Ext3 mount原理
本質上,Ext3 mount的過程實際上是inode被替代的過程。例如,/dev/sdb塊設備被mount到/mnt/alan目錄。那麼mount這個過程所需要解決的問題就是將/mnt/alan的dentry目錄項所指向的inode屏蔽掉,然後重新定位到/dev/sdb所表示的inode索引節點。在沒有分析閱讀linux vfs mount代碼的時候,我的想法是修改dentry所指向的inode索引節點,以此實現mount文件系統的訪問。經過分析,在實際的vfs mount實現過程中,還是和我原始的想法略有差別,但是,基本目標還是相同的。
Linux VFS的mount過程基本原理如下圖所示:
當用戶輸入”mount /dev/sdb /mnt/alan”命令後,Linux會解析/mnt/alan字符串,並且從Dentry Hash表中獲取相關的dentry目錄項,然後將該目錄項標識成DCACHE_MOUNTED。一旦該dentry被標識成DCACHE_MOUNTED,也就意味著在訪問路徑上對其進行了屏蔽。
在mount /dev/sdb設備上的ext3文件系統時,內核會創建一個該文件系統的superblock對象,並且從/dev/sdb設備上讀取所有的superblock信息,初始化該內存對象。Linux內核維護了一個全局superblock對象鍊錶。s_root是superblock對象所維護的dentry目錄項,該目錄項是該文件系統的根目錄。即新mount的文件系統內容都需要通過該根目錄進行訪問。在mount的過程中,VFS會創建一個非常重要的vfsmount對象,該對象維護了文件系統mount的所有信息。Vfsmount對象通過HASH表進行維護,通過path地址計算HASH值,在這裡vfsmount的HASH值通過“/mnt/alan”路徑字符串進行計算得到。Vfsmount中的mnt_root指向superblock對象的s_root根目錄項。因此,通過/mnt/alan地址可以檢索VFSMOUNT Hash Table得到被mount的vfsmount對象,進而得到mnt_root根目錄項。
例如,/dev/sdb被mount之後,用戶想要訪問該設備上的一個文件ab.c,假設該文件的地址為:/mnt/alan/ab.c。在打開該文件的時候,首先需要進行path解析。在解析到/mnt/alan的時候,得到/mnt/alan的dentry目錄項,並且發現該目錄項已經被標識為DCACHE_MOUNTED。之後,會採用/mnt/alan計算HASH值去檢索VFSMOUNT Hash Table,得到對應的vfsmount對象,然後採用vfsmount指向的mnt_root目錄項替代/mnt/alan原來的dentry,從而實現了dentry和inode的重定向。在新的dentry的基礎上,解析程序繼續執行,最終得到表示ab.c文件的inode對象。
關鍵數據結構說明
Linux VFS mount所涉及的關鍵數據結構分析如下。
Vfsmount數據結構
Vfsmount數據結構是vfs mount最為重要的數據結構,其維護了一個mount點的所有信息。該數據結構描述如下:
- struct vfsmount {
- struct list_head mnt_hash; /* 連接到VFSMOUNT Hash Table */
- struct vfsmount *mnt_parent; /* 指向mount樹中的父節點*/
- struct dentry *mnt_mountpoint; /* 指向mount點的目錄項*/
- struct dentry *mnt_root; /* 被mount的文件系統根目錄項*/
- struct super_block *mnt_sb; /* 指向被mount的文件系統superblock */
- #ifdef CONFIG_SMP
- struct mnt_pcp __percpu *mnt_pcp;
- atomic_t mnt_longterm; /* how many of the refs are longterm */
- #else
- int mnt_count;
- int mnt_writers;
- #endif
- struct list_head mnt_mounts; /* 下級(child)vfsmount對象鍊錶*/
- struct list_head mnt_child; /* 鏈入上級vfsmount對象的鍊錶點*/
- int mnt_flags;
- /* 4 bytes hole on 64bits arches without fsnotify */
- #ifdef CONFIG_FSNOTIFY
- __u32 mnt_fsnotify_mask;
- struct hlist_head mnt_fsnotify_marks;
- #endif
- const char *mnt_devname; /* 文件系統所在的設備名字,例如/dev/sdb */
- struct list_head mnt_list;
- struct list_head mnt_expire; /* link in fs-specific expiry list */
- struct list_head mnt_share; /* circular list of shared mounts */
- struct list_head mnt_slave_list;/* list of slave mounts */
- struct list_head mnt_slave; /* slave list entry */
- struct vfsmount *mnt_master; /* slave is on master- > mnt_slave_list */
- struct mnt_namespace *mnt_ns; /* containing namespace */
