#include "print-tree.h"
#include "xattr.h"
#include "volumes.h"
+#include "version.h"
+#include "export.h"
+#include "compression.h"
#define BTRFS_SUPER_MAGIC 0x9123683E
}
enum {
- Opt_subvol, Opt_nodatasum, Opt_nodatacow, Opt_max_extent,
- Opt_max_inline, Opt_alloc_start, Opt_nobarrier, Opt_ssd, Opt_err,
+ Opt_degraded, Opt_subvol, Opt_device, Opt_nodatasum, Opt_nodatacow,
+ Opt_max_extent, Opt_max_inline, Opt_alloc_start, Opt_nobarrier,
+ Opt_ssd, Opt_thread_pool, Opt_noacl, Opt_compress, Opt_err,
};
static match_table_t tokens = {
+ {Opt_degraded, "degraded"},
{Opt_subvol, "subvol=%s"},
+ {Opt_device, "device=%s"},
{Opt_nodatasum, "nodatasum"},
{Opt_nodatacow, "nodatacow"},
{Opt_nobarrier, "nobarrier"},
{Opt_max_extent, "max_extent=%s"},
{Opt_max_inline, "max_inline=%s"},
{Opt_alloc_start, "alloc_start=%s"},
+ {Opt_thread_pool, "thread_pool=%d"},
+ {Opt_compress, "compress"},
{Opt_ssd, "ssd"},
- {Opt_err, NULL}
+ {Opt_noacl, "noacl"},
+ {Opt_err, NULL},
};
u64 btrfs_parse_size(char *str)
return res;
}
-static int parse_options (char * options,
- struct btrfs_root *root,
- char **subvol_name)
+/*
+ * Regular mount options parser. Everything that is needed only when
+ * reading in a new superblock is parsed here.
+ */
+int btrfs_parse_options(struct btrfs_root *root, char *options)
{
- char * p;
- struct btrfs_fs_info *info = NULL;
+ struct btrfs_fs_info *info = root->fs_info;
substring_t args[MAX_OPT_ARGS];
+ char *p, *num;
+ int intarg;
if (!options)
- return 1;
+ return 0;
/*
* strsep changes the string, duplicate it because parse_options
if (!options)
return -ENOMEM;
- if (root)
- info = root->fs_info;
- while ((p = strsep (&options, ",")) != NULL) {
+ while ((p = strsep(&options, ",")) != NULL) {
int token;
if (!*p)
continue;
token = match_token(p, tokens, args);
switch (token) {
+ case Opt_degraded:
+ printk(KERN_INFO "btrfs: allowing degraded mounts\n");
+ btrfs_set_opt(info->mount_opt, DEGRADED);
+ break;
case Opt_subvol:
- if (subvol_name) {
- *subvol_name = match_strdup(&args[0]);
- }
+ case Opt_device:
+ /*
+ * These are parsed by btrfs_parse_early_options
+ * and can be happily ignored here.
+ */
break;
case Opt_nodatasum:
- if (info) {
- printk("btrfs: setting nodatacsum\n");
- btrfs_set_opt(info->mount_opt, NODATASUM);
- }
+ printk(KERN_INFO "btrfs: setting nodatacsum\n");
+ btrfs_set_opt(info->mount_opt, NODATASUM);
break;
case Opt_nodatacow:
- if (info) {
- printk("btrfs: setting nodatacow\n");
- btrfs_set_opt(info->mount_opt, NODATACOW);
- btrfs_set_opt(info->mount_opt, NODATASUM);
- }
+ printk(KERN_INFO "btrfs: setting nodatacow\n");
+ btrfs_set_opt(info->mount_opt, NODATACOW);
+ btrfs_set_opt(info->mount_opt, NODATASUM);
+ break;
+ case Opt_compress:
+ printk(KERN_INFO "btrfs: use compression\n");
+ btrfs_set_opt(info->mount_opt, COMPRESS);
break;
case Opt_ssd:
- if (info) {
- printk("btrfs: use ssd allocation scheme\n");
- btrfs_set_opt(info->mount_opt, SSD);
- }
+ printk(KERN_INFO "btrfs: use ssd allocation scheme\n");
+ btrfs_set_opt(info->mount_opt, SSD);
break;
case Opt_nobarrier:
- if (info) {
- printk("btrfs: turning off barriers\n");
- btrfs_set_opt(info->mount_opt, NOBARRIER);
