ops-math/math/sign/tests/ut/op_api/test_aclnn_sign.cpp-代码预览-ops-math:基于 CANN 的数学类算子库项目 - AtomGit

cann-robot补充exp、sign、log、ones_like、zeros_like、equal、not_equal算子的ut，并优化include
/**
 * 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 <float.h>

#include <array>
#include <iostream>
#include <vector>

#include "gtest/gtest.h"
#include "../../../op_api/aclnn_sign.h"
#include "op_api_ut_common/op_api_ut.h"
#include "op_api_ut_common/scalar_desc.h"
#include "op_api_ut_common/tensor_desc.h"

using namespace std;

class sign_test : public testing::Test {
protected:
    static void SetUpTestCase()
    {
        cout << "sign_test SetUp" << endl;
    }

    static void TearDownTestCase()
    {
        cout << "sign_test TeastDown" << endl;
    }
};

// TEST_F(sign_test, test_sign_dataType_error)
// {
//     vector<aclDataType> ValidList = {ACL_DT_UNDEFINED, ACL_INT8,   ACL_UINT8,  ACL_UINT16,
//                                      ACL_UINT32,       ACL_UINT64, ACL_STRING, ACL_BF16};

//     int length = ValidList.size();
//     vector<int64_t> input_dim = {2, 16, 32, 16};
//     vector<int64_t> result_dim = {2, 16, 32, 16};

//     for (int i = 0; i < length; i++) {
//         auto inputDesc = TensorDesc(input_dim, ValidList[i], ACL_FORMAT_ND).ValueRange(-2, 2);
//         auto resultDesc = TensorDesc(result_dim, ValidList[i], ACL_FORMAT_ND).Precision(0.0001, 0.0001);

//         auto ut = OP_API_UT(aclnnSign, INPUT(inputDesc), OUTPUT(resultDesc));
//         uint64_t workspaceSize = 0;
//         aclnnStatus aclRet = ut.TestGetWorkspaceSize(&workspaceSize);
//         EXPECT_EQ(aclRet, ACLNN_ERR_PARAM_INVALID);
//     }
// }

TEST_F(sign_test, test_sign_format)
{
    vector<aclFormat> ValidList = {
        ACL_FORMAT_UNDEFINED, ACL_FORMAT_NCHW,        ACL_FORMAT_NHWC,  ACL_FORMAT_ND,    ACL_FORMAT_NC1HWC0,
        ACL_FORMAT_FRACTAL_Z, ACL_FORMAT_NC1HWC0_C04, ACL_FORMAT_HWCN,  ACL_FORMAT_NDHWC, ACL_FORMAT_FRACTAL_NZ,
        ACL_FORMAT_NCDHW,     ACL_FORMAT_NDC1HWC0,    ACL_FRACTAL_Z_3D, ACL_FORMAT_NC,    ACL_FORMAT_NCL};

    int length = ValidList.size();
    vector<int64_t> input_dim = {2, 16, 32, 16};
    vector<int64_t> result_dim = {2, 16, 32, 16};

    for (int i = 0; i < length; i++) {
        auto inputDesc = TensorDesc(input_dim, ACL_FLOAT, ValidList[i]).ValueRange(-1, 1);
        auto resultDesc = TensorDesc(result_dim, ACL_FLOAT, ValidList[i]).Precision(0.0001, 0.0001);

        auto ut = OP_API_UT(aclnnSign, INPUT(inputDesc), OUTPUT(resultDesc));
        uint64_t workspaceSize = 0;
        aclnnStatus aclRet = ut.TestGetWorkspaceSize(&workspaceSize);
        EXPECT_EQ(aclRet, ACLNN_SUCCESS);
        // ut.TestPrecision();  // comment bcz of timeout in model tests (277860 ms)
    }
}

TEST_F(sign_test, test_sign_inconsistent_shape)
{
    auto inputDesc = TensorDesc({2, 16, 32, 16}, ACL_FLOAT, ACL_FORMAT_ND);
    auto resultDesc = TensorDesc({2, 16, 32, 18}, ACL_FLOAT, ACL_FORMAT_ND);

    auto ut = OP_API_UT(aclnnSign, INPUT(inputDesc), OUTPUT(resultDesc));

    uint64_t workspaceSize = 0;
    aclnnStatus aclRT = ut.TestGetWorkspaceSize(&workspaceSize);
    EXPECT_EQ(aclRT, ACLNN_ERR_PARAM_INVALID);
}

TEST_F(sign_test, test_sign_empty_tensor)
{
    auto inputDesc = TensorDesc({2, 16, 0, 16}, ACL_FLOAT, ACL_FORMAT_ND).ValueRange(-1, 1);
    auto resultDesc = TensorDesc({2, 16, 0, 16}, ACL_FLOAT, ACL_FORMAT_ND).Precision(0.0001, 0.0001);

    auto ut = OP_API_UT(aclnnSign, INPUT(inputDesc), OUTPUT(resultDesc));

    uint64_t workspaceSize = 0;
    aclnnStatus aclRT = ut.TestGetWorkspaceSize(&workspaceSize);
    EXPECT_EQ(aclRT, ACLNN_SUCCESS);
    // ut.TestPrecision();
}

