* 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.
*/
#include "aclnn_erfc.h"
#include "erfc.h"
#include "aclnn_kernels/contiguous.h"
#include "aclnn_kernels/cast.h"
#include "opdev/op_dfx.h"
#include "opdev/make_op_executor.h"
#include "aclnn_kernels/common/op_error_check.h"
#include "opdev/platform.h"
#include "op_api/level2_base.h"
using namespace op;
constexpr size_t MAX_DIM_LEN = 8;
static const std::initializer_list<DataType> ASCEND910_DTYPE_DTYPE_SUPPORT_LIST = {
op::DataType::DT_DOUBLE, op::DataType::DT_FLOAT, op::DataType::DT_FLOAT16, op::DataType::DT_BOOL,
op::DataType::DT_INT64};
static const std::initializer_list<DataType> ASCEND910B_DTYPE_DTYPE_SUPPORT_LIST = {
op::DataType::DT_DOUBLE, op::DataType::DT_FLOAT, op::DataType::DT_FLOAT16,
op::DataType::DT_BOOL, op::DataType::DT_INT64, op::DataType::DT_BF16};
static const std::initializer_list<DataType> ASCEND910_DTYPE_SELFREF_LIST = {
op::DataType::DT_DOUBLE, op::DataType::DT_FLOAT, op::DataType::DT_FLOAT16};
static const std::initializer_list<DataType> ASCEND910B_DTYPE_SELFREF_LIST = {
op::DataType::DT_DOUBLE, op::DataType::DT_FLOAT, op::DataType::DT_FLOAT16, op::DataType::DT_BF16};
static bool CheckDtypeValid(const aclTensor* self, const aclTensor* out)
{
auto supportList = GetDtypeSupportListV2(ASCEND910B_DTYPE_DTYPE_SUPPORT_LIST, ASCEND910_DTYPE_DTYPE_SUPPORT_LIST);
OP_CHECK_DTYPE_NOT_SUPPORT(self, supportList, return false);
if (self->GetDataType() == op::DataType::DT_BOOL || self->GetDataType() == op::DataType::DT_INT64) {
OP_CHECK_RESULT_DTYPE_CAST_FAILED(op::DataType::DT_FLOAT, out->GetDataType(), return false);
}
if (self->GetDataType() != op::DataType::DT_BOOL && self->GetDataType() != op::DataType::DT_INT64) {
OP_CHECK_RESULT_DTYPE_CAST_FAILED(self->GetDataType(), out->GetDataType(), return false);
}
return true;
}
static bool CheckInplaceDtypeValid(const aclTensor* selfRef)
{
auto inplaceSupportList = GetDtypeSupportListV2(ASCEND910B_DTYPE_SELFREF_LIST, ASCEND910_DTYPE_SELFREF_LIST);
OP_CHECK_DTYPE_NOT_SUPPORT(selfRef, inplaceSupportList, return false);
return true;
}
static aclnnStatus CheckParamsErfc(const aclTensor* self, const aclTensor* out)
{
CHECK_RET(CheckNotNull2Tensor(self, out), ACLNN_ERR_PARAM_NULLPTR);
CHECK_RET(CheckDtypeValid(self, out), ACLNN_ERR_PARAM_INVALID);
CHECK_RET(CheckSameShape1In1Out(self, out), ACLNN_ERR_PARAM_INVALID);
return ACLNN_SUCCESS;
}
static aclnnStatus CheckInplaceParamsErfc(const aclTensor* selfRef)
{
OP_CHECK_NULL(selfRef, return ACLNN_ERR_PARAM_NULLPTR);
CHECK_RET(CheckInplaceDtypeValid(selfRef), ACLNN_ERR_PARAM_INVALID);
return ACLNN_SUCCESS;
}
aclnnStatus aclnnErfcGetWorkspaceSize(
const aclTensor* self, aclTensor* out, uint64_t* workspaceSize, aclOpExecutor** executor)
{
L2_DFX_PHASE_1(aclnnErfc, DFX_IN(self), DFX_OUT(out));
auto uniqueExecutor = CREATE_EXECUTOR();
CHECK_RET(uniqueExecutor.get() != nullptr, ACLNN_ERR_INNER_CREATE_EXECUTOR);
auto ret = CheckParamsErfc(self, out);
CHECK_RET(ret == ACLNN_SUCCESS, ret);
if (self->IsEmpty()) {
*workspaceSize = 0;
uniqueExecutor.ReleaseTo(executor);
return ACLNN_SUCCESS;
}
auto selfContiguous = l0op::Contiguous(self, uniqueExecutor.get());
CHECK_RET(selfContiguous != nullptr, ACLNN_ERR_INNER_NULLPTR);
auto castSelf = selfContiguous;
if (self->GetDataType() == op::DataType::DT_BOOL || self->GetDataType() == op::DataType::DT_INT64) {
castSelf = l0op::Cast(selfContiguous, op::DataType::DT_FLOAT, uniqueExecutor.get());
CHECK_RET(castSelf != nullptr, ACLNN_ERR_INNER_NULLPTR);
}
auto erfcResult = l0op::Erfc(castSelf, uniqueExecutor.get());
CHECK_RET(erfcResult != nullptr, ACLNN_ERR_INNER_NULLPTR);
auto castErfcResult = l0op::Cast(erfcResult, out->GetDataType(), uniqueExecutor.get());
CHECK_RET(castErfcResult != nullptr, ACLNN_ERR_INNER_NULLPTR);
auto viewCopyResult = l0op::ViewCopy(castErfcResult, out, uniqueExecutor.get());
CHECK_RET(viewCopyResult != nullptr, ACLNN_ERR_INNER_NULLPTR);
*workspaceSize = uniqueExecutor->GetWorkspaceSize();
uniqueExecutor.ReleaseTo(executor);
return ACLNN_SUCCESS;
}
aclnnStatus aclnnInplaceErfcGetWorkspaceSize(
const aclTensor* selfRef, uint64_t* workspaceSize, aclOpExecutor** executor)
{
auto out = const_cast<aclTensor*>(selfRef);
auto ret = CheckInplaceParamsErfc(selfRef);
CHECK_RET(ret == ACLNN_SUCCESS, ret);
return aclnnErfcGetWorkspaceSize(selfRef, out, workspaceSize, executor);
}
aclnnStatus aclnnErfc(void* workspace, uint64_t workspaceSize, aclOpExecutor* executor, const aclrtStream stream)
{
L2_DFX_PHASE_2(aclnnErfc);
return CommonOpExecutorRun(workspace, workspaceSize, executor, stream);
}
aclnnStatus aclnnInplaceErfc(void* workspace, uint64_t workspaceSize, aclOpExecutor* executor, aclrtStream stream)
{
L2_DFX_PHASE_2(aclnnInplaceErfc);
return CommonOpExecutorRun(workspace, workspaceSize, executor, stream);
}