log_key(trans, log, path, key_type,
- inode->i_ino, first_offset,
- last_offset);
+ ino, first_offset, last_offset);
if (ret)
err = ret;
}
ret = btrfs_del_item(trans, log, path);
BUG_ON(ret);
- btrfs_release_path(log, path);
+ btrfs_release_path(path);
}
- btrfs_release_path(log, path);
+ btrfs_release_path(path);
return ret;
}
ret = btrfs_lookup_csums_range(
log->fs_info->csum_root,
ds + cs, ds + cs + cl - 1,
- &ordered_sums);
+ &ordered_sums, 0);
BUG_ON(ret);
}
}
}
btrfs_mark_buffer_dirty(dst_path->nodes[0]);
- btrfs_release_path(log, dst_path);
+ btrfs_release_path(dst_path);
kfree(ins_data);
/*
int nritems;
int ins_start_slot = 0;
int ins_nr;
+ u64 ino = btrfs_ino(inode);
log = root->log_root;
return -ENOMEM;
}
- min_key.objectid = inode->i_ino;
+ min_key.objectid = ino;
min_key.type = BTRFS_INODE_ITEM_KEY;
min_key.offset = 0;
- max_key.objectid = inode->i_ino;
+ max_key.objectid = ino;
/* today the code can only do partial logging of directories */
if (!S_ISDIR(inode->i_mode))
max_key.type = (u8)-1;
max_key.offset = (u64)-1;
+ ret = btrfs_commit_inode_delayed_items(trans, inode);
+ if (ret) {
+ btrfs_free_path(path);
+ btrfs_free_path(dst_path);
+ return ret;
+ }
+
mutex_lock(&BTRFS_I(inode)->log_mutex);
/*
if (inode_only == LOG_INODE_EXISTS)
max_key_type = BTRFS_XATTR_ITEM_KEY;
- ret = drop_objectid_items(trans, log, path,
- inode->i_ino, max_key_type);
+ ret = drop_objectid_items(trans, log, path, ino, max_key_type);
} else {
ret = btrfs_truncate_inode_items(trans, log, inode, 0, 0);
}
break;
again:
/* note, ins_nr might be > 0 here, cleanup outside the loop */
- if (min_key.objectid != inode->i_ino)
+ if (min_key.objectid != ino)
break;
if (min_key.type > max_key.type)
break;
}
ins_nr = 0;
}
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
if (min_key.offset < (u64)-1)
min_key.offset++;
}
WARN_ON(ins_nr);
if (inode_only == LOG_INODE_ALL && S_ISDIR(inode->i_mode)) {
- btrfs_release_path(root, path);
- btrfs_release_path(log, dst_path);
+ btrfs_release_path(path);
+ btrfs_release_path(dst_path);
ret = log_directory_changes(trans, root, inode, path, dst_path);
if (ret) {
err = ret;
}
btrfs_item_key_to_cpu(path->nodes[0], &found_key,
path->slots[0]);
- btrfs_release_path(log_root_tree, path);
+ btrfs_release_path(path);
if (found_key.objectid != BTRFS_TREE_LOG_OBJECTID)
break;
if (found_key.offset == 0)
break;
}
- btrfs_release_path(log_root_tree, path);
+ btrfs_release_path(path);
/* step one is to pin it all, step two is to replay just inodes */
if (wc.pin) {
struct btrfs_device *device);
static int btrfs_relocate_sys_chunks(struct btrfs_root *root);
- #define map_lookup_size(n) (sizeof(struct map_lookup) + \
- (sizeof(struct btrfs_bio_stripe) * (n)))
-
static DEFINE_MUTEX(uuid_mutex);
static LIST_HEAD(fs_uuids);
-void btrfs_lock_volumes(void)
-{
- mutex_lock(&uuid_mutex);
-}
-
-void btrfs_unlock_volumes(void)
-{
- mutex_unlock(&uuid_mutex);
-}
-
static void lock_chunks(struct btrfs_root *root)
{
mutex_lock(&root->fs_info->chunk_mutex);
struct bio *cur;
int again = 0;
unsigned long num_run;
- unsigned long num_sync_run;
unsigned long batch_run = 0;
unsigned long limit;
unsigned long last_waited = 0;
int force_reg = 0;
+ struct blk_plug plug;
+
+ /*
+ * this function runs all the bios we've collected for
+ * a particular device. We don't want to wander off to
+ * another device without first sending all of these down.
