Enum MemoryPoolKind 

pub enum MemoryPoolKind {
    Sliced {
        page_size: u64,
        max_slice_size: u64,
        max_pool_size: Option<u64>,
    },
    Exclusive {
        max_alloc_size: u64,
    },
    Direct,
    Persistent,
}

The pool shape a MemoryPoolReport describes, carrying the pool’s effective configuration (after alignment rounding and page-size shrinking).

Variants

Sliced

Allocations are slices carved from shared pages.

Fields

page_size: u64

The size of each device page.

max_slice_size: u64

The largest allocation the pool accepts.

max_pool_size: Option<u64>

The pool’s byte cap (None grows unbounded).

Exclusive

Every allocation is its own device page.

Fields

max_alloc_size: u64

The largest allocation the pool accepts.

Direct

One device allocation per reservation, sized to the request, reused by exact size and returned to the driver only under memory pressure. Wastes only alignment padding, and pays a driver allocation per distinct size rather than per page.

Persistent

Exact-fit slices that are reused only by identical size.

Trait Implementations

impl Clone for MemoryPoolKind

fn clone(&self) -> MemoryPoolKind

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for MemoryPoolKind

impl Debug for MemoryPoolKind

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

Formats the value using the given formatter.

impl Eq for MemoryPoolKind

impl PartialEq for MemoryPoolKind

fn eq(&self, other: &MemoryPoolKind) -> bool

Equality operator ==.

fn ne(&self, other: &Rhs) -> bool

Inequality operator !=.

impl StructuralPartialEq for MemoryPoolKind