TEST_F(sign_test, test_sign_nullptr_input)
{
    auto resultDesc = TensorDesc({2, 16, 32, 18}, ACL_FLOAT, ACL_FORMAT_ND).Precision(0.0001, 0.0001);

    auto ut = OP_API_UT(aclnnSign, INPUT((aclTensor*)nullptr), OUTPUT(resultDesc));

    uint64_t workspaceSize = 0;
    aclnnStatus aclRT = ut.TestGetWorkspaceSize(&workspaceSize);
    EXPECT_EQ(aclRT, ACLNN_ERR_PARAM_NULLPTR);
}

TEST_F(sign_test, test_sign_nullptr_result)
{
    auto inputDesc = TensorDesc({2, 16, 0, 16}, ACL_FLOAT, ACL_FORMAT_ND).ValueRange(-1, 1);

    auto ut = OP_API_UT(aclnnSign, INPUT(inputDesc), OUTPUT((aclTensor*)nullptr));

    uint64_t workspaceSize = 0;
    aclnnStatus aclRT = ut.TestGetWorkspaceSize(&workspaceSize);
    EXPECT_EQ(aclRT, ACLNN_ERR_PARAM_NULLPTR);
}

TEST_F(sign_test, test_sign_aicore_dataType_1)
{
    vector<aclDataType> ValidList = {ACL_FLOAT, ACL_INT32, ACL_BOOL};

    int length = ValidList.size();
    vector<int64_t> input_dim = {2, 16, 32, 16};
    vector<int64_t> result_dim = {2, 16, 32, 16};

    for (int i = 0; i < length; i++) {
        auto inputDesc = TensorDesc(input_dim, ValidList[i], ACL_FORMAT_NCHW).ValueRange(-100, 100);
        auto resultDesc = TensorDesc(result_dim, ValidList[i], ACL_FORMAT_NCHW).Precision(0.0001, 0.0001);

        auto ut = OP_API_UT(aclnnSign, INPUT(inputDesc), OUTPUT(resultDesc));
        uint64_t workspaceSize = 0;
        aclnnStatus aclRet = ut.TestGetWorkspaceSize(&workspaceSize);
        EXPECT_EQ(aclRet, ACLNN_SUCCESS);
        // ut.TestPrecision();
    }
}

TEST_F(sign_test, test_sign_aicore_dataType_2)
{
    auto inputDesc = TensorDesc({2, 16, 32, 16}, ACL_FLOAT16, ACL_FORMAT_ND).ValueRange(-1, 1);
    auto resultDesc = TensorDesc({2, 16, 32, 16}, ACL_FLOAT16, ACL_FORMAT_ND).Precision(0.001, 0.001);

    auto ut = OP_API_UT(aclnnSign, INPUT(inputDesc), OUTPUT(resultDesc));

    uint64_t workspaceSize = 0;
    aclnnStatus aclRT = ut.TestGetWorkspaceSize(&workspaceSize);
    EXPECT_EQ(aclRT, ACLNN_SUCCESS);
    // ut.TestPrecision();
}

// TEST_F(sign_test, test_sign_aicpu_dataType_1)
// {
//     vector<aclDataType> ValidList = {ACL_DOUBLE, ACL_COMPLEX64, ACL_COMPLEX128};

//     int length = ValidList.size();
//     vector<int64_t> input_dim = {2, 16, 32, 16};
//     vector<int64_t> result_dim = {2, 16, 32, 16};

//     for (int i = 0; i < length; i++) {
//         auto inputDesc = TensorDesc(input_dim, ValidList[i], ACL_FORMAT_ND).ValueRange(-1, 1);
//         auto resultDesc = TensorDesc(result_dim, ValidList[i], ACL_FORMAT_ND).Precision(0.001, 0.001);

//         auto ut = OP_API_UT(aclnnSign, INPUT(inputDesc), OUTPUT(resultDesc));
//         uint64_t workspaceSize = 0;
//         aclnnStatus aclRet = ut.TestGetWorkspaceSize(&workspaceSize);
//         EXPECT_EQ(aclRet, ACLNN_SUCCESS);
//         // ut.TestPrecision();
//     }
// }

// TEST_F(sign_test, test_sign_aicpu_dataType_2)
// {
//     auto inputDesc = TensorDesc({2, 16, 32, 16}, ACL_INT64, ACL_FORMAT_ND).ValueRange(-1, 1);
//     auto resultDesc = TensorDesc({2, 16, 32, 16}, ACL_INT64, ACL_FORMAT_ND).Precision(0.0001, 0.0001);

//     auto ut = OP_API_UT(aclnnSign, INPUT(inputDesc), OUTPUT(resultDesc));

//     uint64_t workspaceSize = 0;
//     aclnnStatus aclRT = ut.TestGetWorkspaceSize(&workspaceSize);
//     EXPECT_EQ(aclRT, ACLNN_SUCCESS);
//     // ut.TestPrecision();
// }