+ printk(KERN_INFO "btrfs: turning off barriers\n");
+ btrfs_set_opt(info->mount_opt, NOBARRIER);
+ break;
+ case Opt_thread_pool:
+ intarg = 0;
+ match_int(&args[0], &intarg);
+ if (intarg) {
+ info->thread_pool_size = intarg;
+ printk(KERN_INFO "btrfs: thread pool %d\n",
+ info->thread_pool_size);
}
break;
case Opt_max_extent:
- if (info) {
- char *num = match_strdup(&args[0]);
- if (num) {
- info->max_extent =
- btrfs_parse_size(num);
- kfree(num);
-
- info->max_extent = max_t(u64,
- info->max_extent,
- root->sectorsize);
- printk("btrfs: max_extent at %Lu\n",
- info->max_extent);
- }
+ num = match_strdup(&args[0]);
+ if (num) {
+ info->max_extent = btrfs_parse_size(num);
+ kfree(num);
+
+ info->max_extent = max_t(u64,
+ info->max_extent, root->sectorsize);
+ printk(KERN_INFO "btrfs: max_extent at %llu\n",
+ info->max_extent);
}
break;
case Opt_max_inline:
- if (info) {
- char *num = match_strdup(&args[0]);
- if (num) {
- info->max_inline =
- btrfs_parse_size(num);
- kfree(num);
+ num = match_strdup(&args[0]);
+ if (num) {
+ info->max_inline = btrfs_parse_size(num);
+ kfree(num);
+ if (info->max_inline) {
info->max_inline = max_t(u64,
- info->max_inline,
- root->sectorsize);
- printk("btrfs: max_inline at %Lu\n",
- info->max_inline);
+ info->max_inline,
+ root->sectorsize);
}
+ printk(KERN_INFO "btrfs: max_inline at %llu\n",
+ info->max_inline);
}
break;
case Opt_alloc_start:
- if (info) {
- char *num = match_strdup(&args[0]);
- if (num) {
- info->alloc_start =
- btrfs_parse_size(num);
- kfree(num);
- printk("btrfs: allocations start at "
- "%Lu\n", info->alloc_start);
- }
+ num = match_strdup(&args[0]);
+ if (num) {
+ info->alloc_start = btrfs_parse_size(num);
+ kfree(num);
+ printk(KERN_INFO
+ "btrfs: allocations start at %llu\n",
+ info->alloc_start);
}
break;
+ case Opt_noacl:
+ root->fs_info->sb->s_flags &= ~MS_POSIXACL;
+ break;
default:
break;
}
}
kfree(options);
- return 1;
+ return 0;
+}
+
+/*
+ * Parse mount options that are required early in the mount process.
+ *
+ * All other options will be parsed on much later in the mount process and
+ * only when we need to allocate a new super block.
+ */
+static int btrfs_parse_early_options(const char *options, int flags,
+ void *holder, char **subvol_name,
+ struct btrfs_fs_devices **fs_devices)
+{
+ substring_t args[MAX_OPT_ARGS];
+ char *opts, *p;
+ int error = 0;
+
+ if (!options)
+ goto out;
+
+ /*
+ * strsep changes the string, duplicate it because parse_options
+ * gets called twice
+ */
+ opts = kstrdup(options, GFP_KERNEL);
+ if (!opts)
+ return -ENOMEM;
+
+ while ((p = strsep(&opts, ",")) != NULL) {
+ int token;
+ if (!*p)
+ continue;
+
+ token = match_token(p, tokens, args);
+ switch (token) {
+ case Opt_subvol:
+ *subvol_name = match_strdup(&args[0]);
+ break;
+ case Opt_device:
+ error = btrfs_scan_one_device(match_strdup(&args[0]),
+ flags, holder, fs_devices);
+ if (error)
+ goto out_free_opts;
+ break;
+ default:
+ break;
+ }
+ }
+
+ out_free_opts:
+ kfree(opts);
+ out:
+ /*
+ * If no subvolume name is specified we use the default one. Allocate
+ * a copy of the string "default" here so that code later in the
+ * mount path doesn't care if it's the default volume or another one.
