Mojo function

mla_sm100_prefill

mla_sm100_prefill[output_type: DType, q_type: DType, KVType: MHAOperand, KRopeType: MHAOperand, MaskType: MHAMask, ScoreModType: ScoreModTrait, MaxPromptLenType: OptionallyStaticInt, //, config: MHAConfig[dtype], group: Int, q_depth: Int, cache_depth: Int, use_score_mod: Bool, _is_cache_length_accurate: Bool](output: LayoutTensor[output_type, layout, origin, element_layout=element_layout, layout_int_type=layout_int_type, linear_idx_type=linear_idx_type, masked=masked, alignment=alignment], q: LayoutTensor[q_type, layout, origin, element_layout=element_layout, layout_int_type=layout_int_type, linear_idx_type=linear_idx_type, masked=masked, alignment=alignment], k: KVType, v: KVType, k_rope: KRopeType, mask_functor: MaskType, score_mod_functor: ScoreModType, valid_length: LayoutTensor[DType.uint32, layout, origin, element_layout=element_layout, layout_int_type=layout_int_type, linear_idx_type=linear_idx_type, masked=masked, alignment=alignment], max_prompt_len: MaxPromptLenType, scale: Float32, batch_size: Int, ctx: DeviceContext)