Nothing Is Secret

血性

今天肚子不舒服，时不时就要上厕所，所以潘采夫的《贰时代》今天就被我“二”完了。书中皆短篇评论，正适合在厕所阅读，哈哈。

其中涉及不少文人学者之间的矛盾及口水混战，或许是文人相轻吧，并没有出现面对面的激烈交锋，所以也就失去了将真理大白于天下的机会。

俗语说得好，是骡子是马出来溜溜！只是现代人头脑发达，早已不屑于这种简单的四肢运动。也许还有另一层原因，比如，我就一直渴望奔赴沙场去拼杀一阵，但又担心自己能力不够，如果一上去就被无名小卒给乱刀砍死，那“英名何存”！？

Ext2Fsd 0.51

新版本终于测试完毕，可以发布了！

ChangeLog:

1, FIXME: Data corruption issue, especially for multiple-thread
   writing on XP system
2, FIXME: Set ValidDataLength only for FileEndOfFileInformation
   w/ AdvanceOnly   
3, Support writing to ext4 volumes w/ flex_bg
4, Code cleanup and enhancement for global/volume property settings

Flexible Block Groups (flex_bg)

昨天晚上研究了一下e2fsprogs中关于flex_bg的实现。flex_bg，即Flexible Block Groups，是EXT4文件系统引入的一个feature。简言之就是将之前EXT3中分散在各个group中的bitmaps (block bitmap, inode bitmap)及inode table分别集中起来管理。集中起来至少有两个好处：

  • 减少了磁盘寻道操作：将频繁访问的block group资源放到有磁盘上一块连续区域
  • 可以一次性分配更多block给一个extent/run：以前的group将磁盘空间划分为众多不连续的空间片段，从而导致一个分配请求最大能只申请到一个group所管理的blocks。以最常用的4K BLOCK_SIZE来说，一个group最多能管理4K*8=32K个block（128M），但除去group本身的metadata (bitmap blocks: 2  inodes table: (32768 * 128  + 4095)/4096 = 1024)，还能剩下31742个block空闲，如果此group包含spare_super，还要减去super_block所占用的一个block。

明白了上面的问题，再来看flex_bg的实现就比较容易了。在创建EXT4卷时，mke2fs会根据用户指定的flex block group大小（flex_bg_size必须为，2的幂，单位为group），将最前的flex_bg_size个groups集中起来管理。看下面的例子：

实验用得是320G的硬盘，只有一个分区：

[root@srv ~]# fdisk -l /dev/sdc

Disk /dev/sdc: 320.1 GB, 320072931328 bytes
255 heads, 63 sectors/track, 38913 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0xe4afe4af

   Device Boot      Start         End      Blocks   Id  System
/dev/sdc1               1       38914   312568832    7  HPFS/NTFS

指定flex_bg_size为256个group:

[root@srv ~]# mke2fs -j -O flex_bg,extents,uninit_bg -G 256 -I 256 /dev/sdc

再用debugfs来查看新建EXT4卷group descriptions:

[root@srv ~]# debugfs /dev/sdc1

debugfs:  stats
Filesystem volume name:   <none>
Last mounted on:          <not available>
Filesystem UUID:          5be014f5-5a27-4cf1-81dc-d1f55e71dfdd
Filesystem magic number:  0xEF53
Filesystem revision #:    1 (dynamic)
Filesystem features:      has_journal ext_attr resize_inode dir_index filetype extent flex_bg sparse_super large_file uninit_bg
Filesystem flags:         signed_directory_hash
Default mount options:    (none)
Filesystem state:         clean
Errors behavior:          Continue
Filesystem OS type:       Linux
Inode count:              19537920
Block count:              78142208
Reserved block count:     3907110
Free blocks:              76867144
Free inodes:              19537909
First block:              0
Block size:               4096
Fragment size:            4096
Reserved GDT blocks:      1005
Blocks per group:         32768
Fragments per group:      32768
Inodes per group:         8192
Inode blocks per group:   512
Flex block group size:    256
Filesystem created:       Fri Jul  8 23:02:47 2011
Last mount time:          n/a
Last write time:          Fri Jul  8 23:06:48 2011
Mount count:              0
Maximum mount count:      20
Last checked:             Fri Jul  8 23:02:47 2011
Check interval:           15552000 (6 months)
Next check after:         Wed Jan  4 23:02:47 2012
Lifetime writes:          4904 MB
Reserved blocks uid:      0 (user root)
Reserved blocks gid:      0 (group root)
First inode:              11
Inode size:              256
Required extra isize:     28
Desired extra isize:      28
Journal inode:            8
Default directory hash:   half_md4
Directory Hash Seed:      63bc1c54-cf76-4546-99a3-7aca37c86fc1
Journal backup:           inode blocks
Directories:              2
Group  0: block bitmap at 1025, inode bitmap at 1281, inode table at 1537
           4089 free blocks, 8181 free inodes, 2 used directories, 8181 unused inodes
           [Checksum 0x2e4a]
Group  1: block bitmap at 1026, inode bitmap at 1282, inode table at 2049
           0 free blocks, 8192 free inodes, 0 used directories, 8192 unused inodes
           [Inode not init, Checksum 0x4578]
Group  2: block bitmap at 1027, inode bitmap at 1283, inode table at 2561
           4095 free blocks, 8192 free inodes, 0 used directories, 8192 unused inodes
           [Inode not init, Checksum 0xa897]

