super.c是ext4里非常重要的文件，挂载时候运行的代码基本上都在super.c里边，如果super.c有一点代码上的bug的话，那么一个块设备就不能被以ext4文件系统挂载。我在比较重要的地方都加了注释讲解，如果有认为我没写详细或者写错的朋友们欢迎指出。
/*
* linux/fs/ext4/super.c
*
* Copyright (C) 1992, 1993, 1994, 1995
* Remy Card ([email protected])
* Laboratoire MASI - Institut Blaise Pascal
* Universite Pierre et Marie Curie (Paris VI)
* 好熟悉的人，我记得当初看ext2的时候作者也是这个人，巴黎第五大学的god
* from
*
* linux/fs/minix/inode.c
*
* Copyright (C) 1991, 1992 Linus Torvalds
*
* Big-endian to little-endian byte-swapping/bitmaps by
* David S. Miller ([email protected]), 1995
*/
#include <linux/module.h>
#include <linux/string.h>
#include <linux/fs.h>
#include <linux/time.h>
#include <linux/jbd2.h>
#include <linux/ext4_fs.h>
#include <linux/ext4_jbd2.h>
#include <linux/slab.h>
#include <linux/init.h>
#include <linux/blkdev.h>
#include <linux/parser.h>
#include <linux/smp_lock.h>
#include <linux/buffer_head.h>
#include <linux/vfs.h>
#include <linux/random.h>
#include <linux/mount.h>
#include <linux/namei.h>
#include <linux/quotaops.h>
#include <linux/seq_file.h>
#include <asm/uaccess.h>
#include "xattr.h"
#include "acl.h"
#include "namei.h"
/* super.c里会使用到的函数的声明 */
static int ext4_load_journal(struct super_block *, struct ext4_super_block *,
unsigned long journal_devnum);
static int ext4_create_journal(struct super_block *, struct ext4_super_block *,
unsigned int);
static void ext4_commit_super (struct super_block * sb,
struct ext4_super_block * es,
int sync);
static void ext4_mark_recovery_complete(struct super_block * sb,
struct ext4_super_block * es);
static void ext4_clear_journal_err(struct super_block * sb,
struct ext4_super_block * es);
static int ext4_sync_fs(struct super_block *sb, int wait);
static const char *ext4_decode_error(struct super_block * sb, int errno,
char nbuf[16]);
static int ext4_remount (struct super_block * sb, int * flags, char * data);
static int ext4_statfs (struct dentry * dentry, struct kstatfs * buf);
static void ext4_unlockfs(struct super_block *sb);
static void ext4_write_super (struct super_block * sb);
static void ext4_write_super_lockfs(struct super_block *sb);
/* 根据ext4的超级块的块号以及对应组的组描述符获得数据块位图的所在块号 */
ext4_fsblk_t ext4_block_bitmap(struct super_block *sb,
struct ext4_group_desc *bg)
{
/* ext4是用的小端字节序，所以需要先转换成本机的cpu字节序然后才能和内存里的数进行比较加减等操作，下同。 */
/* 组描述符结构体里的bg_block_bitmap字段表示数据块所在的位置，EXT4_DESC_SIZE这个宏返回组描述符的大小，如果大于EXT4_MIN_DESC_SIZE_64BIT，就是大于64位的最小大小，说明是用64位来表示的，就把高32位也或上，这主要是为了应对有的文件系统块的数目太多32位已经表述不下来的情况 */
return le32_to_cpu(bg->bg_block_bitmap) |
(EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT ?
(ext4_fsblk_t)le32_to_cpu(bg->bg_block_bitmap_hi) << 32 : 0);
}
/* 根据ext4的超级块的块号以及对应组的组描述符获得inode位图的所在块号 */
ext4_fsblk_t ext4_inode_bitmap(struct super_block *sb,
struct ext4_group_desc *bg)
{
/* 组描述符结构体里的bg_inode_bitmap字段表示存放inode的块所在的位置，EXT4_DESC_SIZE这个宏返回组描述符的大小，如果大于EXT4_MIN_DESC_SIZE_64BIT，就是大于64位的最小大小，说明是用64位来表示的，就把高32位也或上，这主要是为了应对有的文件系统块的数目太多32位已经表述不下来的情况 */
return le32_to_cpu(bg->bg_inode_bitmap) |
(EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT ?
(ext4_fsblk_t)le32_to_cpu(bg->bg_inode_bitmap_hi) << 32 : 0);
}
/* 根据ext4的超级块的块号以及对应组的组描述符获得inode table的开始块号 */
ext4_fsblk_t ext4_inode_table(struct super_block *sb,
struct ext4_group_desc *bg)
{
/* 组描述符结构体里的bg_inode_table字段表示inode table的开始块所在的位置，EXT4_DESC_SIZE这个宏返回组描述符的大小，如果大于EXT4_MIN_DESC_SIZE_64BIT，就是大于64位的最小大小，说明是用64位来表示的，就把高32位也或上，这主要是为了应对有的文件系统块的数目太多32位已经表述不下来的情况 */
return le32_to_cpu(bg->bg_inode_table) |
(EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT ?
(ext4_fsblk_t)le32_to_cpu(bg->bg_inode_table_hi) << 32 : 0);
}
/* 对块组描述符结构体的bg_block_bitmap字段赋值为blk */
void ext4_block_bitmap_set(struct super_block *sb,
struct ext4_group_desc *bg, ext4_fsblk_t blk)
{
/* 把blk转换成cpu字节序赋值给块组描述符的bg_block_bitmap，如果文件系统支持64位的，就把bg_block_bitmap_hi也就是高32位也赋值 */
bg->bg_block_bitmap = cpu_to_le32((u32)blk);
if (EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT)
bg->bg_block_bitmap_hi = cpu_to_le32(blk >> 32);
}
/* 对块组描述符结构体的bg_inode_bitmap字段赋值为blk，就是赋值inode位图所在块的赋值 */
void ext4_inode_bitmap_set(struct super_block *sb,
struct ext4_group_desc *bg, ext4_fsblk_t blk)
{
/* 把blk转换成cpu字节序赋值给块组描述符的bg_inode_bitmap，如果文件系统支持64位的，就把bg_inode_bitmap_hi也就是高32位也赋值 */
bg->bg_inode_bitmap = cpu_to_le32((u32)blk);
if (EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT)
bg->bg_inode_bitmap_hi = cpu_to_le32(blk >> 32);
}
/* 对块组描述符结构体的bg_inode_table字段赋值为blk，就是inode table开始块的赋值 */
void ext4_inode_table_set(struct super_block *sb,
struct ext4_group_desc *bg, ext4_fsblk_t blk)
{
/* 把blk转换成cpu字节序赋值给块组描述符的bg_inode_bitmap，如果文件系统支持64位的，就把bg_inode_bitmap_hi也就是高32位也赋值 */
bg->bg_inode_table = cpu_to_le32((u32)blk);
if (EXT4_DESC_SIZE(sb) >= EXT4_MIN_DESC_SIZE_64BIT)
bg->bg_inode_table_hi = cpu_to_le32(blk >> 32);
}
/*
* ext4对于jbd2_journal_start/end的封装，参数nblocks表示这次日志记录可能修改的块的数目
* journal其实是内核里的另外一个模块实现的，ext4这里仅仅是调用了journal模块的接口，在2.6.19内核版本里使用的是journal2，也就是jbd2。
* 对于ext4封装的日志函数而言，我们唯一需要做的特殊的事情就是确保journal_end 的调用函数的结果在superblock里，需要标记为脏所以sync()就在合适的时候会调用文件系统的write_super回调函数。
*/
handle_t *ext4_journal_start_sb(struct super_block *sb, int nblocks)
{
journal_t *journal;
/* 如果文件系统是制度挂载的话，不支持日志，一般只读挂载都是出错了 */
if (sb->s_flags & MS_RDONLY)
return ERR_PTR(-EROFS);
/* 这里如果日志aborted的，我们需要报错，如果日志没有aborted状态，我们直接调用jbd2模块的函数jbd2_journal_start 来货的日志句柄 */
journal = EXT4_SB(sb)->s_journal;
if (is_journal_aborted(journal)) {
ext4_abort(sb, __FUNCTION__,
"Detected aborted journal");
return ERR_PTR(-EROFS);
}
return jbd2_journal_start(journal, nblocks);
}
/* 对于jbd2_journal_stop的封装，做一些基本的检查以后就调用jbd2_journal_stop函数。
* 我们唯一需要做的特殊的事情就是确保journal_end 的调用函数的结果在superblock里，需要标记为脏所以sync()就在合适的时候会调用文件系统的write_super回调函数。
*/
int __ext4_journal_stop(const char *where, handle_t *handle)
{
struct super_block *sb;
int err;
int rc;
sb = handle->h_transaction->t_journal->j_private;
err = handle->h_err;
rc = jbd2_journal_stop(handle);
if (!err)
err = rc;
if (err)
__ext4_std_error(sb, where, err);
return err;
}
/* 中途停止日志首先打印出错误信息，然后调用jbd2的jbd2_journal_abort_handle函数来停止日志 */
void ext4_journal_abort_handle(const char *caller, const char *err_fn,
struct buffer_head *bh, handle_t *handle, int err)
{
/* 从err获得错误信息 */
char nbuf[16];
const char *errstr = ext4_decode_error(NULL, err, nbuf);
/* befferhead记录错误信息 */
if (bh)
BUFFER_TRACE(bh, "abort");
/* 记录错误值在句柄中 */
if (!handle->h_err)
handle->h_err = err;
/* 如果句柄所在的事务已经被终止了，我们就不必再次终止了 */
if (is_handle_aborted(handle))
return;
printk(KERN_ERR "%s: aborting transaction: %s in %s\n",
caller, errstr, err_fn);
jbd2_journal_abort_handle(handle);
}
/* 当ext4报错误的时候的处理函数，比如一些一致性错误或者是读写的IO错误。
* 在ext2上，我们把一些错误状态字存储在超级块里。但是在ext4，这不可能，因为我们可能有写入顺序限制在超级块上，这个限制将会阻止我们写入，考虑到日志将要终止的情况，我们不能依赖以后。
* 我们使用jbd2_journal_abort()获得的错误码来记录错误，在恢复的时候日志将会报错误，直到我们清除这个错误记录。 */
static void ext4_handle_error(struct super_block *sb)
{
/* 首先从超级块结构体获得磁盘上存储的ext4超级块结构体 */
struct ext4_super_block *es = EXT4_SB(sb)->s_es;
/* 挂载状态标记为错误发生状态 */
EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS;
es->s_state |= cpu_to_le16(EXT4_ERROR_FS);
/* 如果已经是只读挂载了，我们就不必继续操作了 */
if (sb->s_flags & MS_RDONLY)
return;
/* 如果不是挂载选项是出错以后继续，那么就停止一切事务 */
if (!test_opt (sb, ERRORS_CONT)) {
journal_t *journal = EXT4_SB(sb)->s_journal;
EXT4_SB(sb)->s_mount_opt |= EXT4_MOUNT_ABORT;
if (journal)
jbd2_journal_abort(journal, -EIO);
}
/* 如果fs是标记未，如果有错误，重新挂载为只读的情况，我们就重新挂载为只读 */
if (test_opt (sb, ERRORS_RO)) {
printk (KERN_CRIT "Remounting filesystem read-only\n");
sb->s_flags |= MS_RDONLY;
}
/* 提交超级块的修改 */
ext4_commit_super(sb, es, 1);
/* 如果挂载选项是出错就panic，打印出panic信息 */
if (test_opt(sb, ERRORS_PANIC))
panic("EXT4-fs (device %s): panic forced after error\n",
sb->s_id);
}
/* 调用上边的函数处理错误，在此之前打印出错误信息 */
void ext4_error (struct super_block * sb, const char * function,
const char * fmt, ...)
{
va_list args;
/* 先打印出错误信息 */
va_start(args, fmt);
printk(KERN_CRIT "EXT4-fs error (device %s): %s: ",sb->s_id, function);
vprintk(fmt, args);
printk("\n");
va_end(args);
/* 然后处理错误 */
ext4_handle_error(sb);
}
/* 根据错误编号返回相应的错误信息字符串 */
static const char *ext4_decode_error(struct super_block * sb, int errno,
char nbuf[16])
{
char *errstr = NULL;
switch (errno) {
/* IO读写错误 */
case -EIO:
errstr = "IO failure";
break;
/* 内存满了 */
case -ENOMEM:
errstr = "Out of memory";
break;
/* 只读的文件系统，如果日志系统也崩溃了的话，就返回日志终止的信息 */
case -EROFS:
if (!sb || EXT4_SB(sb)->s_journal->j_flags & JBD2_ABORT)
errstr = "Journal has aborted";
else
errstr = "Readonly filesystem";
break;
default:
/* 如果用户传入的错误码是未知的我们就直接返回error:错误码的值,但是如果用户传入的缓冲区都是空的，那就直接返回NULL */
if (nbuf) {
if (snprintf(nbuf, 16, "error %d", -errno) >= 0)
errstr = nbuf;
}
break;
}
return errstr;
}
/* __ext4_std_error函数自动的从日志里接受一些希望得到的错误码，然后做出合理的反应. */
void __ext4_std_error (struct super_block * sb, const char * function,
int errno)
{
char nbuf[16];
const char *errstr;
/* 特殊情况，如果错误码是EROFS,并且我忙呢又不在事务里，就没必要记录这个错误 */
if (errno == -EROFS && journal_current_handle() == NULL &&
(sb->s_flags & MS_RDONLY))
return;
/* 解码返回错误值 */
errstr = ext4_decode_error(sb, errno, nbuf);
printk (KERN_CRIT "EXT4-fs error (device %s) in %s: %s\n",
sb->s_id, function, errstr);
/* 处理错误 */
ext4_handle_error(sb);
}
/*
* ext4_abort是一个比ext4_error健壮很多的错误处理函数.这个函数可以用来处理一些不可恢复的错误，比如to或者一些严重的日志错误
* 我们将会强制ext4进入ABORT|READONLY状态，除非文件系统的挂载选项是要求我们立刻panic当遇到错误的时候，这时我们就直接panic
*/
void ext4_abort (struct super_block * sb, const char * function,
const char * fmt, ...)
