orthotope 0.1.5

Arena-based memory allocator with per-thread caches and fixed size classes, designed for allocation-heavy workloads like ML inference and tensor pipelines
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29

use orthotope::{Allocator, AllocatorConfig, ThreadCache};

const fn custom_alignment_config() -> AllocatorConfig {
    AllocatorConfig {
        arena_size: 1 << 20,
        alignment: 128,
        refill_target_bytes: 640,
        local_cache_target_bytes: 640,
    }
}

#[test]
fn small_block_from_custom_aligned_allocator_can_be_freed() {
    let allocator = match Allocator::new(custom_alignment_config()) {
        Ok(allocator) => allocator,
        Err(error) => panic!("expected allocator to initialize: {error}"),
    };
    let mut cache = ThreadCache::new(*allocator.config());

    let ptr = match allocator.allocate_with_cache(&mut cache, 65) {
        Ok(ptr) => ptr,
        Err(error) => panic!("expected 65-byte allocation to succeed: {error}"),
    };

    // SAFETY: `ptr` is the currently live allocation returned above.
    let result = unsafe { allocator.deallocate_with_cache(&mut cache, ptr) };

    assert_eq!(result, Ok(()));
}