+ */
+ if (!*subvol_name) {
+ *subvol_name = kstrdup("default", GFP_KERNEL);
+ if (!*subvol_name)
+ return -ENOMEM;
+ }
+ return error;
}
static int btrfs_fill_super(struct super_block * sb,
sb->s_maxbytes = MAX_LFS_FILESIZE;
sb->s_magic = BTRFS_SUPER_MAGIC;
sb->s_op = &btrfs_super_ops;
+ sb->s_export_op = &btrfs_export_ops;
sb->s_xattr = btrfs_xattr_handlers;
sb->s_time_gran = 1;
+ sb->s_flags |= MS_POSIXACL;
- tree_root = open_ctree(sb, fs_devices);
+ tree_root = open_ctree(sb, fs_devices, (char *)data);
if (IS_ERR(tree_root)) {
printk("btrfs: open_ctree failed\n");
goto fail_close;
}
- parse_options((char *)data, tree_root, NULL);
-
/* this does the super kobj at the same time */
err = btrfs_sysfs_add_super(tree_root->fs_info);
if (err)
goto fail_close;
sb->s_root = root_dentry;
- btrfs_transaction_queue_work(tree_root, HZ * 30);
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,25)
save_mount_options(sb, data);
-#endif
-
return 0;
fail_close:
return err;
}
-static int btrfs_sync_fs(struct super_block *sb, int wait)
+int btrfs_sync_fs(struct super_block *sb, int wait)
{
struct btrfs_trans_handle *trans;
struct btrfs_root *root;
int ret;
root = btrfs_sb(sb);
+ if (sb->s_flags & MS_RDONLY)
+ return 0;
+
sb->s_dirt = 0;
if (!wait) {
filemap_flush(root->fs_info->btree_inode->i_mapping);
return 0;
}
+
+ btrfs_start_delalloc_inodes(root);
+ btrfs_wait_ordered_extents(root, 0);
+
btrfs_clean_old_snapshots(root);
- mutex_lock(&root->fs_info->fs_mutex);
- btrfs_defrag_dirty_roots(root->fs_info);
trans = btrfs_start_transaction(root, 1);
ret = btrfs_commit_transaction(trans, root);
sb->s_dirt = 0;
- mutex_unlock(&root->fs_info->fs_mutex);
return ret;
}
sb->s_dirt = 0;
}
-/*
- * This is almost a copy of get_sb_bdev in fs/super.c.
- * We need the local copy to allow direct mounting of
- * subvolumes, but this could be easily integrated back
- * into the generic version. --hch
- */
-
-/* start copy & paste */
-static int set_bdev_super(struct super_block *s, void *data)
+static int btrfs_test_super(struct super_block *s, void *data)
{
- s->s_bdev = data;
- s->s_dev = s->s_bdev->bd_dev;
- return 0;
-}
+ struct btrfs_fs_devices *test_fs_devices = data;
+ struct btrfs_root *root = btrfs_sb(s);
-static int test_bdev_super(struct super_block *s, void *data)
-{
- return (void *)s->s_bdev == data;
+ return root->fs_info->fs_devices == test_fs_devices;
}
-int btrfs_get_sb_bdev(struct file_system_type *fs_type,
- int flags, const char *dev_name, void *data,
- struct vfsmount *mnt, const char *subvol)
+/*
+ * Find a superblock for the given device / mount point.
+ *
+ * Note: This is based on get_sb_bdev from fs/super.c with a few additions
+ * for multiple device setup. Make sure to keep it in sync.