{
/* 首先打印用户传入的错误信息 */
va_list args;
printk (KERN_CRIT "ext4_abort called.\n");
va_start(args, fmt);
printk(KERN_CRIT "EXT4-fs error (device %s): %s: ",sb->s_id, function);
vprintk(fmt, args);
printk("\n");
va_end(args);
/* 如果挂载选项是出现错误就panic，打印panic信息 */
if (test_opt(sb, ERRORS_PANIC))
panic("EXT4-fs panic from previous error\n");
/* 已经是只读，就没必要接下来的操作了 */
if (sb->s_flags & MS_RDONLY)
return;
/* 如果不是只读挂载的，我们需要在超级块上记录错误，然后只读挂载，最后终止日志 */
printk(KERN_CRIT "Remounting filesystem read-only\n");
EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS;
sb->s_flags |= MS_RDONLY;
EXT4_SB(sb)->s_mount_opt |= EXT4_MOUNT_ABORT;
jbd2_journal_abort(EXT4_SB(sb)->s_journal, -EIO);
}
/* 简单的打印警告信息 */
void ext4_warning (struct super_block * sb, const char * function,
const char * fmt, ...)
{
va_list args;
va_start(args, fmt);
printk(KERN_WARNING "EXT4-fs warning (device %s): %s: ",
sb->s_id, function);
vprintk(fmt, args);
printk("\n");
va_end(args);
}
/* 动态更新到ext4的新版 */
void ext4_update_dynamic_rev(struct super_block *sb)
{
struct ext4_super_block *es = EXT4_SB(sb)->s_es;
/* 如果已经是新版，就不用更新了 */
if (le32_to_cpu(es->s_rev_level) > EXT4_GOOD_OLD_REV)
return;
/* 动态更新是一件很危险的事情，因为要修改inode编号以及大小 */
ext4_warning(sb, __FUNCTION__,
"updating to rev %d because of new feature flag, "
"running e2fsck is recommended",
EXT4_DYNAMIC_REV);
/* 修改文件系统的第一个inode号码，inode大小，版本号修改 */
es->s_first_ino = cpu_to_le32(EXT4_GOOD_OLD_FIRST_INO);
es->s_inode_size = cpu_to_le16(EXT4_GOOD_OLD_INODE_SIZE);
es->s_rev_level = cpu_to_le32(EXT4_DYNAMIC_REV);
/*superblock的其他字段应该是0，如果不是，说明他们正在被使用，不管他们就行了。交给e2fsck来清理不一致的状态。 */
}
/*
* 打开外部的日志块设备，参数是设备号
*/
static struct block_device *ext4_blkdev_get(dev_t dev)
{
struct block_device *bdev;
char b[BDEVNAME_SIZE];
/* 调用其他层给的接口函数来货的block device */
bdev = open_by_devnum(dev, FMODE_READ|FMODE_WRITE);
if (IS_ERR(bdev))
goto fail;
return bdev;
fail:
printk(KERN_ERR "EXT4: failed to open journal device %s: %ld\n",
__bdevname(dev, b), PTR_ERR(bdev));
return NULL;
}
/*
* 释放日志设备，也是调用其他层给的接口函数来释放
*/
static int ext4_blkdev_put(struct block_device *bdev)
{
bd_release(bdev);
return blkdev_put(bdev);
}
/* 移除超级块信息结构体里的块设备 */
static int ext4_blkdev_remove(struct ext4_sb_info *sbi)
{
struct block_device *bdev;
int ret = -ENODEV;
bdev = sbi->journal_bdev;
if (bdev) {
/* ext4_blkdev_put来释放，然后赋值为NULL */
ret = ext4_blkdev_put(bdev);
sbi->journal_bdev = NULL;
}
return ret;
}
/* 由list_head来获得vfs_inode */
static inline struct inode *orphan_list_entry(struct list_head *l)
{
return &list_entry(l, struct ext4_inode_info, i_orphan)->vfs_inode;
}
/* 打印ext4_sb_info结构体上的s_orphan链表所有项 */
static void dump_orphan_list(struct super_block *sb, struct ext4_sb_info *sbi)
{
struct list_head *l;
printk(KERN_ERR "sb orphan head is %d\n",
le32_to_cpu(sbi->s_es->s_last_orphan));
printk(KERN_ERR "sb_info orphan list:\n");
/* 遍历所有项，对于每一个项打印 */
list_for_each(l, &sbi->s_orphan) {
struct inode *inode = orphan_list_entry(l);
printk(KERN_ERR " "
"inode %s:%lu at %p: mode %o, nlink %d, next %d\n",
inode->i_sb->s_id, inode->i_ino, inode,
inode->i_mode, inode->i_nlink,
NEXT_ORPHAN(inode));
}
}
/* 卸载文件系统的时候调用的函数 */
static void ext4_put_super (struct super_block * sb)
{
struct ext4_sb_info *sbi = EXT4_SB(sb);
struct ext4_super_block *es = sbi->s_es;
int i;
/* 释放超级块，以及超级块附着的扩展属性结构体，还有ext4的日志结构体 */
ext4_ext_release(sb);
ext4_xattr_put_super(sb);
jbd2_journal_destroy(sbi->s_journal);
/*如果现在是只读挂载，说明现在在恢复，清理掉恢复中的标记，然后把超级块的buffer head写入磁盘设备*/
if (!(sb->s_flags & MS_RDONLY)) {
EXT4_CLEAR_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER);
es->s_state = cpu_to_le16(sbi->s_mount_state);
BUFFER_TRACE(sbi->s_sbh, "marking dirty");
mark_buffer_dirty(sbi->s_sbh);
ext4_commit_super(sb, es, 1);
}
/* 对于每一个组，减少对应的块组描述符引用计数 */
for (i = 0; i < sbi->s_gdb_count; i++)
brelse(sbi->s_group_desc[i]);
/* 释放块组描述符占用的内存 */
kfree(sbi->s_group_desc);
/* 释放ext4_sb_info里的单cpu变量 */
percpu_counter_destroy(&sbi->s_freeblocks_counter);
percpu_counter_destroy(&sbi->s_freeinodes_counter);
percpu_counter_destroy(&sbi->s_dirs_counter);
/* 释放ext4_sb_info结构体 */
brelse(sbi->s_sbh);
#ifdef CONFIG_QUOTA
/* 如果配置了配额，我们还需要释放所有的配额相关的资源 */
for (i = 0; i < MAXQUOTAS; i++)
kfree(sbi->s_qf_names[i]);
#endif
/* 调试打印信息 */
if (!list_empty(&sbi->s_orphan))
dump_orphan_list(sb, sbi);
J_ASSERT(list_empty(&sbi->s_orphan));
invalidate_bdev(sb->s_bdev, 0);
if (sbi->journal_bdev && sbi->journal_bdev != sb->s_bdev) {
/*
* 遇到了非法的日志设备缓冲区
*/
sync_blockdev(sbi->journal_bdev);
invalidate_bdev(sbi->journal_bdev, 0);
ext4_blkdev_remove(sbi);
}
sb->s_fs_info = NULL;
kfree(sbi);
return;
}
/* inode缓存分配的地方 */
static kmem_cache_t *ext4_inode_cachep;
/*
* 分配inode的函数，因为调用了内部的事务程序，所以使用GFP_NOFS标志
*/
static struct inode *ext4_alloc_inode(struct super_block *sb)
{
struct ext4_inode_info *ei;
/* 分配一部分缓冲区 */
ei = kmem_cache_alloc(ext4_inode_cachep, SLAB_NOFS);
if (!ei)
return NULL;
/* 如果配置了权限控制列表，初始化这些字段 */
#ifdef CONFIG_EXT4DEV_FS_POSIX_ACL
ei->i_acl = EXT4_ACL_NOT_CACHED;
ei->i_default_acl = EXT4_ACL_NOT_CACHED;
#endif
/* 初始化一些inode的字段 */
ei->i_block_alloc_info = NULL;
ei->vfs_inode.i_version = 1;
memset(&ei->i_cached_extent, 0, sizeof(struct ext4_ext_cache));
/* 最后返回分配的inode */
return &ei->vfs_inode;
}
/* 释放inode回内存的缓冲区 */
static void ext4_destroy_inode(struct inode *inode)
{
kmem_cache_free(ext4_inode_cachep, EXT4_I(inode));
}
/* 初始化ext4_inode_info结构体 */
static void init_once(void * foo, kmem_cache_t * cachep, unsigned long flags)
{
struct ext4_inode_info *ei = (struct ext4_inode_info *) foo;
if ((flags & (SLAB_CTOR_VERIFY|SLAB_CTOR_CONSTRUCTOR)) ==
SLAB_CTOR_CONSTRUCTOR) {
INIT_LIST_HEAD(&ei->i_orphan);
#ifdef CONFIG_EXT4DEV_FS_XATTR
init_rwsem(&ei->xattr_sem);
#endif
mutex_init(&ei->truncate_mutex);
inode_init_once(&ei->vfs_inode);
}
}
/* 初始化inode缓存slab分配器，如果是啊比返回ENOMEM */
static int init_inodecache(void)
{
ext4_inode_cachep = kmem_cache_create("ext4_inode_cache",
sizeof(struct ext4_inode_info),
0, (SLAB_RECLAIM_ACCOUNT|
SLAB_MEM_SPREAD),
init_once, NULL);
if (ext4_inode_cachep == NULL)
return -ENOMEM;
return 0;
}
/* 销毁inode缓存 */
static void destroy_inodecache(void)
{
kmem_cache_destroy(ext4_inode_cachep);
}
/* 清理inode结构体以及相关的资源 */
static void ext4_clear_inode(struct inode *inode)
{
struct ext4_block_alloc_info *rsv = EXT4_I(inode)->i_block_alloc_info;
/* 如果配置了posix标准的权限控制列表，就是放acl相关的资源 */
#ifdef CONFIG_EXT4DEV_FS_POSIX_ACL
if (EXT4_I(inode)->i_acl &&
EXT4_I(inode)->i_acl != EXT4_ACL_NOT_CACHED) {
posix_acl_release(EXT4_I(inode)->i_acl);
EXT4_I(inode)->i_acl = EXT4_ACL_NOT_CACHED;
}
if (EXT4_I(inode)->i_default_acl &&
EXT4_I(inode)->i_default_acl != EXT4_ACL_NOT_CACHED) {
posix_acl_release(EXT4_I(inode)->i_default_acl);
EXT4_I(inode)->i_default_acl = EXT4_ACL_NOT_CACHED;
}
#endif
/* 把inode保留的块都释放了，一些字段赋值为NULL或释放掉 */
ext4_discard_reservation(inode);
EXT4_I(inode)->i_block_alloc_info = NULL;
if (unlikely(rsv))
kfree(rsv);
}
/* 打印配额选项 */
static inline void ext4_show_quota_options(struct seq_file *seq, struct super_block *sb)
{
#if defined(CONFIG_QUOTA)
struct ext4_sb_info *sbi = EXT4_SB(sb);
if (sbi->s_jquota_fmt)
seq_printf(seq, ",jqfmt=%s",
(sbi->s_jquota_fmt == QFMT_VFS_OLD) ? "vfsold": "vfsv0");
if (sbi->s_qf_names[USRQUOTA])
seq_printf(seq, ",usrjquota=%s", sbi->s_qf_names[USRQUOTA]);
if (sbi->s_qf_names[GRPQUOTA])
seq_printf(seq, ",grpjquota=%s", sbi->s_qf_names[GRPQUOTA]);
if (sbi->s_mount_opt & EXT4_MOUNT_USRQUOTA)
seq_puts(seq, ",usrquota");
if (sbi->s_mount_opt & EXT4_MOUNT_GRPQUOTA)
seq_puts(seq, ",grpquota");
#endif
}
/* 打印挂载时候的数据模式 */
static int ext4_show_options(struct seq_file *seq, struct vfsmount *vfs)
{
struct super_block *sb = vfs->mnt_sb;
if (test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_JOURNAL_DATA)
seq_puts(seq, ",data=journal");
else if (test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_ORDERED_DATA)
seq_puts(seq, ",data=ordered");
else if (test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_WRITEBACK_DATA)
seq_puts(seq, ",data=writeback");
ext4_show_quota_options(seq, sb);
return 0;
}
/* 从ext4获取一个dentry结构体，vobjp[0]是inode编号，vobjp[1]是版本号 */
static struct dentry *ext4_get_dentry(struct super_block *sb, void *vobjp)
{
__u32 *objp = vobjp;
unsigned long ino = objp[0];
__u32 generation = objp[1];
struct inode *inode;
struct dentry *result;
/* 首先检查inode编号的合法性 */
if (ino < EXT4_FIRST_INO(sb) && ino != EXT4_ROOT_INO)
return ERR_PTR(-ESTALE);
if (ino > le32_to_cpu(EXT4_SB(sb)->s_es->s_inodes_count))
return ERR_PTR(-ESTALE);
/* 根据inode编号返回对应的inode结构体 */
inode = iget(sb, ino);
if (inode == NULL)
return ERR_PTR(-ENOMEM);
/* 说明inode已经被删除了，我们返回对应的错误码 */
if (is_bad_inode(inode) ||
(generation && inode->i_generation != generation)) {
iput(inode);
return ERR_PTR(-ESTALE);
}
/* 根据inode获得dentry */
result = d_alloc_anon(inode);
if (!result) {
iput(inode);
return ERR_PTR(-ENOMEM);
}
return result;
}
#ifdef CONFIG_QUOTA
#define QTYPE2NAME(t) ((t)==USRQUOTA?"user":"group")