TEST_F(sign_test, test_sign_aicore_uncontiguous)
{
    auto inputDesc = TensorDesc({2, 16}, ACL_FLOAT, ACL_FORMAT_ND, {1, 2}, 0, {16, 2}).ValueRange(-1, 1);
    auto resultDesc = TensorDesc({2, 16}, ACL_FLOAT, ACL_FORMAT_ND, {1, 2}, 0, {16, 2}).Precision(0.0001, 0.0001);

    auto ut = OP_API_UT(aclnnSign, INPUT(inputDesc), OUTPUT(resultDesc));

    uint64_t workspaceSize = 0;
    aclnnStatus aclRT = ut.TestGetWorkspaceSize(&workspaceSize);
    EXPECT_EQ(aclRT, ACLNN_SUCCESS);
    // ut.TestPrecision();
}

// TEST_F(sign_test, test_sign_aicpu_uncontiguous)
// {
//     auto inputDesc = TensorDesc({2, 16}, ACL_INT64, ACL_FORMAT_ND, {1, 2}, 0, {16, 2}).ValueRange(-1, 1);
//     auto resultDesc = TensorDesc({2, 16}, ACL_INT64, ACL_FORMAT_ND, {1, 2}, 0, {16, 2}).Precision(0.0001, 0.0001);

//     auto ut = OP_API_UT(aclnnSign, INPUT(inputDesc), OUTPUT(resultDesc));

//     uint64_t workspaceSize = 0;
//     aclnnStatus aclRT = ut.TestGetWorkspaceSize(&workspaceSize);
//     EXPECT_EQ(aclRT, ACLNN_SUCCESS);
//     // ut.TestPrecision();
// }

TEST_F(sign_test, test_sign_aicore_diff_dtype)
{
    vector<int64_t> input_dim = {2, 16, 32, 16};
    vector<int64_t> result_dim = {2, 16, 32, 16};

    auto inputDesc = TensorDesc(input_dim, ACL_FLOAT, ACL_FORMAT_ND).ValueRange(-1, 1);
    auto resultDesc = TensorDesc(result_dim, ACL_FLOAT16, ACL_FORMAT_ND).Precision(0.0001, 0.0001);

    auto ut = OP_API_UT(aclnnSign, INPUT(inputDesc), OUTPUT(resultDesc));

    uint64_t workspaceSize = 0;
    aclnnStatus aclRet = ut.TestGetWorkspaceSize(&workspaceSize);
    EXPECT_EQ(aclRet, ACLNN_ERR_PARAM_INVALID);
}

TEST_F(sign_test, test_sign_aicpu_diff_dtype)
{
    vector<int64_t> input_dim = {2, 16, 32, 16};
    vector<int64_t> result_dim = {2, 16, 32, 16};

    auto inputDesc = TensorDesc(input_dim, ACL_COMPLEX64, ACL_FORMAT_ND).ValueRange(-1, 1);
    auto resultDesc = TensorDesc(result_dim, ACL_COMPLEX128, ACL_FORMAT_ND).Precision(0.0001, 0.0001);

    auto ut = OP_API_UT(aclnnSign, INPUT(inputDesc), OUTPUT(resultDesc));

    uint64_t workspaceSize = 0;
    aclnnStatus aclRet = ut.TestGetWorkspaceSize(&workspaceSize);
    EXPECT_EQ(aclRet, ACLNN_ERR_PARAM_INVALID);
}

TEST_F(sign_test, test_sign_shape_larger_8)
{
    vector<aclFormat> ValidList = {
        ACL_FORMAT_UNDEFINED, ACL_FORMAT_NCHW,        ACL_FORMAT_NHWC,  ACL_FORMAT_ND,    ACL_FORMAT_NC1HWC0,
        ACL_FORMAT_FRACTAL_Z, ACL_FORMAT_NC1HWC0_C04, ACL_FORMAT_HWCN,  ACL_FORMAT_NDHWC, ACL_FORMAT_FRACTAL_NZ,
        ACL_FORMAT_NCDHW,     ACL_FORMAT_NDC1HWC0,    ACL_FRACTAL_Z_3D, ACL_FORMAT_NC,    ACL_FORMAT_NCL};

    int length = ValidList.size();

    for (int i = 0; i < length; i++) {
        auto inputDesc = TensorDesc({1, 1, 1, 1, 2, 1, 1, 1, 2}, ACL_FLOAT, ACL_FORMAT_ND).ValueRange(-1, 1);
        auto resultDesc = TensorDesc({1, 1, 1, 1, 2, 1, 1, 1, 2}, ACL_FLOAT, ACL_FORMAT_ND).Precision(0.0001, 0.0001);
        auto ut = OP_API_UT(aclnnSign, INPUT(inputDesc), OUTPUT(resultDesc));

        uint64_t workspaceSize = 0;
        aclnnStatus aclRet = ut.TestGetWorkspaceSize(&workspaceSize);
        EXPECT_EQ(aclRet, ACLNN_SUCCESS);
        // ut.TestPrecision(); // comment bcz of timeout in model tests (702051 ms)
    }
}