+ * So, setup a plug here and finish it off before we return
+ */
+ blk_start_plug(&plug);
bdi = blk_get_backing_dev_info(device->bdev);
fs_info = device->dev_root->fs_info;
limit = btrfs_async_submit_limit(fs_info);
limit = limit * 2 / 3;
- /* we want to make sure that every time we switch from the sync
- * list to the normal list, we unplug
- */
- num_sync_run = 0;
-
loop:
spin_lock(&device->io_lock);
spin_unlock(&device->io_lock);
- /*
- * if we're doing the regular priority list, make sure we unplug
- * for any high prio bios we've sent down
- */
- if (pending_bios == &device->pending_bios && num_sync_run > 0) {
- num_sync_run = 0;
- blk_run_backing_dev(bdi, NULL);
- }
-
while (pending) {
rmb();
BUG_ON(atomic_read(&cur->bi_cnt) == 0);
- if (cur->bi_rw & REQ_SYNC)
- num_sync_run++;
-
submit_bio(cur->bi_rw, cur);
num_run++;
batch_run++;
- if (need_resched()) {
- if (num_sync_run) {
- blk_run_backing_dev(bdi, NULL);
- num_sync_run = 0;
- }
+ if (need_resched())
cond_resched();
- }
/*
* we made progress, there is more work to do and the bdi
* against it before looping
*/
last_waited = ioc->last_waited;
- if (need_resched()) {
- if (num_sync_run) {
- blk_run_backing_dev(bdi, NULL);
- num_sync_run = 0;
- }
+ if (need_resched())
cond_resched();
- }
continue;
}
spin_lock(&device->io_lock);
}
}
- if (num_sync_run) {
- num_sync_run = 0;
- blk_run_backing_dev(bdi, NULL);
- }
- /*
- * IO has already been through a long path to get here. Checksumming,
- * async helper threads, perhaps compression. We've done a pretty
- * good job of collecting a batch of IO and should just unplug
- * the device right away.
- *
- * This will help anyone who is waiting on the IO, they might have
- * already unplugged, but managed to do so before the bio they
- * cared about found its way down here.
- */
- blk_run_backing_dev(bdi, NULL);
-
cond_resched();
if (again)
goto loop;
spin_unlock(&device->io_lock);
done:
+ blk_finish_plug(&plug);
return 0;
}
/* we don't want to overwrite the superblock on the drive,
* so we make sure to start at an offset of at least 1MB
*/
- search_start = 1024 * 1024;
-
- if (root->fs_info->alloc_start + num_bytes <= search_end)
- search_start = max(root->fs_info->alloc_start, search_start);
+ search_start = max(root->fs_info->alloc_start, 1024ull * 1024);
max_hole_start = search_start;
max_hole_size = 0;
goto error_undo;
device->in_fs_metadata = 0;
+ btrfs_scrub_cancel_dev(root, device);
/*
* the device list mutex makes sure that we don't change
goto error;
leaf = path->nodes[0];
btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
continue;
}
chunk = btrfs_item_ptr(leaf, path->slots[0],
struct btrfs_chunk);
chunk_type = btrfs_chunk_type(leaf, chunk);
- btrfs_release_path(chunk_root, path);
+ btrfs_release_path(path);
if (chunk_type & BTRFS_BLOCK_GROUP_SYSTEM) {
ret = btrfs_relocate_chunk(chunk_root, chunk_tree,
if (found_key.offset == 0)
break;
- btrfs_release_path(chunk_root, path);
+ btrfs_release_path(path);
ret = btrfs_relocate_chunk(chunk_root,
chunk_root->root_key.objectid,
found_key.objectid,
goto done;
if (ret) {
ret = 0;
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
break;
}
btrfs_item_key_to_cpu(l, &key, path->slots[0]);
if (key.objectid != device->devid) {
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
break;
}
length = btrfs_dev_extent_length(l, dev_extent);
if (key.offset + length <= new_size) {
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
break;
}
chunk_tree = btrfs_dev_extent_chunk_tree(l, dev_extent);
chunk_objectid = btrfs_dev_extent_chunk_objectid(l, dev_extent);
chunk_offset = btrfs_dev_extent_chunk_offset(l, dev_extent);
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