- int mnt_id; /* mount identifier */
- int mnt_group_id; /* peer group identifier */
- int mnt_expiry_mark; /* true if marked for expiry */
- int mnt_pinned;
- int mnt_ghosts;
- };
複製代碼
在Linux內核中不僅存在VFSMOUNT的Hash Table,而且還維護了一棵Mount對象樹,通過該mount樹,我們可以了解到各個文件系統之間的關係。該mount樹描述如下:
 上圖所示為三層mount文件系統樹。第一層為系統根目錄“/”;第二層有兩個mount點,一個為/mnt/a,另一個是/mnt/b;第三層在/mnt/a的基礎上又創建了兩個mount點,分別為/mnt/a/c和/mnt/a/d。通過mount樹,可以對整個系統的mount結構一目了然。
Superblock數據結構
每個文件系統都會擁有一個superblock對像對其基本信息進行描述。對於像ext3之類的文件系統而言,在磁盤上會持久化存儲一份superblock元數據信息,內存的superblock對象由磁盤上的信息初始化。對於像block device 之類的“偽文件系統”而言,在mount的時候也會創建superblock對象,只不過很多信息都是臨時生成的,沒有持久化信息。Vfs superblock數據結構定義如下:
- struct super_block {
- struct list_head s_list; /* 鏈入全局鍊錶的對象*/
- dev_t s_dev; /* search index; _not_ kdev_t */
- unsigned char s_dirt;
- unsigned char s_blocksize_bits;
- unsigned long s_blocksize;
- loff_t s_maxbytes; /* Max file size */
- struct file_system_type *s_type;
- const struct super_operations *s_op; /* superblock操作函數集*/
- const struct dquot_operations *dq_op;
- const struct quotactl_ops *s_qcop;
- const struct export_operations *s_export_op;
- unsigned long s_flags;
- unsigned long s_magic;
- struct dentry *s_root; /* 文件系統根目錄項*/
- struct rw_semaphore s_umount;
- struct mutex s_lock;
- int s_count;
- atomic_t s_active;
- #ifdef CONFIG_SECURITY
- void *s_security;
- #endif
- const struct xattr_handler **s_xattr;
- struct list_head s_inodes; /* all inodes */
- struct hlist_bl_head s_anon; /* anonymous dentries for (nfs) exporting */
- #ifdef CONFIG_SMP
- struct list_head __percpu *s_files;
- #else
- struct list_head s_files;
- #endif
- /* s_dentry_lru, s_nr_dentry_unused protected by dcache.c lru locks */
- struct list_head s_dentry_lru; /* unused dentry lru */
- int s_nr_dentry_unused; /* # of dentry on lru */
- /* s_inode_lru_lock protects s_inode_lru and s_nr_inodes_unused */
- spinlock_t s_inode_lru_lock ____cacheline_aligned_in_smp;
- struct list_head s_inode_lru; /* unused inode lru */
- int s_nr_inodes_unused; /* # of inodes on lru */
- struct block_device *s_bdev;
- struct backing_dev_info *s_bdi;
- struct mtd_info *s_mtd;
- struct list_head s_instances;
- struct quota_info s_dquot; /* Diskquota specific options */
- int s_frozen;
- wait_queue_head_t s_wait_unfrozen;
- char s_id[32]; /* Informational name */
- u8 s_uuid[16]; /* UUID */
- void *s_fs_info; /* Filesystem private info */
- fmode_t s_mode;
- /* Granularity of c/m/atime in ns.
- Cannot be worse than a second */
- u32 s_time_gran;
- /*
- * The next field is for VFS *only*. No filesystems have any business
- * even looking at it. You had been warned.