#define QTYPE2MOPT(on, t) ((t)==USRQUOTA?((on)##USRJQUOTA):((on)##GRPJQUOTA))
/* 如果定义了CONFIG_QUOTA宏，就定义和配额相关的操作 */
static int ext4_dquot_initialize(struct inode *inode, int type);
static int ext4_dquot_drop(struct inode *inode);
static int ext4_write_dquot(struct dquot *dquot);
static int ext4_acquire_dquot(struct dquot *dquot);
static int ext4_release_dquot(struct dquot *dquot);
static int ext4_mark_dquot_dirty(struct dquot *dquot);
static int ext4_write_info(struct super_block *sb, int type);
static int ext4_quota_on(struct super_block *sb, int type, int format_id, char *path);
static int ext4_quota_on_mount(struct super_block *sb, int type);
static ssize_t ext4_quota_read(struct super_block *sb, int type, char *data,
size_t len, loff_t off);
static ssize_t ext4_quota_write(struct super_block *sb, int type,
const char *data, size_t len, loff_t off);
/* 配额操作结构体 */
static struct dquot_operations ext4_quota_operations = {
.initialize = ext4_dquot_initialize,
.drop = ext4_dquot_drop,
.alloc_space = dquot_alloc_space,
.alloc_inode = dquot_alloc_inode,
.free_space = dquot_free_space,
.free_inode = dquot_free_inode,
.transfer = dquot_transfer,
.write_dquot = ext4_write_dquot,
.acquire_dquot = ext4_acquire_dquot,
.release_dquot = ext4_release_dquot,
.mark_dirty = ext4_mark_dquot_dirty,
.write_info = ext4_write_info
};
/* 响应用户传来的配额相关的请求的函数结构体 */
static struct quotactl_ops ext4_qctl_operations = {
.quota_on = ext4_quota_on,
.quota_off = vfs_quota_off,
.quota_sync = vfs_quota_sync,
.get_info = vfs_get_dqinfo,
.set_info = vfs_set_dqinfo,
.get_dqblk = vfs_get_dqblk,
.set_dqblk = vfs_set_dqblk
};
#endif
/* ext4文件系统的超级块的操作函数结构体 */
static struct super_operations ext4_sops = {
.alloc_inode = ext4_alloc_inode,
.destroy_inode = ext4_destroy_inode,
.read_inode = ext4_read_inode,
.write_inode = ext4_write_inode,
.dirty_inode = ext4_dirty_inode,
.delete_inode = ext4_delete_inode,
.put_super = ext4_put_super,
.write_super = ext4_write_super,
.sync_fs = ext4_sync_fs,
.write_super_lockfs = ext4_write_super_lockfs,
.unlockfs = ext4_unlockfs,
.statfs = ext4_statfs,
.remount_fs = ext4_remount,
.clear_inode = ext4_clear_inode,
.show_options = ext4_show_options,
#ifdef CONFIG_QUOTA
.quota_read = ext4_quota_read,
.quota_write = ext4_quota_write,
#endif
};
/* 提供给vfs层调用的函数 */
static struct export_operations ext4_export_ops = {
.get_parent = ext4_get_parent,
.get_dentry = ext4_get_dentry,
};
/* 文件系统挂载的所有选项，以枚举变量的形式展现 */
enum {
Opt_bsd_df, Opt_minix_df, Opt_grpid, Opt_nogrpid,
Opt_resgid, Opt_resuid, Opt_sb, Opt_err_cont, Opt_err_panic, Opt_err_ro,
Opt_nouid32, Opt_nocheck, Opt_debug, Opt_oldalloc, Opt_orlov,
Opt_user_xattr, Opt_nouser_xattr, Opt_acl, Opt_noacl,
Opt_reservation, Opt_noreservation, Opt_noload, Opt_nobh, Opt_bh,
Opt_commit, Opt_journal_update, Opt_journal_inum, Opt_journal_dev,
Opt_abort, Opt_data_journal, Opt_data_ordered, Opt_data_writeback,
Opt_usrjquota, Opt_grpjquota, Opt_offusrjquota, Opt_offgrpjquota,
Opt_jqfmt_vfsold, Opt_jqfmt_vfsv0, Opt_quota, Opt_noquota,
Opt_ignore, Opt_barrier, Opt_err, Opt_resize, Opt_usrquota,
Opt_grpquota, Opt_extents,
};
/* 挂载选项和字符串的对应关系 */
static match_table_t tokens = {
{Opt_bsd_df, "bsddf"},
{Opt_minix_df, "minixdf"},
{Opt_grpid, "grpid"},
{Opt_grpid, "bsdgroups"},
{Opt_nogrpid, "nogrpid"},
{Opt_nogrpid, "sysvgroups"},
{Opt_resgid, "resgid=%u"},
{Opt_resuid, "resuid=%u"},
{Opt_sb, "sb=%u"},
{Opt_err_cont, "errors=continue"},
{Opt_err_panic, "errors=panic"},
{Opt_err_ro, "errors=remount-ro"},
{Opt_nouid32, "nouid32"},
{Opt_nocheck, "nocheck"},
{Opt_nocheck, "check=none"},
{Opt_debug, "debug"},
{Opt_oldalloc, "oldalloc"},
{Opt_orlov, "orlov"},
{Opt_user_xattr, "user_xattr"},
{Opt_nouser_xattr, "nouser_xattr"},
{Opt_acl, "acl"},
{Opt_noacl, "noacl"},
{Opt_reservation, "reservation"},
{Opt_noreservation, "noreservation"},
{Opt_noload, "noload"},
{Opt_nobh, "nobh"},
{Opt_bh, "bh"},
{Opt_commit, "commit=%u"},
{Opt_journal_update, "journal=update"},
{Opt_journal_inum, "journal=%u"},
{Opt_journal_dev, "journal_dev=%u"},
{Opt_abort, "abort"},
{Opt_data_journal, "data=journal"},
{Opt_data_ordered, "data=ordered"},
{Opt_data_writeback, "data=writeback"},
{Opt_offusrjquota, "usrjquota="},
{Opt_usrjquota, "usrjquota=%s"},
{Opt_offgrpjquota, "grpjquota="},
{Opt_grpjquota, "grpjquota=%s"},
{Opt_jqfmt_vfsold, "jqfmt=vfsold"},
{Opt_jqfmt_vfsv0, "jqfmt=vfsv0"},
{Opt_grpquota, "grpquota"},
{Opt_noquota, "noquota"},
{Opt_quota, "quota"},
{Opt_usrquota, "usrquota"},
{Opt_barrier, "barrier=%u"},
{Opt_extents, "extents"},
{Opt_err, NULL},
{Opt_resize, "resize"},
};
/* data参数是挂载选项的字符串，这个函数的作用是返回超级块所在的物理块号，如果用户指定了
超级块的位置比如sb=xxx，就返回超级块的位置，如果没有指定，就返回1，说明在默认的位置 */
static ext4_fsblk_t get_sb_block(void **data)
{
ext4_fsblk_t sb_block;
/* 用char指向传入的挂载选项字符串 */
char *options = (char *) *data;
/* 如果参数是NULL或者挂载选项没有指定超级块所在的位置，就说明超级块在默认的位置 */
if (!options || strncmp(options, "sb=", 3) != 0)
return 1;
/* 运行到这里就说明用户指定了超级块的位置 */
options += 3;
/*把字符串转换成超级块数位置 */
sb_block = simple_strtoul(options, &options, 0);
if (*options && *options != ',') {
printk("EXT4-fs: Invalid sb specification: %s\n",
(char *) *data);
return 1;
}
if (*options == ',')
options++;
*data = (void *) options;
return sb_block;
}
/* 把用户传入的字符串形式的挂载参数转化成s_mount_opt上的对应位图的表现形式 */
static int parse_options (char *options, struct super_block *sb,
unsigned int *inum, unsigned long *journal_devnum,
ext4_fsblk_t *n_blocks_count, int is_remount)
{
struct ext4_sb_info *sbi = EXT4_SB(sb);
char * p;
substring_t args[MAX_OPT_ARGS];
int data_opt = 0;
int option;
#ifdef CONFIG_QUOTA
int qtype;
char *qname;
#endif
/* 检查字符串参数 */
if (!options)
return 1;
/* 逐个解析挂载选项，并在s_mount_opt上对相应位置位 */
while ((p = strsep (&options, ",")) != NULL) {
int token;
if (!*p)
continue;
/* 把字符串选项转换成枚举体里的变量 */
token = match_token(p, tokens, args);
switch (token) {
/* 对于bsddf和minixdf都认为是minix文件系统 */
case Opt_bsd_df:
clear_opt (sbi->s_mount_opt, MINIX_DF);
break;
case Opt_minix_df:
set_opt (sbi->s_mount_opt, MINIX_DF);
break;
/* 设置组有没有默认的组ID，如果有就置位，没有就清除对应位 */
case Opt_grpid:
set_opt (sbi->s_mount_opt, GRPID);
break;
/* 新创建的文件的组ID是创建者 */
case Opt_nogrpid:
clear_opt (sbi->s_mount_opt, GRPID);
break;
/* 设置可能是用保留块的用户ID */
case Opt_resuid:
if (match_int(&args[0], &option))
return 0;
sbi->s_resuid = option;
break;
/* 设置可能使用保留块的组ID */
case Opt_resgid:
if (match_int(&args[0], &option))
return 0;
sbi->s_resgid = option;
break;
case Opt_sb:
/* 已经移动到了get_sb_block()函数中处理这里不再处理 */
break;
/* 这三个宏表示出现错误的时候是继续还是重新挂载成为只读，亦或是报PANIC，在相应位上置位或者清除位 */
case Opt_err_panic:
clear_opt (sbi->s_mount_opt, ERRORS_CONT);
clear_opt (sbi->s_mount_opt, ERRORS_RO);
set_opt (sbi->s_mount_opt, ERRORS_PANIC);
break;
case Opt_err_ro:
clear_opt (sbi->s_mount_opt, ERRORS_CONT);
clear_opt (sbi->s_mount_opt, ERRORS_PANIC);
set_opt (sbi->s_mount_opt, ERRORS_RO);
break;
case Opt_err_cont:
clear_opt (sbi->s_mount_opt, ERRORS_RO);
clear_opt (sbi->s_mount_opt, ERRORS_PANIC);
set_opt (sbi->s_mount_opt, ERRORS_CONT);
break;
/* 设置没有默认文件uid，于是创建者就是文件的uid */
case Opt_nouid32:
set_opt (sbi->s_mount_opt, NO_UID32);
break;
/* 挂载的时候不会进行位图的检查 */
case Opt_nocheck:
clear_opt (sbi->s_mount_opt, CHECK);
break;
/* 调试信息会不会在syslog中打印出来 */
case Opt_debug:
set_opt (sbi->s_mount_opt, DEBUG);
break;
/* 不使用Orlov块分配器 */
case Opt_oldalloc:
set_opt (sbi->s_mount_opt, OLDALLOC);
break;
/* 使用Orlov块分配器 */
case Opt_orlov:
clear_opt (sbi->s_mount_opt, OLDALLOC);
break;
#ifdef CONFIG_EXT4DEV_FS_XATTR
/* 配置文件扩展属性 */
case Opt_user_xattr:
set_opt (sbi->s_mount_opt, XATTR_USER);
break;
/* 不配置文件扩展属性 */
case Opt_nouser_xattr:
clear_opt (sbi->s_mount_opt, XATTR_USER);
break;
#else
case Opt_user_xattr:
case Opt_nouser_xattr:
printk("EXT4 (no)user_xattr options not supported\n");
break;
#endif
/* 文件控制权限列表的配置与否 */
#ifdef CONFIG_EXT4DEV_FS_POSIX_ACL
case Opt_acl:
set_opt(sbi->s_mount_opt, POSIX_ACL);
break;
case Opt_noacl:
clear_opt(sbi->s_mount_opt, POSIX_ACL);
break;
#else
case Opt_acl:
case Opt_noacl:
printk("EXT4 (no)acl options not supported\n");
break;
#endif
case Opt_reservation:
set_opt(sbi->s_mount_opt, RESERVATION);
break;
case Opt_noreservation:
clear_opt(sbi->s_mount_opt, RESERVATION);
break;
case Opt_journal_update:
/* 将ext4的日志格式变成当前的格式 */
if (is_remount) {
printk(KERN_ERR "EXT4-fs: cannot specify "
"journal on remount\n");
return 0;
}
set_opt (sbi->s_mount_opt, UPDATE_JOURNAL);
break;
/* 如果是重新挂载不会理会这些journal选项，inum指定有多少个日志文件 */
case Opt_journal_inum:
if (is_remount) {
printk(KERN_ERR "EXT4-fs: cannot specify "
"journal on remount\n");
return 0;
}
if (match_int(&args[0], &option))
return 0;
*inum = option;
break;
/* 指定日志设备的主次设备号 */
case Opt_journal_dev:
if (is_remount) {
printk(KERN_ERR "EXT4-fs: cannot specify "
"journal on remount\n");
return 0;
}
if (match_int(&args[0], &option))
return 0;
*journal_devnum = option;
break;
/* 挂载的时候不会加载日志 */
case Opt_noload:
set_opt (sbi->s_mount_opt, NOLOAD);
break;
/* 每过多少秒会把脏数据写入到磁盘上 */
case Opt_commit:
