Struct AttnFa128Kernels 

pub struct AttnFa128Kernels { /* private fields */ }

2026-09-27: The two twins’ handles; 0 when the target lacks the module or METRALE_NO_ATTN_FA128 is present (the originals then run, for A/B measurement).

Implementations

impl AttnFa128Kernels

pub fn resolve(gpu: &dyn GpuBackend) -> Self

pub fn paged( &self, gpu: &dyn GpuBackend, q: DevicePtr, k_cache: DevicePtr, v_cache: DevicePtr, output: DevicePtr, block_table: DevicePtr, q_len: u32, kv_len: u32, q_offset: u32, num_q_heads: u32, num_kv_heads: u32, head_dim: u32, cache_block_size: u32, sliding_window: u32, inv_sqrt_d: f32, stream: u64, ) -> Result<bool>

2026-09-27: In place of ops::prefill_attention_paged_64 (causal, BF16 paged cache).

pub fn contiguous( &self, gpu: &dyn GpuBackend, q: DevicePtr, k: DevicePtr, v: DevicePtr, output: DevicePtr, seq_len: u32, batch: u32, num_q_heads: u32, num_kv_heads: u32, head_dim: u32, inv_sqrt_d: f32, causal: bool, sliding_window: u32, stream: u64, ) -> Result<bool>

2026-09-27: In place of ops::prefill_attention_64 (contiguous BF16 Q/K/V).

Trait Implementations

impl Clone for AttnFa128Kernels

fn clone(&self) -> AttnFa128Kernels

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for AttnFa128Kernels

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

Formats the value using the given formatter.

impl Copy for AttnFa128Kernels