+ */
+static int btrfs_get_sb(struct file_system_type *fs_type, int flags,
+ const char *dev_name, void *data, struct vfsmount *mnt)
{
+ char *subvol_name = NULL;
struct block_device *bdev = NULL;
struct super_block *s;
struct dentry *root;
struct btrfs_fs_devices *fs_devices = NULL;
int error = 0;
+ error = btrfs_parse_early_options(data, flags, fs_type,
+ &subvol_name, &fs_devices);
+ if (error)
+ goto error;
+
error = btrfs_scan_one_device(dev_name, flags, fs_type, &fs_devices);
if (error)
- return error;
+ goto error_free_subvol_name;
error = btrfs_open_devices(fs_devices, flags, fs_type);
if (error)
- return error;
+ goto error_free_subvol_name;
- bdev = fs_devices->lowest_bdev;
- /*
- * once the super is inserted into the list by sget, s_umount
- * will protect the lockfs code from trying to start a snapshot
- * while we are mounting
- */
- down(&bdev->bd_mount_sem);
- s = sget(fs_type, test_bdev_super, set_bdev_super, bdev);
- up(&bdev->bd_mount_sem);
+ if (!(flags & MS_RDONLY) && fs_devices->rw_devices == 0) {
+ error = -EACCES;
+ goto error_close_devices;
+ }
+
+ bdev = fs_devices->latest_bdev;
+ s = sget(fs_type, btrfs_test_super, set_anon_super, fs_devices);
if (IS_ERR(s))
goto error_s;
up_write(&s->s_umount);
deactivate_super(s);
error = -EBUSY;
- goto error_bdev;
+ goto error_close_devices;
}
- close_bdev_excl(bdev);
+ btrfs_close_devices(fs_devices);
} else {
char b[BDEVNAME_SIZE];
s->s_flags = flags;
strlcpy(s->s_id, bdevname(bdev, b), sizeof(s->s_id));
- sb_set_blocksize(s, block_size(bdev));
error = btrfs_fill_super(s, fs_devices, data,
flags & MS_SILENT ? 1 : 0);
if (error) {
s->s_flags |= MS_ACTIVE;
}
- if (subvol) {
- root = lookup_one_len(subvol, s->s_root, strlen(subvol));
+ if (!strcmp(subvol_name, "."))
+ root = dget(s->s_root);
+ else {
+ mutex_lock(&s->s_root->d_inode->i_mutex);
+ root = lookup_one_len(subvol_name, s->s_root, strlen(subvol_name));
+ mutex_unlock(&s->s_root->d_inode->i_mutex);
if (IS_ERR(root)) {
up_write(&s->s_umount);
deactivate_super(s);
error = -ENXIO;
goto error;
}
- } else {
- root = dget(s->s_root);
}
mnt->mnt_sb = s;
mnt->mnt_root = root;
+
+ kfree(subvol_name);
return 0;
error_s:
error = PTR_ERR(s);
-error_bdev:
+error_close_devices:
btrfs_close_devices(fs_devices);
+error_free_subvol_name:
+ kfree(subvol_name);
error:
return error;
}
-/* end copy & paste */
-static int btrfs_get_sb(struct file_system_type *fs_type,
- int flags, const char *dev_name, void *data, struct vfsmount *mnt)
+static int btrfs_remount(struct super_block *sb, int *flags, char *data)
{
+ struct btrfs_root *root = btrfs_sb(sb);
int ret;
- char *subvol_name = NULL;
- parse_options((char *)data, NULL, &subvol_name);
- ret = btrfs_get_sb_bdev(fs_type, flags, dev_name, data, mnt,
- subvol_name ? subvol_name : "default");
- if (subvol_name)
- kfree(subvol_name);
- return ret;
+ if ((*flags & MS_RDONLY) == (sb->s_flags & MS_RDONLY))
+ return 0;
+
+ if (*flags & MS_RDONLY) {
+ sb->s_flags |= MS_RDONLY;
+
+ ret = btrfs_commit_super(root);
+ WARN_ON(ret);
+ } else {
+ if (root->fs_info->fs_devices->rw_devices == 0)
+ return -EACCES;
+
+ if (btrfs_super_log_root(&root->fs_info->super_copy) != 0)
+ return -EINVAL;
+
+ ret = btrfs_cleanup_reloc_trees(root);
+ WARN_ON(ret);
+
+ ret = btrfs_cleanup_fs_roots(root->fs_info);
+ WARN_ON(ret);
+
+ sb->s_flags &= ~MS_RDONLY;
+ }
+
+ return 0;
}
static int btrfs_statfs(struct dentry *dentry, struct kstatfs *buf)