if (match_int(&args[0], &option))
return 0;
if (option < 0)
return 0;
if (option == 0)
option = JBD_DEFAULT_MAX_COMMIT_AGE;
sbi->s_commit_interval = HZ * option;
break;
/* 三种数据模式的置为和清除位 */
case Opt_data_journal:
data_opt = EXT4_MOUNT_JOURNAL_DATA;
goto datacheck;
case Opt_data_ordered:
data_opt = EXT4_MOUNT_ORDERED_DATA;
goto datacheck;
case Opt_data_writeback:
data_opt = EXT4_MOUNT_WRITEBACK_DATA;
datacheck:
/* 对数据位的检查 */
if (is_remount) {
if ((sbi->s_mount_opt & EXT4_MOUNT_DATA_FLAGS)
!= data_opt) {
printk(KERN_ERR
"EXT4-fs: cannot change data "
"mode on remount\n");
return 0;
}
} else {
sbi->s_mount_opt &= ~EXT4_MOUNT_DATA_FLAGS;
sbi->s_mount_opt |= data_opt;
}
break;
#ifdef CONFIG_QUOTA
/* 用户配额和组配额的应对函数 */
case Opt_usrjquota:
qtype = USRQUOTA;
goto set_qf_name;
case Opt_grpjquota:
qtype = GRPQUOTA;
set_qf_name:
/* 判断是不是启用了配额限制，如果已经启用了，返回错误，现在应当尚未启动 */
if (sb_any_quota_enabled(sb)) {
printk(KERN_ERR
"EXT4-fs: Cannot change journalled "
"quota options when quota turned on.\n");
return 0;
}
/* 配额限制的文件名字 */
qname = match_strdup(&args[0]);
if (!qname) {
printk(KERN_ERR
"EXT4-fs: not enough memory for "
"storing quotafile name.\n");
return 0;
}
if (sbi->s_qf_names[qtype] &&
strcmp(sbi->s_qf_names[qtype], qname)) {
printk(KERN_ERR
"EXT4-fs: %s quota file already "
"specified.\n", QTYPE2NAME(qtype));
kfree(qname);
return 0;
}
sbi->s_qf_names[qtype] = qname;
/* 检查配额文件名是不是在根目录下 */
if (strchr(sbi->s_qf_names[qtype], '/')) {
printk(KERN_ERR
"EXT4-fs: quotafile must be on "
"filesystem root.\n");
kfree(sbi->s_qf_names[qtype]);
sbi->s_qf_names[qtype] = NULL;
return 0;
}
/* 设置配额对应位已经启动 */
set_opt(sbi->s_mount_opt, QUOTA);
break;
/* 去除配额限制 */
case Opt_offusrjquota:
qtype = USRQUOTA;
goto clear_qf_name;
case Opt_offgrpjquota:
qtype = GRPQUOTA;
clear_qf_name:
if (sb_any_quota_enabled(sb)) {
printk(KERN_ERR "EXT4-fs: Cannot change "
"journalled quota options when "
"quota turned on.\n");
return 0;
}
sbi->s_qf_names[qtype] = NULL;
break;
case Opt_jqfmt_vfsold:
sbi->s_jquota_fmt = QFMT_VFS_OLD;
break;
case Opt_jqfmt_vfsv0:
sbi->s_jquota_fmt = QFMT_VFS_V0;
break;
/* 用户和组的配额限制对应的位操作，已经无配额限制的位操作 */
case Opt_quota:
case Opt_usrquota:
set_opt(sbi->s_mount_opt, QUOTA);
set_opt(sbi->s_mount_opt, USRQUOTA);
break;
case Opt_grpquota:
set_opt(sbi->s_mount_opt, QUOTA);
set_opt(sbi->s_mount_opt, GRPQUOTA);
break;
case Opt_noquota:
if (sb_any_quota_enabled(sb)) {
printk(KERN_ERR "EXT4-fs: Cannot change quota "
"options when quota turned on.\n");
return 0;
}
clear_opt(sbi->s_mount_opt, QUOTA);
clear_opt(sbi->s_mount_opt, USRQUOTA);
clear_opt(sbi->s_mount_opt, GRPQUOTA);
break;
#else
case Opt_quota:
case Opt_usrquota:
case Opt_grpquota:
case Opt_usrjquota:
case Opt_grpjquota:
case Opt_offusrjquota:
case Opt_offgrpjquota:
case Opt_jqfmt_vfsold:
case Opt_jqfmt_vfsv0:
printk(KERN_ERR
"EXT4-fs: journalled quota options not "
"supported.\n");
break;
case Opt_noquota:
break;
#endif
case Opt_abort:
set_opt(sbi->s_mount_opt, ABORT);
break;
case Opt_barrier:
if (match_int(&args[0], &option))
return 0;
if (option)
set_opt(sbi->s_mount_opt, BARRIER);
else
clear_opt(sbi->s_mount_opt, BARRIER);
break;
case Opt_ignore:
break;
/* 改变inode大小 */
case Opt_resize:
if (!is_remount) {
printk("EXT4-fs: resize option only available "
"for remount\n");
return 0;
}
if (match_int(&args[0], &option) != 0)
return 0;
*n_blocks_count = option;
break;
/* 缓冲区首部是不是启用，一般都会启用 */
case Opt_nobh:
set_opt(sbi->s_mount_opt, NOBH);
break;
case Opt_bh:
clear_opt(sbi->s_mount_opt, NOBH);
break;
/* extents文件格式 */
case Opt_extents:
set_opt (sbi->s_mount_opt, EXTENTS);
break;
default:
printk (KERN_ERR
"EXT4-fs: Unrecognized mount option \"%s\" "
"or missing value\n", p);
return 0;
}
}
#ifdef CONFIG_QUOTA
if (sbi->s_qf_names[USRQUOTA] || sbi->s_qf_names[GRPQUOTA]) {
if ((sbi->s_mount_opt & EXT4_MOUNT_USRQUOTA) &&
sbi->s_qf_names[USRQUOTA])
clear_opt(sbi->s_mount_opt, USRQUOTA);
if ((sbi->s_mount_opt & EXT4_MOUNT_GRPQUOTA) &&
sbi->s_qf_names[GRPQUOTA])
clear_opt(sbi->s_mount_opt, GRPQUOTA);
/* 检验配额格式是不是正确 */
if ((sbi->s_qf_names[USRQUOTA] &&
(sbi->s_mount_opt & EXT4_MOUNT_GRPQUOTA)) ||
(sbi->s_qf_names[GRPQUOTA] &&
(sbi->s_mount_opt & EXT4_MOUNT_USRQUOTA))) {
printk(KERN_ERR "EXT4-fs: old and new quota "
"format mixing.\n");
return 0;
}
if (!sbi->s_jquota_fmt) {
printk(KERN_ERR "EXT4-fs: journalled quota format "
"not specified.\n");
return 0;
}
} else {
/* 如果日志配额限制没有指定会打印警告 */
if (sbi->s_jquota_fmt) {
printk(KERN_ERR "EXT4-fs: journalled quota format "
"specified with no journalling "
"enabled.\n");
return 0;
}
}
#endif
return 1;
}
/* 填充ext4_super_block结构体里的一些变量 */
static int ext4_setup_super(struct super_block *sb, struct ext4_super_block *es,
int read_only)
{
struct ext4_sb_info *sbi = EXT4_SB(sb);
int res = 0;
/* 首先检查，如果有问题会打印错误 */
if (le32_to_cpu(es->s_rev_level) > EXT4_MAX_SUPP_REV) {
/* 当前es版本大于最高版本，打印错误，重新以只读格式挂载 */
printk (KERN_ERR "EXT4-fs warning: revision level too high, "
"forcing read-only mode\n");
res = MS_RDONLY;
}
/* 已经只读挂载直接返回 */
if (read_only)
return res;
/* 不是合法的文件系统，提示需要运行e2fsck检查修复 */
if (!(sbi->s_mount_state & EXT4_VALID_FS))
printk (KERN_WARNING "EXT4-fs warning: mounting unchecked fs, "
"running e2fsck is recommended\n");
else if ((sbi->s_mount_state & EXT4_ERROR_FS))
printk (KERN_WARNING
"EXT4-fs warning: mounting fs with errors, "
"running e2fsck is recommended\n");
/* 挂载次数超出最大限制 */
else if ((__s16) le16_to_cpu(es->s_max_mnt_count) >= 0 &&
le16_to_cpu(es->s_mnt_count) >=
(unsigned short) (__s16) le16_to_cpu(es->s_max_mnt_count))
printk (KERN_WARNING
"EXT4-fs warning: maximal mount count reached, "
"running e2fsck is recommended\n");
else if (le32_to_cpu(es->s_checkinterval) &&
(le32_to_cpu(es->s_lastcheck) +
le32_to_cpu(es->s_checkinterval) <= get_seconds()))
printk (KERN_WARNING
"EXT4-fs warning: checktime reached, "
"running e2fsck is recommended\n");
/* 如果es结构体的s_max_mnt_count没有设置，我们就设置为默认的最大值 */
if (!(__s16) le16_to_cpu(es->s_max_mnt_count))
es->s_max_mnt_count = cpu_to_le16(EXT4_DFL_MAX_MNT_COUNT);
/* 挂载数目加一，时间记录，更新版本 */
es->s_mnt_count=cpu_to_le16(le16_to_cpu(es->s_mnt_count) + 1);
es->s_mtime = cpu_to_le32(get_seconds());
ext4_update_dynamic_rev(sb);
EXT4_SET_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER);
/* 把对es结构体的修改写入磁盘 */
ext4_commit_super(sb, es, 1);
/* 如果挂载选项里有debug还需要打印调试信息 */
if (test_opt(sb, DEBUG))
printk(KERN_INFO "[EXT4 FS bs=%lu, gc=%lu, "
"bpg=%lu, ipg=%lu, mo=%04lx]\n",
sb->s_blocksize,
sbi->s_groups_count,
EXT4_BLOCKS_PER_GROUP(sb),
EXT4_INODES_PER_GROUP(sb),
sbi->s_mount_opt);
printk(KERN_INFO "EXT4 FS on %s, ", sb->s_id);
if (EXT4_SB(sb)->s_journal->j_inode == NULL) {
char b[BDEVNAME_SIZE];
printk("external journal on %s\n",
bdevname(EXT4_SB(sb)->s_journal->j_dev, b));
} else {
printk("internal journal\n");
}
return res;
}
/* 挂载时候调用，调用之前super_block会被locked，用来检查组描述符 */
static int ext4_check_descriptors (struct super_block * sb)
{
struct ext4_sb_info *sbi = EXT4_SB(sb);
ext4_fsblk_t first_block = le32_to_cpu(sbi->s_es->s_first_data_block);
ext4_fsblk_t last_block;
ext4_fsblk_t block_bitmap;
ext4_fsblk_t inode_bitmap;
ext4_fsblk_t inode_table;
struct ext4_group_desc * gdp = NULL;
int desc_block = 0;
int i;
ext4_debug ("Checking group descriptors");
/* 对每一个组进行检查 */
for (i = 0; i < sbi->s_groups_count; i++)
{
/* 最后一组要进行块对齐，其实一般情况所有组的大小都是一样的 */
if (i == sbi->s_groups_count - 1)
last_block = ext4_blocks_count(sbi->s_es) - 1;
else
last_block = first_block +
(EXT4_BLOCKS_PER_GROUP(sb) - 1);
/* 如果组描述符是所在位置是一个块的开始 */
if ((i % EXT4_DESC_PER_BLOCK(sb)) == 0)
gdp = (struct ext4_group_desc *)
sbi->s_group_desc[desc_block++]->b_data;
/* 逐个进行数据块位图，inode位图，inode table的检查 */
block_bitmap = ext4_block_bitmap(sb, gdp);
if (block_bitmap < first_block || block_bitmap > last_block)
{
ext4_error (sb, "ext4_check_descriptors",
"Block bitmap for group %d"
" not in group (block %llu)!",
i, block_bitmap);
return 0;
}
inode_bitmap = ext4_inode_bitmap(sb, gdp);
if (inode_bitmap < first_block || inode_bitmap > last_block)
{
ext4_error (sb, "ext4_check_descriptors",
"Inode bitmap for group %d"
" not in group (block %llu)!",
i, inode_bitmap);
return 0;
}
inode_table = ext4_inode_table(sb, gdp);
if (inode_table < first_block ||
inode_table + sbi->s_itb_per_group > last_block)
{
ext4_error (sb, "ext4_check_descriptors",
"Inode table for group %d"
" not in group (block %llu)!",
i, inode_table);
return 0;
}
first_block += EXT4_BLOCKS_PER_GROUP(sb);
gdp = (struct ext4_group_desc *)
((__u8 *)gdp + EXT4_DESC_SIZE(sb));
}
/* 更新free blocks和free inode的值 */
ext4_free_blocks_count_set(sbi->s_es, ext4_count_free_blocks(sb));
sbi->s_es->s_free_inodes_count=cpu_to_le32(ext4_count_free_inodes(sb));
return 1;
}
/* ext4_orphan_cleanup()遍历在超级块上的inode链表，寻找被删除的inode.努力在recovery的时候删除这些inode.
* 我们仅仅是对每一个inode执行iget()和iput()，如果我们碰巧指向了一个正在使用的inode或者是已经被删除的inode都是很安全的，最坏的情况也不过是得到一条"bit already cleared”的消息来自ext4_free_inode()函数.
