/**
 * Copyright (c) Huawei Technologies Co., Ltd. 2025-2025. All rights reserved.
 * MindIE is licensed under Mulan PSL v2.
 * You can use this software according to the terms and conditions of the Mulan PSL v2.
 * You may obtain a copy of Mulan PSL v2 at:
 *          http://license.coscl.org.cn/MulanPSL2
 * THIS SOFTWARE IS PROVIDED ON AN "AS IS" BASIS, WITHOUT WARRANTIES OF ANY KIND,
 * EITHER EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO NON-INFRINGEMENT,
 * MERCHANTABILITY OR FIT FOR A PARTICULAR PURPOSE.
 * See the Mulan PSL v2 for more details.
 */

#ifndef BLOCK_SPARSE_ATTENTION_MINDIE_SD_IMPL_H
#define BLOCK_SPARSE_ATTENTION_MINDIE_SD_IMPL_H

#include <ATen/Tensor.h>
#include <c10/util/Optional.h>
#include <string>
#include <tuple>

// Block sparse attention via aclnnBlockSparseAttentionV2 (BF16/FP16/FP8).
// When dequant scales are not provided (BF16/FP16), nullptr is passed to the kernel.
// When dequant scales are provided (FP8), FLOAT32 scale tensors are used.
// Takes block_sparse_mask (int8). Supports TND and BNSD layouts.
// Returns (attention_out, softmax_lse).
std::tuple<at::Tensor, at::Tensor> block_sparse_attention_impl_npu(const at::Tensor &query, const at::Tensor &key,
    const at::Tensor &value, const c10::optional<at::Tensor> &block_sparse_mask, at::IntArrayRef block_shape,
    std::string q_input_layout, std::string kv_input_layout, int64_t num_key_value_heads, double scale_value,
    int64_t inner_precise, c10::OptionalIntArrayRef actual_seq_lengths, c10::OptionalIntArrayRef actual_seq_lengths_kv,
    int64_t softmax_lse_flag, const c10::optional<at::Tensor> &q_dequant_scale = c10::nullopt,
    const c10::optional<at::Tensor> &k_dequant_scale = c10::nullopt,
    const c10::optional<at::Tensor> &v_dequant_scale = c10::nullopt);

#endif // BLOCK_SPARSE_ATTENTION_MINDIE_SD_IMPL_H