Filesystem

Virtual Filesystem (VFS)

MerlionOS uses a unified VFS layer that abstracts six filesystem backends behind a common interface. All filesystems implement the same trait for open, read, write, close, stat, and directory operations.

Supported Filesystems

Filesystem	Type	Description
devfs	Virtual	Device nodes (/dev/null, /dev/serial, /dev/fb0, /dev/audio)
procfs	Virtual	Kernel state with 28 entries (/proc/uptime, /proc/meminfo, /proc/cpuinfo, etc.)
sysfs	Virtual	Device model and bus topology (/sys/devices/, /sys/bus/, /sys/power/)
tmpfs	In-memory	Writable temporary files, backed by page cache
MF16	On-disk	FAT16-like filesystem for block devices
ext4	On-disk	Linux-compatible filesystem with extents, journaling, and htree directories

Mount Hierarchy

/
├── bin/            # user-space ELF programs
├── dev/            # devfs — device nodes
│   ├── null        # discard sink
│   ├── serial      # COM1 serial port
│   ├── fb0         # framebuffer device
│   └── audio       # audio mixer
├── proc/           # procfs — 28 kernel entries
│   ├── uptime      # system uptime
│   ├── meminfo     # heap statistics
│   ├── tasks       # running task list
│   ├── cpuinfo     # CPU topology & features
│   ├── mounts      # active mount table
│   ├── modules     # loaded kernel modules
│   ├── interrupts  # IRQ counters
│   ├── net/        # network statistics
│   └── ...         # additional entries
├── sys/            # sysfs — device model
│   ├── devices/    # device hierarchy
│   ├── bus/        # bus types (PCI, virtio)
│   └── power/      # P/C-states, thermal
├── mnt/            # ext4 / MF16 block mounts
└── tmp/            # tmpfs — writable temp files

MF16 Filesystem

A FAT16-like filesystem (MerlionFS-16) for block devices:

• Cluster-based allocation with chain tracking
• Directory entries with 8.3 filenames
• Block device abstraction layer
• RAM disk and Virtio block device backends

ext4 Filesystem

Linux-compatible ext4 implementation (v51):

• Extents — efficient mapping of large contiguous regions
• Journaling — crash recovery via write-ahead log
• Htree directories — hash-indexed directory lookup for large directories
• Superblock and group descriptors — standard ext4 on-disk layout
• Read and write support for block devices

tmpfs

In-memory filesystem (v58):

• Page-cache-backed storage, no disk I/O
• Full read/write/directory support
• Automatic cleanup on unmount
• Used for /tmp and scratch storage

procfs

Comprehensive process and kernel information filesystem (v56):

• 28 entries exposing kernel state
• Per-process directories (/proc/[pid]/)
• Network, memory, CPU, module, and interrupt statistics
• Dynamic content generated on read

sysfs

Device model filesystem (v57):

• Hierarchical device tree mirroring kernel device model
• Bus enumeration (PCI, virtio)
• Power management attributes (P-states, C-states, thermal)
• Read/write attributes for device configuration

File Descriptors

POSIX-like file descriptor table per process:

• open, read, write, close syscall semantics
• stdin (fd 0), stdout (fd 1), stderr (fd 2) initialization
• Named and bidirectional pipes (v58)
• lsof command to inspect open file descriptors
• POSIX ACLs for per-user/group access control (v59)