ret = btrfs_relocate_chunk(root, chunk_tree, chunk_objectid,
chunk_offset);
return 0;
}
-static noinline u64 chunk_bytes_by_type(u64 type, u64 calc_size,
- int num_stripes, int sub_stripes)
+/*
+ * sort the devices in descending order by max_avail, total_avail
+ */
+static int btrfs_cmp_device_info(const void *a, const void *b)
{
- if (type & (BTRFS_BLOCK_GROUP_RAID1 | BTRFS_BLOCK_GROUP_DUP))
- return calc_size;
- else if (type & BTRFS_BLOCK_GROUP_RAID10)
- return calc_size * (num_stripes / sub_stripes);
- else
- return calc_size * num_stripes;
-}
+ const struct btrfs_device_info *di_a = a;
+ const struct btrfs_device_info *di_b = b;
-/* Used to sort the devices by max_avail(descending sort) */
-int btrfs_cmp_device_free_bytes(const void *dev_info1, const void *dev_info2)
-{
- if (((struct btrfs_device_info *)dev_info1)->max_avail >
- ((struct btrfs_device_info *)dev_info2)->max_avail)
+ if (di_a->max_avail > di_b->max_avail)
return -1;
- else if (((struct btrfs_device_info *)dev_info1)->max_avail <
- ((struct btrfs_device_info *)dev_info2)->max_avail)
+ if (di_a->max_avail < di_b->max_avail)
return 1;
- else
- return 0;
+ if (di_a->total_avail > di_b->total_avail)
+ return -1;
+ if (di_a->total_avail < di_b->total_avail)
+ return 1;
+ return 0;
}
-static int __btrfs_calc_nstripes(struct btrfs_fs_devices *fs_devices, u64 type,
- int *num_stripes, int *min_stripes,
- int *sub_stripes)
+static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans,
+ struct btrfs_root *extent_root,
+ struct map_lookup **map_ret,
+ u64 *num_bytes_out, u64 *stripe_size_out,
+ u64 start, u64 type)
{
- *num_stripes = 1;
- *min_stripes = 1;
- *sub_stripes = 0;
+ struct btrfs_fs_info *info = extent_root->fs_info;
+ struct btrfs_fs_devices *fs_devices = info->fs_devices;
+ struct list_head *cur;
+ struct map_lookup *map = NULL;
+ struct extent_map_tree *em_tree;
+ struct extent_map *em;
+ struct btrfs_device_info *devices_info = NULL;
+ u64 total_avail;
+ int num_stripes; /* total number of stripes to allocate */
+ int sub_stripes; /* sub_stripes info for map */
+ int dev_stripes; /* stripes per dev */
+ int devs_max; /* max devs to use */
+ int devs_min; /* min devs needed */
+ int devs_increment; /* ndevs has to be a multiple of this */
+ int ncopies; /* how many copies to data has */
+ int ret;
+ u64 max_stripe_size;
+ u64 max_chunk_size;
+ u64 stripe_size;
+ u64 num_bytes;
+ int ndevs;
+ int i;
+ int j;
- if (type & (BTRFS_BLOCK_GROUP_RAID0)) {
- *num_stripes = fs_devices->rw_devices;
- *min_stripes = 2;
- }
- if (type & (BTRFS_BLOCK_GROUP_DUP)) {
- *num_stripes = 2;
- *min_stripes = 2;
- }
- if (type & (BTRFS_BLOCK_GROUP_RAID1)) {
- if (fs_devices->rw_devices < 2)
- return -ENOSPC;
- *num_stripes = 2;
- *min_stripes = 2;
- }
- if (type & (BTRFS_BLOCK_GROUP_RAID10)) {
- *num_stripes = fs_devices->rw_devices;
- if (*num_stripes < 4)
- return -ENOSPC;
- *num_stripes &= ~(u32)1;
- *sub_stripes = 2;
- *min_stripes = 4;
+ if ((type & BTRFS_BLOCK_GROUP_RAID1) &&
+ (type & BTRFS_BLOCK_GROUP_DUP)) {
+ WARN_ON(1);
+ type &= ~BTRFS_BLOCK_GROUP_DUP;
}
- return 0;
-}
+ if (list_empty(&fs_devices->alloc_list))
+ return -ENOSPC;
-static u64 __btrfs_calc_stripe_size(struct btrfs_fs_devices *fs_devices,
- u64 proposed_size, u64 type,
- int num_stripes, int small_stripe)
-{
- int min_stripe_size = 1 * 1024 * 1024;
- u64 calc_size = proposed_size;
- u64 max_chunk_size = calc_size;
- int ncopies = 1;
+ sub_stripes = 1;
+ dev_stripes = 1;
+ devs_increment = 1;
+ ncopies = 1;
+ devs_max = 0; /* 0 == as many as possible */
+ devs_min = 1;
- if (type & (BTRFS_BLOCK_GROUP_RAID1 |
- BTRFS_BLOCK_GROUP_DUP |
- BTRFS_BLOCK_GROUP_RAID10))
+ /*
+ * define the properties of each RAID type.