- */
- struct mutex s_vfs_rename_mutex; /* Kludge */
- /*
- * Filesystem subtype. If non-empty the filesystem type field
- * in /proc/mounts will be "type.subtype"
- */
- char *s_subtype;
- /*
- * Saved mount options for lazy filesystems using
- * generic_show_options()
- */
- char __rcu *s_options;
- const struct dentry_operations *s_d_op; /* default d_op for dentries */
- /*
- * Saved pool identifier for cleancache (-1 means none)
- */
- int cleancache_poolid;
- struct shrinker s_shrink; /* per-sb shrinker handle */
- };
複製代碼
代碼流程分析
Linux中實現mount操作需要一定的代碼量,下面對Linux VFS Mount代碼進行分析說明,整個分析過程按照mount操作函數調用流程進行。代碼分析基於Linux-3.2版本。
當用戶在用戶層執行mount命令時,會執行系統調用從用戶態陷入linux內核,執行如下函數(namespace.c):
- SYSCALL_DEFINE5(mount, char __user *, dev_name, char __user *, dir_name,
- char __user *, type, unsigned long, flags, void __user *, data)
- {
- int ret;
- char *kernel_type;
- char *kernel_dir;
- char *kernel_dev;
- unsigned long data_page;
- /* 獲取mount類型 */
- ret = copy_mount_string (type, &kernel_type);
- if (ret < 0 )
- goto out_type;
- /* 獲取mount點目錄字符串 */
- kernel_dir = getname (dir_name);
- if (IS_ERR(kernel_dir)) {
- ret = PTR_ERR (kernel_dir);
- goto out_dir;
- }
- /* 獲取設備名稱字符串 */
- ret = copy_mount_string (dev_name, &kernel_dev);
- if (ret < 0 )
- goto out_dev;
- /* 獲取其它選項 */
- ret = copy_mount_options (data, &data_page);
- if (ret < 0 )
- goto out_data;
- /* 主要函數,執行掛載文件系統的具體操作*/
- ret = do_mount (kernel_dev, kernel_dir, kernel_type, flags,
- (void *) data_page);
- free_page(data_page);
- out_data:
- kfree(kernel_dev);
- out_dev:
- putname(kernel_dir);
- out_dir:
- kfree(kernel_type);
- out_type:
- return ret;
- }
複製代碼
do_mount()函數是mount操作過程中的核心函數,在該函數中,通過mount的目錄字符串找到對應的dentry目錄項,然後通過do_new_mount()函數完成具體的mount操作。do_mount()函數分析如下:
- long do_mount(char *dev_name, char *dir_name, char *type_page,
- unsigned long flags, void *data_page)
- {
- struct path path;
- int retval = 0 ;
- int mnt_flags = 0 ;
- 。。。
- /* 通過mount目錄字符串獲取path,path結構中包含有mount目錄的dentry目錄對象*/
- retval = kern_path (dir_name, LOOKUP_FOLLOW, &path);
- if (retval)
- return retval;
- 。。。
- /* Separate the per-mountpoint flags */
- if (flags & MS_NOSUID)
- mnt_flags |= MNT_NOSUID;
- if (flags & MS_NODEV)
- mnt_flags |= MNT_NODEV;
- if (flags & MS_NOEXEC)
- mnt_flags |= MNT_NOEXEC;
- if (flags & MS_NOATIME)
- mnt_flags |= MNT_NOATIME;
- if (flags & MS_NODIRATIME)
- mnt_flags |= MNT_NODIRATIME;
- if (flags & MS_STRICTATIME)
- mnt_flags &= ~(MNT_RELATIME | MNT_NOATIME);
- if (flags & MS_RDONLY)
- mnt_flags |= MNT_READONLY;
- flags &= ~(MS_NOSUID | MS_NOEXEC | MS_NODEV | MS_ACTIVE | MS_BORN |
- MS_NOATIME | MS_NODIRATIME | MS_RELATIME| MS_KERNMOUNT |
- MS_STRICTATIME);
- /* remount操作 */
- if (flags & MS_REMOUNT)
- retval = do_remount (&path, flags & ~MS_REMOUNT, mnt_flags,
- data_page);
- else if (flags & MS_BIND)
- retval = do_loopback (&path, dev_name, flags & MS_REC);
- else if (flags & (MS_SHARED | MS_PRIVATE | MS_SLAVE | MS_UNBINDABLE))
- retval = do_change_type (&path, flags);
- else if (flags & MS_MOVE)
- retval = do_move_mount (&path, dev_name);
- else
- /* 正常的mount操作,完成具體的mount操作*/
- retval = do_new_mount (&path, type_page, flags, mnt_flags,
- dev_name, data_page);
- dput_out:
- path_put(&path);
- return retval;
- }
複製代碼
do_new_mount()函數主要分成兩大部分:第一部分建立vfsmount對象和superblock對象,必要時從設備上獲取文件系統元數據;第二部分將vfsmount對象加入到mount樹和Hash Table中,並且將原來的dentry對象無效掉。do_new_mount函數說明如下:
