* Copyright (c) 2025 Huawei Technologies Co., Ltd.
* This program is free software, you can redistribute it and/or modify it under the terms and conditions of
* CANN Open Software License Agreement Version 2.0 (the "License").
* Please refer to the License for details. You may not use this file except in compliance with the License.
* 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 FITNESS FOR A PARTICULAR PURPOSE.
* See LICENSE in the root of the software repository for the full text of the License.
*/
* \file conv_util.h
* \brief
*/
#ifndef API_CONV_UTIL_H
#define API_CONV_UTIL_H
#include "kernel_utils.h"
namespace ConvApi {
constexpr static uint32_t K0_BIAS = 8;
constexpr static uint32_t NUM_TWO = 2;
constexpr static uint32_t BLOCK_L0_N = 16;
constexpr static uint32_t BLOCK_L0_M = 16;
constexpr static uint32_t DATA_COPY_OP_LEN = 16;
constexpr static uint64_t C0_SIZE = 32;
constexpr static uint64_t FRACTAL_SIZE = 512;
constexpr static uint64_t BT_SIZE = 64;
constexpr static uint64_t SIZE_OF_FP16 = 2;
constexpr static uint64_t PAD_IDX_T = 2;
constexpr static uint64_t PAD_IDX_B = 3;
constexpr static uint64_t PAD_IDX_L = 0;
constexpr static uint64_t PAD_IDX_R = 1;
constexpr static uint64_t MAX_PAD_R = 255;
constexpr static uint64_t FMAP_BATCH_DIM = 0;
constexpr static uint64_t FMAP_CIN_DIM = 1;
constexpr static uint64_t FMAP_H_DIM = 2;
constexpr static uint64_t FMAP_W_DIM = 3;
constexpr static uint64_t KERNEL_COUT_DIM = 0;
constexpr static uint64_t KERNEL_H_DIM = 2;
constexpr static uint64_t KERNEL_W_DIM = 3;
constexpr static uint64_t OUTPUT_H_DIM = 2;
constexpr static uint64_t OUTPUT_W_DIM = 3;
constexpr static uint32_t BLOCK_SIZE = 512;
constexpr static uint32_t AL1_BLOCK_SIZE = 32;
constexpr static uint32_t BT_BLOCK_SIZE = 64;
constexpr static uint64_t F8_NUM_IN_F16 = 2;
constexpr static uint64_t MAX_UINT16 = 65535;
constexpr static uint32_t TRANS_DATA_C0 = 16;
static __aicore__ inline size_t AlignB(uint64_t a, uint64_t b)
{
return ((a + b - 1) / b) * b;
}
static __aicore__ inline size_t CeilDIV(uint64_t a, uint64_t b)
{
return (a + b - 1) / b;
}
template <class Intf>
static __aicore__ inline size_t GetInputkInOneC0Block()
{
return C0_SIZE / sizeof(typename Intf::InputT);
}
enum class IterateOrder : uint32_t {
ORDER_MTERFIRST = 0,
ORDER_NTERFIRST,
UNDEF,
};
}
#endif