+ * FIXME: move this to a global table and use it in all RAID
+ * calculation code
+ */
+ if (type & (BTRFS_BLOCK_GROUP_DUP)) {
+ dev_stripes = 2;
+ ncopies = 2;
+ devs_max = 1;
+ } else if (type & (BTRFS_BLOCK_GROUP_RAID0)) {
+ devs_min = 2;
+ } else if (type & (BTRFS_BLOCK_GROUP_RAID1)) {
+ devs_increment = 2;
ncopies = 2;
+ devs_max = 2;
+ devs_min = 2;
+ } else if (type & (BTRFS_BLOCK_GROUP_RAID10)) {
+ sub_stripes = 2;
+ devs_increment = 2;
+ ncopies = 2;
+ devs_min = 4;
+ } else {
+ devs_max = 1;
+ }
if (type & BTRFS_BLOCK_GROUP_DATA) {
- max_chunk_size = 10 * calc_size;
- min_stripe_size = 64 * 1024 * 1024;
+ max_stripe_size = 1024 * 1024 * 1024;
+ max_chunk_size = 10 * max_stripe_size;
} else if (type & BTRFS_BLOCK_GROUP_METADATA) {
- max_chunk_size = 256 * 1024 * 1024;
- min_stripe_size = 32 * 1024 * 1024;
+ max_stripe_size = 256 * 1024 * 1024;
+ max_chunk_size = max_stripe_size;
} else if (type & BTRFS_BLOCK_GROUP_SYSTEM) {
- calc_size = 8 * 1024 * 1024;
- max_chunk_size = calc_size * 2;
- min_stripe_size = 1 * 1024 * 1024;
+ max_stripe_size = 8 * 1024 * 1024;
+ max_chunk_size = 2 * max_stripe_size;
+ } else {
+ printk(KERN_ERR "btrfs: invalid chunk type 0x%llx requested\n",
+ type);
+ BUG_ON(1);
}
/* we don't want a chunk larger than 10% of writeable space */
max_chunk_size = min(div_factor(fs_devices->total_rw_bytes, 1),
max_chunk_size);
- if (calc_size * num_stripes > max_chunk_size * ncopies) {
- calc_size = max_chunk_size * ncopies;
- do_div(calc_size, num_stripes);
- do_div(calc_size, BTRFS_STRIPE_LEN);
- calc_size *= BTRFS_STRIPE_LEN;
- }
+ devices_info = kzalloc(sizeof(*devices_info) * fs_devices->rw_devices,
+ GFP_NOFS);
+ if (!devices_info)
+ return -ENOMEM;
- /* we don't want tiny stripes */
- if (!small_stripe)
- calc_size = max_t(u64, min_stripe_size, calc_size);
+ cur = fs_devices->alloc_list.next;
/*
- * we're about to do_div by the BTRFS_STRIPE_LEN so lets make sure
- * we end up with something bigger than a stripe
+ * in the first pass through the devices list, we gather information
+ * about the available holes on each device.
*/
- calc_size = max_t(u64, calc_size, BTRFS_STRIPE_LEN);
-
- do_div(calc_size, BTRFS_STRIPE_LEN);
- calc_size *= BTRFS_STRIPE_LEN;
+ ndevs = 0;
+ while (cur != &fs_devices->alloc_list) {
+ struct btrfs_device *device;
+ u64 max_avail;
+ u64 dev_offset;
- return calc_size;
-}
-
-static struct map_lookup *__shrink_map_lookup_stripes(struct map_lookup *map,
- int num_stripes)
-{
- struct map_lookup *new;
- size_t len = map_lookup_size(num_stripes);
-
- BUG_ON(map->num_stripes < num_stripes);
-
- if (map->num_stripes == num_stripes)
- return map;
-
- new = kmalloc(len, GFP_NOFS);
- if (!new) {
- /* just change map->num_stripes */
- map->num_stripes = num_stripes;
- return map;
- }
-
- memcpy(new, map, len);
- new->num_stripes = num_stripes;
- kfree(map);
- return new;
-}
-
-/*
- * helper to allocate device space from btrfs_device_info, in which we stored
- * max free space information of every device. It is used when we can not
- * allocate chunks by default size.
- *
- * By this helper, we can allocate a new chunk as larger as possible.