*唯一的我们可能指向一个错误的inode原因是e2fsck正在文件系统上运行，它一定已经执行过孤儿inode的清理工作了，所以我们就直接终止就可以了。
*/
static void ext4_orphan_cleanup (struct super_block * sb,
struct ext4_super_block * es)
{
unsigned int s_flags = sb->s_flags;
int nr_orphans = 0, nr_truncates = 0;
#ifdef CONFIG_QUOTA
int i;
#endif
/* 没有孤儿inode需要清理 */
if (!es->s_last_orphan) {
jbd_debug(4, "no orphan inodes to clean up\n");
return;
}
/* 有错误，停止清理inode */
if (EXT4_SB(sb)->s_mount_state & EXT4_ERROR_FS) {
if (es->s_last_orphan)
jbd_debug(1, "Errors on filesystem, "
"clearing orphan list.\n");
es->s_last_orphan = 0;
jbd_debug(1, "Skipping orphan recovery on fs with errors.\n");
return;
}
/* 只读文件系统，首先变成可读写的，清理完成以后再变只读的 */
if (s_flags & MS_RDONLY) {
printk(KERN_INFO "EXT4-fs: %s: orphan cleanup on readonly fs\n",
sb->s_id);
sb->s_flags &= ~MS_RDONLY;
}
#ifdef CONFIG_QUOTA
/* 打开配额限制 */
sb->s_flags |= MS_ACTIVE;
for (i = 0; i < MAXQUOTAS; i++) {
if (EXT4_SB(sb)->s_qf_names[i]) {
int ret = ext4_quota_on_mount(sb, i);
if (ret < 0)
printk(KERN_ERR
"EXT4-fs: Cannot turn on journalled "
"quota: error %d\n", ret);
}
}
#endif
/* 遍历每一个inode进行清理工作 */
while (es->s_last_orphan) {
struct inode *inode;
/* 首先获取inode */
if (!(inode =
ext4_orphan_get(sb, le32_to_cpu(es->s_last_orphan)))) {
es->s_last_orphan = 0;
break;
}
list_add(&EXT4_I(inode)->i_orphan, &EXT4_SB(sb)->s_orphan);
DQUOT_INIT(inode);
/* 记录删除的和截断的数目 */
if (inode->i_nlink) {
printk(KERN_DEBUG
"%s: truncating inode %lu to %Ld bytes\n",
__FUNCTION__, inode->i_ino, inode->i_size);
jbd_debug(2, "truncating inode %lu to %Ld bytes\n",
inode->i_ino, inode->i_size);
ext4_truncate(inode);
nr_truncates++;
} else {
printk(KERN_DEBUG
"%s: deleting unreferenced inode %lu\n",
__FUNCTION__, inode->i_ino);
jbd_debug(2, "deleting unreferenced inode %lu\n",
inode->i_ino);
nr_orphans++;
}
/* 删除inode如果引用计数为0 */
iput(inode);
}
#define PLURAL(x) (x), ((x)==1) ? "" : "s"
/* 打印调试信息 */
if (nr_orphans)
printk(KERN_INFO "EXT4-fs: %s: %d orphan inode%s deleted\n",
sb->s_id, PLURAL(nr_orphans));
if (nr_truncates)
printk(KERN_INFO "EXT4-fs: %s: %d truncate%s cleaned up\n",
sb->s_id, PLURAL(nr_truncates));
#ifdef CONFIG_QUOTA
/* 关闭配额限制 */
for (i = 0; i < MAXQUOTAS; i++) {
if (sb_dqopt(sb)->files[i])
vfs_quota_off(sb, i);
}
#endif
/* 重新加载只读状态 */
sb->s_flags = s_flags;
}
#define log2(n) ffz(~(n))
/*
* 最大文件大小. 这是一个直接索引块，简洁索引块，双重间接索引块的限制.
*/
static loff_t ext4_max_size(int bits)
{
loff_t res = EXT4_NDIR_BLOCKS;
const loff_t upper_limit = 0x1ff7fffd000LL;
res += 1LL << (bits-2);
res += 1LL << (2*(bits-2));
res += 1LL << (3*(bits-2));
res <<= bits;
if (res > upper_limit)
res = upper_limit;
return res;
}
/* 返回这个块组描述符所在的块的位置 */
static ext4_fsblk_t descriptor_loc(struct super_block *sb,
ext4_fsblk_t logical_sb_block, int nr)
{
struct ext4_sb_info *sbi = EXT4_SB(sb);
unsigned long bg, first_meta_bg;
int has_super = 0;
/* 第一个元数据块组 */
first_meta_bg = le32_to_cpu(sbi->s_es->s_first_meta_bg);
/* 如果ext4没有元数据块组，就直接返回超级块所在的位置的后边nr个 */
if (!EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_META_BG) ||
nr < first_meta_bg)
return logical_sb_block + nr + 1;
/* 第bg哥块组描述符 */
bg = sbi->s_desc_per_block * nr;
/* 有超级块的话需要跳过第一个超级块 */
if (ext4_bg_has_super(sb, bg))
has_super = 1;
return (has_super + ext4_group_first_block_no(sb, bg));
}
/* 本文件最重要的函数，填充super_block结构体以及ext4_sb_info结构体的各个字段，挂载过程中最重要的函数 */
static int ext4_fill_super (struct super_block *sb, void *data, int silent)
{
struct buffer_head * bh;
struct ext4_super_block *es = NULL;
struct ext4_sb_info *sbi;
ext4_fsblk_t block;
ext4_fsblk_t sb_block = get_sb_block(&data);
ext4_fsblk_t logical_sb_block;
unsigned long offset = 0;
unsigned int journal_inum = 0;
unsigned long journal_devnum = 0;
unsigned long def_mount_opts;
struct inode *root;
int blocksize;
int hblock;
int db_count;
int i;
int needs_recovery;
__le32 features;
__u64 blocks_count;
/* 为ext4_sb_info分配空间，包括一些简单的缺省值的初始赋值 */
sbi = kzalloc(sizeof(*sbi), GFP_KERNEL);
if (!sbi)
return -ENOMEM;
sb->s_fs_info = sbi;
sbi->s_mount_opt = 0;
sbi->s_resuid = EXT4_DEF_RESUID;
sbi->s_resgid = EXT4_DEF_RESGID;
unlock_kernel();
/* 最小的块大小 */
blocksize = sb_min_blocksize(sb, EXT4_MIN_BLOCK_SIZE);
if (!blocksize) {
printk(KERN_ERR "EXT4-fs: unable to set blocksize\n");
goto out_fail;
}
/*
* ext4只有在1KB大小的块情况下才会buffer对齐，如果不是1KB大小的块我们需要计算在块内的偏移值 */
if (blocksize != EXT4_MIN_BLOCK_SIZE) {
logical_sb_block = sb_block * EXT4_MIN_BLOCK_SIZE;
offset = do_div(logical_sb_block, blocksize);
} else {
logical_sb_block = sb_block;
}
/* 读取超级块所在的块 */
if (!(bh = sb_bread(sb, logical_sb_block))) {
printk (KERN_ERR "EXT4-fs: unable to read superblock\n");
goto out_fail;
}
/*
* s_es必须被尽快的初始化，因为很多的ext4的宏会依赖这个结构体的值
*/
es = (struct ext4_super_block *) (((char *)bh->b_data) + offset);
sbi->s_es = es;
/* 超级块的魔数，检查是否合法 */
sb->s_magic = le16_to_cpu(es->s_magic);
if (sb->s_magic != EXT4_SUPER_MAGIC)
goto cantfind_ext4;
/* 在转换挂载选项之前，先设置初始的值 */
def_mount_opts = le32_to_cpu(es->s_default_mount_opts);
if (def_mount_opts & EXT4_DEFM_DEBUG)
set_opt(sbi->s_mount_opt, DEBUG);
if (def_mount_opts & EXT4_DEFM_BSDGROUPS)
set_opt(sbi->s_mount_opt, GRPID);
if (def_mount_opts & EXT4_DEFM_UID16)
set_opt(sbi->s_mount_opt, NO_UID32);
#ifdef CONFIG_EXT4DEV_FS_XATTR
if (def_mount_opts & EXT4_DEFM_XATTR_USER)
set_opt(sbi->s_mount_opt, XATTR_USER);
#endif
#ifdef CONFIG_EXT4DEV_FS_POSIX_ACL
if (def_mount_opts & EXT4_DEFM_ACL)
set_opt(sbi->s_mount_opt, POSIX_ACL);
#endif
if ((def_mount_opts & EXT4_DEFM_JMODE) == EXT4_DEFM_JMODE_DATA)
sbi->s_mount_opt |= EXT4_MOUNT_JOURNAL_DATA;
else if ((def_mount_opts & EXT4_DEFM_JMODE) == EXT4_DEFM_JMODE_ORDERED)
sbi->s_mount_opt |= EXT4_MOUNT_ORDERED_DATA;
else if ((def_mount_opts & EXT4_DEFM_JMODE) == EXT4_DEFM_JMODE_WBACK)
sbi->s_mount_opt |= EXT4_MOUNT_WRITEBACK_DATA;
if (le16_to_cpu(sbi->s_es->s_errors) == EXT4_ERRORS_PANIC)
set_opt(sbi->s_mount_opt, ERRORS_PANIC);
else if (le16_to_cpu(sbi->s_es->s_errors) == EXT4_ERRORS_RO)
set_opt(sbi->s_mount_opt, ERRORS_RO);
else
set_opt(sbi->s_mount_opt, ERRORS_CONT);
sbi->s_resuid = le16_to_cpu(es->s_def_resuid);
sbi->s_resgid = le16_to_cpu(es->s_def_resgid);
set_opt(sbi->s_mount_opt, RESERVATION);
/* 缺省挂载选项设置完毕以后，根据用户传入的挂载选项来设置 */
if (!parse_options ((char *) data, sb, &journal_inum, &journal_devnum,
NULL, 0))
goto failed_mount;
/* 超级块结构体的flags和权限控制列表 */
sb->s_flags = (sb->s_flags & ~MS_POSIXACL) |
((sbi->s_mount_opt & EXT4_MOUNT_POSIX_ACL) ? MS_POSIXACL : 0);
if (le32_to_cpu(es->s_rev_level) == EXT4_GOOD_OLD_REV &&
(EXT4_HAS_COMPAT_FEATURE(sb, ~0U) ||
EXT4_HAS_RO_COMPAT_FEATURE(sb, ~0U) ||
EXT4_HAS_INCOMPAT_FEATURE(sb, ~0U)))
printk(KERN_WARNING
"EXT4-fs warning: feature flags set on rev 0 fs, "
"running e2fsck is recommended\n");
/*
* 检查feature flags有没有不支持的flag.
*/
features = EXT4_HAS_INCOMPAT_FEATURE(sb, ~EXT4_FEATURE_INCOMPAT_SUPP);
if (features) {
printk(KERN_ERR "EXT4-fs: %s: couldn't mount because of "
"unsupported optional features (%x).\n",
sb->s_id, le32_to_cpu(features));
goto failed_mount;
}
features = EXT4_HAS_RO_COMPAT_FEATURE(sb, ~EXT4_FEATURE_RO_COMPAT_SUPP);
if (!(sb->s_flags & MS_RDONLY) && features) {
printk(KERN_ERR "EXT4-fs: %s: couldn't mount RDWR because of "
"unsupported optional features (%x).\n",
sb->s_id, le32_to_cpu(features));
goto failed_mount;
}
/* blocksize设置成为用户传入的blocksize */
blocksize = BLOCK_SIZE << le32_to_cpu(es->s_log_block_size);
/* 检查blocksize是不是合法 */
if (blocksize < EXT4_MIN_BLOCK_SIZE ||
blocksize > EXT4_MAX_BLOCK_SIZE) {
printk(KERN_ERR
"EXT4-fs: Unsupported filesystem blocksize %d on %s.\n",
blocksize, sb->s_id);
goto failed_mount;
}
/* 得到块设备上的块大小 */
hblock = bdev_hardsect_size(sb->s_bdev);
/* 块设备的块大小和文件系统的块大小不一致 */
if (sb->s_blocksize != blocksize) {
/*
* 确保文件系统块大小大于硬件设备上的块大小.