- static int do_new_mount(struct path *path, char *type, int flags,
- int mnt_flags, char *name, void *data)
- {
- struct vfsmount *mnt;
- int err;
- 。。。
- /* 在內核建立vfsmount對象和superblock對象*/
- mnt = do_kern_mount (type, flags, name, data);
- if (IS_ERR(mnt))
- return PTR_ERR(mnt);
- /* 將vfsmount對象加入系統,屏蔽原有dentry對象*/
- err = do_add_mount (mnt, path, mnt_flags);
- if (err)
- mntput(mnt);
- return err;
- }
複製代碼
do_new_mount()中的第一步調用do_kern_mount()函數,該函數的主幹調用路徑如下: do_kern_mount--> vfs_kern_mount--> mount_fs
在mount_fs()函數中會調用特定文件系統的mount方法,如果mount是ext3文件系統,那麼在mount_fs函數中最終會調用ext3的mount方法。Ext3的mount方法定義在super.c文件中:
- static struct file_system_type ext3_fs_type = {
- .owner = THIS_MODULE ,
- .name = "ext3" ,
- .mount = ext3_mount , /* ext3文件系統mount方法*/
- .kill_sb = kill_block_super ,
- .fs_flags = FS_REQUIRES_DEV ,
- };
複製代碼
Ext3 mount函數主幹調用路徑為:ext3_mount--> mount_bdev。Mount_bdev()函數主要完成superblock對象的內存初始化,並且加入到全局superblock鍊錶中。該函數說明如下:
- struct dentry *mount_bdev(struct file_system_type *fs_type,
- int flags, const char *dev_name, void *data,
- int (*fill_super)(struct super_block *, void *, int))
- {
- struct block_device *bdev;
- struct super_block *s;
- fmode_t mode = FMODE_READ | FMODE_EXCL;
- int error = 0 ;
- if (!(flags & MS_RDONLY))
- mode |= FMODE_WRITE;
- /* 通過設備名字獲取被mount設備的bdev對象*/
- bdev = blkdev_get_by_path (dev_name, mode, fs_type);
- if (IS_ERR(bdev))
- return ERR_CAST(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
- */
- mutex_lock(&bdev- > bd_fsfreeze_mutex);
- if (bdev- > bd_fsfreeze_count > 0) {
- mutex_unlock(&bdev- > bd_fsfreeze_mutex);
- error = -EBUSY;
- goto error_bdev;
- }
- /* 查找或者創建superblock對象 */
- s = sget (fs_type, test_bdev_super, set_bdev_super, bdev);
- mutex_unlock(&bdev- > bd_fsfreeze_mutex);
- if (IS_ERR(s))
- goto error_s;
- if (s- > s_root) {
- /* 被mount文件系統的根目錄項已經存在*/
- if ((flags ^ s- > s_flags) & MS_RDONLY) {
- deactivate_locked_super(s);
- error = -EBUSY;
- goto error_bdev;
- }
- /*
- * s_umount nests inside bd_mutex during
- * __invalidate_device(). blkdev_put() acquires
- * bd_mutex and can't be called under s_umount. Drop
- * s_umount temporarily. This is safe as we're
- * holding an active reference.
- */
- up_write(&s- > s_umount);
- blkdev_put(bdev, mode);
- down_write(&s- > s_umount);
- } else {
- /* 文件系統根目錄項不存在,通過filler_super函數讀取磁盤上的superblock元數據信息,並且初始化superblock內存結構*/
- char b[BDEVNAME_SIZE];
- s- > s_flags = flags | MS_NOSEC;
- s- > s_mode = mode;
- strlcpy(s- > s_id, bdevname(bdev, b), sizeof(s- > s_id));
- sb_set_blocksize(s, block_size(bdev));
- /* 對於ext3文件系統,調用ext3_fill_super函數*/
- error = fill_super (s, data, flags & MS_SILENT ? 1 : 0);
- if (error) {
- deactivate_locked_super(s);
- goto error;
- }
- s- > s_flags |= MS_ACTIVE;
- bdev- > bd_super = s;
- }
- /* 正常返回被mount文件系統根目錄項*/
- return dget(s- > s_root);
- error_s:
- error = PTR_ERR (s);
- error_bdev:
- blkdev_put(bdev, mode);
- error:
- return ERR_PTR(error);
- }
複製代碼
do_new_mount()函數的第二步是將創建的vfsmount對象加入到mount樹和VFSMOUNT Hash Table中,並且將老的dentry目錄項無效掉。該過程主幹函數調用過程如下所示: do_new_mount--> do_add_mount--> graft_tree--> attach_recursive_mnt
attach_recursive_mnt()函數完成第二步過程的主要操作。至此,文件系統的mount操作已經完成。Mount完成之後,如果用戶想要訪問新mount文件系統中的文件,那麼需要在path解析過程中重定位dentry,該過程主要在follow_managed()函數中完成。在該函數中會判斷一個dentry是否已經被標識成DCACHE_MOUNTED,如果該標誌位已經被設置,那麼通過VFSMOUNT Hash Table可以重定位dentry。
如有不對之處,敬請指出更正(tl_wzj@yahoo.com.cn )。
|