- */
-static int __btrfs_alloc_tiny_space(struct btrfs_trans_handle *trans,
- struct btrfs_fs_devices *fs_devices,
- struct btrfs_device_info *devices,
- int nr_device, u64 type,
- struct map_lookup **map_lookup,
- int min_stripes, u64 *stripe_size)
-{
- int i, index, sort_again = 0;
- int min_devices = min_stripes;
- u64 max_avail, min_free;
- struct map_lookup *map = *map_lookup;
- int ret;
+ device = list_entry(cur, struct btrfs_device, dev_alloc_list);
- if (nr_device < min_stripes)
- return -ENOSPC;
+ cur = cur->next;
- btrfs_descending_sort_devices(devices, nr_device);
+ if (!device->writeable) {
+ printk(KERN_ERR
+ "btrfs: read-only device in alloc_list\n");
+ WARN_ON(1);
+ continue;
+ }
- max_avail = devices[0].max_avail;
- if (!max_avail)
- return -ENOSPC;
+ if (!device->in_fs_metadata)
+ continue;
- for (i = 0; i < nr_device; i++) {
- /*
- * if dev_offset = 0, it means the free space of this device
- * is less than what we need, and we didn't search max avail
- * extent on this device, so do it now.
+ if (device->total_bytes > device->bytes_used)
+ total_avail = device->total_bytes - device->bytes_used;
+ else
+ total_avail = 0;
+ /* avail is off by max(alloc_start, 1MB), but that is the same
+ * for all devices, so it doesn't hurt the sorting later on
*/
- if (!devices[i].dev_offset) {
- ret = find_free_dev_extent(trans, devices[i].dev,
- max_avail,
- &devices[i].dev_offset,
- &devices[i].max_avail);
- if (ret != 0 && ret != -ENOSPC)
- return ret;
- sort_again = 1;
- }
- }
- /* we update the max avail free extent of each devices, sort again */
- if (sort_again)
- btrfs_descending_sort_devices(devices, nr_device);
-
- if (type & BTRFS_BLOCK_GROUP_DUP)
- min_devices = 1;
+ ret = find_free_dev_extent(trans, device,
+ max_stripe_size * dev_stripes,
+ &dev_offset, &max_avail);
+ if (ret && ret != -ENOSPC)
+ goto error;
- if (!devices[min_devices - 1].max_avail)
- return -ENOSPC;
+ if (ret == 0)
+ max_avail = max_stripe_size * dev_stripes;
- max_avail = devices[min_devices - 1].max_avail;
- if (type & BTRFS_BLOCK_GROUP_DUP)
- do_div(max_avail, 2);
+ if (max_avail < BTRFS_STRIPE_LEN * dev_stripes)
+ continue;
- max_avail = __btrfs_calc_stripe_size(fs_devices, max_avail, type,
- min_stripes, 1);
- if (type & BTRFS_BLOCK_GROUP_DUP)
- min_free = max_avail * 2;
- else
- min_free = max_avail;
+ devices_info[ndevs].dev_offset = dev_offset;
+ devices_info[ndevs].max_avail = max_avail;
+ devices_info[ndevs].total_avail = total_avail;
+ devices_info[ndevs].dev = device;
+ ++ndevs;
+ }
- if (min_free > devices[min_devices - 1].max_avail)
- return -ENOSPC;
+ /*
+ * now sort the devices by hole size / available space
+ */
+ sort(devices_info, ndevs, sizeof(struct btrfs_device_info),
+ btrfs_cmp_device_info, NULL);
- map = __shrink_map_lookup_stripes(map, min_stripes);
- *stripe_size = max_avail;
+ /* round down to number of usable stripes */
+ ndevs -= ndevs % devs_increment;
- index = 0;
- for (i = 0; i < min_stripes; i++) {
- map->stripes[i].dev = devices[index].dev;
- map->stripes[i].physical = devices[index].dev_offset;
- if (type & BTRFS_BLOCK_GROUP_DUP) {
- i++;
- map->stripes[i].dev = devices[index].dev;
- map->stripes[i].physical = devices[index].dev_offset +
- max_avail;
- }
- index++;
+ if (ndevs < devs_increment * sub_stripes || ndevs < devs_min) {
+ ret = -ENOSPC;
+ goto error;
}
- *map_lookup = map;
- return 0;
-}
-
-static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans,
- struct btrfs_root *extent_root,
- struct map_lookup **map_ret,
- u64 *num_bytes, u64 *stripe_size,
- u64 start, u64 type)
-{
- struct btrfs_fs_info *info = extent_root->fs_info;
- struct btrfs_device *device = NULL;
- struct btrfs_fs_devices *fs_devices = info->fs_devices;
- struct list_head *cur;
- struct map_lookup *map;
- struct extent_map_tree *em_tree;
- struct extent_map *em;
- struct btrfs_device_info *devices_info;
- struct list_head private_devs;
- u64 calc_size = 1024 * 1024 * 1024;
- u64 min_free;
- u64 avail;
- u64 dev_offset;
- int num_stripes;
- int min_stripes;
- int sub_stripes;
- int min_devices; /* the min number of devices we need */
- int i;
- int ret;
- int index;
+ if (devs_max && ndevs > devs_max)
+ ndevs = devs_max;
+ /*
+ * the primary goal is to maximize the number of stripes, so use as many
+ * devices as possible, even if the stripes are not maximum sized.