*/
if (blocksize < hblock) {
printk(KERN_ERR "EXT4-fs: blocksize %d too small for "
"device blocksize %d.\n", blocksize, hblock);
goto failed_mount;
}
brelse (bh);
/* 根据新的块大小，得到新的超级块和es结构体 */
sb_set_blocksize(sb, blocksize);
logical_sb_block = sb_block * EXT4_MIN_BLOCK_SIZE;
offset = do_div(logical_sb_block, blocksize);
bh = sb_bread(sb, logical_sb_block);
if (!bh) {
printk(KERN_ERR
"EXT4-fs: Can't read superblock on 2nd try.\n");
goto failed_mount;
}
es = (struct ext4_super_block *)(((char *)bh->b_data) + offset);
sbi->s_es = es;
if (es->s_magic != cpu_to_le16(EXT4_SUPER_MAGIC)) {
printk (KERN_ERR
"EXT4-fs: Magic mismatch, very weird !\n");
goto failed_mount;
}
}
/* 文件系统的最大文件大小，调用了前边的函数 */
sb->s_maxbytes = ext4_max_size(sb->s_blocksize_bits);
/* inode版本如果是老版本，inode和first ino都是老的值 */
if (le32_to_cpu(es->s_rev_level) == EXT4_GOOD_OLD_REV) {
sbi->s_inode_size = EXT4_GOOD_OLD_INODE_SIZE;
sbi->s_first_ino = EXT4_GOOD_OLD_FIRST_INO;
} else {
/* 指定的inode大小的偏移值 */
sbi->s_inode_size = le16_to_cpu(es->s_inode_size);
sbi->s_first_ino = le32_to_cpu(es->s_first_ino);
if ((sbi->s_inode_size < EXT4_GOOD_OLD_INODE_SIZE) ||
(sbi->s_inode_size & (sbi->s_inode_size - 1)) ||
(sbi->s_inode_size > blocksize)) {
printk (KERN_ERR
"EXT4-fs: unsupported inode size: %d\n",
sbi->s_inode_size);
goto failed_mount;
}
}
/* 文件碎片的大小设置和检查 */
sbi->s_frag_size = EXT4_MIN_FRAG_SIZE <<
le32_to_cpu(es->s_log_frag_size);
if (blocksize != sbi->s_frag_size) {
printk(KERN_ERR
"EXT4-fs: fragsize %lu != blocksize %u (unsupported)\n",
sbi->s_frag_size, blocksize);
goto failed_mount;
}
/* 组描述符大小 */
sbi->s_desc_size = le16_to_cpu(es->s_desc_size);
if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_64BIT)) {
if (sbi->s_desc_size < EXT4_MIN_DESC_SIZE_64BIT ||
sbi->s_desc_size > EXT4_MAX_DESC_SIZE ||
sbi->s_desc_size & (sbi->s_desc_size - 1)) {
printk(KERN_ERR
"EXT4-fs: unsupported descriptor size %lu\n",
sbi->s_desc_size);
goto failed_mount;
}
} else
sbi->s_desc_size = EXT4_MIN_DESC_SIZE;
/* 每组的块，inode，碎片数目等字段赋值 */
sbi->s_blocks_per_group = le32_to_cpu(es->s_blocks_per_group);
sbi->s_frags_per_group = le32_to_cpu(es->s_frags_per_group);
sbi->s_inodes_per_group = le32_to_cpu(es->s_inodes_per_group);
if (EXT4_INODE_SIZE(sb) == 0)
goto cantfind_ext4;
sbi->s_inodes_per_block = blocksize / EXT4_INODE_SIZE(sb);
if (sbi->s_inodes_per_block == 0)
goto cantfind_ext4;
sbi->s_itb_per_group = sbi->s_inodes_per_group /
sbi->s_inodes_per_block;
sbi->s_desc_per_block = blocksize / EXT4_DESC_SIZE(sb);
sbi->s_sbh = bh;
sbi->s_mount_state = le16_to_cpu(es->s_state);
sbi->s_addr_per_block_bits = log2(EXT4_ADDR_PER_BLOCK(sb));
sbi->s_desc_per_block_bits = log2(EXT4_DESC_PER_BLOCK(sb));
for (i=0; i < 4; i++)
sbi->s_hash_seed[i] = le32_to_cpu(es->s_hash_seed[i]);
sbi->s_def_hash_version = es->s_def_hash_version;
/* 赋值过后的检查，检查结果非法就会跳转到failed_mount */
if (sbi->s_blocks_per_group > blocksize * 8) {
printk (KERN_ERR
"EXT4-fs: #blocks per group too big: %lu\n",
sbi->s_blocks_per_group);
goto failed_mount;
}
if (sbi->s_frags_per_group > blocksize * 8) {
printk (KERN_ERR
"EXT4-fs: #fragments per group too big: %lu\n",
sbi->s_frags_per_group);
goto failed_mount;
}
if (sbi->s_inodes_per_group > blocksize * 8) {
printk (KERN_ERR
"EXT4-fs: #inodes per group too big: %lu\n",
sbi->s_inodes_per_group);
goto failed_mount;
}
if (ext4_blocks_count(es) >
(sector_t)(~0ULL) >> (sb->s_blocksize_bits - 9)) {
printk(KERN_ERR "EXT4-fs: filesystem on %s:"
" too large to mount safely\n", sb->s_id);
if (sizeof(sector_t) < 8)
printk(KERN_WARNING "EXT4-fs: CONFIG_LBD not "
"enabled\n");
goto failed_mount;
}
if (EXT4_BLOCKS_PER_GROUP(sb) == 0)
goto cantfind_ext4;
/* 组的数目 */
blocks_count = (ext4_blocks_count(es) -
le32_to_cpu(es->s_first_data_block) +
EXT4_BLOCKS_PER_GROUP(sb) - 1);
do_div(blocks_count, EXT4_BLOCKS_PER_GROUP(sb));
sbi->s_groups_count = blocks_count;
/* 有多少块存放块组描述符 */
db_count = (sbi->s_groups_count + EXT4_DESC_PER_BLOCK(sb) - 1) /
EXT4_DESC_PER_BLOCK(sb);
/* 赋值sbi的块组描述符结构体数组 */
sbi->s_group_desc = kmalloc(db_count * sizeof (struct buffer_head *),
GFP_KERNEL);
if (sbi->s_group_desc == NULL) {
printk (KERN_ERR "EXT4-fs: not enough memory\n");
goto failed_mount;
}
/* 块组描述符数组的锁 */
bgl_lock_init(&sbi->s_blockgroup_lock);
/* 逐个对块组描述符赋值 */
for (i = 0; i < db_count; i++) {
block = descriptor_loc(sb, logical_sb_block, i);
sbi->s_group_desc[i] = sb_bread(sb, block);
if (!sbi->s_group_desc[i]) {
printk (KERN_ERR "EXT4-fs: "
"can't read group descriptor %d\n", i);
db_count = i;
goto failed_mount2;
}
}
/* 检查 */
if (!ext4_check_descriptors (sb)) {
printk(KERN_ERR "EXT4-fs: group descriptors corrupted!\n");
goto failed_mount2;
}
/* sbi的存放块组描述符的块数目和几个单cpu计数器 */
sbi->s_gdb_count = db_count;
get_random_bytes(&sbi->s_next_generation, sizeof(u32));
spin_lock_init(&sbi->s_next_gen_lock);
percpu_counter_init(&sbi->s_freeblocks_counter,
ext4_count_free_blocks(sb));
percpu_counter_init(&sbi->s_freeinodes_counter,
ext4_count_free_inodes(sb));
percpu_counter_init(&sbi->s_dirs_counter,
ext4_count_dirs(sb));
/* 每个文件系统的保留快列表和锁初始化 */
spin_lock_init(&sbi->s_rsv_window_lock);
sbi->s_rsv_window_root = RB_ROOT;
/* 保留块窗口变量初始化. */
sbi->s_rsv_window_head.rsv_start = EXT4_RESERVE_WINDOW_NOT_ALLOCATED;
sbi->s_rsv_window_head.rsv_end = EXT4_RESERVE_WINDOW_NOT_ALLOCATED;
sbi->s_rsv_window_head.rsv_alloc_hit = 0;
sbi->s_rsv_window_head.rsv_goal_size = 0;
ext4_rsv_window_add(sb, &sbi->s_rsv_window_head);
/*
* 初始化能够读取一个inode的相关资源
*/
sb->s_op = &ext4_sops;
sb->s_export_op = &ext4_export_ops;
sb->s_xattr = ext4_xattr_handlers;
#ifdef CONFIG_QUOTA
sb->s_qcop = &ext4_qctl_operations;
sb->dq_op = &ext4_quota_operations;
#endif
INIT_LIST_HEAD(&sbi->s_orphan);
sb->s_root = NULL;
/* 有没有错误需要recovery的 */
needs_recovery = (es->s_last_orphan != 0 ||
EXT4_HAS_INCOMPAT_FEATURE(sb,
EXT4_FEATURE_INCOMPAT_RECOVER));
/*
* 第一个创建的inode是journal的inode
*/
if (!test_opt(sb, NOLOAD) &&
EXT4_HAS_COMPAT_FEATURE(sb, EXT4_FEATURE_COMPAT_HAS_JOURNAL)) {
if (ext4_load_journal(sb, es, journal_devnum))
goto failed_mount3;
} else if (journal_inum) {
if (ext4_create_journal(sb, es, journal_inum))
goto failed_mount3;
} else {
if (!silent)
printk (KERN_ERR
"ext4: No journal on filesystem on %s\n",
sb->s_id);
goto failed_mount3;
}
/* 如果有要求的话，我们会更新日志，然后检查日志模式. */
switch (test_opt(sb, DATA_FLAGS)) {
case 0:
/* 没有日志模式设置，根据日志能力设置
*/
if (jbd2_journal_check_available_features
(sbi->s_journal, 0, 0, JBD2_FEATURE_INCOMPAT_REVOKE))
set_opt(sbi->s_mount_opt, ORDERED_DATA);
else
set_opt(sbi->s_mount_opt, JOURNAL_DATA);
break;
case EXT4_MOUNT_ORDERED_DATA:
case EXT4_MOUNT_WRITEBACK_DATA:
if (!jbd2_journal_check_available_features
(sbi->s_journal, 0, 0, JBD2_FEATURE_INCOMPAT_REVOKE)) {
printk(KERN_ERR "EXT4-fs: Journal does not support "
"requested data journaling mode\n");
goto failed_mount4;
}
default:
break;
}
/* 只有writeback模式下才能NOBH */
if (test_opt(sb, NOBH)) {
if (!(test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_WRITEBACK_DATA)) {
printk(KERN_WARNING "EXT4-fs: Ignoring nobh option - "
"its supported only with writeback mode\n");
clear_opt(sbi->s_mount_opt, NOBH);
}
}
/*
* 现在可以挂载文件系统了.分配根节点的inode和dentry
*/
root = iget(sb, EXT4_ROOT_INO);
sb->s_root = d_alloc_root(root);
if (!sb->s_root) {
printk(KERN_ERR "EXT4-fs: get root inode failed\n");
iput(root);
goto failed_mount4;
}
if (!S_ISDIR(root->i_mode) || !root->i_blocks || !root->i_size) {
dput(sb->s_root);
sb->s_root = NULL;
printk(KERN_ERR "EXT4-fs: corrupt root inode, run e2fsck\n");
goto failed_mount4;
}
ext4_setup_super (sb, es, sb->s_flags & MS_RDONLY);
EXT4_SB(sb)->s_mount_state |= EXT4_ORPHAN_FS;
ext4_orphan_cleanup(sb, es);
EXT4_SB(sb)->s_mount_state &= ~EXT4_ORPHAN_FS;
if (needs_recovery)
printk (KERN_INFO "EXT4-fs: recovery complete.\n");
ext4_mark_recovery_complete(sb, es);
printk (KERN_INFO "EXT4-fs: mounted filesystem with %s data mode.\n",
test_opt(sb,DATA_FLAGS) == EXT4_MOUNT_JOURNAL_DATA ? "journal":
test_opt(sb,DATA_FLAGS) == EXT4_MOUNT_ORDERED_DATA ? "ordered":
"writeback");
ext4_ext_init(sb);
lock_kernel();
return 0;
cantfind_ext4:
if (!silent)
printk(KERN_ERR "VFS: Can't find ext4 filesystem on dev %s.\n",
sb->s_id);
goto failed_mount;
failed_mount4:
jbd2_journal_destroy(sbi->s_journal);
failed_mount3:
percpu_counter_destroy(&sbi->s_freeblocks_counter);
percpu_counter_destroy(&sbi->s_freeinodes_counter);
percpu_counter_destroy(&sbi->s_dirs_counter);
failed_mount2:
for (i = 0; i < db_count; i++)
brelse(sbi->s_group_desc[i]);
kfree(sbi->s_group_desc);
failed_mount:
#ifdef CONFIG_QUOTA
for (i = 0; i < MAXQUOTAS; i++)
kfree(sbi->s_qf_names[i]);
#endif
ext4_blkdev_remove(sbi);
brelse(bh);
out_fail:
sb->s_fs_info = NULL;
kfree(sbi);
lock_kernel();
return -EINVAL;
}
/*
* 初始化日志参数，这个操作不仅仅会在挂载的时候执行，在recovery和remount的时候也会执行.