struct btrfs_root *root = btrfs_sb(dentry->d_sb);
struct btrfs_super_block *disk_super = &root->fs_info->super_copy;
int bits = dentry->d_sb->s_blocksize_bits;
+ __be32 *fsid = (__be32 *)root->fs_info->fsid;
buf->f_namelen = BTRFS_NAME_LEN;
buf->f_blocks = btrfs_super_total_bytes(disk_super) >> bits;
buf->f_bavail = buf->f_bfree;
buf->f_bsize = dentry->d_sb->s_blocksize;
buf->f_type = BTRFS_SUPER_MAGIC;
+ /* We treat it as constant endianness (it doesn't matter _which_)
+ because we want the fsid to come out the same whether mounted
+ on a big-endian or little-endian host */
+ buf->f_fsid.val[0] = be32_to_cpu(fsid[0]) ^ be32_to_cpu(fsid[2]);
+ buf->f_fsid.val[1] = be32_to_cpu(fsid[1]) ^ be32_to_cpu(fsid[3]);
+ /* Mask in the root object ID too, to disambiguate subvols */
+ buf->f_fsid.val[0] ^= BTRFS_I(dentry->d_inode)->root->objectid >> 32;
+ buf->f_fsid.val[1] ^= BTRFS_I(dentry->d_inode)->root->objectid;
+
return 0;
}
.owner = THIS_MODULE,
.name = "btrfs",
.get_sb = btrfs_get_sb,
- .kill_sb = kill_block_super,
+ .kill_sb = kill_anon_super,
.fs_flags = FS_REQUIRES_DEV,
};
+/*
+ * used by btrfsctl to scan devices when no FS is mounted
+ */
static long btrfs_control_ioctl(struct file *file, unsigned int cmd,
unsigned long arg)
{
struct btrfs_ioctl_vol_args *vol;
struct btrfs_fs_devices *fs_devices;
- int ret;
+ int ret = 0;
int len;
vol = kmalloc(sizeof(*vol), GFP_KERNEL);
}
out:
kfree(vol);
- return 0;
+ return ret;
}
static void btrfs_write_super_lockfs(struct super_block *sb)
{
struct btrfs_root *root = btrfs_sb(sb);
- btrfs_transaction_flush_work(root);
+ mutex_lock(&root->fs_info->transaction_kthread_mutex);
+ mutex_lock(&root->fs_info->cleaner_mutex);
}
static void btrfs_unlockfs(struct super_block *sb)
{
struct btrfs_root *root = btrfs_sb(sb);
- btrfs_transaction_queue_work(root, HZ * 30);
+ mutex_unlock(&root->fs_info->cleaner_mutex);
+ mutex_unlock(&root->fs_info->transaction_kthread_mutex);
}
static struct super_operations btrfs_super_ops = {
.delete_inode = btrfs_delete_inode,
- .put_inode = btrfs_put_inode,
.put_super = btrfs_put_super,
.write_super = btrfs_write_super,
.sync_fs = btrfs_sync_fs,
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,25)
- .read_inode = btrfs_read_locked_inode,
-#else
.show_options = generic_show_options,
-#endif
.write_inode = btrfs_write_inode,
.dirty_inode = btrfs_dirty_inode,
.alloc_inode = btrfs_alloc_inode,
.destroy_inode = btrfs_destroy_inode,
.statfs = btrfs_statfs,
+ .remount_fs = btrfs_remount,
.write_super_lockfs = btrfs_write_super_lockfs,
.unlockfs = btrfs_unlockfs,
};
if (err)
return err;
- btrfs_init_transaction_sys();
err = btrfs_init_cachep();
if (err)
- goto free_transaction_sys;
+ goto free_sysfs;
err = extent_io_init();
if (err)
err = btrfs_interface_init();
if (err)
goto free_extent_map;
+
err = register_filesystem(&btrfs_fs_type);
if (err)
goto unregister_ioctl;
+
+ printk(KERN_INFO "%s loaded\n", BTRFS_BUILD_VERSION);
return 0;
unregister_ioctl:
extent_io_exit();
free_cachep:
btrfs_destroy_cachep();
-free_transaction_sys:
- btrfs_exit_transaction_sys();
+free_sysfs:
btrfs_exit_sysfs();
return err;
}
static void __exit exit_btrfs_fs(void)
{
- btrfs_exit_transaction_sys();
btrfs_destroy_cachep();
extent_map_exit();
extent_io_exit();
unregister_filesystem(&btrfs_fs_type);
btrfs_exit_sysfs();
btrfs_cleanup_fs_uuids();
+ btrfs_zlib_exit();
}
module_init(init_btrfs_fs)