+ */
+ stripe_size = devices_info[ndevs-1].max_avail;
+ num_stripes = ndevs * dev_stripes;
- if ((type & BTRFS_BLOCK_GROUP_RAID1) &&
- (type & BTRFS_BLOCK_GROUP_DUP)) {
- WARN_ON(1);
- type &= ~BTRFS_BLOCK_GROUP_DUP;
+ if (stripe_size * num_stripes > max_chunk_size * ncopies) {
+ stripe_size = max_chunk_size * ncopies;
+ do_div(stripe_size, num_stripes);
}
- if (list_empty(&fs_devices->alloc_list))
- return -ENOSPC;
-
- ret = __btrfs_calc_nstripes(fs_devices, type, &num_stripes,
- &min_stripes, &sub_stripes);
- if (ret)
- return ret;
- devices_info = kzalloc(sizeof(*devices_info) * fs_devices->rw_devices,
- GFP_NOFS);
- if (!devices_info)
- return -ENOMEM;
+ do_div(stripe_size, dev_stripes);
+ do_div(stripe_size, BTRFS_STRIPE_LEN);
+ stripe_size *= BTRFS_STRIPE_LEN;
map = kmalloc(map_lookup_size(num_stripes), GFP_NOFS);
if (!map) {
}
map->num_stripes = num_stripes;
- cur = fs_devices->alloc_list.next;
- index = 0;
- i = 0;
-
- calc_size = __btrfs_calc_stripe_size(fs_devices, calc_size, type,
- num_stripes, 0);
-
- if (type & BTRFS_BLOCK_GROUP_DUP) {
- min_free = calc_size * 2;
- min_devices = 1;
- } else {
- min_free = calc_size;
- min_devices = min_stripes;
- }
-
- INIT_LIST_HEAD(&private_devs);
- while (index < num_stripes) {
- device = list_entry(cur, struct btrfs_device, dev_alloc_list);
- BUG_ON(!device->writeable);
- if (device->total_bytes > device->bytes_used)
- avail = device->total_bytes - device->bytes_used;
- else
- avail = 0;
- cur = cur->next;
-
- if (device->in_fs_metadata && avail >= min_free) {
- ret = find_free_dev_extent(trans, device, min_free,
- &devices_info[i].dev_offset,
- &devices_info[i].max_avail);
- if (ret == 0) {
- list_move_tail(&device->dev_alloc_list,
- &private_devs);
- map->stripes[index].dev = device;
- map->stripes[index].physical =
- devices_info[i].dev_offset;
- index++;
- if (type & BTRFS_BLOCK_GROUP_DUP) {
- map->stripes[index].dev = device;
- map->stripes[index].physical =
- devices_info[i].dev_offset +
- calc_size;
- index++;
- }
- } else if (ret != -ENOSPC)
- goto error;
-
- devices_info[i].dev = device;
- i++;
- } else if (device->in_fs_metadata &&
- avail >= BTRFS_STRIPE_LEN) {
- devices_info[i].dev = device;
- devices_info[i].max_avail = avail;
- i++;
- }
-
- if (cur == &fs_devices->alloc_list)
- break;
- }
-
- list_splice(&private_devs, &fs_devices->alloc_list);
- if (index < num_stripes) {
- if (index >= min_stripes) {
- num_stripes = index;
- if (type & (BTRFS_BLOCK_GROUP_RAID10)) {
- num_stripes /= sub_stripes;
- num_stripes *= sub_stripes;
- }
-
- map = __shrink_map_lookup_stripes(map, num_stripes);
- } else if (i >= min_devices) {
- ret = __btrfs_alloc_tiny_space(trans, fs_devices,
- devices_info, i, type,
- &map, min_stripes,
- &calc_size);
- if (ret)
- goto error;
- } else {
- ret = -ENOSPC;
- goto error;
+ for (i = 0; i < ndevs; ++i) {
+ for (j = 0; j < dev_stripes; ++j) {
+ int s = i * dev_stripes + j;
+ map->stripes[s].dev = devices_info[i].dev;
+ map->stripes[s].physical = devices_info[i].dev_offset +
+ j * stripe_size;
}
}
map->sector_size = extent_root->sectorsize;
map->sub_stripes = sub_stripes;
*map_ret = map;
- *stripe_size = calc_size;
- *num_bytes = chunk_bytes_by_type(type, calc_size,
- map->num_stripes, sub_stripes);
+ num_bytes = stripe_size * (num_stripes / ncopies);
- trace_btrfs_chunk_alloc(info->chunk_root, map, start, *num_bytes);