*/
static void ext4_init_journal_params(struct super_block *sb, journal_t *journal)
{
struct ext4_sb_info *sbi = EXT4_SB(sb);
/* 如果用户传入的commit参数是0的话，就设置为默认的commit时间 */
if (sbi->s_commit_interval)
journal->j_commit_interval = sbi->s_commit_interval;
spin_lock(&journal->j_state_lock);
if (test_opt(sb, BARRIER))
journal->j_flags |= JBD2_BARRIER;
else
journal->j_flags &= ~JBD2_BARRIER;
spin_unlock(&journal->j_state_lock);
}
/* 返回ext4的日志结构体 */
static journal_t *ext4_get_journal(struct super_block *sb,
unsigned int journal_inum)
{
struct inode *journal_inode;
journal_t *journal;
/* 获得journal的inode，然后检验是不是引用计数为0，这意味着被删除了 */
journal_inode = iget(sb, journal_inum);
if (!journal_inode) {
printk(KERN_ERR "EXT4-fs: no journal found.\n");
return NULL;
}
if (!journal_inode->i_nlink) {
make_bad_inode(journal_inode);
iput(journal_inode);
printk(KERN_ERR "EXT4-fs: journal inode is deleted.\n");
return NULL;
}
jbd_debug(2, "Journal inode found at %p: %Ld bytes\n",
journal_inode, journal_inode->i_size);
/* 检查inode是否合法 */
if (is_bad_inode(journal_inode) || !S_ISREG(journal_inode->i_mode)) {
printk(KERN_ERR "EXT4-fs: invalid journal inode.\n");
iput(journal_inode);
return NULL;
}
/* 调用jbd2模块的初始化函数 */
journal = jbd2_journal_init_inode(journal_inode);
if (!journal) {
printk(KERN_ERR "EXT4-fs: Could not load journal inode\n");
iput(journal_inode);
return NULL;
}
journal->j_private = sb;
ext4_init_journal_params(sb, journal);
return journal;
}
/* 根据用户传入的主次设备号来获得日志结构体 */
static journal_t *ext4_get_dev_journal(struct super_block *sb,
dev_t j_dev)
{
struct buffer_head * bh;
journal_t *journal;
ext4_fsblk_t start;
ext4_fsblk_t len;
int hblock, blocksize;
ext4_fsblk_t sb_block;
unsigned long offset;
struct ext4_super_block * es;
struct block_device *bdev;
/* 根据主次设备号，获得块设备结构体 */
bdev = ext4_blkdev_get(j_dev);
if (bdev == NULL)
return NULL;
if (bd_claim(bdev, sb)) {
printk(KERN_ERR
"EXT4: failed to claim external journal device.\n");
blkdev_put(bdev);
return NULL;
}
/* 检查ext4和日志块设备的块大小是不是匹配，要求ext4块大小一定要大于日志块设备的块大小 */
blocksize = sb->s_blocksize;
hblock = bdev_hardsect_size(bdev);
if (blocksize < hblock) {
printk(KERN_ERR
"EXT4-fs: blocksize too small for journal device.\n");
goto out_bdev;
}
/* 首先从日志块设备上获得超级块 */
sb_block = EXT4_MIN_BLOCK_SIZE / blocksize;
offset = EXT4_MIN_BLOCK_SIZE % blocksize;
set_blocksize(bdev, blocksize);
if (!(bh = __bread(bdev, sb_block, blocksize))) {
printk(KERN_ERR "EXT4-fs: couldn't read superblock of "
"external journal\n");
goto out_bdev;
}
/* 然后获得es结构体 */
es = (struct ext4_super_block *) (((char *)bh->b_data) + offset);
if ((le16_to_cpu(es->s_magic) != EXT4_SUPER_MAGIC) ||
!(le32_to_cpu(es->s_feature_incompat) &
EXT4_FEATURE_INCOMPAT_JOURNAL_DEV)) {
printk(KERN_ERR "EXT4-fs: external journal has "
"bad superblock\n");
brelse(bh);
goto out_bdev;
}
/* 检验日志的uuid是不是符合 */
if (memcmp(EXT4_SB(sb)->s_es->s_journal_uuid, es->s_uuid, 16)) {
printk(KERN_ERR "EXT4-fs: journal UUID does not match\n");
brelse(bh);
goto out_bdev;
}
len = ext4_blocks_count(es);
start = sb_block + 1;
brelse(bh); /* 超级块使用完毕 */
/* jbd2模块的初始化操作 */
journal = jbd2_journal_init_dev(bdev, sb->s_bdev,
start, len, blocksize);
if (!journal) {
printk(KERN_ERR "EXT4-fs: failed to create device journal\n");
goto out_bdev;
}
journal->j_private = sb;
/* 经典函数，读j_sb_buffer块 */
ll_rw_block(READ, 1, &journal->j_sb_buffer);
/* 等待读取完毕 */
wait_on_buffer(journal->j_sb_buffer);
/* 检查是不是uptodate以及是不是仅仅一个user */
if (!buffer_uptodate(journal->j_sb_buffer)) {
printk(KERN_ERR "EXT4-fs: I/O error on journal device\n");
goto out_journal;
}
if (be32_to_cpu(journal->j_superblock->s_nr_users) != 1) {
printk(KERN_ERR "EXT4-fs: External journal has more than one "
"user (unsupported) - %d\n",
be32_to_cpu(journal->j_superblock->s_nr_users));
goto out_journal;
}
EXT4_SB(sb)->journal_bdev = bdev;
ext4_init_journal_params(sb, journal);
return journal;
out_journal:
jbd2_journal_destroy(journal);
out_bdev:
ext4_blkdev_put(bdev);
return NULL;
}
/* 根据日志的主次设备号来进行加载日志操作 */
static int ext4_load_journal(struct super_block *sb,
struct ext4_super_block *es,
unsigned long journal_devnum)
{
journal_t *journal;
unsigned int journal_inum = le32_to_cpu(es->s_journal_inum);
dev_t journal_dev;
int err = 0;
int really_read_only;
/* 如果指定了日志设备就用指定的日志设备 */
if (journal_devnum &&
journal_devnum != le32_to_cpu(es->s_journal_dev)) {
printk(KERN_INFO "EXT4-fs: external journal device major/minor "
"numbers have changed\n");
journal_dev = new_decode_dev(journal_devnum);
} else
/* 使用es的默认日志块设备 */
journal_dev = new_decode_dev(le32_to_cpu(es->s_journal_dev));
really_read_only = bdev_read_only(sb->s_bdev);
/*
* 如果我们在recovery，我们需要检查是不是拥有对设备的读写权限.
*/
if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER)) {
if (sb->s_flags & MS_RDONLY) {
printk(KERN_INFO "EXT4-fs: INFO: recovery "
"required on readonly filesystem.\n");
if (really_read_only) {
printk(KERN_ERR "EXT4-fs: write access "
"unavailable, cannot proceed.\n");
return -EROFS;
}
printk (KERN_INFO "EXT4-fs: write access will "
"be enabled during recovery.\n");
}
}
/* journalinum和日志块设备不能共存 */
if (journal_inum && journal_dev) {
printk(KERN_ERR "EXT4-fs: filesystem has both journal "
"and inode journals!\n");
return -EINVAL;
}
/* 获得日志结构体 */
if (journal_inum) {
if (!(journal = ext4_get_journal(sb, journal_inum)))
return -EINVAL;
} else {
if (!(journal = ext4_get_dev_journal(sb, journal_dev)))
return -EINVAL;
}
/* 获得日志结构体以后，检查是不是把日志需要更新为当前的格式 */
if (!really_read_only && test_opt(sb, UPDATE_JOURNAL)) {
err = jbd2_journal_update_format(journal);
if (err) {
printk(KERN_ERR "EXT4-fs: error updating journal.\n");
jbd2_journal_destroy(journal);
return err;
}
}
if (!EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER))
err = jbd2_journal_wipe(journal, !really_read_only);
if (!err)
err = jbd2_journal_load(journal);
if (err) {
printk(KERN_ERR "EXT4-fs: error loading journal.\n");
jbd2_journal_destroy(journal);
return err;
}
EXT4_SB(sb)->s_journal = journal;
ext4_clear_journal_err(sb, es);
if (journal_devnum &&
journal_devnum != le32_to_cpu(es->s_journal_dev)) {
es->s_journal_dev = cpu_to_le32(journal_devnum);
sb->s_dirt = 1;
/* 把recovery的flag刷新到磁盘 */
ext4_commit_super(sb, es, 1);
}
return 0;
}
/* 创建日志结构体 */
static int ext4_create_journal(struct super_block * sb,
struct ext4_super_block * es,
unsigned int journal_inum)
{
journal_t *journal;
/* 检查，只读情况下返回错误 */
if (sb->s_flags & MS_RDONLY) {
printk(KERN_ERR "EXT4-fs: readonly filesystem when trying to "
"create journal.\n");
return -EROFS;
}
/* 通过日志的inode号码返回日志结构体 */
if (!(journal = ext4_get_journal(sb, journal_inum)))
return -EINVAL;
printk(KERN_INFO "EXT4-fs: creating new journal on inode %u\n",
journal_inum);
/* jbd2的真正创建的函数 */
if (jbd2_journal_create(journal)) {
printk(KERN_ERR "EXT4-fs: error creating journal.\n");
jbd2_journal_destroy(journal);
return -EIO;
}
/* 存储日志结构体 */
EXT4_SB(sb)->s_journal = journal;
ext4_update_dynamic_rev(sb);
/* feature的设置 */
EXT4_SET_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER);
EXT4_SET_COMPAT_FEATURE(sb, EXT4_FEATURE_COMPAT_HAS_JOURNAL);
es->s_journal_inum = cpu_to_le32(journal_inum);
sb->s_dirt = 1;
/* 把recovery的flag写到磁盘. */
ext4_commit_super(sb, es, 1);
return 0;
}
/* 把超级块写入到磁盘 */
static void ext4_commit_super (struct super_block * sb,
struct ext4_super_block * es,
int sync)
{
struct buffer_head *sbh = EXT4_SB(sb)->s_sbh;
if (!sbh)
return;
/* 写入时间的修改 */
es->s_wtime = cpu_to_le32(get_seconds());
ext4_free_blocks_count_set(es, ext4_count_free_blocks(sb));
es->s_free_inodes_count = cpu_to_le32(ext4_count_free_inodes(sb));
BUFFER_TRACE(sbh, "marking dirty");
/* 标记缓冲区首部为脏 */
mark_buffer_dirty(sbh);
/* 然后同步 */
if (sync)
sync_dirty_buffer(sbh);
}
/*
* recovry结束了以后，如果我们在mount或者remount为只读，我们就需要在日志上记录 */
static void ext4_mark_recovery_complete(struct super_block * sb,
struct ext4_super_block * es)
{
journal_t *journal = EXT4_SB(sb)->s_journal;
jbd2_journal_lock_updates(journal);
jbd2_journal_flush(journal);
if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER) &&
sb->s_flags & MS_RDONLY) {
/* 正在recovery并且还是只读，把super的修改写入到磁盘 */
EXT4_CLEAR_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER);
sb->s_dirt = 0;
ext4_commit_super(sb, es, 1);
}
jbd2_journal_unlock_updates(journal);
}
/*
* 如果我们挂载的fs记录了一个错误，我们需要把错误放到主文件系统中
*/
static void ext4_clear_journal_err(struct super_block * sb,
struct ext4_super_block * es)
{
journal_t *journal;
int j_errno;
const char *errstr;
journal = EXT4_SB(sb)->s_journal;
/*
* 检查状态，如果有错误，用ext4_error()或者ext4_abort()
*/
j_errno = jbd2_journal_errno(journal);
if (j_errno) {
char nbuf[16];
/* 把错误转换成字符串 */
errstr = ext4_decode_error(sb, j_errno, nbuf);
/* 打印错误 */
ext4_warning(sb, __FUNCTION__, "Filesystem error recorded "
"from previous mount: %s", errstr);
ext4_warning(sb, __FUNCTION__, "Marking fs in need of "
"filesystem check.");
/* es和sbi结构体都要记录错误，然后把超级块写入到磁盘 */
EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS;
es->s_state |= cpu_to_le16(EXT4_ERROR_FS);
ext4_commit_super (sb, es, 1);
/* 记录结束以后，清理错误字段 */
jbd2_journal_clear_err(journal);
}
}
/*
* 强制提交事务，等待结束.
*/
int ext4_force_commit(struct super_block *sb)
{
journal_t *journal;
int ret;
if (sb->s_flags & MS_RDONLY)
return 0;
journal = EXT4_SB(sb)->s_journal;
sb->s_dirt = 0;
ret = ext4_journal_force_commit(journal);
return ret;
}
/*
* Ext4通过日志写入超级块,我们不必手动写入，仅仅是启动writeback即可
*/
static void ext4_write_super (struct super_block * sb)
{
if (mutex_trylock(&sb->s_lock) != 0)
BUG();
sb->s_dirt = 0;
}
static int ext4_sync_fs(struct super_block *sb, int wait)
{
tid_t target;
sb->s_dirt = 0;
if (jbd2_journal_start_commit(EXT4_SB(sb)->s_journal, &target)) {
if (wait)
jbd2_log_wait_commit(EXT4_SB(sb)->s_journal, target);
}
return 0;
}
/*
* LVM会调用这个函数，在一个只读快照创建之前。这个函数会把日志全部写入磁盘.
*/
static void ext4_write_super_lockfs(struct super_block *sb)
{
sb->s_dirt = 0;
if (!(sb->s_flags & MS_RDONLY)) {
journal_t *journal = EXT4_SB(sb)->s_journal;
/* 启动日志barriar. */
jbd2_journal_lock_updates(journal);
jbd2_journal_flush(journal);
EXT4_CLEAR_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER);
ext4_commit_super(sb, EXT4_SB(sb)->s_es, 1);
}
}
/*
* 当LVM的日志创建操作结束以后调用，重新设置RECOVER的flag
*/
static void ext4_unlockfs(struct super_block *sb)
{
if (!(sb->s_flags & MS_RDONLY)) {
lock_super(sb);
EXT4_SET_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER);
ext4_commit_super(sb, EXT4_SB(sb)->s_es, 1);
unlock_super(sb);
jbd2_journal_unlock_updates(EXT4_SB(sb)->s_journal);
}
}
/* remount的函数 */
static int ext4_remount (struct super_block * sb, int * flags, char * data)
{
struct ext4_super_block * es;
struct ext4_sb_info *sbi = EXT4_SB(sb);
ext4_fsblk_t n_blocks_count = 0;
unsigned long old_sb_flags;
struct ext4_mount_options old_opts;
int err;
#ifdef CONFIG_QUOTA
int i;
#endif
/* 存储原先的挂载选项 */
old_sb_flags = sb->s_flags;
old_opts.s_mount_opt = sbi->s_mount_opt;
old_opts.s_resuid = sbi->s_resuid;
old_opts.s_resgid = sbi->s_resgid;
old_opts.s_commit_interval = sbi->s_commit_interval;
#ifdef CONFIG_QUOTA
old_opts.s_jquota_fmt = sbi->s_jquota_fmt;
for (i = 0; i < MAXQUOTAS; i++)
old_opts.s_qf_names[i] = sbi->s_qf_names[i];
#endif
/*
* 检查选项从而使用新的挂载选项.
