Struct InodeRaw 

#[repr(C, align(8))]
pub struct InodeRaw {

    pub i_ino: u64,
    pub i_atime: u64,
    pub i_ctime: u64,
    pub i_mtime: u64,
    pub i_atime_nsec: u32,
    pub i_ctime_nsec: u32,
    pub i_mtime_nsec: u32,
    pub i_meta_config: u32,
    pub i_uid: u32,
    pub i_gid: u32,
    pub i_mode: u32,
    pub i_flags: u32,
    pub i_nlink: u64,
    pub i_size: u64,
    pub i_blocks: u64,
    pub i_last_seq: u64,
    pub i_last_cno: u64,
    pub i_bmap: BMapRawType,
}

Fields

i_ino: u64

i_atime: u64

i_ctime: u64

i_mtime: u64

i_atime_nsec: u32

i_ctime_nsec: u32

i_mtime_nsec: u32

i_meta_config: u32

i_uid: u32

i_gid: u32

i_mode: u32

i_flags: u32

i_nlink: u64

i_size: u64

i_blocks: u64

i_last_seq: u64

i_last_cno: u64

i_bmap: BMapRawType

Implementations

impl InodeRaw

pub const FLAG_INLINE: u32 = 0x1

i_flags bit marking a fully-inlined small file: its bytes live in this inode’s tail region (i_blocks..i_bmap) instead of in data blocks, and the file has NO separate FILE/<uuid> object. i_size is the length.

pub const FLAG_KEEP_OPEN: u32 = 0x2

i_flags bit marking a file that opts in to cross-mount open-but-unlinked semantics: while any mount holds it open (tracked by an S3 open-lease), an unlink from another mount silly-renames it instead of tombstoning, so its storage survives until the last close. Set via the user.hyperfs.keep_open xattr; off by default (no extra coordination cost). Independent of FLAG_INLINE.

pub fn as_mut_u8_slice(&mut self) -> &mut [u8]

pub fn as_u8_slice(&self) -> &[u8]

pub fn from_u8_slice(buf: &[u8]) -> Self

pub fn is_keep_open(&self) -> bool

True if this inode opts in to cross-mount open-but-unlinked (see [FLAG_KEEP_OPEN]).

pub fn set_keep_open(&mut self, on: bool)

Set/clear [FLAG_KEEP_OPEN].

pub const fn inline_offset() -> usize

Byte offset of the inline-data region within the raw inode (the start of the contiguous tail i_blocks, i_last_seq, i_last_cno, i_bmap).

pub const fn inline_cap() -> usize

Maximum inline payload (the size of that tail region).

pub fn is_inline(&self) -> bool

True if this inode is a fully-inlined small file.

pub fn inline_data(&self) -> &[u8]

The inlined bytes (i_size of them); empty if not inlined.

pub fn set_inline(&mut self, data: &[u8])

Mark this inode inline and store data (must fit inline_cap()): sets the flag, i_size = data.len(), zeroes the tail region, copies in. Caller is responsible for there being no FILE/<uuid> object/segments.

pub fn clear_inline(&mut self)

Clear the inline flag (e.g. when spilling to a real segmented file). Does not touch the tail region; the caller rebuilds the bmap.

pub fn set_rdev(&mut self, rdev: u64)

Persist a char/block device node’s rdev. A device node has no data segments, so the otherwise-unused i_last_cno slot holds it. Call after set_inline(&[]) (which zeroes the tail). Inode::to_stat reads it back for device-mode inodes.

pub fn rdev(&self) -> u64

The persisted device rdev (see [set_rdev]); meaningful only for a char/block device-mode inode.

Trait Implementations

impl Clone for InodeRaw

fn clone(&self) -> InodeRaw

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Copy for InodeRaw

impl Debug for InodeRaw

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for InodeRaw

fn default() -> Self

Returns the “default value” for a type.