+ *stripe_size_out = stripe_size;
+ *num_bytes_out = num_bytes;
- em = alloc_extent_map(GFP_NOFS);
+ trace_btrfs_chunk_alloc(info->chunk_root, map, start, num_bytes);
+
+ em = alloc_extent_map();
if (!em) {
ret = -ENOMEM;
goto error;
}
em->bdev = (struct block_device *)map;
em->start = start;
- em->len = *num_bytes;
+ em->len = num_bytes;
em->block_start = 0;
em->block_len = em->len;
ret = btrfs_make_block_group(trans, extent_root, 0, type,
BTRFS_FIRST_CHUNK_TREE_OBJECTID,
- start, *num_bytes);
+ start, num_bytes);
BUG_ON(ret);
- index = 0;
- while (index < map->num_stripes) {
- device = map->stripes[index].dev;
- dev_offset = map->stripes[index].physical;
+ for (i = 0; i < map->num_stripes; ++i) {
+ struct btrfs_device *device;
+ u64 dev_offset;
+
+ device = map->stripes[i].dev;
+ dev_offset = map->stripes[i].physical;
ret = btrfs_alloc_dev_extent(trans, device,
info->chunk_root->root_key.objectid,
BTRFS_FIRST_CHUNK_TREE_OBJECTID,
- start, dev_offset, calc_size);
+ start, dev_offset, stripe_size);
BUG_ON(ret);
- index++;
}
kfree(devices_info);
void btrfs_mapping_init(struct btrfs_mapping_tree *tree)
{
- extent_map_tree_init(&tree->map_tree, GFP_NOFS);
+ extent_map_tree_init(&tree->map_tree);
}
void btrfs_mapping_tree_free(struct btrfs_mapping_tree *tree)
static int __btrfs_map_block(struct btrfs_mapping_tree *map_tree, int rw,
u64 logical, u64 *length,
struct btrfs_multi_bio **multi_ret,
- int mirror_num, struct page *unplug_page)
+ int mirror_num)
{
struct extent_map *em;
struct map_lookup *map;
em = lookup_extent_mapping(em_tree, logical, *length);
read_unlock(&em_tree->lock);
- if (!em && unplug_page) {
- kfree(multi);
- return 0;
- }
-
if (!em) {
printk(KERN_CRIT "unable to find logical %llu len %llu\n",
(unsigned long long)logical,
*length = em->len - offset;
}
- if (!multi_ret && !unplug_page)
+ if (!multi_ret)
goto out;
num_stripes = 1;
stripe_nr_end - stripe_nr_orig);
stripe_index = do_div(stripe_nr, map->num_stripes);
} else if (map->type & BTRFS_BLOCK_GROUP_RAID1) {
- if (unplug_page || (rw & (REQ_WRITE | REQ_DISCARD)))
+ if (rw & (REQ_WRITE | REQ_DISCARD))
num_stripes = map->num_stripes;
else if (mirror_num)
stripe_index = mirror_num - 1;
stripe_index = do_div(stripe_nr, factor);
stripe_index *= map->sub_stripes;
- if (unplug_page || (rw & REQ_WRITE))
+ if (rw & REQ_WRITE)
num_stripes = map->sub_stripes;
else if (rw & REQ_DISCARD)
num_stripes = min_t(u64, map->sub_stripes *
}
} else {
for (i = 0; i < num_stripes; i++) {
- if (unplug_page) {
- struct btrfs_device *device;
- struct backing_dev_info *bdi;
-
- device = map->stripes[stripe_index].dev;
- if (device->bdev) {
- bdi = blk_get_backing_dev_info(device->
- bdev);
- if (bdi->unplug_io_fn)
- bdi->unplug_io_fn(bdi,
- unplug_page);
- }
- } else {
- multi->stripes[i].physical =
- map->stripes[stripe_index].physical +
- stripe_offset +
- stripe_nr * map->stripe_len;
- multi->stripes[i].dev =
- map->stripes[stripe_index].dev;
- }
+ multi->stripes[i].physical =
+ map->stripes[stripe_index].physical +
+ stripe_offset +
+ stripe_nr * map->stripe_len;
+ multi->stripes[i].dev =
+ map->stripes[stripe_index].dev;
stripe_index++;
}
}
struct btrfs_multi_bio **multi_ret, int mirror_num)
{
return __btrfs_map_block(map_tree, rw, logical, length, multi_ret,
- mirror_num, NULL);
+ mirror_num);
}