*/
if (!parse_options(data, sb, NULL, NULL, &n_blocks_count, 1)) {
err = -EINVAL;
goto restore_opts;
}
if (sbi->s_mount_opt & EXT4_MOUNT_ABORT)
ext4_abort(sb, __FUNCTION__, "Abort forced by user");
sb->s_flags = (sb->s_flags & ~MS_POSIXACL) |
((sbi->s_mount_opt & EXT4_MOUNT_POSIX_ACL) ? MS_POSIXACL : 0);
es = sbi->s_es;
/* 初始化新的日志参数 */
ext4_init_journal_params(sb, sbi->s_journal);
if ((*flags & MS_RDONLY) != (sb->s_flags & MS_RDONLY) ||
n_blocks_count > ext4_blocks_count(es)) {
if (sbi->s_mount_opt & EXT4_MOUNT_ABORT) {
err = -EROFS;
goto restore_opts;
}
/* 要只读挂载 */
if (*flags & MS_RDONLY) {
/*
* 首先设置超级块的flag为只读的
*/
sb->s_flags |= MS_RDONLY;
if (!(es->s_state & cpu_to_le16(EXT4_VALID_FS)) &&
(sbi->s_mount_state & EXT4_VALID_FS))
es->s_state = cpu_to_le16(sbi->s_mount_state);
ext4_mark_recovery_complete(sb, es);
} else {
__le32 ret;
/* 如果原先是只读的，并且不支持非只读挂载，打印错误 */
if ((ret = EXT4_HAS_RO_COMPAT_FEATURE(sb,
~EXT4_FEATURE_RO_COMPAT_SUPP))) {
printk(KERN_WARNING "EXT4-fs: %s: couldn't "
"remount RDWR because of unsupported "
"optional features (%x).\n",
sb->s_id, le32_to_cpu(ret));
err = -EROFS;
goto restore_opts;
}
/*
* 挂载文件系统为可读写的
*/
ext4_clear_journal_err(sb, es);
sbi->s_mount_state = le16_to_cpu(es->s_state);
if ((err = ext4_group_extend(sb, es, n_blocks_count)))
goto restore_opts;
if (!ext4_setup_super (sb, es, 0))
sb->s_flags &= ~MS_RDONLY;
}
}
#ifdef CONFIG_QUOTA
/* 释放原先的配额限制字符串 */
for (i = 0; i < MAXQUOTAS; i++)
if (old_opts.s_qf_names[i] &&
old_opts.s_qf_names[i] != sbi->s_qf_names[i])
kfree(old_opts.s_qf_names[i]);
#endif
return 0;
restore_opts:
sb->s_flags = old_sb_flags;
sbi->s_mount_opt = old_opts.s_mount_opt;
sbi->s_resuid = old_opts.s_resuid;
sbi->s_resgid = old_opts.s_resgid;
sbi->s_commit_interval = old_opts.s_commit_interval;
#ifdef CONFIG_QUOTA
/* 新的配额字符串 */
sbi->s_jquota_fmt = old_opts.s_jquota_fmt;
for (i = 0; i < MAXQUOTAS; i++) {
if (sbi->s_qf_names[i] &&
old_opts.s_qf_names[i] != sbi->s_qf_names[i])
kfree(sbi->s_qf_names[i]);
sbi->s_qf_names[i] = old_opts.s_qf_names[i];
}
#endif
return err;
}
/* 返回ext4的文件系统状态，存放在buf里 */
static int ext4_statfs (struct dentry * dentry, struct kstatfs * buf)
{
struct super_block *sb = dentry->d_sb;
struct ext4_sb_info *sbi = EXT4_SB(sb);
struct ext4_super_block *es = sbi->s_es;
ext4_fsblk_t overhead;
int i;
/* 如果是minix的文件系统，开始就是状态信息，不需要向后移动 */
if (test_opt (sb, MINIX_DF))
overhead = 0;
else {
unsigned long ngroups;
/* 组的数目 */
ngroups = EXT4_SB(sb)->s_groups_count;
smp_rmb();
/*
* 计算偏移
*/
/*
* first_data_block之前的都是数据偏移*/
overhead = le32_to_cpu(es->s_first_data_block);
/*
* Add the overhead attributed to the superblock and
* block group descriptors. If the sparse superblocks
* feature is turned on, then not all groups have this.
*/
for (i = 0; i < ngroups; i++) {
overhead += ext4_bg_has_super(sb, i) +
ext4_bg_num_gdb(sb, i);
cond_resched();
}
/*
* 每一个块组都有inode位图，数据块位图，inode table
*/
overhead += (ngroups * (2 + EXT4_SB(sb)->s_itb_per_group));
}
/* buf的赋值 */
buf->f_type = EXT4_SUPER_MAGIC;
buf->f_bsize = sb->s_blocksize;
buf->f_blocks = ext4_blocks_count(es) - overhead;
buf->f_bfree = percpu_counter_sum(&sbi->s_freeblocks_counter);
buf->f_bavail = buf->f_bfree - ext4_r_blocks_count(es);
if (buf->f_bfree < ext4_r_blocks_count(es))
buf->f_bavail = 0;
buf->f_files = le32_to_cpu(es->s_inodes_count);
buf->f_ffree = percpu_counter_sum(&sbi->s_freeinodes_counter);
buf->f_namelen = EXT4_NAME_LEN;
return 0;
}
/* 帮助配额同步的函数，在配额文件写入之前我们必须锁住 */
#ifdef CONFIG_QUOTA
/* 返回配额文件的inode */
static inline struct inode *dquot_to_inode(struct dquot *dquot)
{
return sb_dqopt(dquot->dq_sb)->files[dquot->dq_type];
}
/* 配额限制的初始化 */
static int ext4_dquot_initialize(struct inode *inode, int type)
{
handle_t *handle;
int ret, err;
/* 我们创建配额结构体，所以需要足够的保留块 */
handle = ext4_journal_start(inode, 2*EXT4_QUOTA_INIT_BLOCKS(inode->i_sb));
if (IS_ERR(handle))
return PTR_ERR(handle);
/* 配额初始化函数 */
ret = dquot_initialize(inode, type);
err = ext4_journal_stop(handle);
if (!ret)
ret = err;
return ret;
}
/* 释放配额结构体 */
static int ext4_dquot_drop(struct inode *inode)
{
handle_t *handle;
int ret, err;
handle = ext4_journal_start(inode, 2*EXT4_QUOTA_DEL_BLOCKS(inode->i_sb));
if (IS_ERR(handle))
return PTR_ERR(handle);
ret = dquot_drop(inode);
err = ext4_journal_stop(handle);
if (!ret)
ret = err;
return ret;
}
/* 写入配额结构体 */
static int ext4_write_dquot(struct dquot *dquot)
{
int ret, err;
handle_t *handle;
struct inode *inode;
/* 配额文件的inode */
inode = dquot_to_inode(dquot);
handle = ext4_journal_start(inode,
EXT4_QUOTA_TRANS_BLOCKS(dquot->dq_sb));
if (IS_ERR(handle))
return PTR_ERR(handle);
/* 写入函数 */
ret = dquot_commit(dquot);
err = ext4_journal_stop(handle);
if (!ret)
ret = err;
return ret;
}
/* 获得配额结构体 */
static int ext4_acquire_dquot(struct dquot *dquot)
{
int ret, err;
handle_t *handle;
handle = ext4_journal_start(dquot_to_inode(dquot),
EXT4_QUOTA_INIT_BLOCKS(dquot->dq_sb));
if (IS_ERR(handle))
return PTR_ERR(handle);
ret = dquot_acquire(dquot);
err = ext4_journal_stop(handle);
if (!ret)
ret = err;
return ret;
}
/* 释放配额结构体 */
static int ext4_release_dquot(struct dquot *dquot)
{
int ret, err;
handle_t *handle;
handle = ext4_journal_start(dquot_to_inode(dquot),
EXT4_QUOTA_DEL_BLOCKS(dquot->dq_sb));
if (IS_ERR(handle))
return PTR_ERR(handle);
ret = dquot_release(dquot);
err = ext4_journal_stop(handle);
if (!ret)
ret = err;
return ret;
}
/* 标记配额结构体脏 */
static int ext4_mark_dquot_dirty(struct dquot *dquot)
{
/* 判断我们会不会对配额进行日志操作 */
if (EXT4_SB(dquot->dq_sb)->s_qf_names[USRQUOTA] ||
EXT4_SB(dquot->dq_sb)->s_qf_names[GRPQUOTA]) {
dquot_mark_dquot_dirty(dquot);
return ext4_write_dquot(dquot);
} else {
return dquot_mark_dquot_dirty(dquot);
}
}
/* 配额信息写入 */
static int ext4_write_info(struct super_block *sb, int type)
{
int ret, err;
handle_t *handle;
/* 数据块+inode块一共两个 */
handle = ext4_journal_start(sb->s_root->d_inode, 2);
if (IS_ERR(handle))
return PTR_ERR(handle);
/* 调用dquot的函数提交信息 */
ret = dquot_commit_info(sb, type);
err = ext4_journal_stop(handle);
if (!ret)
ret = err;
return ret;
}
/*
* 打开配额
*/
static int ext4_quota_on_mount(struct super_block *sb, int type)
{
return vfs_quota_on_mount(sb, EXT4_SB(sb)->s_qf_names[type],
EXT4_SB(sb)->s_jquota_fmt, type);
}
/*
* 配额打开的时候调用的标准函数
*/
static int ext4_quota_on(struct super_block *sb, int type, int format_id,
char *path)
{
int err;
struct nameidata nd;
/* 不支持配额就直接退出 */
if (!test_opt(sb, QUOTA))
return -EINVAL;
/* 没有找到配额限制 */
if (!EXT4_SB(sb)->s_qf_names[USRQUOTA] &&
!EXT4_SB(sb)->s_qf_names[GRPQUOTA])
return vfs_quota_on(sb, type, format_id, path);
err = path_lookup(path, LOOKUP_FOLLOW, &nd);
if (err)
return err;
/* 配额文件不在这个文件系统里 */
if (nd.mnt->mnt_sb != sb) {
path_release(&nd);
return -EXDEV;
}
/* 配额文件不在根目录 */
if (nd.dentry->d_parent->d_inode != sb->s_root->d_inode)
printk(KERN_WARNING
"EXT4-fs: Quota file not on filesystem root. "
"Journalled quota will not work.\n");
path_release(&nd);
return vfs_quota_on(sb, type, format_id, path);
}
/* 从配额文件里读取配额限制，不从pagecache中读取，这样就避免了获取锁，因为配额文件不会被截断 */
static ssize_t ext4_quota_read(struct super_block *sb, int type, char *data,
size_t len, loff_t off)
{
struct inode *inode = sb_dqopt(sb)->files[type];
/* 计算出来配额文件的块数和块内偏移 */
sector_t blk = off >> EXT4_BLOCK_SIZE_BITS(sb);
int err = 0;
int offset = off & (sb->s_blocksize - 1);
int tocopy;
size_t toread;
struct buffer_head *bh;
loff_t i_size = i_size_read(inode);
if (off > i_size)
return 0;
if (off+len > i_size)
len = i_size-off;
toread = len;
/* 循环读取配额文件内容 */
while (toread > 0) {
tocopy = sb->s_blocksize - offset < toread ?
sb->s_blocksize - offset : toread;
bh = ext4_bread(NULL, inode, blk, 0, &err);
if (err)
return err;
if (!bh) /* A hole? */
memset(data, 0, tocopy);
else
memcpy(data, bh->b_data+offset, tocopy);
brelse(bh);
offset = 0;
toread -= tocopy;
data += tocopy;
blk++;
}
return len;
}
/* 写入配额文件 */
static ssize_t ext4_quota_write(struct super_block *sb, int type,
const char *data, size_t len, loff_t off)
{
struct inode *inode = sb_dqopt(sb)->files[type];
/* 计算出来配额文件的块数和块内偏移 */
sector_t blk = off >> EXT4_BLOCK_SIZE_BITS(sb);
int err = 0;
int offset = off & (sb->s_blocksize - 1);
int tocopy;
int journal_quota = EXT4_SB(sb)->s_qf_names[type] != NULL;
size_t towrite = len;
struct buffer_head *bh;
handle_t *handle = journal_current_handle();
/* 写入之前先锁，然后循环读取配额，然后修改内容 */
mutex_lock_nested(&inode->i_mutex, I_MUTEX_QUOTA);
while (towrite > 0) {
tocopy = sb->s_blocksize - offset < towrite ?
sb->s_blocksize - offset : towrite;
/* 读取至缓冲区首部 */
bh = ext4_bread(handle, inode, blk, 1, &err);
if (!bh)
goto out;
if (journal_quota) {
err = ext4_journal_get_write_access(handle, bh);
if (err) {
brelse(bh);
goto out;
}
}
lock_buffer(bh);
/* 写入修改 */
memcpy(bh->b_data+offset, data, tocopy);
flush_dcache_page(bh->b_page);
unlock_buffer(bh);
/* 如果有配额日志的话就要日志修改 */
if (journal_quota)
err = ext4_journal_dirty_metadata(handle, bh);
else {
/* Always do at least ordered writes for quotas */
err = ext4_journal_dirty_data(handle, bh);
mark_buffer_dirty(bh);
}
brelse(bh);
if (err)
goto out;
offset = 0;
towrite -= tocopy;
data += tocopy;
blk++;
}
out:
if (len == towrite)
return err;
if (inode->i_size < off+len-towrite) {
i_size_write(inode, off+len-towrite);
EXT4_I(inode)->i_disksize = inode->i_size;
}
inode->i_version++;
inode->i_mtime = inode->i_ctime = CURRENT_TIME;
ext4_mark_inode_dirty(handle, inode);
mutex_unlock(&inode->i_mutex);
return len - towrite;
}
#endif
/* 获取超级块 */
static int ext4_get_sb(struct file_system_type *fs_type,
int flags, const char *dev_name, void *data, struct vfsmount *mnt)
{
return get_sb_bdev(fs_type, flags, dev_name, data, ext4_fill_super, mnt);
}
/* 注册文件系统的结构体 */
static struct file_system_type ext4dev_fs_type = {
.owner = THIS_MODULE,
.name = "ext4dev",
.get_sb = ext4_get_sb,
.kill_sb = kill_block_super,
.fs_flags = FS_REQUIRES_DEV,
};
/* ext4的模块初始化函数 */
static int __init init_ext4_fs(void)
{
/* 一次初始化扩展属性，inode缓存 */
int err = init_ext4_xattr();
if (err)
return err;
err = init_inodecache();
if (err)
goto out1;
/* 注册ext4文件系统 */
err = register_filesystem(&ext4dev_fs_type);
if (err)
goto out;
return 0;
out:
destroy_inodecache();
out1:
exit_ext4_xattr();
return err;
}
static void __exit exit_ext4_fs(void)
{
unregister_filesystem(&ext4dev_fs_type);
destroy_inodecache();
exit_ext4_xattr();
}
MODULE_AUTHOR("Remy Card, Stephen Tweedie, Andrew Morton, Andreas Dilger, Theodore Ts'o and others");
MODULE_DESCRIPTION("Fourth Extended Filesystem with extents");
MODULE_LICENSE("GPL");
module_init(init_ext4_fs)
module_exit(exit_ext4_fs)