ext4 Data Structures and Algorithms

• 1. About this Book
  • 1.1. License
  • 1.2. Terminology
  • 1.3. Other References
• 2. High Level Design
  • 2.1. Blocks
  • 2.2. Layout
  • 2.3. Flexible Block Groups
  • 2.4. Meta Block Groups
  • 2.5. Lazy Block Group Initialization
  • 2.6. Special inodes
  • 2.7. Block and Inode Allocation Policy
  • 2.8. Checksums
  • 2.9. Bigalloc
  • 2.10. Inline Data
    • 2.10.1. Inline Directories
  • 2.11. Large Extended Attribute Values
  • 2.12. Verity files
  • 2.13. Atomic Block Writes
    • 2.13.1. Introduction
    • 2.13.2. Requirements
    • 2.13.3. Multi-fsblock Implementation Details
    • 2.13.4. Handling Split Extents Across Leaf Blocks
    • 2.13.5. Handling Journal transactions
  • 2.14. How to
    • 2.14.1. Creating Filesystems with Atomic Write Support
    • 2.14.2. Application Interface
  • 2.15. Hardware Support
  • 2.16. See Also
• 3. Global Structures
  • 3.1. Super Block
  • 3.2. Block Group Descriptors
  • 3.3. Block and inode Bitmaps
  • 3.4. Inode Table
  • 3.5. Multiple Mount Protection
  • 3.6. Journal (jbd2)
    • 3.6.1. Layout
    • 3.6.2. External Journal
    • 3.6.3. Block Header
    • 3.6.4. Super Block
    • 3.6.5. Descriptor Block
    • 3.6.6. Data Block
    • 3.6.7. Revocation Block
    • 3.6.8. Commit Block
    • 3.6.9. Fast commits
    • 3.6.10. Fast Commit Replay Idempotence
    • 3.6.11. Journal Checkpoint
  • 3.7. Orphan file
• 4. Dynamic Structures
  • 4.1. Index Nodes
    • 4.1.1. Inode Size
    • 4.1.2. Finding an Inode
    • 4.1.3. Inode Timestamps
  • 4.2. The Contents of inode.i_block
    • 4.2.1. Symbolic Links
    • 4.2.2. Direct/Indirect Block Addressing
    • 4.2.3. Extent Tree
    • 4.2.4. Inline Data
  • 4.3. Directory Entries
    • 4.3.1. Linear (Classic) Directories
    • 4.3.2. Hash Tree Directories
  • 4.4. Extended Attributes
    • 4.4.1. Attribute Name Indices
    • 4.4.2. POSIX ACLs