int btrfs_rmap_block(struct btrfs_mapping_tree *map_tree,
return 0;
}
-int btrfs_unplug_page(struct btrfs_mapping_tree *map_tree,
- u64 logical, struct page *page)
-{
- u64 length = PAGE_CACHE_SIZE;
- return __btrfs_map_block(map_tree, READ, logical, &length,
- NULL, 0, page);
-}
-
static void end_bio_multi_stripe(struct bio *bio, int err)
{
struct btrfs_multi_bio *multi = bio->bi_private;
free_extent_map(em);
}
- em = alloc_extent_map(GFP_NOFS);
+ em = alloc_extent_map();
if (!em)
return -ENOMEM;
num_stripes = btrfs_chunk_num_stripes(leaf, chunk);
return ret;
}
-int btrfs_read_super_device(struct btrfs_root *root, struct extent_buffer *buf)
-{
- struct btrfs_dev_item *dev_item;
-
- dev_item = (struct btrfs_dev_item *)offsetof(struct btrfs_super_block,
- dev_item);
- return read_one_dev(root, buf, dev_item);
-}
-
int btrfs_read_sys_array(struct btrfs_root *root)
{
struct btrfs_super_block *super_copy = &root->fs_info->super_copy;
}
if (key.objectid == BTRFS_DEV_ITEMS_OBJECTID) {
key.objectid = 0;
- btrfs_release_path(root, path);
+ btrfs_release_path(path);
goto again;
}
ret = 0;
/* physical drive uuid (or lvm uuid) */
u8 uuid[BTRFS_UUID_SIZE];
+ /* per-device scrub information */
+ struct scrub_dev *scrub_device;
+
struct btrfs_work work;
};
struct btrfs_device *dev;
u64 dev_offset;
u64 max_avail;
+ u64 total_avail;
};
struct map_lookup {
struct btrfs_bio_stripe stripes[];
};
-/* Used to sort the devices by max_avail(descending sort) */
-int btrfs_cmp_device_free_bytes(const void *dev_info1, const void *dev_info2);
-
-/*
- * sort the devices by max_avail, in which max free extent size of each device
- * is stored.(Descending Sort)
- */
-static inline void btrfs_descending_sort_devices(
- struct btrfs_device_info *devices,
- size_t nr_devices)
-{
- sort(devices, nr_devices, sizeof(struct btrfs_device_info),
- btrfs_cmp_device_free_bytes, NULL);
-}
-
+ #define map_lookup_size(n) (sizeof(struct map_lookup) + \
+ (sizeof(struct btrfs_bio_stripe) * (n)))
+
int btrfs_account_dev_extents_size(struct btrfs_device *device, u64 start,
u64 end, u64 *length);
void btrfs_mapping_tree_free(struct btrfs_mapping_tree *tree);
int btrfs_map_bio(struct btrfs_root *root, int rw, struct bio *bio,
int mirror_num, int async_submit);
-int btrfs_read_super_device(struct btrfs_root *root, struct extent_buffer *buf);
int btrfs_open_devices(struct btrfs_fs_devices *fs_devices,
fmode_t flags, void *holder);
int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder,
int btrfs_rm_device(struct btrfs_root *root, char *device_path);
int btrfs_cleanup_fs_uuids(void);
int btrfs_num_copies(struct btrfs_mapping_tree *map_tree, u64 logical, u64 len);
-int btrfs_unplug_page(struct btrfs_mapping_tree *map_tree,
- u64 logical, struct page *page);
int btrfs_grow_device(struct btrfs_trans_handle *trans,
struct btrfs_device *device, u64 new_size);
struct btrfs_device *btrfs_find_device(struct btrfs_root *root, u64 devid,
int btrfs_shrink_device(struct btrfs_device *device, u64 new_size);
int btrfs_init_new_device(struct btrfs_root *root, char *path);
int btrfs_balance(struct btrfs_root *dev_root);
-void btrfs_unlock_volumes(void);
-void btrfs_lock_volumes(void);
int btrfs_chunk_readonly(struct btrfs_root *root, u64 chunk_offset);
int find_free_dev_extent(struct btrfs_trans_handle *trans,
struct btrfs_device *device, u64 num_bytes,