……

Group 255: block bitmap at 1280, inode bitmap at 1536, inode table at 142337
           32768 free blocks, 8192 free inodes, 0 used directories, 8192 unused inodes
           [Inode not init, Block not init, Checksum 0xcd10]
Group 256: block bitmap at 8388608, inode bitmap at 8388864, inode table at 8389120
           0 free blocks, 8192 free inodes, 0 used directories, 8192 unused inodes
           [Inode not init, Checksum 0x42cd]

……

Group 2384: block bitmap at 75497552, inode bitmap at 75497808, inode table at 75538944
          23296 free blocks, 8192 free inodes, 0 used directories, 8192 unused inodes
          [Inode not init, Checksum 0xd1d2]

可以看出，group 0-255的block bitmap，inode bitmap及inodes table是连在一起的，如block bitmap是从1025开始到1280，inode bitmap则从1281开始，直到1536结束。inodes talbe也同样。后面从group 256开始，又是常规的不连续方式。

不妨再做个实验，将指定flex_bg_size设为4096,将所有group（共2385个）都包含进flexible block group:

[root@srv ~]# mke2fs -j -O flex_bg,extents,uninit_bg -G 4096 -I 256 /dev/sdc

启动debugfs来查看group descriptions:

[root@srv ~]# debugfs /dev/sdc1

Filesystem volume name:   <none>
Last mounted on:          <not available>
Filesystem UUID:          ab2057e4-2510-4c25-bd72-c2867bebb294
Filesystem magic number:  0xEF53
Filesystem revision #:    1 (dynamic)
Filesystem features:      has_journal ext_attr resize_inode dir_index filetype extent flex_bg sparse_super large_file uninit_bg
Filesystem flags:         signed_directory_hash
Default mount options:    (none)
Filesystem state:         clean
Errors behavior:          Continue
Filesystem OS type:       Linux
Inode count:              19537920
Block count:              78142208
Reserved block count:     3907110
Free blocks:              76867144
Free inodes:              19537909
First block:              0
Block size:               4096
Fragment size:            4096
Reserved GDT blocks:      1005
Blocks per group:         32768
Fragments per group:      32768
Inodes per group:         8192
Inode blocks per group:   512
Flex block group size:    4096
Filesystem created:       Fri Jul  8 23:14:02 2011
Last mount time:          n/a
Last write time:          Fri Jul  8 23:17:59 2011
Mount count:              0
Maximum mount count:      39
Last checked:             Fri Jul  8 23:14:02 2011
Check interval:           15552000 (6 months)
Next check after:         Wed Jan  4 23:14:02 2012
Lifetime writes:          4904 MB
Reserved blocks uid:      0 (user root)
Reserved blocks gid:      0 (group root)
First inode:              11
Inode size:              256
Required extra isize:     28
Desired extra isize:      28
Journal inode:            8
Default directory hash:   half_md4
Directory Hash Seed:      f6f4dbbc-3673-4b58-b11c-2fbae02d7ee3
Journal backup:           inode blocks
Directories:              2
Group  0: block bitmap at 1025, inode bitmap at 5121, inode table at 9217
           7511 free blocks, 8181 free inodes, 2 used directories, 8181 unused inodes
           [Checksum 0xee4f]
Group  1: block bitmap at 1026, inode bitmap at 5122, inode table at 9729
           0 free blocks, 8192 free inodes, 0 used directories, 8192 unused inodes
           [Inode not init, Checksum 0x255d]
Group  2: block bitmap at 1027, inode bitmap at 5123, inode table at 10241
           4095 free blocks, 8192 free inodes, 0 used directories, 8192 unused inodes
           [Inode not init, Checksum 0x047e]

……

Group 2384: block bitmap at 3409, inode bitmap at 7505, inode table at 1265665
           23296 free blocks, 8192 free inodes, 0 used directories, 8192 unused inodes
           [Inode not init, Checksum 0x8406]

看得出，整个flexible block group被分隔三部分，第一部分是所有的block bitmap，第二部分是所有的inode bitmap，最后是所有的inodes table。所有的表项都是连续存放于磁盘上的。